WO2021249197A1 - 数据处理方法、数据处理装置和健康管理装置 - Google Patents

数据处理方法、数据处理装置和健康管理装置 Download PDF

Info

Publication number
WO2021249197A1
WO2021249197A1 PCT/CN2021/096383 CN2021096383W WO2021249197A1 WO 2021249197 A1 WO2021249197 A1 WO 2021249197A1 CN 2021096383 W CN2021096383 W CN 2021096383W WO 2021249197 A1 WO2021249197 A1 WO 2021249197A1
Authority
WO
WIPO (PCT)
Prior art keywords
data
keywords
health management
management device
data processing
Prior art date
Application number
PCT/CN2021/096383
Other languages
English (en)
French (fr)
Inventor
王胄
刘子刚
高原
王同波
杨贵龙
齐峰
韩阳
Original Assignee
京东方科技集团股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 京东方科技集团股份有限公司 filed Critical 京东方科技集团股份有限公司
Priority to US17/772,783 priority Critical patent/US20220383997A1/en
Publication of WO2021249197A1 publication Critical patent/WO2021249197A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H15/00—ICT specially adapted for medical reports, e.g. generation or transmission thereof
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/35—Clustering; Classification
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/22—Matching criteria, e.g. proximity measures
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment

Definitions

  • the embodiments of the present disclosure relate to a data processing method, a data processing device, and a health management device.
  • Common chronic diseases mainly include cardiovascular and cerebrovascular diseases, cancer, diabetes, and chronic respiratory diseases.
  • Cardiovascular and cerebrovascular diseases include hypertension, stroke and coronary heart disease.
  • Statistics show that the number of Chinese citizens suffering from chronic diseases such as hypertension, diabetes, and dyslipidemia has increased.
  • At least one embodiment of the present disclosure provides a data processing method, which includes: performing data processing on device-related data originating from a device associated with a health management device.
  • the performing data processing on the device-related data derived from the device associated with the health management apparatus includes: information based on the first set of keywords related to the template data and Obtaining the structural similarity between the first set of keywords and the second set of keywords by the information of the second set of keywords involved in the device-related data; and calculating the device-related data based on the structural similarity At least part of is transferred to the template data.
  • the performing data processing on the device-related data derived from the device associated with the health management apparatus further includes: receiving a first set of keywords related to the template data And the device-related data; and the information of the first set of keywords includes the feature vector of each keyword of the first set of keywords, and the information of the second set of keywords includes the second set of keywords The feature vector of each keyword of the group keyword.
  • the performing data processing on the device-related data originating from the device associated with the health management apparatus further includes: extracting the second data from the device-related data Group keywords and the hierarchical information of each keyword of the second group of keywords in the device-related data; based on the second group of keywords and each keyword of the second group of keywords The hierarchical information in the device-related data generates a feature vector of each keyword of the second group of keywords.
  • the structural similarity between the first set of keywords and the second set of keywords includes the characteristics of each keyword of the first set of keywords The similarity between the vector and the feature vector of each keyword of the second group of keywords.
  • the first group of keys is obtained based on the information of the first group of keywords related to the template data and the information of the second group of keywords related to the device-related data.
  • the structural similarity between the word and the second set of keywords includes: obtaining the structural similarity between the first set of keywords and the second set of keywords by calculating a similarity matrix S; and the The similarity matrix S satisfies the following expression:
  • v 1 , v 2 ,...v m are the feature vectors of the first set of keywords
  • u 1 , u 2 ,...u b are the feature vectors of the second set of keywords
  • m is the first set of keywords.
  • the number of keywords in the group of keywords, and b is the number of keywords in the second group of keywords.
  • the transmitting at least part of the device-related data to the template data based on the structural similarity includes: transmitting each of the second set of keywords
  • the similarity between the feature vector of each keyword and the feature vector of each keyword of the first group of keywords is the similarity with a value greater than the preset similarity threshold as the correlation similarity; each correlation similarity Among the two corresponding keywords, the keyword belonging to the second group of keywords is used as the first keyword, and the keyword belonging to the first group of keywords among the two keywords corresponding to each of the relevance similarities is used as the keyword belonging to the first group of keywords A second keyword; and associating data associated with the first keyword in the device-related data with the second keyword related to the template data.
  • the performing data processing on the device-related data derived from the device associated with the health management apparatus further includes: receiving an interpolated data set, wherein the interpolated data set The data set includes at least part of the data of other objects in the group of the object corresponding to the data to be supplemented; and the interpolated data set is interpolated to obtain the data to be supplemented.
  • the performing data processing on the device-related data derived from the device associated with the health management apparatus further includes: calculating the data to be supplemented and the interpolated data Collecting the nominal distance of each piece of data; and the interpolating the interpolated data set to obtain the data to be supplemented includes: performing distance inverse interpolation on the interpolated data set based at least on the nominal distance to obtain the to-be supplemented data.
  • the calculating the nominal distance between the data to be supplemented and each piece of data in the interpolated data set includes: based on the data to be supplemented and the interpolated data set The time distance and geographic distance of each piece of data calculate the nominal distance between the data to be supplemented and each piece of data in the interpolated data set.
  • the data processing method further includes receiving time information and geographic location information corresponding to each piece of data in the interpolated data set, and time information and time information corresponding to the data to be supplemented. Geographical location information.
  • the calculating the nominal distance between the data to be supplemented and each piece of data in the interpolated data set includes: calculating the time information corresponding to each piece of data in the interpolated data set and the time information corresponding to the data to be supplemented The time distance between the data to be supplemented and each piece of data in the interpolated data set; the data to be supplemented and the data to be supplemented are calculated based on the geographic information corresponding to each piece of data in the interpolated data set and the geographic information corresponding to the data to be supplemented The geographic distance of each piece of data in the interpolated data set; and the spatial distance between the data to be supplemented and each piece of data in the interpolated data set, and the time distance between the data to be supplemented and each piece of data in the interpolated data set The weighted sum of is used as the nominal distance between the data to be supplemented and each data in the interpolated data set.
  • the data processing method further includes: receiving a damping item of the group of the object corresponding to the data to be supplemented.
  • the interpolating the interpolated data set to obtain the data to be supplemented includes: performing inverse distance interpolation on the interpolated data set based on the nominal distance and the damping term to obtain the data to be supplemented.
  • the data to be supplemented x p satisfies the following expression:
  • the data processing method further includes: receiving information of an object associated with the health management device; based on the information of the object associated with the health management device and a query table Obtain the scores of the object associated with the health management device on a plurality of scoring items; and based on the scores of the object associated with the health management device on the plurality of scoring items as the object associated with the health management device Assign group.
  • the assigning a group to the object associated with the health management device based on the scores of the object associated with the health management device on the plurality of scoring items includes : Obtain the overall score of the object associated with the health management device based on the scores on the multiple scoring items and the weights of the multiple scoring items; and assign the object associated with the health management device based on the overall score Group.
  • the obtaining the overall score of the object associated with the health management device based on the scores on the plurality of scoring items and the weight of the plurality of scoring items includes : Taking the weighted sum of the scores of the object associated with the health management device on the multiple scoring items as the overall score of the object associated with the health management device;
  • the object allocation group associated with the health management device includes: rounding up the overall score as the group serial number of the object associated with the health management device.
  • the information of the object associated with the health management device includes objective information and subjective information; at least part of the objective information is matched with a portable medical device or a wearable medical device.
  • the objects associated with the health management device are monitored and acquired; and at least part of the subjective information is obtained by inquiries by medical workers or electronic questionnaires filled out by the objects associated with the health management device.
  • the multiple scoring items include: age, gender, BMI, systolic blood pressure, diastolic blood pressure, occupational work intensity, number of exacerbations, number of hospitalizations, number of surgical treatments, Normal breathing rate, normal blood oxygen saturation, mental state and exercise state.
  • the data processing method further includes: receiving an overall score of the object associated with the health management device, and checking the overall score of the object associated with the health management device. Output warning information when it is greater than the score threshold.
  • the performing data processing on the device-related data derived from the device associated with the health management device includes: performing the data processing on the device-related data derived from the device associated with the health management device Data is processed in a distributed manner.
  • At least one embodiment of the present disclosure also provides a data processing device, which includes a processor and a memory.
  • the memory stores computer program instructions suitable for execution by the processor, and when the computer program instructions are executed by the processor, the processor executes the following methods including: Data processing of the device-related data of the device.
  • At least one embodiment of the present disclosure further provides a health management device, which includes any data processing device provided in at least one embodiment of the present disclosure.
  • At least one embodiment of the present disclosure also provides a non-transitory storage medium, the non-transitory storage medium stores computer program instructions, and when the computer program instructions are executed by a processor, the computer executes the following method including: Data processing is performed on equipment-related data from equipment associated with the health management device.
  • At least one embodiment of the present disclosure also provides another data processing device, which includes a data processing module.
  • the data processing module is configured to perform data processing (for example, data preprocessing) on equipment-related data originating from equipment associated with the health management device. For example, data after data processing is provided to a memory associated with the health management device.
  • the data processing module includes a data transfer module; the data transfer module includes a structural similarity calculation sub-module and a transfer sub-module; the structural similarity calculation sub-module is It is configured to obtain the structural similarity between the first set of keywords and the second set of keywords based on the information of the first set of keywords related to the template data and the information of the second set of keywords related to the device-related data And the transfer sub-module is configured to transfer at least part of the device-related data to the template data based on the structural similarity.
  • the data transfer module further includes a first data receiving submodule; the first data receiving submodule is configured to receive the first set of key data related to the template data Word information and the device-related data; the first set of keyword information includes the feature vector of each keyword of the first set of keywords; the second set of keyword information includes the second set of keywords The feature vector of each keyword in the group of keywords.
  • the data transfer module further includes an information extraction sub-module and a feature vector generation sub-module;
  • the information extraction sub-module is configured to receive the sub-module from the first data Acquire the device-related data, and extract the second set of keywords and the hierarchical information of each keyword of the second set of keywords in the device-related data from the device-related data;
  • the vector generation sub-module is configured to generate each key of the second set of keywords based on the second set of keywords and the level information of each keyword of the second set of keywords in the device-related data And provide the feature vector of each keyword of the second group of keywords to the structural similarity calculation sub-module.
  • the structural similarity between the first set of keywords and the second set of keywords includes the characteristics of each keyword of the first set of keywords The similarity between the vector and the feature vector of each keyword of the second group of keywords.
  • the structural similarity calculation sub-module is configured to obtain the relationship between the first set of keywords and the second set of keywords by calculating a similarity matrix S The structural similarity of; and the similarity matrix S satisfies the following expression:
  • v 1 , v 2 ,...v m are the feature vectors of the first set of keywords
  • u 1 , u 2 ,...u b are the feature vectors of the second set of keywords
  • m is the first set of keywords.
  • the number of keywords in the group of keywords, and b is the number of keywords in the second group of keywords.
  • the transmitting at least part of the device-related data to the template data based on the structural similarity includes: transmitting each of the second set of keywords
  • the similarity between the feature vector of each keyword and the feature vector of each keyword of the first group of keywords is the similarity with a value greater than the preset similarity threshold as the correlation similarity; each correlation similarity Among the two corresponding keywords, the keyword belonging to the second group of keywords is used as the first keyword, and the keyword belonging to the first group of keywords among the two keywords corresponding to each of the relevance similarities is used as the keyword belonging to the first group of keywords A second keyword; and associating data associated with the first keyword in the device-related data with the second keyword related to the template data.
  • the data processing module further includes a data supplement module.
  • the data supplement module includes a second data receiving submodule and an interpolation calculation submodule; the second data receiving submodule is configured to receive an interpolated data set, and the interpolated data set includes a group of objects corresponding to the data to be supplemented And the interpolation calculation sub-module is configured to perform interpolation on the interpolated data set to obtain the data to be supplemented.
  • the data supplement module further includes a distance calculation module; the distance calculation module is configured to calculate the value of each piece of data in the data to be supplemented and the interpolated data set.
  • the distance calculation module is configured to calculate the data to be supplemented based on the time distance and geographic distance of each piece of data in the interpolated data set. And the nominal distance of each piece of data in the interpolated data set.
  • the second data receiving submodule is further configured to receive time information and geographic location information corresponding to each piece of data in the interpolated data set and the data to be supplemented Corresponding time information and geographic location information; the distance calculation module is also configured to calculate the data to be supplemented and the data to be supplemented based on the time information corresponding to each piece of data in the interpolated data set and the time information corresponding to the data to be supplemented The time distance of each piece of data in the interpolated data set; the distance calculation module is further configured to calculate the to-be-added based on the geographic information corresponding to each piece of data in the interpolated data set and the geographic information corresponding to the data to be supplemented The geographic distance between the data and each piece of data in the interpolated data set; and the nominal distance between the data to be supplemented and each piece of data in the interpolated data set is equal to the data to be supplemented and each piece of data in the interpolated data set
  • the second data receiving submodule is further configured to receive the damping item of the group of the object corresponding to the data to be supplemented; and the interpolation calculation submodule is configured To obtain the data to be supplemented by performing distance inversely proportional interpolation on the interpolated data set based on the nominal distance and the damping term.
  • the data to be supplemented x p satisfies the following expression:
  • the data processing module includes an object grouping module.
  • the object grouping module includes a third data receiving submodule, a score obtaining submodule, and a grouping submodule; the third data receiving submodule is configured to receive information about objects associated with the health management device; the score obtaining The sub-module is configured to obtain the scores of the objects associated with the health management device on a plurality of scoring items based on the information of the objects associated with the health management device and the query table; and the grouping sub-module is configured to obtain scores based on the The scores of the objects associated with the health management device on the multiple scoring items are the object assignment groups associated with the health management device.
  • the grouping sub-module is configured to obtain objects associated with the health management device based on the scores on the plurality of scoring items and the weights of the plurality of scoring items Based on the overall score, the objects associated with the health management device are assigned groups.
  • the overall score of the object associated with the health management device is equal to the weighted sum of the scores of the object associated with the health management device on the multiple scoring items; And the group number of the object associated with the health management device is equal to the overall score rounded up.
  • the information of the object associated with the health management device includes objective information and subjective information; at least part of the objective information is matched by a portable medical device or a wearable medical device.
  • the objects associated with the health management device are monitored and acquired; and at least part of the subjective information is obtained by inquiries by medical workers or electronic questionnaires filled out by the objects associated with the health management device.
  • the multiple scoring items include: age, gender, BMI, systolic blood pressure, diastolic blood pressure, occupational work intensity, number of exacerbations, number of hospitalizations, number of surgical treatments, Normal breathing rate, normal blood oxygen saturation, mental state and exercise state.
  • the data processing module further includes an early warning sub-module, wherein the early warning sub-module is configured to receive information associated with the health management device provided by the grouping sub-module And output early warning information when the overall score of the object associated with the health management device is greater than the score threshold.
  • the early warning sub-module is configured to receive information associated with the health management device provided by the grouping sub-module And output early warning information when the overall score of the object associated with the health management device is greater than the score threshold.
  • the data processing device further includes a pre-cache module and a post-cache module.
  • the pre-caching module is configured to receive equipment-related data generated by equipment associated with the health management apparatus, and to provide the equipment-related data to the data processing module; and the post-caching module is configured to The processed data output by the data processing module is received, and the processed data is provided to the memory associated with the health management device.
  • the pre-caching module is configured to cooperate with a distributed system to allow the data processing module to correlate the equipment generated by the equipment associated with the health management apparatus Data is processed in a distributed manner.
  • Fig. 1 is an exemplary block diagram of a data processing device provided by at least one embodiment of the present disclosure
  • Fig. 2 is an exemplary block diagram of a first example of a data processing module provided by at least one embodiment of the present disclosure
  • Fig. 3 is an exemplary block diagram of a data transfer module provided by at least one embodiment of the present disclosure
  • Fig. 4 is another exemplary block diagram of a data transfer module provided by at least one embodiment of the present disclosure.
  • Fig. 5 is an exemplary block diagram of a second example of a data processing module provided by at least one embodiment of the present disclosure
  • Fig. 6 is an exemplary block diagram of a data transfer module provided by at least one embodiment of the present disclosure
  • Fig. 7 is an exemplary block diagram of a third example of a data processing module provided by at least one embodiment of the present disclosure.
  • Fig. 8 is an exemplary block diagram of a fourth example of a data processing module provided by at least one embodiment of the present disclosure.
  • Fig. 9 is an exemplary block diagram of a fifth example of a data processing module provided by at least one embodiment of the present disclosure.
  • Fig. 10 is an exemplary block diagram of the data supplement module shown in Fig. 9;
  • FIG. 11 is another exemplary block diagram of the fifth example of the data processing module provided by at least one embodiment of the present disclosure.
  • Fig. 12 is an exemplary block diagram of a data agile warehousing technology system provided by at least one embodiment of the present disclosure
  • FIG. 13 is an exemplary block diagram of the COPD baseline scoring module of the data agile warehousing technology system shown in FIG. 12;
  • FIG. 14 is an exemplary block diagram of the inverse distance hybrid interpolation module of the data agile warehousing technology system shown in FIG. 12;
  • FIG. 15 is an exemplary block diagram of the cross-keyword mutual transfer module of the data agile warehousing technology system shown in FIG. 12;
  • FIG. 16 is an exemplary flowchart of a data processing method provided by at least one embodiment of the present disclosure.
  • FIG. 17 shows an exemplary block diagram of a non-transitory storage medium provided by at least one embodiment of the present disclosure
  • FIG. 18 shows an exemplary block diagram of another data processing device provided by at least one embodiment of the present disclosure.
  • FIG. 19 is an exemplary block diagram of a health management device provided by at least one embodiment of the present disclosure.
  • word vector technology can be divided into word vector technology based on statistical methods and word vector technology based on language models.
  • word vector technology based on statistical methods can be divided into word vector technology based on co-occurrence matrix and word vector technology based on singular value decomposition.
  • the language model generates the word vector by training the neural network language model NNLM (neural network language model), and the word vector is an incidental output of the language model.
  • word vector technology based on language models includes word vector technology based on word2vec.
  • word2vec is implemented by borrowing neural networks, using skip-gram and continuous bag of words (CBOW).
  • the skip-gram model uses a word as input to predict the context around it
  • the CBOW model uses the context of a word as input to predict the word itself.
  • the training methods of the jump word model model and the continuous bag of words model can be referred to related technologies, which will not be repeated here.
  • the inventor of the present disclosure has noticed in research that with the development of technology, a large number of portable devices and wearable devices are adopted by patients. Therefore, the amount of data received by the management device for managing the devices used by patients has increased, which may lead to The management device cannot receive all the data, especially the time period during which the equipment used by the patient centrally uploads the data (for example, a period of time after the patient comes home from get off work at night, and the equipment is connected to the WIFI).
  • the aforementioned portable devices and wearable devices may involve multiple manufacturers, the format and content of the data generated by the device used by the patient, the keywords involved in the data, and the frequency of the device collection device may not be consistent, which may lead to differences in the management device. Data that does not match the associated database is difficult to be used directly by subsequent data processing modules.
  • the embodiments of the present disclosure provide a data processing method, a data processing device, and a health management device.
  • the data processing device includes a data processing module.
  • the data processing module is configured to perform data processing (for example, data preprocessing) on equipment-related data originating from equipment associated with the health management device.
  • the processed data is provided, for example, to a memory associated with the health management device.
  • the uniformity of the memory provided to the health management device can be improved.
  • Fig. 1 is an exemplary block diagram of a data processing device provided by at least one embodiment of the present disclosure.
  • the data processing device includes a data processing module.
  • the data processing module is configured to perform data processing (for example, data preprocessing) on equipment-related data derived from the equipment associated with the health management device, and provide the processed data to the storage or health management associated with the health management device.
  • the data analysis module or data processing module included in the device is configured to perform data processing (for example, data preprocessing) on equipment-related data derived from the equipment associated with the health management device, and provide the processed data to the storage or health management associated with the health management device.
  • a device associated (e.g., bound) with a health management device allows data generated by the aforementioned device (e.g., raw data) or data processed based on the aforementioned device-related data to be provided (e.g., directly or indirectly) Provided to) the memory associated with the health management device.
  • the device-related data derived from the device associated with the health management device refers to data generated by the device associated with the health management device and received by the data processing module.
  • the memory associated with the health management device refers to a memory (or a database carried by the memory) for storing data received by the health management device or data generated by the health management device.
  • the equipment associated with the health management device can be set according to actual application requirements, which is not specifically limited in at least one embodiment of the present disclosure.
  • the equipment associated with the health management device may be a detection equipment or a treatment equipment.
  • the detection device is used to monitor the physical sign data of the user who uses the device (for example, blood pressure, blood oxygen saturation, respiratory rate, body temperature, etc.).
  • the detection equipment may be a pulmonary function meter, a pulse oximeter, a ventilator, a non-invasive multi-parameter detector, a blood pressure meter, a smart bracelet, a forehead thermometer, and the like.
  • treatment equipment is used in the treatment and rehabilitation of subjects.
  • an oxygen generator can be used to perform oxygen therapy on patients who use the oxygen generator.
  • the equipment associated with the health management device may be a portable medical device, a wearable medical device, or a stationary device.
  • the portable medical device may be a device belonging to a subject (patient) or a device belonging to a hospital and used by doctors for outpatient treatment.
  • the wearable medical device may be a device worn by a subject (e.g., a patient) for monitoring the physical signs of the subject.
  • a stationary device is an instrument arranged in a fixed location (for example, a hospital, a rehabilitation center, or a nursing home).
  • each piece of device-related data may include identification information and substantial information.
  • the identification information of the equipment associated with the health management apparatus may be at least one of the identifier (for example, serial number) of the equipment and the identifier of the user of the equipment.
  • the identifier of the device is selected from the serial number of the device, the platform serial number of the device (for example, the number of the device in the health management platform), and the internal number of the device (for example, the number of the device in a database associated with the memory).
  • the device user's identifier is selected from the device user's identity information (for example, ID card number or passport number), mobile phone number, WeChat ID, and Alipay number.
  • the substantial information includes at least one of detection data and usage data.
  • the substantial information may be the physical sign data of the object using the detection device acquired by the detection device.
  • the substantial information may include vital capacity.
  • the substantial information may be at least one of the setting parameter data of the treatment device or the rehabilitation device and the time data of the subject using the treatment device or the rehabilitation device.
  • the substantial information may include at least one of the ventilation frequency of the ventilator and the time the subject uses the ventilator.
  • the format and time frequency (corresponding to the frequency at which the equipment collects data) of the equipment-related data generated by the different equipment associated with the health management device may be different from each other.
  • the format of the device-related data derived from the device associated with the health management device may be different from the data storage format and time frequency of the database carried by the memory associated with the health management device.
  • the format of the data includes: the number of keywords involved in the data, the fields of the keywords involved in the data, the specific names of the keywords involved in the data (keywords with the same fields), and the structure of the data (for example, the keywords involved in the data Level in the data).
  • the keywords involved in the data refer to the characters used to indicate the meaning of the values in the data.
  • the data storage format of the database carried by the memory associated with the health management device may be ⁇ "equipment identifier": “", “substantial information”: ⁇ "respiratory rate”: “”, “blood oxygen saturation”: “ “, “systolic blood pressure”: “” ⁇ ; in this case, the keywords involved in the data storage format of the database include: “device identifier”, “respiratory rate”, “blood oxygen saturation” and “systolic blood pressure” , And the above four key levels are level 1, level 2, level 2, and level 2.
  • device-related data can be ⁇ "device identifier”: “01010202", “substantial information”: ⁇ "respiratory rate”: “15”, “blood oxygen saturation”: “95%” ⁇ ; in this case
  • the keywords involved in the device-related data include: “device identifier”, “respiratory rate” and “blood oxygen saturation”, and the above four key levels are level 1, level 2, and level 2. Therefore, the number of keywords involved in the device-related data is smaller than the number of keywords involved in the data storage format of the database.
  • the device-related data can be ⁇ device identifier": "01010202", “respiratory rate”: “15”, “blood oxygen saturation”: “95%”, “systolic blood pressure”: “100” ⁇ ; this
  • the keywords involved in the device-related data include: “device identifier”, “respiratory rate”, “blood oxygen saturation” and “systolic blood pressure”, and the above four key levels are level 1, level 1. Level, Level 1, and Level 1. Therefore, the level of some keywords involved in the device-related data is inconsistent with the level of the corresponding keywords involved in the data storage format of the database (for example, "breathing rate”).
  • device-related data can be ⁇ "device identifier": “01010202", “substantial information”: ⁇ "respiratory rate”: “15”, “blood oxygen saturation”: “95%”, “systolic blood pressure” : “100” ⁇ ; in this case, the keywords involved in the device-related data include: “device identifier”, “respiratory rate” and “blood oxygen saturation” and “systolic blood pressure”, and the above four key points The levels are 1, 2, 2, and 2. Therefore, the specific names of some keywords (for example, "device identifier” and “device identifier”) involved in the device-related data are different.
  • device-related data can be ⁇ "device identifier": “01010202", “substantial information”: ⁇ "respiratory rate”: “15”, “blood oxygen saturation”: “95%”, “weight” : “50kg” ⁇ ; in this case, the keywords involved in the device-related data include: “device identifier”, “respiratory rate” and “blood oxygen saturation” and “weight”, and the above four key points The levels are level 1, level 2, level 2, and level 2. Therefore, the fields of some keywords (for example, "systolic blood pressure” and "weight”) involved in the device-related data are different.
  • the predetermined data format and time frequency may be the data storage format and time frequency of a database carried by a memory associated with the health management device.
  • the predetermined data format and time frequency may also be that most of the multiple devices associated with the health management device adopt the data format and time frequency.
  • the time frequency of device-related data is inconsistent with the predetermined time frequency, including at least one of the following situations: the time frequency of the device-related data generated by the device (for example, the frequency at which the device collects data) is inconsistent with the predetermined time frequency; The time frequency of the device-related data of the device is consistent with the predetermined time frequency, but due to the data loss or defect in the data transmission process, the time frequency of the device-related data transmitted to the memory associated with the health management device is inconsistent with the predetermined time frequency .
  • the above-mentioned data loss or defect includes the missing of the entire piece of data and the missing of numerical values corresponding to some fields in a piece of data.
  • the device-related data generated by the device in a predetermined time period includes ⁇ "device identifier": “01010202”, “substantial information”: ⁇ "respiratory rate”: “15”, “blood oxygen saturation Degree”: “95%”, “systolic blood pressure”: “100” ⁇ and ⁇ "device identifier”: “01010202”, “substantial information”: ⁇ "respiratory rate”: “16”, “blood oxygen saturation “: “96%”, “systolic blood pressure”: “102” ⁇ , but the data received by the data processing module is only ⁇ "device identifier”: “01010202”, “substantial information”: ⁇ "respiratory rate” : “15”, “Spo2”: “95%”, “systolic blood pressure”: “100” ⁇ , there is a problem of missing the entire piece of data; in this case, it will lead to device-related data
  • the device-related data generated by the device in a predetermined time period includes ⁇ "device identifier": “01010202”, “substantial information”: ⁇ "respiratory rate”: “15”, “blood oxygen Saturation”: “95%”, “systolic blood pressure”: “100” ⁇ and ⁇ "device identifier”: “01010202”, “substantial information”: ⁇ "respiratory rate”: “16”, “blood oxygen saturation Degree”: “96%”, “systolic blood pressure”: “102” ⁇ , but the data received by the data processing module is only ⁇ "device identifier”: “01010202”, “substantial information”: ⁇ "respiratory frequency “: “15”, “blood oxygen saturation”: “95%”, “systolic blood pressure”: "100” ⁇ and ⁇ ”device identifier”: “01010202”, “substantial information”: ⁇ "respiratory
  • performing data processing on device-related data may be at least one of unifying the format of device-related data with a predetermined data format, and unifying the time frequency of device-related data with a predetermined time frequency. item.
  • the unity of the data provided to the memory associated with the health management device can be improved, thereby facilitating the follow-up of the data Data processing and analysis, as well as help to improve the utilization rate of data.
  • Fig. 2 is an exemplary block diagram of a first example of a data processing module provided by at least one embodiment of the present disclosure.
  • the data processing module includes a data transfer module.
  • Fig. 3 is an exemplary block diagram of a data transfer module provided by at least one embodiment of the present disclosure.
  • Fig. 4 is another exemplary block diagram of a data transfer module provided by at least one embodiment of the present disclosure.
  • the data transfer module includes a structural similarity calculation sub-module and a transfer sub-module.
  • the structural similarity calculation sub-module is configured to obtain the relationship between the first set of keywords and the second set of keywords based on the information of the first set of keywords related to the template data and the information of the second set of keywords related to the device-related data. The similarity of the structure.
  • the data format and time frequency of the template data may be a predetermined data format and time frequency.
  • the data format and time frequency of the template data may be consistent with the data storage format and time frequency of the database carried by the memory associated with the health management device.
  • the data format and time frequency of the template data may be consistent with the data format and time frequency used by most of the multiple devices associated with the health management device.
  • the template data may be data that does not include specific values or empty data with a data format.
  • the template data may be ⁇ "device identifier": "", “substantial information”: ⁇ "respiratory rate”: “”, “blood oxygen saturation”: “”, “systolic blood pressure”: “" ⁇ .
  • the keyword information is the feature vector of the keyword.
  • the feature vector VE of the keyword can be calculated based on the text t of the keyword and the number of layers L where the keyword is located in the data, and calculated using the text feature extraction algorithm f.
  • the text feature extraction algorithm f can be a word vector technology based on word2vec or other applicable algorithms.
  • the text feature extraction algorithm f can refer to related technologies, which will not be repeated here.
  • the application range of the data transfer module can be improved.
  • the data transfer module is not only suitable for data with a fixed or known data structure.
  • the information of the first set of keywords includes the feature vector of each keyword of the first set of keywords
  • the information of the second set of keywords includes the feature vector of each keyword of the second set of keywords
  • the data transfer module further includes a first data receiving submodule.
  • the first data receiving submodule is configured to receive device-related data.
  • the first data receiving submodule is configured to obtain necessary data from the initial data source (streaming data) of the data processing device.
  • the data transfer module also includes an information extraction sub-module and a feature vector generation sub-module.
  • the information extraction sub-module is configured to obtain device-related data from the first data receiving sub-module, and extract the second set of keywords from the device-related data, and each keyword in the second set of keywords in the device-related data Level information.
  • the following method can be used to extract the second set of keywords and the hierarchical information of each keyword of the second set of keywords in the device-related data from the device-related data.
  • the original data can be unpacked into list data based on the corresponding field standard provided by the supplier of the equipment associated with the health management device. For example, suppose the supplier of the device uses the json format as the transmission format, and the transmission data follows the http protocol. The unpacking process is to extract the message content from the data stream encapsulated by the http protocol, and then put the corresponding fields (such as patient name and patient COPD) The number of times of emphasis) (that is, the header key) is converted to data in a list format.
  • the corresponding fields such as patient name and patient COPD
  • the number of times of emphasis that is, the header key
  • the feature vector generation sub-module is configured to generate a feature vector of each keyword of the second set of keywords based on the second set of keywords and level information, and provide the feature vector of each keyword of the second set of keywords Give the structural similarity calculation sub-module.
  • the feature vector generation sub-module is configured to generate the hierarchical information of each keyword of the second set of keywords in the device-related data, and use the above-mentioned text feature extraction algorithm to generate the information of each keyword of the second set of keywords. Feature vector.
  • the first data receiving submodule is further configured to receive information of the first group of keywords related to the template data, and provide the information of the first group of keywords to the structural similarity calculation submodule.
  • the structural similarity calculation submodule can be reduced The amount of calculation is particularly large when the amount of data received by the memory associated with the health management device is large.
  • the first data receiving submodule is further configured to receive the first group of keywords involved in the template data and the level information of each keyword of the first group of keywords in the template data; in this case, The feature vector generation submodule is also configured to generate a feature vector of each keyword of the first set of keywords based on the first set of keywords and the level information of each keyword of the first set of keywords in the template data, and The feature vector of each keyword of the first group of keywords is provided to the structural similarity calculation sub-module.
  • the first data receiving submodule is also configured to receive template data; in this case, the information extraction submodule is also configured to obtain template data from the first data receiving submodule, and extract template data from the template data The first group of keywords and the level information of each keyword of the first group of keywords in the template data; the feature vector generation sub-module is also configured to be based on the first group of keywords and each keyword of the first group of keywords The hierarchical information in the template data generates a feature vector of each keyword of the first group of keywords, and provides the feature vector of each keyword of the first group of keywords to the structural similarity calculation sub-module.
  • the structural similarity between the first set of keywords and the second set of keywords includes the difference between the feature vector of each keyword of the first set of keywords and the feature vector of each keyword of the second set of keywords. Similarity.
  • the feature vectors of the first group of keywords are v 1 , v 2 , ... v m
  • the feature vectors of the second group of keywords are u 1 , u 2 , ... u b
  • m is the first group of keywords
  • the number of keywords, b is the number of keywords in the second group of keywords.
  • the similarity between the feature vector of each keyword of the first group of keywords and the feature vector of each keyword of the second group of keywords includes u 1 v 1 , u 1 v 2 ,... u 1 v m , u 2 v 1 , u 2 v 2 ,...u 2 v m , u b v 1 , u b v 2 ,...u b v m .
  • the application range of the data transfer module can be improved.
  • the data transfer module can be applied to data with unknown data format.
  • the structural similarity calculation sub-module is configured to obtain the structural similarity between the first group of keywords and the second group of keywords by calculating the following similarity matrix S; the similarity matrix S satisfies the following expression:
  • the amount of calculation for the subsequent transfer sub-modules can be reduced.
  • the similarity between the feature vector of each keyword of the first group of keywords and the feature vector of each keyword of the second group of keywords may be matrixed to obtain the similarity matrix S.
  • the cosine similarity can be used to calculate the value of each element u i v j in the similarity matrix S, where i is greater than or equal to 1 and less than or equal to b, and j is greater than or equal to 1 and less than or equal to m.
  • i is greater than or equal to 1 and less than or equal to b
  • j is greater than or equal to 1 and less than or equal to m.
  • the value of cosine similarity is 1
  • the angle between the two vectors is 90°
  • the value of cosine similarity is 0
  • the cosine similarity value is -1.
  • the cosine similarity (cosine value) of the two vectors is between -1 and 1, and the cosine similarity (cosine value) of the two vectors The larger the value, the closer the two vectors are.
  • the similarity of the two sub-data corresponding to the two vectors is higher.
  • the value of cosine similarity is -1, it means that two vectors are negatively correlated.
  • the values of all elements in the vector corresponding to the first sub-data and the vector corresponding to the second sub-data can be made positive.
  • the cosine of the vector corresponding to the first sub-data and the vector corresponding to the second sub-data The similarity lies between 0-1.
  • the similarity between the feature vector of each keyword of the first group of keywords and the feature vector of each keyword of the second group of keywords can be matrixed to obtain the similarity matrix S, and then the similarity can be calculated The value of each element in the degree matrix S.
  • the value of similarity between the feature vector of each keyword of the first group of keywords and the feature vector of each keyword of the second group of keywords can be calculated first, and then the calculated value of similarity can be calculated Perform matrix arrangement to obtain the calculated similarity matrix S.
  • the transfer sub-module is configured to transfer at least part of the device-related data to the template data based on the structural similarity.
  • transferring at least part of the device-related data to the template data based on the structural similarity includes: converting the feature vector of each keyword of the second set of keywords and the feature vector of each keyword of the first set of keywords Among the similarities of, the similarity whose value is greater than the preset similarity threshold is regarded as the relative similarity; among the two keywords corresponding to each relevant similarity, the keyword belonging to the second group of keywords is regarded as the first keyword, and Take the keyword belonging to the first group of keywords among the two keywords corresponding to each relevance similarity as the second keyword; combine the data associated with the first keyword in the device-related data and the second key related to the template data Word association.
  • associating the data associated with the first keyword in the device-related data with the second keyword related to the template data refers to transferring the data associated with the first keyword in the device-related data to the template data corresponding to the second keyword. Under the key field.
  • the preset similarity threshold may be set according to actual application requirements, which is not specifically limited in the embodiment of the present disclosure.
  • the device-related data includes ⁇ "patient name”: “Zhang San”, “age”: “28”, “weight”: “66” ⁇ ; template data is ⁇ "patient name”: “Li Si”, “ Age”: “48”, “Substantial Information”: ⁇ "Blood Pressure”: “42”, “Weight”: “55” ⁇ ; the data can be transferred based on the following steps.
  • first set of keywords involved in the template data ie, "patient name”, “age”, “blood pressure”, “weight”
  • level of each keyword in the first set of keywords in the template data Information ie, level 1, level 1, level 2, and level 2
  • second set of keywords involved in obtaining device-related data ie, "patient name”, “age”, “weight”
  • the hierarchical information of each keyword of the keyword in the template data that is, level 1, level 1, level 1).
  • the feature vector of the first set of keywords and the feature vector of the second set of keywords involved in the template data are obtained.
  • the preset similarity threshold can be set to 0.95.
  • the following matrix S > 0.95 can be obtained.
  • the words are "age (v 4 , the second keyword)” and "age (u 3 , the first keyword)”.
  • two keywords involved in related similarity can be identified as similar fields and can be merged.
  • the associated template data is as follows: ⁇ "patient name”: “Li Si; Zhang San”, “age”: “48; 28”, “substantial information”: ⁇ "blood pressure”: "42; []”, “Weight”: "55; 66” ⁇ .
  • the transfer sub-module transfers the corresponding value to the corresponding field, that is, completes a cross-keyword mutual transfer.
  • the calculation of structural similarity is to merge and disambiguate similar fields in different standards of different devices.
  • the value corresponding to the "patient name" field of manufacturer A and the "patient name” field of manufacturer B is the value of the keyword.
  • "Patient Name” corresponds to "Zhang San”.
  • the way of transmission can be to merge the keywords of the two data formats into one keyword, and serialize the corresponding values. For example, you can convert the "patient name"-"Zhang San" in the original data format.
  • the data transfer module can effectively convert data in different formats to the same storage form, improving the unity of patient data in the database (database).
  • the data transfer module can also be referred to as a cross-keyword mutual transfer module.
  • the structural similarity it is possible to facilitate transmission based on the relevance of the semantic structure.
  • the data of the corresponding keyword can be converted to the corresponding field for storage based on the above-mentioned structural similarity, and the data corresponding to the keywords with the same substantive meaning in different languages can be transferred to each other, which improves The unity of patient data in the database.
  • Fig. 5 is an exemplary block diagram of a second example of a data processing module provided by at least one embodiment of the present disclosure. As shown in Figure 5, the data processing module includes an object grouping module.
  • Fig. 6 is an exemplary block diagram of a data transfer module provided by at least one embodiment of the present disclosure.
  • the object grouping module includes a third data receiving sub-module, a score obtaining sub-module, and a grouping sub-module.
  • the third data receiving submodule is configured to receive information of objects associated with the health management device.
  • the information of the object associated with the health management device includes objective information and subjective information.
  • At least part of the objective information is acquired by a portable medical device (for example, a portable medical device owned by a patient) or a wearable medical device (for example, a wearable medical device worn by a patient) on an object associated with the health management device.
  • a portable medical device for example, a portable medical device owned by a patient
  • a wearable medical device for example, a wearable medical device worn by a patient
  • the latest physical data of the objects associated with the health management device can be acquired, thereby allowing the When the condition of the object associated with the management device deteriorates or worsens, it promptly reminds the object associated with the health management device and the doctor who provides diagnosis and treatment services for the object associated with the health management device.
  • the object grouping can also be updated in time when the condition of the object associated with the health management device deteriorates or worsens.
  • the grouping result output by the module can thus improve the accuracy of the calculation result of the module that uses the grouping result output by the object grouping module to perform data processing.
  • subjective information can be obtained by at least one of the following methods: medical workers inquiring (inquiry and confirmation through a remote video system); the object associated with the health management device is used for the first time on the object that matches the health management device (for example, the patient end; for example, WeChat applet) is provided (for example, provided by filling in an electronic questionnaire); objects associated with the health management device are updated by refilling the electronic questionnaire when the situation changes.
  • medical workers inquiring inquiry and confirmation through a remote video system
  • the object associated with the health management device is used for the first time on the object that matches the health management device (for example, the patient end; for example, WeChat applet) is provided (for example, provided by filling in an electronic questionnaire); objects associated with the health management device are updated by refilling the electronic questionnaire when the situation changes.
  • the third data receiving submodule is related to the type of disease of the object associated with the health management device.
  • the third data receiving sub-module receives the information of the object associated with the health management device including: age, gender, height and weight (or BMI), systolic blood pressure, diastolic blood pressure, occupation Any one or any combination of work intensity, number of exacerbations, number of hospitalizations, number of surgical treatments, normal breathing rate, normal blood oxygen saturation, mental state, and exercise state.
  • height, weight, blood pressure, respiratory rate, blood oxygen saturation, etc. can be obtained by one or more measurements using portable medical equipment or wearable medical equipment.
  • the object before collecting the information of the object associated with the health management device, the object can be made aware that the information will be collected.
  • the score obtaining sub-module is configured to obtain the scores of the objects associated with the health management device on multiple scoring items based on the information of the objects associated with the health management device and the query table.
  • the multiple scoring items and the specific form of the query table are related to the type of disease suffered by the object associated with the health management device.
  • multiple scoring items include: age, gender, BMI (body mass index), systolic blood pressure, diastolic blood pressure, occupational work intensity, number of exacerbations (number of exacerbations of a specified disease) , Number of hospitalizations, number of surgical treatments, normal breathing rate, normal blood oxygen saturation, mental state and exercise status.
  • the automatic grouping of the objects associated with the health management device can be realized.
  • the scores of the objects associated with the health management device on multiple scoring items can more intuitively indicate the degree of abnormality of the object on the specified scoring items. Therefore, it is helpful for medical workers to conduct back-checks on the results based on the query table, and furthermore: It is helpful for medical workers to give feedback on the matching degree between the score distribution in the questionnaire and the actual medical practice.
  • multiple scoring items and query tables can use non-negative scoring, that is, if the object is normal in the scoring item, the object's score on the scoring item is recorded as 0 points, and the object's score on the scoring item The larger it is, the more abnormal the subject is in the scoring item.
  • non-negative scoring method negative scores can be avoided, which can improve the reliability of the calculation results.
  • Table 1 for an example of a COPD lookup table.
  • the look-up table shown in Table 1 uses a quantitative scoring mechanism and is formulated based on existing clinical observations.
  • the standards are simple, uniform, and easy to implement.
  • the grouping sub-module is configured to assign groups to the objects associated with the health management device based on the scores of the objects associated with the health management device on multiple scoring items.
  • the grouping sub-module is configured to obtain the overall score of the object associated with the health management device based on the scores on the multiple scoring items and the weights of the multiple scoring items.
  • the score of the object associated with the health management device on the i-th scoring item is y i , where i is greater than or equal to 1 and less than or equal to c, and c is the number of multiple scoring items.
  • c is equal to 13.
  • the weight of the i-th scoring item among multiple scoring items is w fi .
  • the overall score of the object associated with the health management device is equal to the weighted sum of the scores of the object associated with the health management device on multiple scoring items.
  • the overall score Sc of the object associated with the health management device can be represented by the following expression.
  • the grouping sub-module is also configured to assign groups to objects associated with the health management device based on the overall score.
  • the group number of the object associated with the health management device is equal to the overall score rounded up.
  • the number of the group where the specified object is located is 2.
  • the objects associated with the health management device can be divided into 6 subgroups, namely, group 0, group 1, group 2, group 3, group 4, and group 5.
  • the personalization factors of the object for example, the patient
  • the personalization factors of the medical treatment can be taken into consideration.
  • a total of 13 weighting coefficients from w f1 to w f13 can be sequentially set according to the actual situation of the medical institution serving the patient.
  • the weight coefficient can be used to adaptively adjust the reference baseline.
  • w f8 can be appropriately adjusted according to the actual bed supply and the number of diagnosis and treatment in the hospital to better fit the current diagnosis and treatment plan of the hospital.
  • the sum of the weight of the 10th scoring item and the weight of the 11th scoring item can be set to a value not less than 45% to improve the quality of the grouping (for example, accuracy) .
  • the health management device by obtaining the overall score of the object associated with the health management device based on the scores on multiple scoring items and the weights of multiple scoring items, it can be flexibly modified according to the standards of medical practices in various regions, thereby improving the stability and grouping. Scalability.
  • Fig. 7 is an exemplary block diagram of a third example of a data processing module provided by at least one embodiment of the present disclosure.
  • the data processing module includes an object grouping module and a storage level allocation sub-module.
  • the specific implementation of the object grouping module can refer to related embodiments, which will not be repeated here.
  • the storage level allocation sub-module is configured to classify the device-related data corresponding to the objects associated with the management device based on at least the grouping result output by the object grouping module (for example, specify the group of the object), and according to the classification result to
  • the device-related data corresponding to the object associated with the management device is allocated to the processor that matches the classification result. For example, you can put high-level data in a high-speed memory (for example, cache), put medium-level data in a medium-speed memory (for example, random access memory), and put low-level data in a low-speed memory.
  • Medium for example, read-only memory or fixed memory).
  • the group of the object indicates the severity of the object’s illness
  • the group of the object can also be used to indicate the importance of the object’s data and the predicted value of the query frequency. Therefore, at least according to the group of the specified object before storing the data
  • the equipment-related data corresponding to the object associated with the management device is classified, and the equipment-related data corresponding to the object associated with the management device is assigned a processor that matches the classification result according to the classification result, which can speed up the query speed of important data and high-frequency data .
  • the storage level allocation sub-module is configured to at least based on the grouping result output by the object grouping module (for example, the group of the specified object), the storage space corresponding to the device-related data of the specified object needs to be occupied, and the device-related data corresponding to the specified object The importance of the data, the query frequency of the past data of the designated object, and the classification of the equipment-related data corresponding to the object associated with the management device.
  • the storage level allocation sub-module can be allowed to generate an autonomous caching decision strategy based on the importance (predicted value) and query frequency (predicted value) of the device-related data generated by the device associated with the management device.
  • Cache recent high-frequency data can speed up query speed and analysis speed, improve the timeliness of the early warning sub-module (see Figure 8), selectively perform write operations on important low-frequency data, improve database communication efficiency, and make important data highly reliable And high availability.
  • Fig. 8 is an exemplary block diagram of a fourth example of a data processing module provided by at least one embodiment of the present disclosure.
  • the data processing module includes an object grouping module and an early warning sub-module.
  • the early warning submodule is configured to receive the overall score of the object associated with the health management device provided by the grouping submodule, and output early warning information when the overall score of the object associated with the health management device is greater than the score threshold.
  • the scoring threshold can be set according to actual application requirements, which is not specifically limited in the embodiments of the present disclosure.
  • the scoring threshold may be set to a fixed value (a numerical value indicating that the condition is moderate to severe); for example, the scoring threshold may be set to a variable value.
  • the score threshold may be the rounded up value of the score value of the object before this calculation.
  • the data processing module include an object grouping module and an early warning sub-module, it is possible to promptly remind the object associated with the health management device and the doctor who provides medical services for the object associated with the health management device when the object's condition deteriorates or worsens. This can improve the experience of the objects (for example, patients) and doctors who use the above-mentioned health management device.
  • the data processing module may include an object grouping module, a storage level allocation sub-module, and an early warning sub-module at the same time.
  • object grouping module a storage level allocation sub-module
  • early warning sub-module a storage level allocation sub-module at the same time.
  • Fig. 9 is an exemplary block diagram of a fifth example of a data processing module provided by at least one embodiment of the present disclosure.
  • the data processing module includes an object grouping module and a data supplement module.
  • the specific implementation of the object grouping module can refer to related embodiments, which will not be repeated here.
  • the data processing module may not include an object grouping module.
  • the data supplement module directly receives information about multiple objects associated with the health management device from the database or memory associated with the health management device. Grouping results.
  • Fig. 10 is an exemplary block diagram of the data supplement module shown in Fig. 9;
  • Fig. 11 is another exemplary block diagram of a fifth example of the data processing module provided by at least one embodiment of the present disclosure.
  • the data supplement module includes a second data receiving sub-module and an interpolation calculation sub-module.
  • the second data receiving submodule is configured to receive an interpolated data set, and the interpolated data set includes at least part of data of other objects in the group where the object corresponding to the data to be supplemented is located.
  • the interpolated data set may include data of all objects among multiple objects in the group corresponding to the object to be supplemented, except for the object corresponding to the data to be supplemented.
  • the key field of each piece of data in the interpolated data set is substantially the same as the key field of the data to be supplemented.
  • multiple pieces of data have a one-to-one correspondence with multiple objects, that is, the number of multiple pieces of data is equal to that of the group where the first object (the object corresponding to the data to be supplemented) is located except for the first object.
  • the number of other objects For another example, in the interpolated data set, each object is related to multiple pieces of data. In this case, the number of multiple pieces of data is greater than that of the first object (the object corresponding to the data to be supplemented) except for the first object. The number of other objects.
  • the aforementioned data missing problem can be caused by data transmission or data acquisition.
  • the aforementioned data missing problem may be caused by the low frequency of data collection by the device.
  • the data to be supplemented may be data corresponding to some keywords in the device-related data.
  • the device-related data generated on the specified device includes ⁇ "device identifier": “01010202”, “substantial information”: ⁇ "respiratory rate”: “15”, “blood oxygen saturation”: “95%”, “ Systolic blood pressure”: “100” ⁇
  • the data received by the data supplement module includes ⁇ "device identifier”: “01010202”, “substantial information”: ⁇ "respiratory rate”: “15”, “blood oxygen saturation” : “95%”, “systolic blood pressure”: “” ⁇
  • it can be determined that the data corresponding to the keyword "systolic blood pressure" in the device-related data is missing; correspondingly; if the object corresponding to the data to be supplemented ( For example, the first object) is located in the third group, and the second data receiving submodule can be used to receive the device-related data corresponding to all objects
  • the interpolated data set includes n pieces of data, x i is the i-th piece of data in the difference data set, and i is a positive integer greater than or equal to 1 and less than or equal to n.
  • the interpolation calculation sub-module is configured to perform interpolation on the interpolated data set to obtain data to be supplemented. For example, by using at least part of the data of other objects in the group corresponding to the object to be supplemented to obtain the data to be supplemented, the credibility of the data to be supplemented can be improved.
  • the data supplement module also includes a distance calculation module; the distance calculation module is configured to calculate the nominal distance of each data in the data set to be supplemented and the interpolated data set, and provide the nominal distance to the interpolation calculation Sub-module; the interpolation calculation sub-module is configured to perform distance inversely proportional interpolation on the interpolated data set based on at least the nominal distance to obtain the data to be supplemented.
  • the inventor of the present disclosure noticed that, compared to the historical data of the object corresponding to the data to be supplemented, the data of the device-related data of other objects in the group corresponding to the object to be supplemented is more close to the data to be supplemented in nominal distance. It is close to the data to be supplemented. Therefore, the data to be supplemented can be obtained by performing distance inverse interpolation on the interpolated data set based at least on the nominal distance, which can make the data to be supplemented more credible and accurate.
  • the distance calculation module is configured to calculate the nominal distance between the data to be supplemented and each data in the interpolated data set based on the time distance and geographic distance of each data in the data to be supplemented and the interpolated data set. For example, instead of calculating the nominal distance between the data to be supplemented and each data in the interpolated data set based only on the time distance or geographic distance of each data in the data to be supplemented and the interpolated data set, The time distance and geographic distance of each piece of data calculate the nominal distance of each piece of data in the data to be supplemented and the interpolated data set, which can make the nominal distance of each piece of data in the data to be supplemented and the interpolated data set more accurate.
  • the second data receiving submodule is further configured to receive time information (e.g., time stamp) and geographic location information corresponding to each piece of data in the interpolated data set, and time information (e.g., time stamp) and geographic location information corresponding to the data to be supplemented. location information.
  • time information e.g., time stamp
  • time information e.g., time stamp
  • the time information (e.g., time stamp) and geographic location information corresponding to the data to be supplemented may be based on the predetermined data collection frequency of the device and the time information (e.g., time stamp) corresponding to the device-related data adjacent to the data to be supplemented. Geographic location information acquisition (presumably obtained).
  • the geographic location information corresponding to the above data may be at least one of the home address and work address provided by the object during registration, the location of the device when the device used by the object generates the data, and the location of the device when the device used by the object uploads the above data.
  • the time information corresponding to the above-mentioned data may be at least one of the time when the device used by the object generates the data and the time when the device used by the object uploads the above-mentioned data.
  • the distance calculation module is also configured to calculate the time distance between the data to be supplemented and each data in the interpolated data set based on the time information corresponding to each piece of data in the interpolated data set and the time information corresponding to the data to be supplemented; the distance calculation module is also configured to It is configured to calculate the geographic distance between the data to be supplemented and each data in the interpolated data set based on the geographic information corresponding to each piece of data in the interpolated data set and the geographic information corresponding to the data to be supplemented.
  • the geographic location information between the patient corresponding to the data to be supplemented and the patient corresponding to the differenced data can be determined to determine the number of intersections in the shortest street between the two patients, and the number of intersections in the shortest street above can be regarded as the number of intersections between the two patients. Geographic distance.
  • the time difference between the time stamp of the patient data corresponding to the data to be supplemented and the time stamp of the inter-patient data corresponding to the difference data may be the time distance between the two patients.
  • the nominal distance between the data to be supplemented and each data in the interpolated data set is equal to the weighted sum of the spatial distance between the data to be supplemented and each data in the interpolated data set and the time distance of each data in the data to be supplemented and the interpolated data set.
  • the nominal distance between the data x p to be supplemented and the ith data x i in the differenced data set is d pi .
  • the weight coefficients ws and wt can be set for the spatial distance and the temporal distance, respectively.
  • the weighting coefficients ws and wt can be set based on medical practice and statistical survey results for a specified disease. For example, by taking the spatial distance between the data to be supplemented and each data in the interpolated data set and the weighted sum of the time distance between the data to be supplemented and each data in the interpolated data set as the nominal value of each data in the data to be supplemented and the interpolated data set The distance can improve the application range of the distance calculation module and the data supplement module and the accuracy of the output result.
  • the second data receiving sub-module is further configured to receive the damping item of the group corresponding to the object to be supplemented; the interpolation calculation sub-module is configured to perform distance inverse interpolation on the interpolated data set based on the nominal distance and the damping item to obtain the to-be-supplemented data set data.
  • e k is the damping term of the group of the object corresponding to the data to be supplemented, k is greater than or equal to 1 and less than or equal to t, and t is the number of the group related to the object associated with the health management device.
  • the inverse distance interpolation of the interpolated data set to obtain the data to be supplemented can increase the degree of interpolation (numerical value) of the data to be supplemented corresponding to the objects in the group with a greater severity of the disease, thereby improving The efficiency of the data obtained by interpolation (for example, the validity of the measurement).
  • the values of the damping term e k of different groups may be different.
  • the values of the damping items e k in different groups can be set based on medical practice and statistical survey results for the specified disease.
  • the systolic blood pressure data of patient A in the first group at time t1 (16.00 on June 3, 2020) is missing, there are three other patients in the first group besides patient A, and the above three patients are at time t1 (16.00 on June 3, 2020).
  • the systolic blood pressure data at point t1 are 110, 120, and 130, respectively.
  • the nominal distances of patient A and the above three patients that can be obtained based on the distance calculation module are 1, 2 and 3, respectively.
  • the damping term e k of the first group is set to 0.1.
  • the interpolated data x p can be calculated by the following expression.
  • the damping term e k is set to 1.1, the degree of interpolation can be enlarged, and the value of x p is 127.
  • the data to be supplemented may be data corresponding to all keywords in the device-related data.
  • the device-related data generated by the designated device in a predetermined period of time includes ⁇ "device identifier": “01010202”, “substantial information”: ⁇ "respiratory rate”: “15”, “blood oxygen saturation”: “95 %”, “systolic blood pressure”: “100” ⁇ and ⁇ "device identifier”: “01010202”, “substantial information”: ⁇ "respiratory rate”: “18”, “blood oxygen saturation”: “94% “, “systolic blood pressure”: “98” ⁇ , but the data received by the data supplement module only includes ⁇ "device identifier”: “01010202”, “substantial information”: ⁇ "respiratory rate”: “15”, " In the case of blood oxygen saturation”: “95%”, “systolic blood pressure”: “100” ⁇ , it can be determined that the
  • the missing data transmission problems, data acquisition problems, etc.
  • frequency can be different
  • the data is interpolated inversely proportional to the distance.
  • the data processing device further includes a front buffer module (not shown in the figure) and a rear buffer module (not shown in the figure).
  • the pre-caching module is configured to receive equipment-related data generated by the equipment associated with the health management device and provide the equipment-related data to the data processing module; and the post-caching module is configured to receive the processed data output by the data processing module , And provide the processed data to the memory of the health management device.
  • the pre-cache module is configured to cooperate with the distributed system to allow the data processing module to perform distributed processing on the device-related data generated by the device associated with the health management device.
  • the data processing module includes a data transfer module, an object grouping module, and a data supplement module;
  • the data processing device includes a data processing module, a pre-cache module, and a post-cache module.
  • the modules and submodules involved in at least one embodiment of the present disclosure can be implemented by software, firmware, hardware, and any combination thereof.
  • the hardware includes a server, a Field Programmable Gate Array (FPGA), and the like.
  • the data processing device provided by at least one embodiment of the present disclosure will be exemplarily described below in conjunction with the examples shown in FIGS. 12-15.
  • FIG. 12 is an exemplary block diagram of a data agile warehousing technology system 1 provided by at least one embodiment of the present disclosure.
  • the agile warehousing technology system 1 includes a COPD baseline scoring module 11, a distance inverse mixed interpolation module 12 and a cross-keyword mutual transfer module 13.
  • the COPD baseline scoring module 11, the inverse distance hybrid interpolation module 12, and the cross-keyword mutual transfer module 13 can be implemented by the object grouping module, data supplement module, and data transfer module shown in Figures 1 to 11, respectively. Therefore, the aforementioned COPD
  • the specific implementation of the baseline scoring module 11, the inverse distance hybrid interpolation module 12, and the cross-keyword mutual transfer module 13 can be seen in the examples shown in FIGS. 1 to 11.
  • the COPD baseline scoring module is mainly based on the basic information provided by the patient (including past medical history) to make a reasonable score, and the heterogeneous grouping of the patients facilitates the grouping of mobile data.
  • the inverse distance hybrid interpolation module performs hybrid interpolation based on the aforementioned scoring module according to the nominal distance distribution (including baseline conditions, geographic distribution, etc.) of the patients in the group.
  • the interpolation algorithm can use known, discrete data points in the range Introduce new data points to improve the differences caused by data gaps and different measurement frequencies.
  • the cross-keyword mutual transfer module is mainly used to normalize and transfer the mobile data transmitted by different patients due to the difference items caused by the equipment and usage methods used, so as to allow unified data.
  • the aforementioned agile data storage technology system 1 can provide an efficient and convenient database storage and processing solution for mobile data collected by mobile devices worn by patients with chronic obstructive pulmonary disease.
  • the agile data storage technology system can also use a front data link buffer 2 as the import end of the entire system.
  • the cache can efficiently store the mobile data that has poured into the system in the recent period of time, and provide the data to the subsequent processing modules in a timely manner according to the first-in-first-out rule (in conjunction with the distributed system, the first-in first-out can improve the consistency of transfer and storage, Keep data synchronization), so that mobile data can enter the system in real time, for example, to prevent data omission.
  • the internal data preprocessing is completed by the coordination of the above three modules. After the data preprocessing is completed, the cleaned mobile data is imported (for example, imported via the rear data link 3) into the subsequent analysis system.
  • this system uses a distributed pool processing system (that is, a thread pool design), with threads as the smallest scheduling unit, which can speed up the parallel speed. For example, after each piece of data is preprocessed, it is handed over to the post-built library program to process and enter the database. For example, this system can efficiently realize the entire process of moving data through preprocessing and entering the database.
  • a distributed pool processing system that is, a thread pool design
  • an efficient first-in, first-out cache can be maintained, which can effectively receive excessive streaming data input and prevent data loss.
  • the real-time computing performance of the system can be improved and the robustness of the system can be improved.
  • FIG. 13 is an exemplary block diagram of the COPD baseline scoring module of the data agile warehousing technology system 1 shown in FIG. 12.
  • the COPD baseline scoring module 11 includes a scoring table query sub-module 111 and a patient grouping sub-module 112.
  • the scoring table query submodule 111 and the patient grouping submodule 112 can be implemented by the score obtaining submodule and the grouping submodule shown in FIG. 6 respectively. Therefore, the specific implementation of the score table query sub-module 111 and the patient grouping sub-module 112 can refer to the example shown in FIG. 6.
  • FIG. 14 is an exemplary block diagram of the inverse distance hybrid interpolation module 12 of the data agile warehousing technology system 1 shown in FIG. 12.
  • the inverse distance hybrid interpolation module 13 includes a data import sub-module 131, a distance calculation sub-module 133 and an interpolation calculation sub-module 132.
  • the data import submodule 131, the distance calculation submodule 133, and the interpolation calculation submodule 132 can be implemented by the second data receiving submodule, the distance calculation submodule, and the difference calculation submodule shown in FIG. 10, respectively. Therefore, the specific implementation of the above-mentioned data import sub-module 131, distance calculation sub-module 133, and interpolation calculation sub-module 132 can refer to the example shown in FIG. 10.
  • the missing part of the input movement data can be interpolated and supplemented, such as streaming data with different frequencies.
  • the distance damping factor is also introduced in the interpolation calculation, which can effectively improve the credibility of the interpolation calculation.
  • a hybrid distance computer system For example, by adopting a hybrid distance computer system, combining patient baseline information and historical data to calculate nominal distance (for example, the definition of distance includes actual distance, geodesic distance, etc.), which can be customized according to actual needs. , And finally input the result into the inverse distance interpolation calculation.
  • nominal distance for example, the definition of distance includes actual distance, geodesic distance, etc.
  • FIG. 15 is an exemplary block diagram of the cross-keyword mutual transfer module of the data agile warehousing technology system 1 shown in FIG. 12.
  • the cross-keyword mutual transfer module 12 includes a data import submodule 121, a keyword similarity calculation submodule 123, a structure matrix calculation submodule 124, and a data transfer submodule 122.
  • the data import sub-module 121, the keyword similarity calculation sub-module 123 + the structural matrix calculation sub-module 124, and the data transmission sub-module 122 may be respectively composed of the first data receiving sub-module and the structural similarity calculation sub-module shown in FIG. And pass the sub-module implementation. Therefore, the specific implementation of the data import sub-module 121, the keyword similarity calculation sub-module 123 + the structural matrix calculation sub-module 124, and the data transmission sub-module 122 can be referred to the example shown in FIG. 3.
  • the aforementioned mobile data agile warehousing solution can realize data preprocessing of these raw data and implement agile warehousing based on the collection of return data from various intelligent devices carried by patients with this sub-type of disease. , Provides a complete, standardized, and tidy high-quality data source, which facilitates further analysis based on the data of this type of disease.
  • the COPD baseline scoring module uses a data table quantitative scoring mechanism, which is formulated based on existing clinical observations.
  • the standards are simple, uniform and very easy to implement. At the same time, it can be flexibly modified according to the standards of medical practices in various regions, thereby improving The logical stability and scalability of this program.
  • the scoring mechanism the patient data in the database can be merged and grouped by simple query, which is conducive to the efficient distributed execution of subsequent preprocessing tasks.
  • the inverse distance hybrid interpolation module adopts a nominal distance distribution mechanism. Based on the grouping of patients, according to the spatiotemporal correlation between the collection location and collection time of mobile data, the credibility of interpolation based on adjacent patient data can be obtained. The magnitude and inverse relationship of its value can be mixed interpolation based on baseline data. According to the interpolation module, the missing data value can be filled, and the data of different frequencies can be interpolated to obtain the stored patient data of the same frequency, which can greatly improve the granularity problem of subsequent data analysis.
  • the cross-keyword mutual transfer module uses a keyword similarity matrix conversion mechanism. After the patient's mobile data transmission arrives, unpack according to the obtained data to obtain list data, and then calculate the structure group similarity matrix according to the similarity of the header keywords and the structural characteristics of this type of data. When the device keywords used by the patient are not consistent with other patient keywords, according to this module, the data can be smoothly converted to the same storage form to ensure the unity of the patient data in the database.
  • At least one embodiment of the present disclosure provides a data processing method, which includes: performing data processing on device-related data originating from a device associated with a health management device.
  • FIG. 16 is an exemplary flowchart of a data processing method provided by at least one embodiment of the present disclosure.
  • the data processing method includes the following step S90.
  • Step S90 Perform data processing on the device-related data originating from the device associated with the health management device.
  • the processed data is provided to a memory or database associated with the health management device.
  • step S90 includes the following steps S911 and S912.
  • Step S911 Acquire the structural similarity between the first group of keywords and the second group of keywords based on the information of the first group of keywords related to the template data and the information of the second group of keywords related to the device-related data.
  • Step S912 Transfer at least part of the device-related data to the template data based on the structural similarity.
  • step S911 and step S912 may be executed sequentially.
  • the structural similarity between the first set of keywords and the second set of keywords includes the difference between the feature vector of each keyword of the first set of keywords and the feature vector of each keyword of the second set of keywords. Similarity. v 1 , v 2 ,...v m are the feature vectors of the first group of keywords, u 1 , u 2 ,...u b are the feature vectors of the second group of keywords, and m is the keyword in the first group of keywords The number of keywords, b is the number of keywords in the second group of keywords.
  • step S911 includes: obtaining the structural similarity between the first group of keywords and the second group of keywords by calculating the similarity matrix S; the similarity matrix S satisfies the following expression.
  • step S912 includes the following steps S9121-step S9123.
  • Step S9121 Regarding the similarity between the feature vector of each keyword of the second group of keywords and the feature vector of each keyword of the first group of keywords, the similarity whose value is greater than the preset similarity threshold is regarded as the correlation Similarity.
  • Step S9122 Take the keyword belonging to the second group of keywords among the two keywords corresponding to each relevance similarity as the first keyword, and set the two keywords corresponding to each relevance similarity to belong to the first keyword.
  • the key of the group key is used as the second key.
  • Step S9123 Associate the data associated with the first keyword in the device-related data with the second keyword associated with the template data.
  • step S9121, step S9122, and step S9123 may be executed sequentially.
  • step S90 further includes the following step S913.
  • Step S913 Receive the information of the first group of keywords and the device-related data related to the template data. For example, step S913 may be executed before step S911.
  • the information of the first set of keywords includes the feature vector of each keyword of the first set of keywords
  • the information of the second set of keywords includes the feature vector of each keyword of the second set of keywords
  • step S90 also includes the following steps S914 and S915.
  • Step S914 Extract the second group of keywords and the level information of each keyword of the second group of keywords in the device-related data from the device-related data.
  • Step S915 Generate a feature vector of each keyword of the second set of keywords based on the second set of keywords and the level information of each keyword of the second set of keywords in the device-related data.
  • step S914 and step S915 may be executed sequentially.
  • step S914 and step S915 may be executed after step S913 and in step S911.
  • step S913, step S914, step S915, step S911, and step S912 may be executed sequentially.
  • the data processing method is not limited to including step S913.
  • the data processing method does not include the aforementioned step S913, and the data processing method includes receiving information about the first set of keywords related to the template data and information about the second set of keywords related to the device-related data; in this case below, the data processing method does not include the above-mentioned steps S914 and S915.
  • the data processing method does not include the aforementioned step S913, and the data processing method provided by at least one embodiment of the present disclosure includes receiving template data and device-related data; in this case, except for steps S911, S912,
  • the data processing method further includes the following step S916 and step S917; step S916 and step S917 can be executed sequentially.
  • step S916 and step S917 may be executed after step S913 and in step S911.
  • Step S916 Extract the first group of keywords and the level information of each keyword of the first group of keywords in the template data from the template data.
  • Step S917 Generate a feature vector of each keyword of the first set of keywords based on the first set of keywords and the hierarchical information of each keyword of the first set of keywords in the device-related data.
  • the data processing method does not include the above-mentioned step S913, but includes the following step S918: receiving the first group of keywords related to the template data and the hierarchical information of the first group of keywords and the first related data related to the device.
  • the second set of keywords and the level information of the second set of keywords in this case, the data processing method further includes step S915, step S917, step S911, and step S912.
  • step S918, step S915+step S917, step S911, and step S912 may be executed sequentially.
  • step S911-step S917 For example, for the specific implementation of step S911-step S917, reference may be made to the related description of the example shown in FIG. 2, which will not be repeated here.
  • step S90 includes the following steps S921 to S923.
  • Step S921 Receive the information of the object associated with the health management device.
  • Step S922 Obtain the scores of the objects associated with the health management device on multiple scoring items based on the information of the objects associated with the health management device and the query table.
  • Step S923 Assign groups to the objects associated with the health management device based on the scores of the objects associated with the health management device on the multiple scoring items.
  • step S921, step S922, and step S923 may be executed sequentially.
  • step S923 includes the following steps S9231 and S9232.
  • step S9231 and step S9232 may be executed sequentially.
  • Step S9231 Obtain the overall score of the object associated with the health management device based on the scores on the multiple scoring items and the weights of the multiple scoring items.
  • Step S9232 Assign groups to objects associated with the health management device based on the overall score.
  • step S9231 obtaining the overall score of the object associated with the health management device based on the scores on the multiple scoring items and the weights of the multiple scoring items includes: placing the objects associated with the health management device on the multiple scoring items The weighted sum of the scores is used as the overall score of the object associated with the health management device.
  • step S9232 assigning groups to objects associated with the health management device based on the overall score includes: rounding up the overall score as the group number of the objects associated with the health management device.
  • the information of the object associated with the health management device includes objective information and subjective information; at least part of the objective information is obtained by monitoring and acquiring the object associated with the health management device by portable medical equipment or wearable medical equipment; and at least part of the subjective information Obtained by an electronic questionnaire filled out by a medical worker or an object associated with the health management device.
  • multiple scoring items include: age, gender, BMI, systolic blood pressure, diastolic blood pressure, occupational work intensity, number of exacerbations, number of hospitalizations, number of surgical treatments, normal respiratory rate, normal blood oxygen saturation, and spirit Status and movement status.
  • step S90 further includes the following step S924.
  • step S924 may be executed after step S9231.
  • Step S924 Receive the overall score of the object associated with the health management device, and output early warning information when the overall score of the object associated with the health management device is greater than the score threshold.
  • step S921 to step S924 reference may be made to the related description of the example shown in FIG. 5, which will not be repeated here.
  • step S90 includes the following steps S931 and S932.
  • Step S931 Receive an interpolated data set, where the interpolated data set includes at least part of data of other objects in the group where the object corresponding to the data to be supplemented is located.
  • Step S932 Perform interpolation on the interpolated data set to obtain data to be supplemented.
  • step S931 and step S932 may be executed sequentially.
  • step S932 performing interpolation on the interpolated data set to obtain the data to be supplemented includes: performing inverse distance interpolation on the interpolated data set based on at least the nominal distance to obtain the data to be supplemented.
  • step S90 further includes the following step S933.
  • Step S933 Calculate the nominal distance between the data to be supplemented and each piece of data in the interpolated data set.
  • step S933 may be executed before step S932.
  • calculating the nominal distance between the data to be supplemented and each data in the interpolated data set includes: calculating the data to be supplemented and the interpolated data based on the time distance and geographic distance of each data in the data to be supplemented and the interpolated data set The nominal distance of each piece of data in the collection.
  • step S90 further includes the following step S934.
  • Step S934 Receive the time information and geographic location information corresponding to each piece of data in the interpolated data set and the time information and geographic location information corresponding to the data to be supplemented.
  • step S934 may be executed before step S933.
  • step S933 calculating the nominal distance between the data to be supplemented and each piece of data in the interpolated data set includes the following steps S9331 to S9333.
  • Step S9331 Calculate the time distance between the data to be supplemented and each data in the interpolated data set based on the time information corresponding to each piece of data in the interpolated data set and the time information corresponding to the data to be supplemented;
  • Step S9332 Calculate the geographic distance between the data to be supplemented and each data in the interpolated data set based on the geographic information corresponding to each piece of data in the interpolated data set and the geographic information corresponding to the data to be supplemented;
  • Step S9333 Use the spatial distance between the data to be supplemented and each data in the interpolated data set and the weighted sum of the time distance between the data to be supplemented and each data in the interpolated data set as the name of each data in the data to be supplemented and the interpolated data set distance.
  • step S9331 to step S9333 can be executed in the order of step S9331, step S9332, and step S9333.
  • step S9331 to step S9333 can be executed in the order of step S9332, step S9331, and step S9333.
  • steps S9331 to S9333 can be executed in the order of step S9331+step S9332 (that is, step S9331 and step S9332 are executed at the same time) and step S9333.
  • step S90 further includes the following step S935.
  • Step S935 Receive the damping item of the group of the object corresponding to the data to be supplemented. For example, step S935 is executed before step S932.
  • step S932 performing interpolation on the interpolated data set to obtain the data to be supplemented includes: performing inverse distance interpolation on the interpolated data set based on the nominal distance and the damping term to obtain the data to be supplemented.
  • the data to be supplemented x p satisfies the following expression:
  • the nominal distance of the ith data x i ; e k is the damping item of the group of the object corresponding to the data to be supplemented, k is greater than or equal to 1 and less than or equal to t, and t is the number of the group involved in the object associated with the health management device.
  • step S931+step S934+step S935 that is, step S931, step S934, and step S935 are executed simultaneously
  • step S933, and step S932 may be executed sequentially.
  • step S931+step S934 that is, step S931 and step S934 are executed at the same time
  • step S933, step S935, and step S932 may be executed sequentially.
  • step S931, step S934, step S933, step S935, and step S932 may be executed sequentially.
  • step S934, step S931, step S933, step S935, and step S932 may be executed sequentially.
  • step S931 to step S935 reference may be made to the related description of the example shown in FIG. 9 and FIG. 10, and details are not described herein again.
  • step S90 includes the following steps S921-step S923 and step S931-step S935.
  • step S921-step S923 are executed before step S931-step S935.
  • step S90 includes the following steps S911-step S915, step S921-step S923, and step S931-step S935.
  • step S921-step S923 are executed before step S931-step S935; step S911-step S915 are executed before step S931-step S935.
  • step S911-step S915 are executed before step S921-step S923.
  • step S911-step S915 are executed after step S921-step S923.
  • step S90 includes the following steps S911-step S915 and step S921-step S923.
  • step S911-step S915 are executed before step S921-step S923.
  • step S911-step S915 are executed after step S921-step S923.
  • performing data processing on the device-related data derived from the device associated with the health management device includes: performing distributed processing on the device-related data derived from the device associated with the health management device.
  • FIG. 17 shows an exemplary block diagram of a non-transitory storage medium provided by at least one embodiment of the present disclosure.
  • the non-transitory storage medium stores computer program instructions, and when the computer program instructions are executed by the processor, the computer executes the following method including: checking device-related data derived from the device associated with the health management device Perform data processing.
  • non-transitory storage media may include read-only memory (ROM), hard disk, flash memory, and so on.
  • ROM read-only memory
  • hard disk hard disk
  • flash memory and so on.
  • FIG. 18 shows an exemplary block diagram of another data processing device provided by at least one embodiment of the present disclosure.
  • the other data processing device includes a processor and a memory.
  • the memory stores computer program instructions suitable for execution by the processor.
  • the processor executes the following method including: performing data processing on device-related data originating from the device associated with the health management device.
  • the processor is, for example, a central processing unit (CPU) or another form of processing unit with data processing capabilities and/or instruction execution capabilities.
  • the processor can be implemented as a general-purpose processor, and also a single-chip microcomputer or micro-processing unit. Device, digital signal processor, dedicated image processing chip, or field programmable logic array, etc.
  • the memory may include, for example, volatile memory and/or nonvolatile memory, and may include, for example, read-only memory (ROM), hard disk, flash memory, and the like.
  • the memory may be implemented as one or more computer program products, and the computer program products may include various forms of computer-readable storage media, on which one or more computer programs may be stored instruction.
  • the processor can run the program instructions to achieve the desired function.
  • the memory can also store various other application programs and various data, as well as various data used and/or generated by the application programs.
  • At least one embodiment of the present disclosure further provides a health management device, which includes any data processing device provided in at least one embodiment of the present disclosure.
  • FIG. 19 is an exemplary block diagram of a health management device provided by at least one embodiment of the present disclosure. As shown in FIG. 19, the health management device includes any data processing device provided by at least one embodiment of the present disclosure. For example, for the specific implementation of the data processing device, reference may be made to the relevant embodiments, which will not be repeated here.

Landscapes

  • Engineering & Computer Science (AREA)
  • Public Health (AREA)
  • Medical Informatics (AREA)
  • Health & Medical Sciences (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Primary Health Care (AREA)
  • General Health & Medical Sciences (AREA)
  • Epidemiology (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Physics & Mathematics (AREA)
  • Biomedical Technology (AREA)
  • Pathology (AREA)
  • Evolutionary Computation (AREA)
  • Evolutionary Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Medical Treatment And Welfare Office Work (AREA)

Abstract

一种数据处理方法、数据处理装置和健康管理装置。该数据处理方法包括：对源于与健康管理装置关联的设备的设备相关数据进行数据处理。该数据处理装置和健康管理装置可以提升提供给与健康管理装置关联的存储器的统一性。

Description

数据处理方法、数据处理装置和健康管理装置
对相关申请的交叉参考
本申请要求于2020年6月9日递交的中国专利申请第202010519612.7号的优先权，出于所有目的，在此全文引用上述中国专利申请公开的内容以作为本申请的一部分。
技术领域
本公开的实施例涉及一种数据处理方法、数据处理装置和健康管理装置。
背景技术
随着经济发展和生活水平的提高，越来越多的人患有不同程度的慢性病。常见的慢性病(或者慢病)主要有心脑血管疾病、癌症、糖尿病、慢性呼吸***疾病等，心脑血管疾病包含高血压、脑卒中和冠心病等。资料显示，中国公民中患有高血压、糖尿病、血脂异常等慢性病的患者变多。
发明内容
本公开的至少一个实施例提供了一种数据处理方法，其包括：对源于与健康管理装置关联的设备的设备相关数据进行数据处理。
例如，在所述数据处理方法的至少一个示例中，所述对源于与所述健康管理装置关联的设备的设备相关数据进行数据处理包括：基于模板数据涉及的第一组关键字的信息以及所述设备相关数据涉及的第二组关键字的信息获取所述第一组关键字和所述第二组关键字之间的结构相似度；以及基于所述结构相似度将所述设备相关数据的至少部分传递至所述模板数据中。
例如，在所述数据处理方法的至少一个示例中，所述对源于与所述健康管理装置关联的设备的设备相关数据进行数据处理还包括：接收所述模板数据涉及的第一组关键字的信息以及所述设备相关数据；以及所述第一组关键字的信息包括所述第一组关键字的每个关键字的特征向量，所述第二组关键字的信息包括所述第二组关键字的每个关键字的特征向量。
例如，在所述数据处理方法的至少一个示例中，所述对源于与所述健康管理装置关联的设备的设备相关数据进行数据处理还包括：从所述设备相关数据提取出所述第二组关键字以及所述第二组关键字的每个关键字在所述设备相关数据中的层级信息；基于所述第二组关键字以及所述第二组关键字的每个关键字在所述设备相关数据中的层级信息生成所述第二组关键字的每个关键字的特征向量。
例如，在所述数据处理方法的至少一个示例中，所述第一组关键字和所述第二组关键 字之间的结构相似度包括所述第一组关键字的每个关键字的特征向量与所述第二组关键字的每个关键字的特征向量之间的相似度。
例如，在所述数据处理方法的至少一个示例中，所述基于模板数据涉及的第一组关键字的信息以及所述设备相关数据涉及的第二组关键字的信息获取所述第一组关键字和所述第二组关键字之间的结构相似度包括：通过计算相似度矩阵S来获取所述第一组关键字和所述第二组关键字之间的结构相似度；以及所述相似度矩阵S满足以下的表达式：
Figure PCTCN2021096383-appb-000001
v 1、v 2、……v m为所述第一组关键字的特征向量，u 1、u 2、……u b为所述第二组关键字的特征向量，m为所述第一组关键字中关键字的数目，b为所述第二组关键字中关键字的数目。
例如，在所述数据处理方法的至少一个示例中，所述基于所述结构相似度将所述设备相关数据的至少部分传递至所述模板数据中包括：将所述第二组关键字的每个关键字的特征向量与所述第一组关键字的每个关键字的特征向量之间的相似度中取值大于预设相似度阈值的相似度作为相关相似度；将每个相关相似度对应的两个关键字中的属于第二组关键字的关键字作为第一关键字，并将所述每个相关相似度对应的两个关键字中的属于第一组关键字的关键字作为第二关键字；以及将所述设备相关数据中与所述第一关键字关联的数据与所述模板数据涉及的所述第二关键字关联。
例如，在所述数据处理方法的至少一个示例中，所述对源于与所述健康管理装置关联的设备的设备相关数据进行数据处理还包括：接收***值数据集，其中，所述***值数据集包括待补充数据对应的对象所在组的其他对象的至少部分的数据；以及对所述***值数据集进行插值获取所述待补充数据。
例如，在所述数据处理方法的至少一个示例中，所述对源于与所述健康管理装置关联的设备的设备相关数据进行数据处理还包括：计算所述待补充数据和所述***值数据集中每条数据的名义距离；以及所述对所述***值数据集进行插值获取所述待补充数据包括：至少基于所述名义距离对所述***值数据集进行距离反比插值获取所述待补充数据。
例如，在所述数据处理方法的至少一个示例中，所述计算所述待补充数据和所述***值数据集中每条数据的名义距离包括：基于所述待补充数据和所述***值数据集中每条数据的时间距离和地理距离计算所述待补充数据和所述***值数据集中每条数据的名义距离。
例如，在所述数据处理方法的至少一个示例中，所述数据处理方法还包括接收所述***值数据集中每条数据对应的时间信息和地理位置信息以及所述待补充数据对应的时间信息和地理位置信息。所述计算所述待补充数据和所述***值数据集中每条数据的名义距离 包括：基于所述***值数据集中每条数据对应的时间信息以及所述待补充数据对应的时间信息计算所述待补充数据和所述***值数据集中每条数据的时间距离；基于所述***值数据集中每条数据对应的地理信息以及所述待补充数据对应的地理信息计算所述待补充数据和所述***值数据集中每条数据的地理距离；以及将所述待补充数据和所述***值数据集中每条数据的空间距离以及所述待补充数据和所述***值数据集中每条数据的时间距离的加权和作为所述待补充数据和所述***值数据集中每条数据的名义距离。
例如，在所述数据处理方法的至少一个示例中，所述数据处理方法还包括：接收所述待补充数据对应的对象所在组的阻尼项。所述对所述***值数据集进行插值获取所述待补充数据包括：基于所述名义距离和所述阻尼项对所述***值数据集进行距离反比插值获取所述待补充数据。
例如，在所述数据处理方法的至少一个示例中，所述待补充数据x p满足以下的表达式：
Figure PCTCN2021096383-appb-000002
x i为所述被差值数据集中第i条数据，i为大于等于1小于等于n的正整数；n为所述***值数据集中数据的条数；d pi为所述待补充数据x p和所述被差值数据集中第i条数据x i的名义距离；e k为所述待补充数据对应的对象所在组的阻尼项，k大于等于1小于等于t，t为与所述健康管理装置关联的对象所涉及的组的编号。
例如，在所述数据处理方法的至少一个示例中，所述数据处理方法还包括：接收与所述健康管理装置关联的对象的信息；基于与所述健康管理装置关联的对象的信息以及查询表获取与所述健康管理装置关联的对象在多个评分项目上的评分；以及基于与所述健康管理装置关联的对象在所述多个评分项目上的评分为与所述健康管理装置关联的对象分配组别。
例如，在所述数据处理方法的至少一个示例中，所述基于与所述健康管理装置关联的对象在所述多个评分项目上的评分为与所述健康管理装置关联的对象分配组别包括：基于在所述多个评分项目上的评分以及所述多个评分项目的权重获取与健康管理装置关联的对象的总体评分；以及基于所述总体评分为与所述健康管理装置关联的对象分配组别。
例如，在所述数据处理方法的至少一个示例中，所述基于在所述多个评分项目上的评分以及所述多个评分项目的权重获取与所述健康管理装置关联的对象的总体评分包括：将与所述健康管理装置关联的对象在所述多个评分项目上的评分的加权和作为与所述健康管理装置关联的对象的总体评分；以及所述基于所述总体评分为与所述健康管理装置关联的对象分配组别包括：将所述总体评分向上取整值作为与所述健康管理装置关联的对象的组别序号。
例如，在所述数据处理方法的至少一个示例中，与所述健康管理装置关联的对象的信 息包括客观信息和主观信息；所述客观信息的至少部分由便携式医疗设备或可穿戴医疗设备对与所述健康管理装置关联的对象进行监测获取；以及所述主观信息的至少部分由医务工作者问询或者与所述健康管理装置关联的对象填写的电子问卷获取。
例如，在所述数据处理方法的至少一个示例中，所述多个评分项目包括：年龄、性别、BMI、收缩压、舒张压、职业工作烈度、病情加重次数、住院治疗次数、手术治疗次数、常态呼吸频率、常态血氧饱和度、精神状态和运动状态。
例如，在所述数据处理方法的至少一个示例中，所述数据处理方法还包括：接收与所述健康管理装置关联的对象的总体评分，并在与所述健康管理装置关联的对象的总体评分大于评分阈值时输出预警信息。
例如，在所述数据处理方法的至少一个示例中，所述对源于与健康管理装置关联的设备的设备相关数据进行数据处理包括：对所述源于与健康管理装置关联的设备的设备相关数据进行分布式处理。
本公开的至少一个实施例还提供了一种数据处理装置，其包括：处理器和存储器。所述存储器中存储有适于所述处理器执行的计算机程序指令，所述计算机程序指令被所述处理器运行时使得所述处理器执行以下的方法包括：对源于与健康管理装置关联的设备的设备相关数据进行数据处理。
本公开的至少一个实施例还提供了一种健康管理装置，其包括本公开的至少一个实施例提供的任一数据处理装置。
本公开的至少一个实施例还提供了一种非暂时性存储介质，所述非暂时性存储介质存储有计算机程序指令，所述计算机程序指令被处理器运行时使得计算机执行以下的方法包括：对源于与健康管理装置关联的设备的设备相关数据进行数据处理。
本公开的至少一个实施例还提供了另一种数据处理装置，其包括数据处理模块。所述数据处理模块被配置为对源于与健康管理装置关联的设备的设备相关数据进行数据处理(例如，数据预处理)。例如，数据处理后的数据被提供给与所述健康管理装置关联的存储器。
例如，在所述数据处理装置的至少一个示例中，所述数据处理模块包括数据传递模块；所述数据传递模块包括结构相似度计算子模块以及传递子模块；所述结构相似度计算子模块被配置为基于模板数据涉及的第一组关键字的信息以及所述设备相关数据涉及的第二组关键字的信息获取所述第一组关键字和所述第二组关键字之间的结构相似度；以及所述传递子模块被配置为基于所述结构相似度将所述设备相关数据的至少部分传递至所述模板数据中。
例如，在所述数据处理装置的至少一个示例中，所述数据传递模块还包括第一数据接收子模块；所述第一数据接收子模块被配置为接收所述模板数据涉及的第一组关键字的信息以及所述设备相关数据；所述第一组关键字的信息包括所述第一组关键字的每个关键字的特征向量；所述第二组关键字的信息包括所述第二组关键字的每个关键字的特征向量。
例如，在所述数据处理装置的至少一个示例中，所述数据传递模块还包括信息提取子 模块和特征向量生成子模块；所述信息提取子模块被配置为从所述第一数据接收子模块获取所述设备相关数据，并从所述设备相关数据提取出所述第二组关键字以及所述第二组关键字的每个关键字在所述设备相关数据中的层级信息；所述特征向量生成子模块被配置为基于所述第二组关键字以及所述第二组关键字的每个关键字在所述设备相关数据中的层级信息生成所述第二组关键字的每个关键字的特征向量，并将所述第二组关键字的每个关键字的特征向量提供给所述结构相似度计算子模块。
例如，在所述数据处理装置的至少一个示例中，所述第一组关键字和所述第二组关键字之间的结构相似度包括所述第一组关键字的每个关键字的特征向量与所述第二组关键字的每个关键字的特征向量之间的相似度。
例如，在所述数据处理装置的至少一个示例中，所述结构相似度计算子模块被配置为通过计算相似度矩阵S来获取所述第一组关键字和所述第二组关键字之间的结构相似度；以及所述相似度矩阵S满足以下的表达式：
Figure PCTCN2021096383-appb-000003
v 1、v 2、……v m为所述第一组关键字的特征向量，u 1、u 2、……u b为所述第二组关键字的特征向量，m为所述第一组关键字中关键字的数目，b为所述第二组关键字中关键字的数目。
例如，在所述数据处理装置的至少一个示例中，所述基于所述结构相似度将所述设备相关数据的至少部分传递至所述模板数据中包括：将所述第二组关键字的每个关键字的特征向量与所述第一组关键字的每个关键字的特征向量之间的相似度中取值大于预设相似度阈值的相似度作为相关相似度；将每个相关相似度对应的两个关键字中的属于第二组关键字的关键字作为第一关键字，并将所述每个相关相似度对应的两个关键字中的属于第一组关键字的关键字作为第二关键字；以及将所述设备相关数据中与所述第一关键字关联的数据与所述模板数据涉及的所述第二关键字关联。
例如，在所述数据处理装置的至少一个示例中，所述数据处理模块还包括数据补充模块。所述数据补充模块包括第二数据接收子模块和插值计算子模块；所述第二数据接收子模块被配置为接收***值数据集，所述***值数据集包括待补充数据对应的对象所在组的其他对象的至少部分的数据；以及所述插值计算子模块被配置为对所述***值数据集进行插值获取所述待补充数据。
例如，在所述数据处理装置的至少一个示例中，所述数据补充模块还包括距离计算模块；所述距离计算模块被配置为计算所述待补充数据和所述***值数据集中每条数据的名义距离，并将所述名义距离提供给所述插值计算子模块；以及所述插值计算子模块被配置为至少基于所述名义距离对所述***值数据集进行距离反比插值获取所述待补充数据。
例如，在所述数据处理装置的至少一个示例中，所述距离计算模块被配置为基于所述待补充数据和所述***值数据集中每条数据的时间距离和地理距离计算所述待补充数据和所述***值数据集中每条数据的名义距离。
例如，在所述数据处理装置的至少一个示例中，所述第二数据接收子模块还被配置为接收所述***值数据集中每条数据对应的时间信息和地理位置信息以及所述待补充数据对应的时间信息和地理位置信息；所述距离计算模块还被配置为基于所述***值数据集中每条数据对应的时间信息以及所述待补充数据对应的时间信息计算所述待补充数据和所述***值数据集中每条数据的时间距离；所述距离计算模块还被配置为基于所述***值数据集中每条数据对应的地理信息以及所述待补充数据对应的地理信息计算所述待补充数据和所述***值数据集中每条数据的地理距离；以及所述待补充数据和所述***值数据集中每条数据的名义距离等于所述待补充数据和所述***值数据集中每条数据的空间距离以及所述待补充数据和所述***值数据集中每条数据的时间距离的加权和。
例如，在所述数据处理装置的至少一个示例中，所述第二数据接收子模块还被配置为接收所述待补充数据对应的对象所在组的阻尼项；以及所述插值计算子模块被配置为基于所述名义距离和所述阻尼项对所述***值数据集进行距离反比插值获取所述待补充数据。
例如，在所述数据处理装置的至少一个示例中，所述待补充数据x p满足以下的表达式：
Figure PCTCN2021096383-appb-000004
x i为所述被差值数据集中第i条数据，i为大于等于1小于等于n的正整数；n为所述***值数据集中数据的条数；d pi为所述待补充数据x p和所述被差值数据集中第i条数据x i的名义距离；e k为所述待补充数据对应的对象所在组的阻尼项，k大于等于1小于等于t，t为与所述健康管理装置关联的对象所涉及的组的编号。
例如，在所述数据处理装置的至少一个示例中，所述数据处理模块包括对象分组模块。所述对象分组模块包括第三数据接收子模块、评分获取子模块和分组子模块；所述第三数据接收子模块被配置为接收与所述健康管理装置关联的对象的信息；所述评分获取子模块被配置为基于与所述健康管理装置关联的对象的信息以及查询表获取与所述健康管理装置关联的对象在多个评分项目上的评分；以及所述分组子模块被配置为基于与所述健康管理装置关联的对象在所述多个评分项目上的评分为与所述健康管理装置关联的对象分配组别。
例如，在所述数据处理装置的至少一个示例中，所述分组子模块被配置为基于在所述多个评分项目上的评分以及所述多个评分项目的权重获取与健康管理装置关联的对象的总体评分，并基于所述总体评分为与所述健康管理装置关联的对象分配组别。
例如，在所述数据处理装置的至少一个示例中，与所述健康管理装置关联的对象的总 体评分等于与所述健康管理装置关联的对象在所述多个评分项目上的评分的加权和；以及与所述健康管理装置关联的对象的组别序号等于所述总体评分向上取整值。
例如，在所述数据处理装置的至少一个示例中，与所述健康管理装置关联的对象的信息包括客观信息和主观信息；所述客观信息的至少部分由便携式医疗设备或可穿戴医疗设备对与所述健康管理装置关联的对象进行监测获取；以及所述主观信息的至少部分由医务工作者问询或者与所述健康管理装置关联的对象填写的电子问卷获取。
例如，在所述数据处理装置的至少一个示例中，所述多个评分项目包括：年龄、性别、BMI、收缩压、舒张压、职业工作烈度、病情加重次数、住院治疗次数、手术治疗次数、常态呼吸频率、常态血氧饱和度、精神状态和运动状态。
例如，在所述数据处理装置的至少一个示例中，所述数据处理模块还包括预警子模块，其中，所述预警子模块被配置为接收所述分组子模块提供的与所述健康管理装置关联的对象的总体评分，并在与所述健康管理装置关联的对象的总体评分大于评分阈值时输出预警信息。
例如，在所述数据处理装置的至少一个示例中，所述数据处理装置还包括前置缓存模块和后置缓存模块。所述前置缓存模块被配置为接收与所述健康管理装置关联的设备生成的设备相关数据，并将所述设备相关数据提供给所述数据处理模块；以及所述后置缓存模块被配置为接收所述数据处理模块输出的处理后的数据，并将所述处理后的数据提供给与所述健康管理装置关联的存储器。
例如，在所述数据处理装置的至少一个示例中，所述前置缓存模块被配置为与分布式***配合，以允许所述数据处理模块对与所述健康管理装置关联的设备生成的设备相关数据进行分布式处理。
附图说明
为了更清楚地说明本公开实施例的技术方案，下面将对实施例的附图作简单地介绍，显而易见地，下面描述中的附图仅仅涉及本公开的一些实施例，而非对本公开的限制。
图1是本公开的至少一个实施例提供的一种数据处理装置的一个示例性框图；
图2是本公开的至少一个实施例提供的数据处理模块的第一个示例的示例性框图；
图3是本公开的至少一个实施例提供的数据传递模块的一个示例性框图；
图4是本公开的至少一个实施例提供的数据传递模块的另一个示例性框图；
图5是本公开的至少一个实施例提供的数据处理模块的第二个示例的示例性框图；
图6是本公开的至少一个实施例提供的数据传递模块的一个示例性框图；
图7是本公开的至少一个实施例提供的数据处理模块的第三个示例的示例性框图；
图8是本公开的至少一个实施例提供的数据处理模块的第四个示例的示例性框图；
图9是本公开的至少一个实施例提供的数据处理模块的第五个示例的一个示例性框图；
图10是图9所示的数据补充模块的一个示例性框图；
图11是本公开的至少一个实施例提供的数据处理模块的第五个示例的另一个示例性框图；
图12是本公开的至少一个实施例提供的数据敏捷入库技术***的示例性框图；
图13是图12所示的数据敏捷入库技术***的COPD基线评分模块的示例性框图；
图14是图12所示的数据敏捷入库技术***的距离反比混合插值模块的示例性框图；
图15是图12所示的数据敏捷入库技术***的跨关键字互传递模块的示例性框图；
图16是本公开的至少一个实施例提供的数据处理方法的示例性流程图；
图17示出了本公开的至少一个实施例提供的非暂时性存储介质的示例性框图；
图18示出了本公开的至少一个实施例提供的另一种数据处理装置的示例性框图；以及
图19是本公开的至少一个实施例提供的健康管理装置的示例性框图。
具体实施方式
为使本公开实施例的目的、技术方案和优点更加清楚，下面将结合本公开实施例的附图，对本公开实施例的技术方案进行清楚、完整地描述。显然，所描述的实施例是本公开的一部分实施例，而不是全部的实施例。基于所描述的本公开的实施例，本领域普通技术人员在无需创造性劳动的前提下所获得的所有其他实施例，都属于本公开保护的范围。
除非另作定义，此处使用的技术术语或者科学术语应当为本公开所属领域内具有一般技能的人士所理解的通常意义。本公开中使用的“第一”、“第二”以及类似的词语并不表示任何顺序、数量或者重要性，而只是用来区分不同的组成部分。同样，“包括”或者“包含”等类似的词语意指出现该词前面的元件或者物件涵盖出现在该词后面列举的元件或者物件及其等同，而不排除其他元件或者物件。“连接”或者“相连”等类似的词语并非限定于物理的或者机械的连接，而是可以包括电性的连接，不管是直接的还是间接的。“上”、“下”、“左”、“右”等仅用于表示相对位置关系，当被描述对象的绝对位置改变后，则该相对位置关系也可能相应地改变。
例如，词向量技术可以划分为基于统计方法的词向量技术以及基于语言模型的词向量技术。例如，基于统计方法的词向量技术可以划分为基于共现矩阵的词向量技术以及基于奇异值分解的词向量技术。例如，在基于语言模型的词向量技术中，语言模型生成词向量是通过训练神经网络语言模型NNLM(neural network language model)，词向量作为语言模型的附带产出。例如，基于语言模型的词向量技术包括基于word2vec的词向量技术。例如，word2vec是借用神经网络的方式实现的，利用了跳字模型(skip-gram)和连续词袋模型(continuous bag of words，CBOW)。例如，skip-gram模型利用一个词语作为输入，来预测它周围的上下文，CBOW模型利用一个词语的上下文作为输入，来预测这个词语本身。例如，跳字模型模型和连续词袋模型的训练方法可以参见相关技术，在此不再赘述。
本公开的发明人在研究中注意到，随着技术的发展，大量的便携式设备、可穿戴设备被患者采用，因此，用于管理患者使用的设备的管理装置接收的数据量增加，这可能导致管理装置无法接收所有的数据，尤其是患者使用的设备集中上传数据的时间段(例如，患 者晚上下班回到家中，设备连接WIFI后的一段时间)。此外，由于上述便携式设备和可穿戴设备可能涉及多个厂商，因此，患者使用的设备生成的数据的格式和内容、数据涉及的关键字以及设备采集设备的频率可能不统一，进而导致与管理装置关联的数据库不匹配的数据难以被后续的数据处理模块直接使用。
例如，目前，对于呼吸慢病医疗过程中的智能化就诊需求，越来越多的医疗大数据***对数据处理变得繁杂，随之对数据入库的质量提出了新的要求；另一方面，进入移动医疗时代，大量的便携式、可穿戴设备进入了普通患者的家庭，这类设备产生了大量的数据，但是现阶段的医疗大数据***缺乏对该类数据进行接收并预处理的手段。
本公开的实施例提供了一种数据处理方法、数据处理装置和健康管理装置。该数据处理装置包括数据处理模块。数据处理模块被配置为对源于与健康管理装置关联的设备的设备相关数据进行数据处理(例如，数据预处理)。数据处理后的数据例如提供给与健康管理装置关联的存储器。例如，通过使用数据处理模块被配置为对源于与健康管理装置关联的设备的设备相关数据进行数据处理，可以提升提供给与健康管理装置关联的存储器的统一性。
下面通过几个示例和实施例对根据本公开实施例提供的数据处理装置进行非限制性的说明，如下面所描述的，在不相互抵触的情况下这些具体示例和实施例中不同特征可以相互组合，从而得到新的示例和实施例，这些新的示例和实施例也都属于本公开保护的范围。
图1是本公开的至少一个实施例提供的数据处理装置的示例性框图。如图1所示，该数据处理装置包括数据处理模块。数据处理模块被配置为对源于与健康管理装置关联的设备的设备相关数据进行数据处理(例如，数据预处理)，并将数据处理后的数据提供给与健康管理装置关联的存储器或健康管理装置包括的数据分析模块或数据处理模块。
例如，与健康管理装置关联(例如，绑定)的设备允许上述设备生成的数据(例如，原始数据)或者基于上述设备相关数据获取的数据处理后的数据被提供给(例如，直接或间接地提供给)与健康管理装置关联的存储器。
例如，源于与健康管理装置关联的设备的设备相关数据是指与健康管理装置关联的设备生成的且被数据处理模块接收的数据。
例如，与健康管理装置关联的存储器是指用于存储健康管理装置接收的数据或者健康管理装置生成的数据的存储器(或者存储器承载的数据库)。
例如，与健康管理装置关联的设备可以根据实际应用需求进行设定，本公开的至少一个实施例对此不作具体限定。
例如，与健康管理装置关联的设备可以是检测设备或治疗设备。例如，检测设备用于监测使用该设备的用户的体征数据(例如，血压、血氧饱和度、呼吸频率、体温等)。例如，检测设备可以是肺功能仪、脉搏血氧仪、呼吸机、无创多参数检测仪、血压计、智能手环、额温枪等。例如，治疗设备用于对象的治疗和康复环节。例如，可以使用制氧机对使用制氧机的患者进行氧疗。
例如，与健康管理装置关联的设备可以是便携式医疗设备、可穿戴医疗设备、固定式 设备。例如，便携式医疗设备可以是属于对象(患者)的设备或者是属于医院且用于医生外出诊疗使用的设备。例如，可穿戴医疗设备可以是被对象(例如，患者)佩戴的用于监测对象的体征的设备。例如，固定式设备是布置在固定的位置(例如，医院、康复中心或养老院)的仪器。
例如，每条设备相关数据(例如，原始数据)可以包括标识信息和实质信息。例如，与健康管理装置关联的设备的标识信息可以是设备的标识符(例如，序列号)和设备使用者的标识符的至少一个。例如，设备的标识符选自设备的序列号、设备的平台序列号(例如，设备在健康管理平台中的编号)以及设备的内部编号(例如，设备在与存储器关联的数据库中的编号)。例如，设备使用者的标识符选自设备使用者的身份信息(例如，身份证号或护照号)、手机号、微信号以及支付宝号。
例如，实质信息包括检测数据和使用数据的至少一类。例如，在与健康管理装置关联的设备是检测设备的情况下，实质信息可以是检测设备获取的使用该检测设备的对象的体征数据。例如，在与健康管理装置关联的设备是肺功能仪的情况下，实质信息可以包括肺活量。例如，在与健康管理装置关联的设备是治疗设备或康复设备的情况下，实质信息可以是治疗设备或康复设备的设置参数数据以及对象使用治疗设备或康复设备的时间数据的至少一类。例如，在与健康管理装置关联的设备是呼吸机的情况下，实质信息可以包括呼吸机的通气频率以及对象使用呼吸机的时间的至少一个。
例如，由于与健康管理装置关联的设备涉及多种类型的设备，因此，与健康管理装置关联的不同的设备生成的设备相关数据的格式和时间频率(对应于设备采集数据的频率)可能彼此不同。例如，源于与健康管理装置关联的设备的设备相关数据的格式可能和与健康管理装置关联的存储器承载的数据库的数据存储格式和时间频率不同。
例如，数据的格式包括：数据涉及的关键字的数目、数据涉及的关键字的字段、数据涉及的关键字的具体名称(字段相同的关键字)、数据的结构(例如，数据涉及的关键字位于数据中的层级)。例如，数据涉及的关键字是指用于指示数据中数值含义的字符。
例如，与健康管理装置关联的存储器承载的数据库的数据存储格式可以是{“设备标识符”：“”，“实质信息”：{“呼吸频率”：“”，“血氧饱和度”：“”，“收缩压”：“”}}；此种情况下，数据库的数据存储格式涉及的关键字包括：“设备标识符”、“呼吸频率”、“血氧饱和度”和“收缩压”，且上述四个关键所在的层级分别是1级、2级、2级和2级。
例如，设备相关数据可以是{“设备标识符”：“01010202”，“实质信息”：{“呼吸频率”：“15”，“血氧饱和度”：“95％”}}；此种情况下，设备相关数据涉及的关键字包括：“设备标识符”、“呼吸频率”和“血氧饱和度”，且上述四个关键所在的层级分别是1级、2级和2级。因此，设备相关数据涉及的关键字的数目小于数据库的数据存储格式涉及的关键字的数目。
又例如，设备相关数据可以是{设备标识符”：“01010202”，“呼吸频率”：“15”，“血氧饱和度”：“95％”，“收缩压”：“100”}；此种情况下，设备相关数据涉及的关 键字包括：“设备标识符”、“呼吸频率”、“血氧饱和度”和“收缩压”，且上述四个关键所在的层级分别是1级、1级、1级和1级。因此，设备相关数据涉及的部分关键字的层级与数据库的数据存储格式涉及的对应的关键字(例如，“呼吸频率”)的层级不一致。
再例如，设备相关数据可以是{“设备识别符”：“01010202”，“实质信息”：{“呼吸频率”：“15”，“血氧饱和度”：“95％”，“收缩压”：“100”}}；此种情况下，设备相关数据涉及的关键字包括：“设备识别符”、“呼吸频率”和“血氧饱和度”和“收缩压”，且上述四个关键所在的层级分别是1级、2级、2级和2级。因此，设备相关数据涉及的部分关键字(例如，“设备标识符”和“设备识别符”)的具体名称不同。
又再例如，设备相关数据可以是{“设备标识符”：“01010202”，“实质信息”：{“呼吸频率”：“15”，“血氧饱和度”：“95％”，“体重”：“50kg”}}；此种情况下，设备相关数据涉及的关键字包括：“设备识别符”、“呼吸频率”和“血氧饱和度”和“体重”，且上述四个关键所在的层级分别是1级、2级、2级和2级。因此，设备相关数据涉及的部分关键字(例如，“收缩压”和“体重”)的字段不同。
例如，预定的数据格式和时间频率可以是与健康管理装置关联的存储器承载的数据库的数据存储格式和时间频率。又例如，预定的数据格式和时间频率还可以是与健康管理装置关联的多个设备的大部分设备采用数据格式和时间频率。
例如，设备相关数据的时间频率均与预定的时间频率不一致包括以下情况的至少一种：设备生成的设备相关数据的时间频率(例如，设备采集数据的频率)与预定的时间频率不一致；设备生成的设备相关数据的时间频率与预定的时间频率一致，但由于数据传输过程中存在的数据丢失或缺损问题，传递至与健康管理装置关联的存储器的设备相关数据的时间频率与预定的时间频率不一致。
例如，上述数据丢失或缺损包括整条数据缺失以及一条数据中的部分字段对应的数值缺失。
例如，在设备在预定时间段(例如，10分钟)生成的设备相关数据包括{“设备识别符”：“01010202”，“实质信息”：{“呼吸频率”：“15”，“血氧饱和度”：“95％”，“收缩压”：“100”}}以及{“设备识别符”：“01010202”，“实质信息”：{“呼吸频率”：“16”，“血氧饱和度”：“96％”，“收缩压”：“102”}}，但被数据处理模块接收后的数据仅为{“设备识别符”：“01010202”，“实质信息”：{“呼吸频率”：“15”，“血氧饱和度”：“95％”，“收缩压”：“100”}}的情况下，则存在整条数据缺失的问题；此种情况下，将导致设备相关数据在特定时间段的时间频率与预定的时间频率不一致。
又例如，在设备在预定时间段(例如，10分钟)生成的设备相关数据包括{“设备识别符”：“01010202”，“实质信息”：{“呼吸频率”：“15”，“血氧饱和度”：“95％”，“收缩压”：“100”}}以及{“设备识别符”：“01010202”，“实质信息”：{“呼吸频率”：“16”，“血氧饱和度”：“96％”，“收缩压”：“102”}}，但被数据处理模块接收后的数据仅为{“设备识别符”：“01010202”，“实质信息”：{“呼吸频率”：“15”，“血氧饱和度”：“95％”，“收缩压”：“100”}}以及{“设备识别符”：“01010202”， “实质信息”：{“呼吸频率”：“16”，“血氧饱和度”：“96％”，“收缩压”：“”}}的情况下，则上述第二条数据存在部分数据缺失的问题；此种情况下，将导致设备相关数据的部分字段(“收缩压”)的数据在特定时间段的时间频率与预定的时间频率不一致。
例如，对设备相关数据进行数据处理(例如，数据预处理)可以是将设备相关数据的格式与预定的数据格式统一化以及将设备相关数据的时间频率与预定的时间频率统一化中的至少一项。
例如，通过使用数据处理模块对源于与健康管理装置关联的设备的设备相关数据进行数据处理，可以提升提供给与健康管理装置关联的存储器的数据的统一性，由此有利于对数据进行后续的数据处理和分析，以及有利于提高数据的利用率。
图2是本公开的至少一个实施例提供的数据处理模块的第一个示例的示例性框图。例如，如图2所示，该数据处理模块包括数据传递模块。
图3是本公开的至少一个实施例提供的数据传递模块的一个示例性框图。图4是本公开的至少一个实施例提供的数据传递模块的另一个示例性框图。例如，如图3和图4所示，该数据传递模块包括结构相似度计算子模块以及传递子模块。
例如，结构相似度计算子模块被配置为基于模板数据涉及的第一组关键字的信息以及设备相关数据涉及的第二组关键字的信息获取第一组关键字和第二组关键字之间的结构相似度。
例如，模板数据的数据格式和时间频率可以是预定的数据格式和时间频率。例如，模板数据的数据格式和时间频率可以与健康管理装置关联的存储器承载的数据库的数据存储格式和时间频率一致。又例如，模板数据的数据格式和时间频率可以和与健康管理装置关联的多个设备的大部分设备采用数据格式和时间频率一致。
例如，模板数据可以是不包括具体数值的数据或者具有数据格式的空数据。例如，模板数据可以是{“设备标识符”：“”，“实质信息”：{“呼吸频率”：“”，“血氧饱和度”：“”，“收缩压”：“”}}。
例如，关键字的信息为关键字的特征向量。例如，关键字的特征向量VE可以基于关键字的文本t以及关键字位于数据的层数L，并使用文本特征提取算法f计算得到。例如，特征向量VE可以采用以下的表达式进行表示：VE＝f(t，L)。例如，文本特征提取算法f可以是基于word2vec的词向量技术或其它适用的算法。例如，文本特征提取算法f可以参见相关技术，在此不再赘述。
例如，通过使得关键字的特征向量基于关键字的文本t以及关键字位于数据的层数L计算获得，可以提升数据传递模块的应用范围。例如，使得数据传递模块不仅仅适用于具有固定或已知数据结构的数据中。
例如，第一组关键字的信息包括第一组关键字的每个关键字的特征向量；第二组关键字的信息包括第二组关键字的每个关键字的特征向量。
例如，如图3和图4所示，数据传递模块还包括第一数据接收子模块。例如，第一数据接收子模块被配置为接收设备相关数据。
例如，第一数据接收子模块被配置为从数据处理装置的初始数据源(流式数据)中获得必须的数据。
例如，如图4所示，数据传递模块还包括信息提取子模块和特征向量生成子模块。
例如，信息提取子模块被配置为从第一数据接收子模块获取设备相关数据，并从设备相关数据提取出第二组关键字以及第二组关键字的每个关键字在设备相关数据中的层级信息。
例如，可以通过以下的方法从设备相关数据提取出第二组关键字以及第二组关键字的每个关键字在设备相关数据中的层级信息。
例如，可以基于与健康管理装置关联的设备的供应商提供的对应字段标准，将原始数据解包为列表类数据。例如，假设设备的供应商采用json格式作为传输格式，传输数据遵循http协议，解包的过程就是从http协议封装的数据流中抽取报文内容，然后把对应的字段(如患者姓名和患者COPD加重次数)(也即，表头关键字)转换为列表格式的数据。
例如，还可以采用其它相关技术从设备相关数据提取出第二组关键字以及第二组关键字的每个关键字在设备相关数据中的层级信息，在此不再赘述。
例如，特征向量生成子模块被配置为基于第二组关键字以及层级信息生成第二组关键字的每个关键字的特征向量，并将第二组关键字的每个关键字的特征向量提供给结构相似度计算子模块。例如，特征向量生成子模块被配置为基于第二组关键字的每个关键字在设备相关数据中的层级信息，并利用上述的文本特征提取算法生成第二组关键字的每个关键字的特征向量。
在一个示例中，第一数据接收子模块还被配置为接收模板数据涉及的第一组关键字的信息，并将第一组关键字的信息提供给结构相似度计算子模块。例如，通过使得第一数据接收子模块直接接收模板数据涉及的第一组关键字的信息，并将第一组关键字的信息提供给结构相似度计算子模块，可以降低结构相似度计算子模块运算量，尤其是在与健康管理装置关联的存储器接收的数据量较大的情况下。
在另一个示例中，第一数据接收子模块还被配置为接收模板数据涉及的第一组关键字以及第一组关键字的每个关键字在模板数据中的层级信息；此种情况下，特征向量生成子模块还被配置为基于第一组关键字以及第一组关键字的每个关键字在模板数据中的层级信息生成第一组关键字的每个关键字的特征向量，并将第一组关键字的每个关键字的特征向量提供给结构相似度计算子模块。
在再一个示例中，第一数据接收子模块还被配置为接收模板数据；此种情况下，信息提取子模块还被配置为从第一数据接收子模块获取模板数据，并从模板数据提取出第一组关键字以及第一组关键字的每个关键字在模板数据中的层级信息；特征向量生成子模块还被配置为基于第一组关键字以及第一组关键字的每个关键字在模板数据中的层级信息生成第一组关键字的每个关键字的特征向量，并将第一组关键字的每个关键字的特征向量提供给结构相似度计算子模块。
例如，第一组关键字和第二组关键字之间的结构相似度包括第一组关键字的每个关键 字的特征向量与第二组关键字的每个关键字的特征向量之间的相似度。例如，第一组关键字的特征向量为v 1、v 2、……v m，第二组关键字的特征向量为u 1、u 2、……u b，m为第一组关键字中关键字的数目，b为第二组关键字中关键字的数目。例如，第一组关键字的每个关键字的特征向量与第二组关键字的每个关键字的特征向量之间的相似度包括u 1v 1，u 1v 2，……u 1v m，u 2v 1，u 2v 2，……u 2v m，u bv 1，u bv 2，……u bv m。
例如，通过计算第一组关键字的每个关键字的特征向量与第二组关键字的每个关键字的特征向量之间的相似度作为结构相似度，可以提升数据传递模块的应用范围。例如，使得数据传递模块可以适用于具有未知数据格式的数据。
例如，结构相似度计算子模块被配置为通过计算以下的相似度矩阵S来获取第一组关键字和第二组关键字之间的结构相似度；相似度矩阵S满足以下的表达式：
Figure PCTCN2021096383-appb-000005
例如，通过计算以下的相似度矩阵S来获取第一组关键字和第二组关键字之间的结构相似度，可以降低后续传递子模块的运算量。例如，可以对第一组关键字的每个关键字的特征向量与第二组关键字的每个关键字的特征向量之间的相似度进行矩阵编排来获得相似度矩阵S。
例如，可以采用余弦相似度计算相似度矩阵S中每个元素u iv j的数值，此处i大于等于1小于等于b，j大于等于1小于等于m。例如，在两个向量(例如，u i和v j)具有相同的指向时，余弦相似度的值为1；两个向量的夹角为90°时，余弦相似度的值为0；两个向量指向完全相反的方向时，余弦相似度的值为-1，因此，两个向量的余弦相似度(余弦值)位于-1到1之间，且两个向量的余弦相似度(余弦值)取值越大，表示两个向量越接近。对应地，两个向量对应的两个子数据的相似度越高。例如，在余弦相似度的值为-1时，表示两个向量负相关。例如，可以使得第一子数据对应的向量和第二子数据对应的向量中所有元素的取值为正，此种情况下，第一子数据对应的向量和第二子数据对应的向量的余弦相似度位于0-1之间。
例如，可以采用以下的表达式计算相似度矩阵S中每个元素的数值：
Figure PCTCN2021096383-appb-000006
例如，可以先对第一组关键字的每个关键字的特征向量与第二组关键字的每个关键字的特征向量之间的相似度进行矩阵编排来获得相似度矩阵S，然后计算相似度矩阵S中每个元素的数值。又例如，可以先计算第一组关键字的每个关键字的特征向量与第二组关键字的每个关键字的特征向量之间的相似度的数值，然后将计算得到的相似度的数值进行矩阵编排来获得计算后的相似度矩阵S。
例如，传递子模块被配置为基于结构相似度将设备相关数据的至少部分传递至模板数据中。例如，基于结构相似度将设备相关数据的至少部分传递至模板数据中包括：将第二组关键字的每个关键字的特征向量与第一组关键字的每个关键字的特征向量之间的相似度中取值大于预设相似度阈值的相似度作为相关相似度；将每个相关相似度对应的两个关键字中的属于第二组关键字的关键字作为第一关键字，并将每个相关相似度对应的两个关键字中的属于第一组关键字的关键字作为第二关键字；将设备相关数据中与第一关键字关联的数据与模板数据涉及的第二关键字关联。
例如，将设备相关数据中与第一关键字关联的数据与模板数据涉及的第二关键字关联是指将将设备相关数据中与第一关键字关联的数据传递到模板数据的对应于第二关键字的字段下。
例如，预设相似度阈值可以根据实际应用需求进行设定，本公开的实施例对此不作具体限定。
下面结合一个示例对数据传递模块的具体实现方式做示例性说明。
例如，假设设备相关数据包括{“患者姓名”：“张三”，“年龄”：“28”，“体重”：“66”}；模板数据为{“患者名字”：“李四”，“年龄”：“48”，“实质信息”：{“血压”：“42”，“体重”：“55”}}；可以基于以下的步骤进行数据传递。
首先，获取模板数据涉及的第一组关键字(也即，“患者名字”，“年龄”，“血压”，“体重”)以及第一组关键字的每个关键字在模板数据中的层级信息(也即，1级，1级，2级和2级)；获取设备相关数据涉及的第二组关键字(也即，“患者姓名”，“年龄”，“体重”)以及第二组关键字的每个关键字在模板数据中的层级信息(也即，1级，1级，1级)。
其次，获取模板数据涉及的第一组关键字的特征向量以及第二组关键字的特征向量。例如，第一组关键字的特征向量v 1，v 2，v 3，v 4分别表示“患者名字”，“年龄”，“血压”，“体重”，且v 1＝[1,0,0.1]，v 2＝[0,0,2]，v 3＝[0.45,0.27,0.41]，v 4＝[3.1,0.2,0.3]。例如，第二组关键字的特征向量u 1，u 2和u 3分别表示“患者姓名”，“年龄”，“体重”，且u 1＝[1,0,0]，u 2＝[0,0,2]以及u 3＝[0,3,0]。
再次，计算相似度矩阵S。
Figure PCTCN2021096383-appb-000007
第四，查找相似度矩阵S中取值大于预设相似度阈值的元素，并将相似度矩阵S中取值大于预设相似度阈值的元素作为相关相似度。例如，预设相似度阈值可以设置为0.95。
例如，执行该操作后，可以得到如下的矩阵S >0.95。矩阵S >0.95中取值为1的元素对应于相似度矩阵S中取值大于预设相似度阈值的元素。因此，相似度矩阵S中的相关相似度包括u 1v 1＝0.99，u 2v 2＝1以及u 3v 4＝0.99。
Figure PCTCN2021096383-appb-000008
第五，找到相关相似度涉及的两个关键字，并确定两个关键字中第一关键字和第二关键字。相关相似度u 1v 1＝0.99涉及的两个关键字为“患者姓名(v 1，第二关键字)”和“患者名字(u 1，第一关键字)”；相关相似度u 2v 2＝1涉及的两个关键字为“年龄(v 2，第二关键字)”和“年龄(u 2，第一关键字)”；相关相似度u 3v 4＝0.99涉及的两个关键字为“年龄(v 4，第二关键字)”和“年龄(u 3，第一关键字)”。例如，相关相似度涉及的两个关键字可以被认定为相似字段，可进行合并。
第六，将设备相关数据中与第一关键字关联的数据与模板数据涉及的第二关键字关联。例如，关联后的模板数据如下：{“患者名字”：“李四；张三”，“年龄”：“48；28”，“实质信息”：{“血压”：“42；[]”，“体重”：“55；66”}}。
例如，基于计算得到的结相似度，传递子模块将对应数值传递给相应字段内，即完成一次跨关键字互传递。例如，计算结构相似度是为了让不同设备的不同标准中类似的字段进行合并和消歧。例如，A厂商的“患者姓名”字段以及B厂商的“患者名字”字段对应的数值为关键字的值。例如，“患者姓名”对应“张三”。例如，传递的方式可以是将两种数据格式的关键字合并为一种关键字，并将对应的值进行序列化操作，比如，可以将原数据格式中的“患者姓名”-“张三”，“患者名字”-“李四”，合并为“患者姓名”-“张三”，“患者姓名”-“李四”。例如，数据传递模块可以有效地将不同格式的数据顺利转换至相同的存储形态，提升入库(数据库)的患者数据的统一性。
例如，数据传递模块也可以被称为跨关键字互传递模块。例如，通过计算结构相似度，可以依据语义结构关联性性进行传递提供便利。例如，通过计算结构相似度，可以基于上述结构相似度将对应关键字的数据转换至相应的字段进行存储，可以让对应于具有不同语言表述相同实质含义的关键字的数据进行互传递，提升了入库患者数据的统一性。
图5是本公开的至少一个实施例提供的数据处理模块的第二个示例的示例性框图。如图5所示，数据处理模块包括对象分组模块。
图6是本公开的至少一个实施例提供的数据传递模块的一个示例性框图。如图6所示，对象分组模块包括第三数据接收子模块、评分获取子模块和分组子模块。
例如，第三数据接收子模块被配置为接收与健康管理装置关联的对象的信息。例如，与健康管理装置关联的对象的信息包括客观信息和主观信息。
例如，客观信息的至少部分由便携式医疗设备(例如，患者拥有的便携式医疗设备)或可穿戴医疗设备(例如，患者佩戴的可穿戴医疗设备)对与健康管理装置关联的对象进行监测获取。例如，通过使得客观信息的至少部分由便携式医疗设备或可穿戴医疗设备对与健康管理装置关联的对象进行监测获取，可以获取与健康管理装置关联的对象的最新体征数据，由此允许在与健康管理装置关联的对象的病情恶化或加重时及时提醒与健康管理装置关联的对象以及为与健康管理装置关联的对象提供诊疗服务的医生。此外，通过使得 客观信息的至少部分由便携式医疗设备或可穿戴医疗设备对与健康管理装置关联的对象进行监测获取，还可以在与健康管理装置关联的对象的病情恶化或加重时及时更新对象分组模块输出的分组结果，由此可以提升利用对象分组模块输出的分组结果进行数据处理的模块的计算结果的准确性。
例如，主观信息可以由以下方式的至少一种获得：医务工作者问询(通过远程视频***进行问询和确认)；与健康管理装置关联的对象在首次使用与健康管理装置匹配的对象端(例如，患者端；例如，微信小程序)时通过提供(例如，通过填写电子问卷的形式提供)；与健康管理装置关联的对象在情况发生改变时，通过重新填写电子问卷的形式更新。
例如，第三数据接收子模块接收的与健康管理装置关联的对象的所患有的疾病种类相关。例如，对于慢性阻塞性肺疾病(COPD)，第三数据接收子模块接收的与健康管理装置关联的对象的信息包括：年龄、性别、身高和体重(或BMI)、收缩压、舒张压、职业工作烈度、病情加重次数、住院治疗次数、手术治疗次数、常态呼吸频率、常态血氧饱和度、精神状态和运动状态中的任意一项或任意组合。
例如，身高、体重、血压、呼吸频率、血氧饱和度等可以使用便携式医疗设备或可穿戴医疗设备进行一次或多次测量获得。例如，在采集与健康管理装置关联的对象的信息之前，可以让对象知晓信息将被采集。
例如，评分获取子模块被配置为基于与健康管理装置关联的对象的信息以及查询表获取与健康管理装置关联的对象在多个评分项目上的评分。
例如，多个评分项目以及查询表的具体形式与健康管理装置关联的对象所患有的疾病种类相关。例如，在数据处理模块的至少一个示例中，多个评分项目包括：年龄、性别、BMI(身体质量指数)、收缩压、舒张压、职业工作烈度、病情加重次数(指定疾病的病情加重次数)、住院治疗次数、手术治疗次数、常态呼吸频率、常态血氧饱和度、精神状态和运动状态。
例如，通过采用基于与健康管理装置关联的对象的信息以及查询表获取与健康管理装置关联的对象在多个评分项目上的评分可以实现对与健康管理装置关联的对象的自动化分组。并且，与健康管理装置关联的对象在多个评分项目上的评分可以更为直观的指示对象在指定评分项目上的异常程度，因此，有利于医务工作者基于查询表进行结果反查，进而有利于医务工作者对查询表中的分数分布与实际的医疗实践的匹配程度给出反馈。
例如，多个评分项目以及查询表可以采用非负记分法，也即，如果对象在该评分项目属于正常则对象在该评分项目上的评分记为0分，且对象在该评分项目上的评分越大，表明对象在该评分项目上越异常。例如，通过采用非负记分法，可以避免评分为负值，这可以提升计算结果的可靠度。
例如，用于COPD查询表的一个示例参见表1。例如，表1所示的查询表采用了量化评分机制，依据现有的临床观察而制定，标准简单、统一且非常易于实施。
表1 用于COPD查询表
Figure PCTCN2021096383-appb-000009
例如，分组子模块被配置为基于与健康管理装置关联的对象在多个评分项目上的评分为与健康管理装置关联的对象分配组别。
例如，分组子模块被配置为基于在多个评分项目上的评分以及多个评分项目的权重获取与健康管理装置关联的对象的总体评分。
例如，与健康管理装置关联的对象在第i个评分项目上的评分为y i，此处，i大于等于1小于等于c，c为多个评分项目的数目。例如，对于表1所示的用于COPD查询表，c等于13。例如，多个评分项目中第i个评分项目的权重为w fi。
例如，与健康管理装置关联的对象的总体评分等于与健康管理装置关联的对象在多个评分项目上的评分的加权和。例如，与健康管理装置关联的对象的总体评分Sc可以采用以下的表达式进行表示。
Figure PCTCN2021096383-appb-000010
例如，分组子模块还被配置为基于总体评分为与健康管理装置关联的对象分配组别。例如，与健康管理装置关联的对象的组别序号等于总体评分向上取整值。例如，在指定对象的总体评分为1.2的情况下，指定对象所在组的编号为2。例如，对于上述的COPD示例，可以将与健康管理装置关联的对象划分为6个亚组，也即第0组、第1组、第2组、第3组、第4组和第5组。
例如，通过基于在多个评分项目上的评分以及多个评分项目的权重获取与健康管理装置关联的对象的总体评分可以将对象(例如，患者)的个性化因素以及医疗的个性化因素纳入考虑。
例如，不同医院的医疗资源存在显著差异，这种差异对于对象的治疗方案会带来很大影响，比如，床位比较紧张可能会显著降低住院治疗次数。例如，对于表1所示的查询表，可以根据为患者服务的医疗机构的实际情况，依次设置从w f1到w f13共计13个权重系数。例如，权重系数可以用来自适应地调整参考基线。例如，可以根据医院实际床位供给和诊疗次数可以适当调整w f8从而更好地拟合医院当前诊疗方案。例如，根据我们已有的数据经验，可以将第10个评分项目的权重与第11个评分项目的权重之和设置为不低于45％的数值，以提升分组的质量(例如，准确性)。
例如，通过基于在多个评分项目上的评分以及多个评分项目的权重获取与健康管理装置关联的对象的总体评分，可以根据各地医疗实践的标准进行灵活修改，从而可以提升分组的稳定性和扩展性。
图7是本公开的至少一个实施例提供的数据处理模块的第三个示例的示例性框图。如图7所示，数据处理模块包括对象分组模块和存储级别分配子模块。例如，对象分组模块的具体实现方式可以参见相关实施例，在此不再赘述。
例如，存储级别分配子模块被配置为至少基于对象分组模块输出的分组结果(例如，指定对象的组别)，将与管理装置关联的对象对应的设备相关数据进行分级，并根据分级结果为与管理装置关联的对象对应的设备相关数据分配与分级结果相匹配的处理器。例如，可以将级别高的数据放入高速存储器中(例如，缓存)中，将级别中等的数据放入中速存储器中(例如，随机存取存储器)中，将级别低的数据放入低速存储器中(例如，只读存储器或固定存储器)中。
例如，由于对象的组别表示对象的病情严重程度，对象的组别也可以用于指示对象的数据的重要程度以及查询频率的预测值，因此，在存储数据之前至少根据指定对象的组别对与管理装置关联的对象对应的设备相关数据进行分级，并根据分级结果为与管理装置关联的对象对应的设备相关数据分配与分级结果相匹配的处理器，可以加快重要数据和高频数据查询速度。对应地，可以在对象的病情恶化或加重时及时提醒与健康管理装置关联的对象以及为与健康管理装置关联的对象提供诊疗服务的医生，由此可以提升使用上述健康管理装置的对象(例如，患者)和医生的使用体验。
例如，存储级别分配子模块被配置为至少基于对象分组模块输出的分组结果(例如，指定对象的组别)，对应于指定对象的设备相关数据需要占用的存储空间，对应于指定对象的设备相关数据的重要程度，指定对象的以往数据的查询频率，将与管理装置关联的对象对应的设备相关数据进行分级。
例如，通过采用患者加权分组机制，可以允许存储级别分配子模块将与管理装置关联的设备生成的设备相关数据根据其重要程度(预测值)和查询频率(预测值)产生自主缓存决策策略，对近期高频数据进行缓存可加快查询速度和分析速度，提高预警子模块(参见图8)的及时性，对重要低频数据选择性地执行写入操作，提高数据库通信效率，使得重要数据具有高可靠性和高可用性。
图8是本公开的至少一个实施例提供的数据处理模块的第四个示例的示例性框图。如 图8所示，数据处理模块包括对象分组模块和预警子模块。例如，预警子模块被配置为接收分组子模块提供的与健康管理装置关联的对象的总体评分，并在与健康管理装置关联的对象的总体评分大于评分阈值时输出预警信息。例如，评分阈值可以根据实际应用需求进行设定，本公开的实施例对此不作具体限定。例如，评分阈值可以设置为固定值(指示病情为中重症的数值)；例如，评分阈值可以设置为变化值。例如，评分阈值可以是本次计算之前，对象的评分值的向上取整值。例如，通过使得数据处理模块包括对象分组模块和预警子模块，可以在对象的病情恶化或加重时及时提醒与健康管理装置关联的对象以及为与健康管理装置关联的对象提供诊疗服务的医生，由此可以提升使用上述健康管理装置的对象(例如，患者)和医生的使用体验。
在一些示例中，数据处理模块可以同时包括对象分组模块、存储级别分配子模块和预警子模块，对象分组模块、存储级别分配子模块和预警子模块的具体实现方式可以参见相关实施例，在此不再赘述。
图9是本公开的至少一个实施例提供的数据处理模块的第五个示例的示例性框图。如图9所示，数据处理模块包括对象分组模块和数据补充模块。例如，对象分组模块的具体实现方式可以参见相关实施例，在此不再赘述。
需要说明的是，在一些示例中，数据处理模块可以不包括对象分组模块，此种情况下，数据补充模块直接从与健康管理装置关联的数据库或存储器接收与健康管理装置关联的多个对象的分组结果。
图10是图9所示的数据补充模块的一个示例性框图；图11是是本公开的至少一个实施例提供的数据处理模块的第五个示例的另一个示例性框图。如图10和图11所示，数据补充模块包括第二数据接收子模块和插值计算子模块。
例如，第二数据接收子模块被配置为接收***值数据集，***值数据集包括待补充数据对应的对象所在组的其他对象的至少部分的数据。例如，***值数据集可以包括待补充数据对应的对象所在组的多个对象中除待补充数据对应的对象之外的所有对象的数据。例如，***值数据集中的每条数据的关键字的字段与待补充数据的关键字的字段实质上相同。
例如，在***值数据集中，多条数据与多个对象一一对应，也即，多条数据的条数等于第一对象(待补充数据对应的对象)所在组的除第一对象之外的其他对象的数目。又例如，在***值数据集中，每个对象与多条数据相关，此种情况下，多条数据的条数大于第一对象(待补充数据对应的对象)所在组的除第一对象之外的其他对象的数目。
例如，可以首先使用数据传递模块将多个对象对应的设备相关数据的数据格式统一，然后可以通过查看设备相关数据在指定时间处对应于指定关键字处的数值是否为空来判定指定对象对应的设备相关数据是否存在数据缺失问题。例如，上述数据缺失问题可以是由数据传输或数据获取导致的。又例如，上述数据缺失问题可以是由设备采集数据的频率较低导致的。
例如，待补充数据可以是设备相关数据中对应于部分关键字的数据。例如，在指定设备生成的设备相关数据包括{“设备识别符”：“01010202”，“实质信息”：{“呼吸频 率”：“15”，“血氧饱和度”：“95％”，“收缩压”：“100”}}，数据补充模块接收到的数据包括{“设备识别符”：“01010202”，“实质信息”：{“呼吸频率”：“15”，“血氧饱和度”：“95％”，“收缩压”：“”}}的情况下，可以判定设备相关数据中对应于关键字“收缩压”的数据存在缺失问题；对应地；如果待补充数据对应的对象(例如，第一对象)位于第三组，可以使用第二数据接收子模块接收第三组的多个对象中除第一对象之外的所有对象对应的设备相关数据中对应于关键字“收缩压”的数据。
例如，***值数据集包括n条数据，x i为被差值数据集中第i条数据，i为大于等于1小于等于n的正整数。
例如，插值计算子模块被配置为对***值数据集进行插值获取待补充数据。例如，通过使用待补充数据对应的对象所在组的其他对象的至少部分的数据获取待补充数据，可以提升待补充数据的可信度。
例如，如图10和图11所示，数据补充模块还包括距离计算模块；距离计算模块被配置为计算待补充数据和***值数据集中每条数据的名义距离，并将名义距离提供给插值计算子模块；插值计算子模块被配置为至少基于名义距离对***值数据集进行距离反比插值获取待补充数据。
例如，本公开的发明人注意到，相比于待补充数据对应的对象的历史数据，待补充数据对应的对象所在组的其他对象的设备相关数据中与待补充数据名义距离近的数据更为接近待补充数据，因此，通过至少基于名义距离对***值数据集进行距离反比插值获取待补充数据，可以使得待补充数据更为可信和准确。
例如，距离计算模块被配置为基于待补充数据和***值数据集中每条数据的时间距离和地理距离计算待补充数据和***值数据集中每条数据的名义距离。例如，相比于仅基于待补充数据和***值数据集中每条数据的时间距离或地理距离计算待补充数据和***值数据集中每条数据的名义距离，通过基于待补充数据和***值数据集中每条数据的时间距离和地理距离计算待补充数据和***值数据集中每条数据的名义距离，可以使得待补充数据和***值数据集中每条数据的名义距离更为准确。
例如，第二数据接收子模块还被配置为接收***值数据集中每条数据对应的时间信息(例如，时间戳)和地理位置信息以及待补充数据对应的时间信息(例如，时间戳)和地理位置信息。
例如，待补充数据对应的时间信息(例如，时间戳)和地理位置信息可以基于设备的预定的数据采集频率以及与待补充数据相邻的设备相关数据对应的时间信息(例如，时间戳)和地理位置信息获取(推测得到)。
例如，上述数据对应的地理位置信息可以是对象在注册时提供的家庭住址和工作地址、对象使用的设备生成数据时设备的位置、对象使用的设备上传上述数据时设备的位置的至少一项。
例如，上述数据对应的时间信息可以是对象使用的设备生成数据的时间以及对象使用的设备上传上述数据的时间的至少一项。
例如，距离计算模块还被配置为基于***值数据集中每条数据对应的时间信息以及待补充数据对应的时间信息计算待补充数据和***值数据集中每条数据的时间距离；距离计算模块还被配置为基于***值数据集中每条数据对应的地理信息以及待补充数据对应的地理信息计算待补充数据和***值数据集中每条数据的地理距离。
例如，可以待补充数据对应的患者与被差值数据对应的患者间的地理位置信息，确定两个患者之间的最短街道的路口数目，并将上述最短街道的路口数目作为两个患者之间的地理距离。例如，可以待补充数据对应的患者的数据的时间戳以及被差值数据对应的患者间的数据的时间戳之间的时间差为两个患者之间的时间距离。
例如，待补充数据和***值数据集中每条数据的名义距离等于待补充数据和***值数据集中每条数据的空间距离以及待补充数据和***值数据集中每条数据的时间距离的加权和。待补充数据x p和被差值数据集中第i条数据x i的名义距离为d pi。
例如，可以为空间距离和时间距离分别设置权重系数ws和wt。例如，权重系数ws和wt可以基于针对指定疾病的医疗实践和统计调查结果进行设定。例如，通过将待补充数据和***值数据集中每条数据的空间距离以及待补充数据和***值数据集中每条数据的时间距离的加权和作为待补充数据和***值数据集中每条数据的名义距离，可以提升距离计算模块和数据补充模块的应用范围以及输出结果的准确度。
例如，第二数据接收子模块还被配置为接收待补充数据对应的对象所在组的阻尼项；插值计算子模块被配置为基于名义距离和阻尼项对***值数据集进行距离反比插值获取待补充数据。
例如，待补充数据x p满足以下的表达式：
Figure PCTCN2021096383-appb-000011
e k为待补充数据对应的对象所在组的阻尼项，k大于等于1小于等于t，t为与健康管理装置关联的对象所涉及的组的编号。
例如，基于名义距离和阻尼项对***值数据集进行距离反比插值获取待补充数据，可以增大疾病严重程度较大的组的对象对应的待补充数据的插值程度(数值)，由此可以提升插值得到数据的有效率(例如，计量有效性)。
例如，不同组的阻尼项e k的数值可以不同。例如，不同组的阻尼项e k的数值可以基于针对指定疾病的医疗实践和统计调查结果进行设定。
例如，下面结合一个示例对本公开的至少一个实施例提供的数据补充模块的实现方式进行示例性说明。
例如，假设第一组的患者A在时间点t1(2020年6月3日16.00)的收缩血压数据缺失，第一组除患者A之外还有其他三个患者，且上述三个患者在时间点t1的收缩血压数据 分别是110、120和130。例如，基于距离计算模块可以得到的患者A和上述三个患者的名义距离分别为1、2和3。例如，将第一组的阻尼项e k设定为0.1。此种情况下，***值数据x p可以由以下的表达式计算得到。
Figure PCTCN2021096383-appb-000012
例如，若设置阻尼项e k为1.1，则可放大插值程度，x p的值为127。
需要说明的是，待补充数据可以是设备相关数据中对应于全部关键字的数据。例如，在指定设备在预定时间段生成的设备相关数据包括{“设备识别符”：“01010202”，“实质信息”：{“呼吸频率”：“15”，“血氧饱和度”：“95％”，“收缩压”：“100”}}和{“设备识别符”：“01010202”，“实质信息”：{“呼吸频率”：“18”，“血氧饱和度”：“94％”，“收缩压”：“98”}}，但数据补充模块接收到的数据仅包括{“设备识别符”：“01010202”，“实质信息”：{“呼吸频率”：“15”，“血氧饱和度”：“95％”，“收缩压”：“100”}}的情况下，可以判定设备相关数据中对应于所有关键字的数据存在缺失问题；此种情况下，可以利用上述数据补充模块以及对应的方法分别获取对应于每个关键字的数据。
例如，依据患者在健康管理装置中注册时提供的生活地理位置信息、移动设备所回传的地理位置信息和时间戳信息，可以将缺失的数据(传输问题、数据获取问题等)和频次不一的数据进行距离反比插值。
例如，数据处理装置还包括前置缓存模块(图中未示出)和后置缓存模块(图中未示出)。前置缓存模块被配置为接收与健康管理装置关联的设备生成的设备相关数据，并将设备相关数据提供给数据处理模块；以及后置缓存模块被配置为接收数据处理模块输出的处理后的数据，并将处理后的数据提供给健康管理装置的存储器。
例如，前置缓存模块被配置为与分布式***配合，以允许数据处理模块对与健康管理装置关联的设备生成的设备相关数据进行分布式处理。
在一些示例中，数据处理模块包括数据传递模块、对象分组模块和数据补充模块；数据处理装置包括数据处理模块、前置缓存模块和后置缓存模块。
例如，本公开的至少一个实施例涉及的模块和子模块可以通过软件、固件、硬件及其任意组合实现，例如，该硬件包括服务器、现场可编程门阵列(Field Programmable Gate Array，FPGA)等。
下面结合图12-图15所示的示例对本公开的至少一个实施例提供的数据处理装置做示例性说明。
本公开的至少一个实施例还提供了一种数据敏捷入库技术***1。图12是本公开的至少一个实施例提供的数据敏捷入库技术***1的示例性框图。
例如，如图12所示，该敏捷入库技术***1包括COPD基线评分模块11、距离反比混合插值模块12和跨关键字互传递模块13。
例如，COPD基线评分模块11、距离反比混合插值模块12和跨关键字互传递模块13可以分别由图1-图11所示的对象分组模块、数据补充模块和数据传递模块实现，因此，上述COPD基线评分模块11、距离反比混合插值模块12和跨关键字互传递模块13的具体实现方式可以参见图1-图11所示的示例。
例如，COPD基线评分模块主要由患者所提供的基本信息(包括既往病史)进行合理的评分，并将患者进行异质性分组利于对其进行移动数据的分组处理。
例如，距离反比混合插值模块在前述评分模块的基础上根据组内患者的名义距离分布(包括基线情况、地理分布等)进行混合插值，插值算法可以通过已知的，离散的数据点，在范围内推求新数据点，从而改善因数据缺损、测量频次不同带来的差异。
例如，跨关键字互传递模块主要用来解决不同的患者所传递的移动数据因为所使用的设备、使用方法等带来的差异项进行归一化和传递，以允许得到统一化数据。
例如，上述数据敏捷入库技术***1可以为面向慢性阻塞性肺疾病的患者所佩戴的移动设备采集的移动数据提供一种高效便捷的数据库入库存储和处理解决方案。
例如，如图12所示，数据敏捷入库技术***还可以使用一个前置的数据链路缓存2作为整个***的导入端。该缓存可以将近一段时间内涌入***的移动数据高效存储，并且按照先进先出的规则将数据及时供给后续处理模块(与分布式***中协同，先进先出可以提升转移和存储的一致性，保持数据的同步)，使得移动数据可以例如实时地进入本***，防止数据遗漏。其内部的数据预处理由以上三个模块一起协同工作完成。待数据预处理完成后，将清洗后的移动数据导入(例如，经由后置的数据链路3导入)后续分析***中。例如，由于数据是分片按时间戳入列，本***使用分布式池处理体系(即采用线程池设计)，以线程为最小调度单位，可以加快并行速度。例如，将每个片数据进行预处理后，就交由后置入库程序处理，进入数据库。例如，本***可以高效地实现移动数据经过预处理并进入数据库的全部过程。
例如，通过使用前置/后置数据链路进行缓存，可以保持一个高效率的先入先出缓存，可以有效接收超量的流式数据输入，防止数据流失。
例如，通过采用分布式池处理机制，按时间片对流式数据进行分片，并分配至分布式计算池进行预处理计算，可以提升***的实时计算效能，提供***的健壮性。
图13是图12所示的数据敏捷入库技术***1的COPD基线评分模块的示例性框图。例如，如图13所示，COPD基线评分模块11包括评分表查询子模块111和患者分组子模块112。
例如，评分表查询子模块111和患者分组子模块112可以分别由图6所示的评分获取子模块和分组子模块实现。因此，上述评分表查询子模块111和患者分组子模块112的具体实现方式可以参见图6所示的示例。
图14是图12所示的数据敏捷入库技术***1的距离反比混合插值模块12的示例性框图。例如，如图13所示，距离反比混合插值模块13包括数据导入子模块131、距离计算子模块133和插值计算子模块132。
例如，数据导入子模块131、距离计算子模块133和插值计算子模块132可以分别由图10所示的第二数据接收子模块、距离计算子模块和差值计算子模块实现。因此，上述数据导入子模块131、距离计算子模块133和插值计算子模块132的具体实现方式可以参见图10所示的示例。
例如，通过采用距离反比插值计算机制，通过在分组的患者内进行按距离反比的插值计算，可以将输入的移动数据中丢失的部分进行插值补足，如频率不一的流式数据等。例如，插值计算中还引入了距离阻尼因素，可以有效提升本插值计算的可信度。
例如，通过采用混合距离计算机制，结合患者基线信息和历史的数据进行名义距离(例如，距离的定义包括实际距离、测地线距离等)的计算，可以根据实际情况的需要进行客制化定制，最终将结果输入距离反比插值计算中。
图15是图12所示的数据敏捷入库技术***1的跨关键字互传递模块的示例性框图。例如，如图14所示，跨关键字互传递模块12包括数据导入子模块121、关键字相似度计算子模块123、结构矩阵计算子模块124、数据传递子模块122。
例如，数据导入子模块121、关键字相似度计算子模块123+结构矩阵计算子模块124以及数据传递子模块122可以分别由图3所示的第一数据接收子模块、结构相似度计算子模块和传递子模块实现。因此，数据导入子模块121、关键字相似度计算子模块123+结构矩阵计算子模块124以及数据传递子模块122的具体实现方式可以参见图3所示的示例。
例如，上述的移动数据敏捷入库方案，在收集该子类疾病的患者所携带的各类智能化设备的回传数据基础上，可以实现对于这些原始数据的数据预处理，并实施敏捷入库，提供了完整、规范、整齐的高质量数据源，为在该类型疾病的数据基础上进行进一步的分析提供了便利。
例如，COPD基线评分模块采用了数据表量化评分机制，依据现有的临床观察而制定，标准简单、统一且非常易于实施，与此同时亦可根据各地医疗实践的标准进行灵活修改，从而提高了本方案的逻辑稳定性和扩展性。依据该评分机制可以对入库患者数据进行简单查询就可以归并分组，利于后续预处理任务高效地分布式执行。
例如，距离反比混合插值模块采用了名义距离分布机制，在患者分组的基础上，依据移动数据的采集地点和采集时间的时空相关性，可以得到按相邻患者数据进行插值的可信度，根据其值的大小和反比关系，可以进行基于基线数据的混合插值。依据该插值模块可以对丢失的数据值进行填充，同时对不同频率的数据可以进行插值进而获得相同频率的入库患者数据，可以极大地改善后续数据分析的粒度难题。
例如，跨关键字互传递模块采用了关键字相似度矩阵转换机制。在患者移动数据传输到达之后，根据所获得的数据进行解包，可获得列表类数据，再根据其表头关键字相似度和该类数据的结构特点，计算结构成组相似度矩阵。当患者所使用的设备关键字与其他患者关键字不统一的时候，依据该模块，可以将数据顺利转换至相同的存储形态，保证入库患者数据的统一性。
本公开的至少一个实施例提供了一种数据处理方法，其包括：对源于与健康管理装置 关联的设备的设备相关数据进行数据处理。
图16是本公开的至少一个实施例提供的数据处理方法的示例性流程图。
例如，如图16所示，该数据处理方法包括以下的步骤S90。
步骤S90：对源于与健康管理装置关联的设备的设备相关数据进行数据处理。
例如，数据处理后的数据被提供给与健康管理装置关联的存储器或数据库中。
在第一个示例中，步骤S90包括以下的步骤S911和步骤S912。
步骤S911：基于模板数据涉及的第一组关键字的信息以及设备相关数据涉及的第二组关键字的信息获取第一组关键字和第二组关键字之间的结构相似度。
步骤S912：基于结构相似度将设备相关数据的至少部分传递至模板数据中。
例如，步骤S911和步骤S912可以被顺次执行。
例如，第一组关键字和第二组关键字之间的结构相似度包括第一组关键字的每个关键字的特征向量与第二组关键字的每个关键字的特征向量之间的相似度。v 1、v 2、……v m为第一组关键字的特征向量，u 1、u 2、……u b为第二组关键字的特征向量，m为第一组关键字中关键字的数目，b为第二组关键字中关键字的数目。
例如，步骤S911包括：通过计算相似度矩阵S来获取第一组关键字和第二组关键字之间的结构相似度；相似度矩阵S满足以下的表达式。
Figure PCTCN2021096383-appb-000013
例如，步骤S912包括以下的步骤S9121-步骤S9123。
步骤S9121：将第二组关键字的每个关键字的特征向量与第一组关键字的每个关键字的特征向量之间的相似度中取值大于预设相似度阈值的相似度作为相关相似度。
步骤S9122：将每个相关相似度对应的两个关键字中的属于第二组关键字的关键字作为第一关键字，并将每个相关相似度对应的两个关键字中的属于第一组关键字的关键字作为第二关键字。
步骤S9123：将设备相关数据中与第一关键字关联的数据与模板数据涉及的第二关键字关联。
例如，步骤S9121、步骤S9122和步骤S9123可以被顺次执行。
例如，在第一个示例中，步骤S90还包括以下的步骤S913。
步骤S913：接收模板数据涉及的第一组关键字的信息以及设备相关数据。例如，步骤S913可以在步骤S911之前执行。
例如，第一组关键字的信息包括第一组关键字的每个关键字的特征向量，第二组关键字的信息包括第二组关键字的每个关键字的特征向量。
例如，在第一个示例中，步骤S90还包括以下的步骤S914和步骤S915。
步骤S914：从设备相关数据提取出第二组关键字以及第二组关键字的每个关键字在设备相关数据中的层级信息。
步骤S915：基于第二组关键字以及第二组关键字的每个关键字在设备相关数据中的层级信息生成第二组关键字的每个关键字的特征向量。
例如，步骤S914和步骤S915可以被顺次执行。例如，步骤S914和步骤S915可以在步骤S913之后且在步骤S911执行。
例如，步骤S913、步骤S914、步骤S915、步骤S911和步骤S912可以被顺次执行。
需要说明的是，在第一个示例中，数据处理方法不限于包括步骤S913。
在一个示例中，该数据处理方法不包括上述的步骤S913，该数据处理方法包括接收模板数据涉及的第一组关键字的信息以及设备相关数据涉及的第二组关键字的信息；此种情况下，数据处理方法不包括上述的步骤S914和步骤S915。
在另一个示例中，该数据处理方法不包括上述的步骤S913，本公开的至少一个实施例提供的数据处理方法包括接收模板数据以及设备相关数据；此种情况下，除步骤S911、步骤S912、步骤S914和步骤S915之外，数据处理方法还包括以下的步骤S916和步骤S917；步骤S916和步骤S917可以被顺次执行。例如，步骤S916和步骤S917可以在步骤S913之后且在步骤S911执行。
步骤S916：从模板数据提取出第一组关键字以及第一组关键字的每个关键字在模板数据中的层级信息。
步骤S917：基于第一组关键字以及第一组关键字的每个关键字在设备相关数据中的层级信息生成第一组关键字的每个关键字的特征向量。
在再一个示例中，该数据处理方法不包括上述的步骤S913，但包括以下的步骤S918：接收模板数据涉及的第一组关键字以及第一组关键字的层级信息以及设备相关数据涉及的第二组关键字以及第二组关键字的层级信息；此种情况下，该数据处理方法还包括步骤S915、步骤S917、步骤S911和步骤S912。例如，步骤S918、步骤S915+步骤S917、步骤S911和步骤S912可以顺次被执行。
例如，步骤S911-步骤S917的具体实现方式可以参见图2所示的示例的相关描述，在此不再赘述。
在第二个示例中，步骤S90包括以下的步骤S921-步骤S923。
步骤S921：接收与健康管理装置关联的对象的信息。
步骤S922：基于与健康管理装置关联的对象的信息以及查询表获取与健康管理装置关联的对象在多个评分项目上的评分。
步骤S923：基于与健康管理装置关联的对象在多个评分项目上的评分为与健康管理装置关联的对象分配组别。
例如，步骤S921、步骤S922和步骤S923可以被顺次执行。
例如，步骤S923包括以下的步骤S9231和步骤S9232。例如，步骤S9231和步骤S9232可以被顺次执行。
步骤S9231：基于在多个评分项目上的评分以及多个评分项目的权重获取与健康管理装置关联的对象的总体评分。
步骤S9232：基于总体评分为与健康管理装置关联的对象分配组别。
例如，在步骤S9231中，基于在多个评分项目上的评分以及多个评分项目的权重获取与健康管理装置关联的对象的总体评分包括：将与健康管理装置关联的对象在多个评分项目上的评分的加权和作为与健康管理装置关联的对象的总体评分。
例如，在步骤S9232中，基于总体评分为与健康管理装置关联的对象分配组别包括：将总体评分向上取整值作为与健康管理装置关联的对象的组别序号。
例如，与健康管理装置关联的对象的信息包括客观信息和主观信息；客观信息的至少部分由便携式医疗设备或可穿戴医疗设备对与健康管理装置关联的对象进行监测获取；以及主观信息的至少部分由医务工作者问询或者与健康管理装置关联的对象填写的电子问卷获取。
例如，对于COPD，多个评分项目包括：年龄、性别、BMI、收缩压、舒张压、职业工作烈度、病情加重次数、住院治疗次数、手术治疗次数、常态呼吸频率、常态血氧饱和度、精神状态和运动状态。
例如，在第二个示例中，步骤S90还包括以下的步骤S924。例如，步骤S924可以在步骤S9231之后执行。
步骤S924：接收与健康管理装置关联的对象的总体评分，并在与健康管理装置关联的对象的总体评分大于评分阈值时输出预警信息。
例如，步骤S921-步骤S924的具体实现方式可以参见图5所示的示例的相关描述，在此不再赘述。
在第三个示例中，步骤S90包括以下的步骤S931和步骤S932。
步骤S931：接收***值数据集，此处，***值数据集包括待补充数据对应的对象所在组的其他对象的至少部分的数据。
步骤S932：对***值数据集进行插值获取待补充数据。
例如，步骤S931和步骤S932可以被顺次执行。
例如，在步骤S932中，对***值数据集进行插值获取待补充数据包括：至少基于名义距离对***值数据集进行距离反比插值获取待补充数据。
例如，在第三个示例中，步骤S90还包括以下的步骤S933。
步骤S933：计算待补充数据和***值数据集中每条数据的名义距离。
例如，步骤S933可以在步骤S932之前执行。
例如，在步骤S933中，计算待补充数据和***值数据集中每条数据的名义距离包括：基于待补充数据和***值数据集中每条数据的时间距离和地理距离计算待补充数据和***值数据集中每条数据的名义距离。
例如，在第三个示例中，步骤S90还包括以下的步骤S934。
步骤S934：接收***值数据集中每条数据对应的时间信息和地理位置信息以及待补充 数据对应的时间信息和地理位置信息。
例如，步骤S934可以在步骤S933之前执行。
例如，在步骤S933中，计算待补充数据和***值数据集中每条数据的名义距离包括以下的步骤S9331-步骤S9333。
步骤S9331：基于***值数据集中每条数据对应的时间信息以及待补充数据对应的时间信息计算待补充数据和***值数据集中每条数据的时间距离；
步骤S9332：基于***值数据集中每条数据对应的地理信息以及待补充数据对应的地理信息计算待补充数据和***值数据集中每条数据的地理距离；
步骤S9333：将待补充数据和***值数据集中每条数据的空间距离以及待补充数据和***值数据集中每条数据的时间距离的加权和作为待补充数据和***值数据集中每条数据的名义距离。
例如，可以按照步骤S9331、步骤S9332和步骤S9333的顺序执行步骤S9331-步骤S9333。又例如，可以按照步骤S9332、步骤S9331和步骤S9333的顺序执行步骤S9331-步骤S9333。再例如，可以按照步骤S9331+步骤S9332(也即，步骤S9331和步骤S9332同时被执行)和步骤S9333的顺序执行步骤S9331-步骤S9333。
例如，在第三个示例中，步骤S90还包括以下的步骤S935。
步骤S935：接收待补充数据对应的对象所在组的阻尼项。例如，步骤S935在步骤S932之前被执行。
例如，在步骤S932中，对***值数据集进行插值获取待补充数据包括：基于名义距离和阻尼项对***值数据集进行距离反比插值获取待补充数据。
例如，在数据处理方法的至少一个示例中，待补充数据x p满足以下的表达式：
Figure PCTCN2021096383-appb-000014
x i为被差值数据集中第i条数据，i为大于等于1小于等于n的正整数；n为***值数据集中数据的条数；d pi为待补充数据x p和被差值数据集中第i条数据x i的名义距离；e k为待补充数据对应的对象所在组的阻尼项，k大于等于1小于等于t，t为与健康管理装置关联的对象所涉及的组的编号。
例如，步骤S931+步骤S934+步骤S935(也即，步骤S931、步骤S934、步骤S935被同时执行)、步骤S933以及步骤S932可以被顺次执行。又例如，步骤S931+步骤S934(也即，步骤S931和步骤S934被同时执行)、步骤S933、步骤S935以及步骤S932可以被顺次执行。再例如，步骤S931、步骤S934、步骤S933、步骤S935以及步骤S932可以被顺次执行。又再例如，步骤S934、步骤S931、步骤S933、步骤S935以及步骤S932可以被顺次执行。
例如，步骤S931-步骤S935的具体实现方式可以参见图9和图10所示的示例的相关描述，在此不再赘述。
在第四个示例中，步骤S90包括以下的步骤S921-步骤S923以及步骤S931-步骤S935。例如，步骤S921-步骤S923在步骤S931-步骤S935之前被执行。
在第五个示例中，步骤S90包括以下的步骤S911-步骤S915、步骤S921-步骤S923以及步骤S931-步骤S935。例如，步骤S921-步骤S923在步骤S931-步骤S935之前被执行；步骤S911-步骤S915在步骤S931-步骤S935之前被执行。例如，步骤S911-步骤S915在步骤S921-步骤S923之前被执行。又例如，步骤S911-步骤S915在步骤S921-步骤S923之后被执行。
在第六个示例中，在第五个示例中，步骤S90包括以下的步骤S911-步骤S915以及步骤S921-步骤S923。例如，步骤S911-步骤S915在步骤S921-步骤S923之前被执行。又例如，步骤S911-步骤S915在步骤S921-步骤S923之后被执行。
例如，在步骤S90中，对源于与健康管理装置关联的设备的设备相关数据进行数据处理包括：对源于与健康管理装置关联的设备的设备相关数据进行分布式处理。
本公开的至少一个实施例还提供了一种非暂时性存储介质。图17示出了本公开的至少一个实施例提供的非暂时性存储介质的示例性框图。例如，如图17所示，该非暂时性存储介质存储有计算机程序指令，计算机程序指令被处理器运行时使得计算机执行以下的方法包括：对源于与健康管理装置关联的设备的设备相关数据进行数据处理。
例如，非暂时性存储介质可以包括只读存储器(ROM)、硬盘、闪存等。例如，对源于与健康管理装置关联的设备的设备相关数据进行数据处理的具体实现方法可以参见本公开的至少一个实施例提供的数据处理方法，在此不再赘述。
本公开的至少一个实施例还提供了另一种数据处理装置。图18示出了本公开的至少一个实施例提供的另一种数据处理装置的示例性框图。例如，如图18所示，该另一种数据处理装置包括：处理器和存储器。存储器中存储有适于处理器执行的计算机程序指令，计算机程序指令被处理器运行时使得处理器执行以下的方法包括：对源于与健康管理装置关联的设备的设备相关数据进行数据处理。
例如，该处理器例如是中央处理单元(CPU)或者具有数据处理能力和/或指令执行能力的其它形式的处理单元，例如，该处理器可以实现为通用处理器，并且也为单片机、微处理器、数字信号处理器、专用的图像处理芯片、或现场可编程逻辑阵列等。存储器例如可以包括易失性存储器和/或非易失性存储器，例如可以包括只读存储器(ROM)、硬盘、闪存等。相应地，该存储器可以实现为一个或多个计算机程序产品，所述计算机程序产品可以包括各种形式的计算机可读存储介质，在所述计算机可读存储介质上可以存储一个或多个计算机程序指令。处理器可以运行所述程序指令，以实现期望的功能。该存储器还可以存储其他各种应用程序和各种数据，以及所述应用程序使用和/或产生的各种数据等。
例如，对源于与健康管理装置关联的设备的设备相关数据进行数据处理的具体实现方法可以参见本公开的至少一个实施例提供的数据处理方法，在此不再赘述。
本公开的至少一个实施例还提供了一种健康管理装置，其包括本公开的至少一个实施例提供的任一数据处理装置。图19是本公开的至少一个实施例提供的健康管理装置的示例性框图。如图19所示，该健康管理装置包括本公开的至少一个实施例提供的任一数据处理装置。例如，数据处理装置的具体实现方式可以参见相关实施例，在此不再赘述。
虽然上文中已经用一般性说明及具体实施方式，对本公开作了详尽的描述，但在本公开实施例基础上，可以对之作一些修改或改进，这对本领域技术人员而言是显而易见的。因此，在不偏离本公开精神的基础上所做的这些修改或改进，均属于本公开要求保护的范围。
以上所述仅是本公开的示范性实施方式，而非用于限制本公开的保护范围，本公开的保护范围由所附的权利要求确定。

Claims (24)

  1. 一种数据处理方法，包括：对源于与健康管理装置关联的设备的设备相关数据进行数据处理。
  2. 根据权利要求1所述的数据处理方法，其中，所述对源于与所述健康管理装置关联的设备的设备相关数据进行数据处理包括：
    基于模板数据涉及的第一组关键字的信息以及所述设备相关数据涉及的第二组关键字的信息获取所述第一组关键字和所述第二组关键字之间的结构相似度；以及
    基于所述结构相似度将所述设备相关数据的至少部分传递至所述模板数据中。
  3. 根据权利要求2所述的数据处理方法，其中，所述对源于与所述健康管理装置关联的设备的设备相关数据进行数据处理还包括：接收所述模板数据涉及的第一组关键字的信息以及所述设备相关数据；以及
    所述第一组关键字的信息包括所述第一组关键字的每个关键字的特征向量，所述第二组关键字的信息包括所述第二组关键字的每个关键字的特征向量。
  4. 根据权利要求3所述的数据处理方法，其中，所述对源于与所述健康管理装置关联的设备的设备相关数据进行数据处理还包括：
    从所述设备相关数据提取出所述第二组关键字以及所述第二组关键字的每个关键字在所述设备相关数据中的层级信息；
    基于所述第二组关键字以及所述第二组关键字的每个关键字在所述设备相关数据中的层级信息生成所述第二组关键字的每个关键字的特征向量。
  5. 根据权利要求3或4所述的数据处理方法，其中，所述第一组关键字和所述第二组关键字之间的结构相似度包括所述第一组关键字的每个关键字的特征向量与所述第二组关键字的每个关键字的特征向量之间的相似度。
  6. 根据权利要求5所述的数据处理方法，其中，所述基于模板数据涉及的第一组关键字的信息以及所述设备相关数据涉及的第二组关键字的信息获取所述第一组关键字和所述第二组关键字之间的结构相似度包括：通过计算相似度矩阵S来获取所述第一组关键字和所述第二组关键字之间的结构相似度；以及
    所述相似度矩阵S满足以下的表达式：
    Figure PCTCN2021096383-appb-100001
    v 1、v 2、……v m为所述第一组关键字的特征向量，
    u 1、u 2、……u b为所述第二组关键字的特征向量，
    m为所述第一组关键字中关键字的数目，
    b为所述第二组关键字中关键字的数目。
  7. 根据权利要求3-6任一项所述的数据处理方法，其中，所述基于所述结构相似度将所述设备相关数据的至少部分传递至所述模板数据中包括：
    将所述第二组关键字的每个关键字的特征向量与所述第一组关键字的每个关键字的特征向量之间的相似度中取值大于预设相似度阈值的相似度作为相关相似度；
    将每个相关相似度对应的两个关键字中的属于第二组关键字的关键字作为第一关键字，并将所述每个相关相似度对应的两个关键字中的属于第一组关键字的关键字作为第二关键字；以及
    将所述设备相关数据中与所述第一关键字关联的数据与所述模板数据涉及的所述第二关键字关联。
  8. 根据权利要求1-7任一项所述的数据处理方法，其中，所述对源于与所述健康管理装置关联的设备的设备相关数据进行数据处理还包括：
    接收***值数据集，其中，所述***值数据集包括待补充数据对应的对象所在组的其他对象的至少部分的数据；以及
    对所述***值数据集进行插值获取所述待补充数据。
  9. 根据权利要求8所述的数据处理方法，其中，所述对源于与所述健康管理装置关联的设备的设备相关数据进行数据处理还包括：计算所述待补充数据和所述***值数据集中每条数据的名义距离；以及
    所述对所述***值数据集进行插值获取所述待补充数据包括：至少基于所述名义距离对所述***值数据集进行距离反比插值获取所述待补充数据。
  10. 根据权利要求9所述的数据处理方法，其中，所述计算所述待补充数据和所述***值数据集中每条数据的名义距离包括：基于所述待补充数据和所述***值数据集中每条数据的时间距离和地理距离计算所述待补充数据和所述***值数据集中每条数据的名义距离。
  11. 根据权利要求10所述的数据处理方法，还包括：接收所述***值数据集中每条数据对应的时间信息和地理位置信息以及所述待补充数据对应的时间信息和地理位置信息，
    其中，所述计算所述待补充数据和所述***值数据集中每条数据的名义距离包括：
    基于所述***值数据集中每条数据对应的时间信息以及所述待补充数据对应的时间信息计算所述待补充数据和所述***值数据集中每条数据的时间距离；
    基于所述***值数据集中每条数据对应的地理信息以及所述待补充数据对应的地理信息计算所述待补充数据和所述***值数据集中每条数据的地理距离；以及
    将所述待补充数据和所述***值数据集中每条数据的空间距离以及所述待补充数据和所述***值数据集中每条数据的时间距离的加权和作为所述待补充数据和所述***值数据集中每条数据的名义距离。
  12. 根据权利要求9-11任一项所述的数据处理方法，还包括：接收所述待补充数据对应的对象所在组的阻尼项，
    其中，所述对所述***值数据集进行插值获取所述待补充数据包括：基于所述名义距离和所述阻尼项对所述***值数据集进行距离反比插值获取所述待补充数据。
  13. 根据权利要求12所述的数据处理方法，其中，所述待补充数据x p满足以下的表达式：
    Figure PCTCN2021096383-appb-100002
    x i为所述被差值数据集中第i条数据，i为大于等于1小于等于n的正整数；
    n为所述***值数据集中数据的条数；
    d pi为所述待补充数据x p和所述被差值数据集中第i条数据x i的名义距离；
    e k为所述待补充数据对应的对象所在组的阻尼项，k大于等于1小于等于t，t为与所述健康管理装置关联的对象所涉及的组的编号。
  14. 根据权利要求1-13任一项所述的数据处理方法，还包括：
    接收与所述健康管理装置关联的对象的信息；
    基于与所述健康管理装置关联的对象的信息以及查询表获取与所述健康管理装置关联的对象在多个评分项目上的评分；以及
    基于与所述健康管理装置关联的对象在所述多个评分项目上的评分为与所述健康管理装置关联的对象分配组别。
  15. 根据权利要求14所述的数据处理方法，其中，所述基于与所述健康管理装置关联的对象在所述多个评分项目上的评分为与所述健康管理装置关联的对象分配组别包括：
    基于在所述多个评分项目上的评分以及所述多个评分项目的权重获取与健康管理装置关联的对象的总体评分；以及
    基于所述总体评分为与所述健康管理装置关联的对象分配组别。
  16. 根据权利要求15所述的数据处理方法，其中，所述基于在所述多个评分项目上的评分以及所述多个评分项目的权重获取与所述健康管理装置关联的对象的总体评分包括：将与所述健康管理装置关联的对象在所述多个评分项目上的评分的加权和作为与所述健康管理装置关联的对象的总体评分；以及
    所述基于所述总体评分为与所述健康管理装置关联的对象分配组别包括：将所述总体评分向上取整值作为与所述健康管理装置关联的对象的组别序号。
  17. 根据权利要求15或16所述的数据处理方法，其中，与所述健康管理装置关联的对象的信息包括客观信息和主观信息；
    所述客观信息的至少部分由便携式医疗设备或可穿戴医疗设备对与所述健康管理装置关联的对象进行监测获取；以及
    所述主观信息的至少部分由医务工作者问询或者与所述健康管理装置关联的对象填写的电子问卷获取。
  18. 根据权利要求17所述的数据处理方法，其中，所述多个评分项目包括：年龄、性别、BMI、收缩压、舒张压、职业工作烈度、病情加重次数、住院治疗次数、手术治疗次数、常态呼吸频率、常态血氧饱和度、精神状态和运动状态。
  19. 根据权利要求17所述的数据处理方法，还包括：接收与所述健康管理装置关联的对象的总体评分，并在与所述健康管理装置关联的对象的总体评分大于评分阈值时输出预警信息。
  20. 根据权利要求1-19任一项所述的数据处理方法，其中，所述对源于与健康管理装置关联的设备的设备相关数据进行数据处理包括：对所述源于与健康管理装置关联的设备的设备相关数据进行分布式处理。
  21. 一种数据处理装置，包括：处理器和存储器，其中，所述存储器中存储有适于所述处理器执行的计算机程序指令，所述计算机程序指令被所述处理器运行时使得所述处理器执行以下的方法包括：对源于与健康管理装置关联的设备的设备相关数据进行数据处理。
  22. 一种健康管理装置，包括如权利要求21所述的数据处理装置。
  23. 一种非暂时性存储介质，所述非暂时性存储介质存储有计算机程序指令，所述计算机程序指令被处理器运行时使得计算机执行以下的方法包括：对源于与健康管理装置关联的设备的设备相关数据进行数据处理。
  24. 一种数据处理装置，包括数据处理模块，其中，所述数据处理模块被配置为对源于与健康管理装置关联的设备的设备相关数据进行数据处理。
PCT/CN2021/096383 2020-06-09 2021-05-27 数据处理方法、数据处理装置和健康管理装置 WO2021249197A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US17/772,783 US20220383997A1 (en) 2020-06-09 2021-05-27 Data Processing Method, Data Processing Apparatus, and Health Management Apparatus

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010519612.7A CN113782194B (zh) 2020-06-09 2020-06-09 数据处理方法、数据处理装置和健康管理装置
CN202010519612.7 2020-06-09

Publications (1)

Publication Number Publication Date
WO2021249197A1 true WO2021249197A1 (zh) 2021-12-16

Family

ID=78834671

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/096383 WO2021249197A1 (zh) 2020-06-09 2021-05-27 数据处理方法、数据处理装置和健康管理装置

Country Status (3)

Country Link
US (1) US20220383997A1 (zh)
CN (1) CN113782194B (zh)
WO (1) WO2021249197A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116820352A (zh) * 2023-08-23 2023-09-29 湖南奔普智能科技有限公司 一种具有数据容灾功能的病区自助结算***

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102647369A (zh) * 2011-02-16 2012-08-22 ***通信集团江苏有限公司 用于实现远程医疗的网关设备和远程医疗的实现方法
CN105590031A (zh) * 2016-02-19 2016-05-18 中国医学科学院生物医学工程研究所 一种应用于智能医疗设备的数据处理装置
CN105956409A (zh) * 2016-05-27 2016-09-21 上海道拓医药科技股份有限公司 智能医疗设备实时数据管理方法及***
CN108597605A (zh) * 2018-03-19 2018-09-28 特斯联（北京）科技有限公司 一种个人健康生活大数据采集与分析***
CN108874928A (zh) * 2018-05-31 2018-11-23 平安科技（深圳）有限公司 简历数据信息解析处理方法、装置、设备及存储介质
US20190034590A1 (en) * 2017-07-28 2019-01-31 Google Inc. System and Method for Predicting and Summarizing Medical Events from Electronic Health Records
CN110767279A (zh) * 2019-10-21 2020-02-07 山东师范大学 基于lstm的电子健康记录缺失数据补全方法及***

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105574042A (zh) * 2014-10-17 2016-05-11 ***通信集团公司 一种电子健康档案数据的处理方法及装置
CN104921718B (zh) * 2015-01-30 2018-09-07 董云鹏 健康监控与管理***
CN108257648B (zh) * 2018-02-08 2021-03-09 杭州医云康网络科技有限公司 一种基于大数据的医疗健康数据管理***

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102647369A (zh) * 2011-02-16 2012-08-22 ***通信集团江苏有限公司 用于实现远程医疗的网关设备和远程医疗的实现方法
CN105590031A (zh) * 2016-02-19 2016-05-18 中国医学科学院生物医学工程研究所 一种应用于智能医疗设备的数据处理装置
CN105956409A (zh) * 2016-05-27 2016-09-21 上海道拓医药科技股份有限公司 智能医疗设备实时数据管理方法及***
US20190034590A1 (en) * 2017-07-28 2019-01-31 Google Inc. System and Method for Predicting and Summarizing Medical Events from Electronic Health Records
CN108597605A (zh) * 2018-03-19 2018-09-28 特斯联（北京）科技有限公司 一种个人健康生活大数据采集与分析***
CN108874928A (zh) * 2018-05-31 2018-11-23 平安科技（深圳）有限公司 简历数据信息解析处理方法、装置、设备及存储介质
CN110767279A (zh) * 2019-10-21 2020-02-07 山东师范大学 基于lstm的电子健康记录缺失数据补全方法及***

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116820352A (zh) * 2023-08-23 2023-09-29 湖南奔普智能科技有限公司 一种具有数据容灾功能的病区自助结算***
CN116820352B (zh) * 2023-08-23 2023-11-10 湖南奔普智能科技有限公司 一种具有数据容灾功能的病区自助结算***

Also Published As

Publication number Publication date
CN113782194B (zh) 2023-05-05
CN113782194A (zh) 2021-12-10
US20220383997A1 (en) 2022-12-01

Similar Documents

Publication Publication Date Title
WO2021143779A1 (zh) 一种基于知识图谱的跨科室慢性肾病早期诊断决策支持***
US20190325995A1 (en) Method and system for predicting patient outcomes using multi-modal input with missing data modalities
CN107242858B (zh) 一种个性化孕婴监护方法及***
Hamid et al. Dempster–Shafer theory for classification and hybridised models of multi-criteria decision analysis for prioritisation: a telemedicine framework for patients with heart diseases
Najjar et al. A two-step approach for mining patient treatment pathways in administrative healthcare databases
US12014285B2 (en) Systems and methods for machine learning in patient placement
CN103690240A (zh) 一种医疗***
US20130231953A1 (en) Method, system and computer program product for aggregating population data
WO2021098534A1 (zh) 相似度确定、网络训练、查找方法及装置、电子装置和存储介质
US20190108264A1 (en) System prototype of intelligent regional mobile healthcare integrated data center based on metadata
KR20160125543A (ko) 사용자 중심의 헬스케어 빅데이터 서비스 방법, 그 방법을 수행하는 컴퓨터프로그램 및 시스템
CN108257648A (zh) 一种基于大数据的医疗健康数据管理***
Wang Construct an optimal triage prediction model: A case study of the emergency department of a teaching hospital in Taiwan
WO2021249197A1 (zh) 数据处理方法、数据处理装置和健康管理装置
CN105022783A (zh) 一种基于Hadoop的用户服务安全***和方法
CN113707323B (zh) 基于机器学习的疾病预测方法、装置、设备及介质
CN111933283B (zh) 一种健康监测方法及装置
US20160371457A1 (en) System and Method for Data Analyzing of Health-Related Data
US10431339B1 (en) Method and system for determining relevant patient information
US20220375618A1 (en) Method and apparatus of calculating comprehensive disease index
Chen et al. [Retracted] Storage Method for Medical and Health Big Data Based on Distributed Sensor Network
Asri IoT and Reality Mining for Real-Time Disease Prediction
Wang et al. Implementation of Real‐Time Medical and Health Data Mining System Based on Machine Learning
Fazio et al. A hybrid storage service for the management of big e-health data: a tele-rehabilitation case of study
CN114373541A (zh) 基于分布式的智能中医诊断方法、***及存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21822589

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21822589

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 07/06/2023)

122 Ep: pct application non-entry in european phase

Ref document number: 21822589

Country of ref document: EP

Kind code of ref document: A1