CN113223674B - Medical image marking method, device, equipment and storage medium - Google Patents
Medical image marking method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN113223674B CN113223674B CN202110597868.4A CN202110597868A CN113223674B CN 113223674 B CN113223674 B CN 113223674B CN 202110597868 A CN202110597868 A CN 202110597868A CN 113223674 B CN113223674 B CN 113223674B
- Authority
- CN
- China
- Prior art keywords
- medical image
- label
- target object
- medical
- current
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
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
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/40—ICT specially adapted for the handling or processing of medical images for processing medical images, e.g. editing
-
- 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
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
The application provides a medical image marking method, a device, equipment and a storage medium, and relates to the technical field of medical images. The method comprises the following steps: acquiring a current medical image of the target object examination and a current medical report corresponding to the current medical image; extracting description information of the medical report; and generating a label of the current medical image according to the description information of the medical report. In the scheme, the description information in the medical report corresponding to each medical image is extracted, and then the labels are further generated for each medical image according to the description information of the medical report, so that the accuracy and the efficiency of generating the medical image labels are effectively improved; and medical staff can also carry out efficient management, more accurate retrieval and reuse on the medical images according to the labels of the medical images.
Description
Technical Field
The present application relates to the field of medical image processing technologies, and in particular, to a medical image marking method, apparatus, device, and storage medium.
Background
In hospital clinic, 100G of data can be added to the image inspection image daily, 30T of data increment can be achieved each year, and at the same time, the image inspection image of a patient is required to be stored for at least 20 years, so that various medical academia have accumulated a huge number of medical images. Therefore, how to manage, retrieve and reuse medical images has been a challenge.
At present, labeling of medical image labels is realized mainly by relying on the professional knowledge of labeling operators, but the manual labeling mode has the problems of high labor cost, long time consumption, low efficiency and the like.
Disclosure of Invention
The application aims to overcome the defects in the prior art and provide a medical image marking method, a device, equipment and a storage medium, so as to improve the accuracy and efficiency of generating medical image labels.
In order to achieve the above purpose, the technical scheme adopted by the embodiment of the application is as follows:
in a first aspect, an embodiment of the present application provides a medical image marking method, including:
acquiring a current medical image of target object examination and a current medical report corresponding to the current medical image;
extracting descriptive information of the medical report;
and generating a label of the current medical image according to the description information of the medical report.
Optionally, after generating the label of the current medical image according to the description information of the medical report, the method further includes:
generating a label of the target object according to the current medical image checked by the target object and the label of the current medical image, wherein the label of the target object is used for indicating whether the target object is an important object or not.
Optionally, the generating the label of the target object according to the current medical image of the target object and the label of the current medical image includes:
extracting basic information of the target object according to the current medical image of the target object, wherein the basic information comprises: the identification of the target object, the age of the target object, the current inspection date and the inspection type;
acquiring a label of at least one historical medical image of the target object according to the identification of the target object;
generating a label of the target object according to the basic information of the target object, the label of the current medical image and the label of the at least one historical medical image.
Optionally, the generating the label of the target object according to the basic information of the target object, the label of the current medical image, and the label of the at least one historical medical image includes:
generating an embedding matrix according to the basic information of the target object, the label of the current medical image and the label of the at least one historical medical image;
and inputting the embedded matrix into a pre-constructed label classification model to generate a label of the target object.
Optionally, the extracting the description information of the medical report includes:
extracting at least one examination site of the medical report and a symptom description of each of the examination sites using a medical text model;
and obtaining the description information of the medical report according to the at least one examination part, the symptom description of each examination part and a pre-constructed relation extraction model.
Optionally, the generating a label of the current medical image according to the description information of the medical report includes:
determining whether descriptive information in the medical report belongs to rare disease descriptions using a pre-constructed disease description sample library for recording a plurality of descriptive information and whether each descriptive information belongs to rare disease descriptive information;
if the descriptive information in the medical report is a rare disease description, a rare disease label is generated for the current medical image.
Optionally, the generating the label of the current medical image according to the description information of the medical report further includes:
if the descriptive information in the medical report is not a rare disease description, a common disease label is generated for the current medical image.
In a second aspect, an embodiment of the present application further provides a medical image marking apparatus, the apparatus including:
the acquisition module is used for acquiring a current medical image of the target object examination and a current medical report corresponding to the current medical image;
an extraction module for extracting descriptive information of the medical report;
and the generation module is used for generating a label of the current medical image according to the description information of the medical report.
Optionally, the generating module is further configured to:
generating a label of the target object according to the current medical image checked by the target object and the label of the current medical image, wherein the label of the target object is used for indicating whether the target object is an important object or not.
Optionally, the extracting module is further configured to extract basic information of the target object according to a current medical image of the target object, where the basic information includes: the identification of the target object, the age of the target object, the current inspection date and the inspection type;
the acquisition module is further used for acquiring a label of at least one historical medical image of the target object according to the identification of the target object;
the generating module is further configured to generate a label of the target object according to the basic information of the target object, the label of the current medical image, and the label of the at least one historical medical image.
Optionally, the generating module is further configured to:
generating an embedding matrix according to the basic information of the target object, the label of the current medical image and the label of the at least one historical medical image;
and inputting the embedded matrix into a pre-constructed label classification model to generate a label of the target object.
Optionally, the extraction module is further configured to:
extracting at least one examination site of the medical report and a symptom description of each of the examination sites using a medical text model;
and obtaining the description information of the medical report according to the at least one examination part, the symptom description of each examination part and a pre-constructed relation extraction model.
Optionally, the generating module is further configured to:
determining whether descriptive information in the medical report belongs to rare disease descriptions using a pre-constructed disease description sample library for recording a plurality of descriptive information and whether each descriptive information belongs to rare disease descriptive information;
if the descriptive information in the medical report is a rare disease description, a rare disease label is generated for the current medical image.
Optionally, the generating module is further configured to:
if the descriptive information in the medical report is not a rare disease description, a common disease label is generated for the current medical image.
In a third aspect, an embodiment of the present application further provides an electronic device, including: a processor, a storage medium storing machine-readable instructions executable by the processor, the processor and the storage medium communicating over a bus when the electronic device is running, the processor executing the machine-readable instructions to perform the steps of the method as provided in the first aspect, and a bus.
In a fourth aspect, embodiments of the present application also provide a computer-readable storage medium having stored thereon a computer program which, when executed by a processor, performs the steps of the method as provided in the first aspect.
The beneficial effects of the application are as follows:
the embodiment of the application provides a medical image marking method, a device, equipment and a storage medium, wherein the method comprises the following steps: acquiring a current medical image of the target object examination and a current medical report corresponding to the current medical image; extracting description information of the medical report; and generating a label of the current medical image according to the description information of the medical report. In the scheme, the description information in the medical report corresponding to each medical image is extracted, and then the labels are further generated for each medical image according to the description information of the medical report, so that the accuracy and the efficiency of generating the medical image labels are effectively improved; and medical staff can also carry out efficient management, more accurate retrieval and reuse on the medical images according to the labels of the medical images.
The medical image label, the target patient label and the basic information of the target patient B can be correlated and written into the database, so that the medical staff can search the target patient more quickly and accurately by combining the multi-label information, the change trend of the diseases in the imaging examination of the target object can be obtained, the unimportant patient data can be compressed, the occupied related hardware resources can be reduced, and the image data such as the medical image can be efficiently managed.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a schematic structural diagram of an electronic device according to an embodiment of the present application;
FIG. 2 is a schematic flow chart of a medical image marking method according to an embodiment of the present application;
FIG. 3 is a flowchart of another medical image marking method according to an embodiment of the present application;
FIG. 4 is a flowchart of another medical image marking method according to an embodiment of the present application;
FIG. 5 is a flowchart of another medical image marking method according to an embodiment of the present application;
FIG. 6 is a flowchart of another medical image marking method according to an embodiment of the present application;
fig. 7 is an overall flowchart of a medical image marking method according to an embodiment of the present application;
fig. 8 is a schematic structural diagram of a medical image marking device according to an embodiment of the present application.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present application more apparent, the technical solutions of the embodiments of the present application will be clearly and completely described with reference to the accompanying drawings in the embodiments of the present application, and it should be understood that the drawings in the present application are for the purpose of illustration and description only and are not intended to limit the scope of the present application. In addition, it should be understood that the schematic drawings are not drawn to scale. A flowchart, as used in this disclosure, illustrates operations implemented according to some embodiments of the present application. It should be understood that the operations of the flow diagrams may be implemented out of order and that steps without logical context may be performed in reverse order or concurrently. Moreover, one or more other operations may be added to or removed from the flow diagrams by those skilled in the art under the direction of the present disclosure.
In addition, the described embodiments are only some, but not all, embodiments of the application. The components of the embodiments of the present application generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the application, as presented in the figures, is not intended to limit the scope of the application, as claimed, but is merely representative of selected embodiments of the application. All other embodiments, which can be made by a person skilled in the art without making any inventive effort, are intended to be within the scope of the present application.
It should be noted that the term "comprising" will be used in embodiments of the application to indicate the presence of the features stated hereafter, but not to exclude the addition of other features.
Fig. 1 is a schematic structural diagram of an electronic device according to an embodiment of the present application; the electronic equipment can be processing equipment such as a tablet, a mobile phone terminal, a computer or a server and the like, has a data processing function, and is used for realizing the medical image marking method. As shown in fig. 1, the electronic device includes: memory 101, processor 102.
Wherein the memory 101 and the processor 102 are electrically connected directly or indirectly to each other for data transmission or interaction. For example, the components may be electrically connected to each other via one or more communication buses or signal lines.
The memory 101 stores therein a software function module stored in the memory 101 in the form of software or firmware (firmware), and the processor 102 executes various functional applications and data processing by running the software program and the module stored in the memory 101, that is, implements the data analysis method in the embodiment of the present application.
The Memory 101 may be, but is not limited to, a random access Memory (Random Access Memory, RAM), a Read Only Memory (ROM), a programmable Read Only Memory (Programmable Read-Only Memory, PROM), an erasable Read Only Memory (Erasable Programmable Read-Only Memory, EPROM), etc. The memory 101 is used for storing a program, and the processor 102 executes the program after receiving an execution instruction.
The processor 102 may be an integrated circuit chip having signal processing capabilities. The processor 202 may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.
The principle of implementation and the corresponding beneficial effects of the medical image marking method provided by the application will be described below with respect to a number of specific embodiments.
FIG. 2 is a schematic flow chart of a medical image marking method according to an embodiment of the present application; optionally, the execution subject of the method is the electronic device shown in fig. 1 and described above, as shown in fig. 2, and the method includes:
s201, acquiring a current medical image of the target object examination and a current medical report corresponding to the current medical image.
Wherein the target object may be a subject, such as a patient suffering from a disease or a normal person.
The medical image is also called a medical image, and refers to a non-invasively acquiring a picture or image of an internal tissue of a human body or a part of the human body for medical treatment or medical research, and may be classified into an electronic Computed Tomography (CT) medical image, a magnetic resonance (Magnetic Resonance, MR) medical image, and the like, without being limited thereto.
The medical report refers to a report in a text form which is issued by a hospital after medical image examination is performed on a human body or a part of the human body.
For example, the current medical image acquired for examination of the target object may be a CT slice of the pancreatic site of the pancreatic patient, and a pancreatic medical report corresponding to the pancreatic CT slice.
S202, extracting description information of the medical report.
The description information is text description information of pathological information about a current medical image in a medical report, and the description information can be understood as imaging characteristics and disease information.
Illustratively, if the "pancreatic medical report" obtained above includes: the pancreas does not see significant site-directed lesions, natural language techniques can be used to extract descriptive information about the current medical image in the medical report: keywords such as "pancreas", "unseen", "occupied lesions", etc., and according to "unseen", "occupied lesions", it can be obtained that the diagnosis conclusion of "pancreas" is negative, and the diagnosis conclusion can also be called descriptive information.
S203, generating a label of the current medical image according to the description information of the medical report.
Wherein the label of the current medical image comprises: rare disease tags, common disease tags.
Based on the embodiment, according to the description information of the negative, common disease labels can be generated for the CT slices of the pancreas of the pancreatic patient, so that the accuracy and the efficiency of generating medical image labels are effectively improved.
In addition, medical staff can also carry out high-efficient management, more accurate retrieval and reuse to medical image according to the label that generates for each medical image.
In summary, an embodiment of the present application provides a medical image marking method, including: acquiring a current medical image of the target object examination and a current medical report corresponding to the current medical image; extracting description information of the medical report; and generating a label of the current medical image according to the description information of the medical report. In the scheme, the description information in the medical report corresponding to each medical image is extracted, and then the labels are further generated for each medical image according to the description information of the medical report, so that the accuracy and the efficiency of generating the medical image labels are effectively improved; and medical staff can also carry out efficient management, more accurate retrieval and reuse on the medical images according to the labels of the medical images.
In the medical image marking method provided by the application, the label of the target object can be generated.
Alternatively, in this embodiment, the tag of the target object may be generated according to the current medical image checked by the target object and the tag of the current medical image, where the tag of the target object is used to indicate whether the target object is an important object.
How to generate a label of a target object from a current medical image of the target object and a label of the current medical image will be specifically explained by the following embodiments.
FIG. 3 is a flowchart of another medical image marking method according to an embodiment of the present application; as shown in fig. 3, generating a label of a target object according to a current medical image of the target object and a label of the current medical image includes:
s301, extracting basic information of the target object according to the current medical image of the target object.
Wherein, the basic information includes: the identity of the target object, the age of the target object, the current date of the examination, the type of examination. May further include: sex information of the target object.
Examination type refers to the source of the target subject, e.g., the target subject may be a patient of a different category, such as hospitalization, physical examination, emergency, clinic, etc.
For example, the current medical image P1 is an MR medical image of a portion of the human body of the target object a, which is a file based on the digital imaging and communications in medicine (Digital Imaging and Communications in Medicine, DICOM for short) type.
Accordingly, the basic information of the target object a contained in the medical image P1 may be read using the image reading function according to the acquired current medical image P1 of the target object a, for example, the read basic information of the target object a includes, but is not limited to: the identification ID100 of the target object A, the age 50 of the target object, the current examination date 2021-2-14, the examination type are the clinic.
S302, according to the identification of the target object, acquiring a label of at least one historical medical image of the target object.
Typically, the same target subject is subjected to image inspection of the same region (or different regions) at different time periods to obtain a plurality of historical medical images.
Therefore, the tags of a plurality of history medical images of the target object a within two years 2020-2021 or the tag of one history medical image can be acquired based on the identification ID100 of the target object a extracted as described above.
S303, generating a label of the target object according to the basic information of the target object, the label of the current medical image and the label of at least one historical medical image.
The following specifically explains how the label of the target object is further generated based on the basic information of the target object, the label of the current medical image, and the label of at least one history medical image.
FIG. 4 is a flowchart of another medical image marking method according to an embodiment of the present application; as shown in fig. 4, optionally, step S303 is described above: generating a tag of the target object based on the basic information of the target object, the tag of the current medical image, and the tag of the at least one historical medical image, comprising:
s401, generating an embedding matrix according to basic information of a target object, a label of a current medical image and a label of at least one historical medical image.
Optionally, the "basic information of the target object", "tag of the current medical image", and "tag of the at least one historical medical image" are formed into a row of T columns of input vectors, and if there are N target objects at the same time, N rows of T columns of input vectors may be formed according to the basic information of each target object, the tag of the current medical image, and the tag of the at least one historical medical image, which may also be denoted as n×t.
Because different target objects have different inspections for a plurality of times, the number of the tags of the extracted historical medical images is also different, and the number of columns of the input vector formed by different target objects is further different. Therefore, the input vector of each target object is changed to a uniform length, for example, the lengths of the input vectors formed by different target objects can be set to 6 columns, the input vectors with the lengths being less than 6 columns are complemented, and the input vectors with the lengths being greater than 6 columns are cut off, so that the embedded processing of the input vector formed by the target object is realized, the column number of the input vector formed by the basic information of the target object, the label of the current medical image and the label of the at least one historical medical image is fixed, and the input vector with the fixed length can be also called an embedded matrix.
S402, inputting the embedded matrix into a pre-constructed label classification model to generate a label of the target object.
The label classification model is trained based on an embedded clustering algorithm and is used for generating labels of target objects.
In this embodiment, weights are set for each input vector in the generated embedding matrix and input to the tag classification model, based on the weights of the last layer of network of the tag classification model, the classes and distances of surrounding M adjacent target objects are calculated, the adjacent target object classes are used as the weights and parameters of the parameters and the last layer of network together, and two classifications are constructed, so as to generate the tag of the target object, such as the classification result of the non-important patient of the tag of the target object.
The description information of how to extract the medical report will be specifically explained by the following examples.
FIG. 5 is a flowchart of another medical image marking method according to an embodiment of the present application; as shown in fig. 5, step S202 described above: extracting descriptive information of the medical report, including:
s501, extracting at least one examination part of a medical report and symptom descriptions of each examination part by using a medical text model.
In this embodiment, the medical text model is a medical text model based on a Pre-trained model (Pre-training of Deep Bidirectional Transformers for Language Understanding, abbreviated as Bert) for chinese medical text processing. For example, at least one examination site and a symptom description of each examination site included in a medical report corresponding to a medical image are extracted by a Bert-based medical text model.
For example, the examination site and the symptom description of each examination site extracted to the target object B are: "about 30% of the middle stenosis of the right coronary artery (Right Coronary Artery, RCA) of the heart", i.e. the examination site is the heart, the imaging is shown as "stenosis", and the corresponding symptoms are described as "atherosclerosis".
S502, extracting a model according to at least one examination part, symptom descriptions of all examination parts and a pre-constructed relation to obtain description information of a medical report.
The relation extraction model is used for judging whether the symptom description of the examination part is negative or positive, and the negative or positive is also called descriptive information.
On the basis of the above embodiment, the model may be extracted according to the relationship, and it is determined that the result of the examination of the heart portion of the target object B is positive, that is, that the description information of the medical report corresponding to the current medical image examined by the target object B is positive.
FIG. 6 is a flowchart of another medical image marking method according to an embodiment of the present application; as shown in fig. 6, step S203 is as follows: generating a label of the current medical image according to the description information of the medical report, wherein the label comprises the following components:
s601, determining whether description information in a medical report belongs to rare disease descriptions or not by using a pre-constructed disease description sample library.
The disease description sample library is used for recording various description information and whether the description information belongs to the description information of rare diseases.
In this embodiment, in order to make a better judgment on the description information in the medical report, a sample library of description information about imaging performance, various description information and whether each disease description belongs to rare diseases needs to be constructed, and the sample library includes common examination sites, including but not limited to: the examination sequence of the medical images comprises common sequences such as plain scan, enhancement, three-dimensional, etc. for the skull, chest, abdomen, liver, head, cervical, thoracic and lumbar vertebrae, etc.
The disease description sample library comprises at least 100 ten thousand pieces of description information, and can cover epidemiological observation based on the description information in the medical report, count each description information, define which description information belongs to rare disease descriptions and which description information belongs to common disease descriptions.
In addition, if there is data of newly added "description information", but the data of newly added "description information" containing manifestation or conclusion is not in the sample library, the newly added "description information" is judged as rare disease description.
In the present embodiment, the description information in the medical report is compared with whether the "disease description sample library" belongs to the "rare disease description" or the "common disease description" using the disease description sample library.
For example, the "disease description sample library" records description information of diseases such as gastric cancer, hepatic cyst, gall bladder stones, hutch disease and the like, and gastric cancer, hepatic cyst, gall bladder stones are common disease descriptions, hutch disease is a rare disease description.
S602, if the description information in the medical report is the rare disease description, generating a rare disease label for the current medical image.
In one implementation, if the descriptive information in the medical report corresponding to the medical image P1 of the target patient a belongs to a rare disease description, a rare disease label is generated for the medical image P1.
S603, if the description information in the medical report is not the rare disease description, generating a common disease label for the current medical image.
In another implementation, if the description information in the medical report corresponding to the medical image P2 of the target patient B does not belong to the rare disease description, a common disease label is generated for the medical image P2.
In addition, the common disease label of the medical image P2, the label of the target patient B and the basic information of the target patient B can be correlated and written into the database, so that the combination of multi-label information is realized, the medical staff can search the target patient more quickly and accurately, the unimportant patient data is compressed, the occupied related hardware resources are reduced, and the image data such as the medical image and the like can be managed efficiently.
Fig. 7 is an overall flowchart of a medical image marking method according to an embodiment of the present application; optionally, as shown in fig. 7, the method includes:
s701, acquiring a current medical image of the target object examination and a current medical report corresponding to the current medical image.
S702, extracting at least one examination part of the medical report and symptom descriptions of each examination part by using the medical text model.
S703, extracting a model according to at least one examination part, symptom descriptions of each examination part and a pre-constructed relation to obtain description information of the medical report.
S704, determining whether description information in the medical report belongs to rare disease descriptions or not by using a pre-constructed disease description sample library. If yes, go to step S705; if not, go to step S706.
S705, generating rare disease labels for the current medical image.
S706, generating a common disease label for the current medical image.
And S707, extracting the basic information of the target object according to the current medical image of the target object.
S708, according to the identification of the target object, acquiring the label of at least one historical medical image of the target object.
S709, generating an embedding matrix according to the basic information of the target object, the label of the current medical image and the label of at least one historical medical image.
S710, inputting the embedded matrix into a pre-constructed label classification model to generate a label of the target object.
Optionally, the overall implementation steps and the beneficial effects of the medical image marking method provided in the embodiment of the present application have been described in detail in the foregoing specific embodiments, and are not described in detail herein.
The following is a description of the medical image marking device, the storage medium, etc. for executing the medical image marking device and the storage medium provided by the present application, and specific implementation processes and technical effects thereof are referred to above, and are not repeated below.
FIG. 8 is a schematic structural view of a medical image marking apparatus according to an embodiment of the present application; as shown in fig. 8, the apparatus includes: an acquisition module 801, an extraction module 802, and a generation module 803.
An obtaining module 801, configured to obtain a current medical image of a target object and a current medical report corresponding to the current medical image;
an extraction module 802 for extracting descriptive information of the medical report;
a generating module 803 is configured to generate a label of the current medical image according to the description information of the medical report.
Optionally, the generating module 803 is further configured to:
and generating a label of the target object according to the current medical image checked by the target object and the label of the current medical image, wherein the label of the target object is used for indicating whether the target object is an important object or not.
Optionally, the extracting module 802 is further configured to extract basic information of the target object according to the current medical image of the target object, where the basic information includes: identification of the target object, age of the target object, current inspection date, inspection type;
the obtaining module 801 is further configured to obtain a label of at least one historical medical image of the target object according to the identifier of the target object;
the generating module 803 is further configured to generate a label of the target object according to the basic information of the target object, the label of the current medical image, and the label of the at least one historical medical image.
Optionally, the generating module 803 is further configured to
Generating an embedding matrix according to the basic information of the target object, the label of the current medical image and the label of at least one historical medical image;
and inputting the embedded matrix into a pre-constructed label classification model to generate a label of the target object.
Optionally, the extracting module 802 is further configured to:
extracting at least one examination site of the medical report and a symptom description of each examination site using the medical text model;
and extracting a model according to at least one examination part, symptom descriptions of each examination part and the pre-established relation to obtain the description information of the medical report.
Optionally, the generating module 803 is further configured to:
determining whether descriptive information in the medical report belongs to rare disease descriptions using a pre-constructed disease description sample library, the disease description sample library being used to record a plurality of descriptive information and whether each descriptive information belongs to rare disease descriptive information; if the descriptive information in the medical report is a rare disease description, a rare disease label is generated for the current medical image.
Optionally, the generating module 803 is further configured to generate a common disease label for the current medical image if the description information in the medical report is not a rare disease description.
The foregoing apparatus is used for executing the method provided in the foregoing embodiment, and its implementation principle and technical effects are similar, and are not described herein again.
The above modules may be one or more integrated circuits configured to implement the above methods, for example: one or more application specific integrated circuits (Application Specific Integrated Circuit, abbreviated as ASIC), or one or more microprocessors (digital singnal processor, abbreviated as DSP), or one or more field programmable gate arrays (Field Programmable Gate Array, abbreviated as FPGA), or the like. For another example, when a module above is implemented in the form of a processing element scheduler code, the processing element may be a general-purpose processor, such as a central processing unit (Central Processing Unit, CPU) or other processor that may invoke the program code. For another example, the modules may be integrated together and implemented in the form of a system-on-a-chip (SOC).
Optionally, the present application also provides a program product, such as a computer readable storage medium, comprising a program for performing the above-described method embodiments when being executed by a processor.
In the several embodiments provided by the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in hardware plus software functional units.
The integrated units implemented in the form of software functional units described above may be stored in a computer readable storage medium. The software functional unit is stored in a storage medium, and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) or a processor (english: processor) to perform some of the steps of the methods according to the embodiments of the application. And the aforementioned storage medium includes: u disk, mobile hard disk, read-Only Memory (ROM), random access Memory (Random Access Memory, RAM), magnetic disk or optical disk, etc.
Claims (7)
1. A medical image marking method, comprising:
acquiring a current medical image of target object examination and a current medical report corresponding to the current medical image;
extracting descriptive information of the medical report;
generating a label of the current medical image according to the description information of the medical report;
wherein the extracting the description information of the medical report includes:
extracting at least one examination site of the medical report and a symptom description of each of the examination sites using a medical text model;
obtaining description information of the medical report according to the at least one examination part, symptom descriptions of the examination parts and a pre-constructed relation extraction model;
generating a label of the current medical image according to the description information of the medical report; comprising the following steps:
determining whether descriptive information in the medical report belongs to rare disease descriptions using a pre-constructed disease description sample library for recording a plurality of descriptive information and whether each descriptive information belongs to rare disease descriptive information;
if the descriptive information in the medical report is a rare disease description, a rare disease label is generated for the current medical image.
2. The method of claim 1, wherein generating the label of the current medical image from the description information of the medical report further comprises:
generating a label of the target object according to the current medical image checked by the target object and the label of the current medical image, wherein the label of the target object is used for indicating whether the target object is an important object or not;
wherein the generating the label of the target object according to the current medical image checked by the target object and the label of the current medical image includes:
extracting basic information of the target object according to the current medical image checked by the target object, wherein the basic information comprises: the identification of the target object, the age of the target object, the current inspection date and the inspection type;
acquiring a label of at least one historical medical image of the target object according to the identification of the target object;
generating a label of the target object according to the basic information of the target object, the label of the current medical image and the label of the at least one historical medical image.
3. The method of claim 2, wherein generating the label of the target object from the base information of the target object, the label of the current medical image, and the label of the at least one historical medical image comprises:
generating an embedding matrix according to the basic information of the target object, the label of the current medical image and the label of the at least one historical medical image;
and inputting the embedded matrix into a pre-constructed label classification model to generate a label of the target object.
4. The method of claim 1, wherein generating a label of the current medical image from the description information of the medical report further comprises:
if the descriptive information in the medical report is not a rare disease description, a common disease label is generated for the current medical image.
5. A medical image marking apparatus, the apparatus comprising:
the acquisition module is used for acquiring a current medical image of the target object examination and a current medical report corresponding to the current medical image;
an extraction module for extracting descriptive information of the medical report;
the generation module is used for generating a label of the current medical image according to the description information of the medical report;
wherein, the extraction module is further used for:
extracting at least one examination site of the medical report and a symptom description of each of the examination sites using a medical text model;
obtaining description information of the medical report according to the at least one examination part, symptom descriptions of the examination parts and a pre-constructed relation extraction model;
the generating module is further configured to:
determining whether descriptive information in the medical report belongs to rare disease descriptions using a pre-constructed disease description sample library for recording a plurality of descriptive information and whether each descriptive information belongs to rare disease descriptive information;
if the descriptive information in the medical report is a rare disease description, a rare disease label is generated for the current medical image.
6. An electronic device comprising a processor, a storage medium and a bus, the storage medium storing program instructions executable by the processor, the processor and the storage medium communicating over the bus when the electronic device is in operation, the processor executing the program instructions to perform the steps of the method of any of claims 1-4 when executed.
7. A computer-readable storage medium, characterized in that the storage medium has stored thereon a computer program which, when executed by a processor, performs the steps of the method according to any of claims 1-4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110597868.4A CN113223674B (en) | 2021-05-28 | 2021-05-28 | Medical image marking method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110597868.4A CN113223674B (en) | 2021-05-28 | 2021-05-28 | Medical image marking method, device, equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113223674A CN113223674A (en) | 2021-08-06 |
| CN113223674B true CN113223674B (en) | 2023-10-20 |
Family
ID=77099417
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110597868.4A Active CN113223674B (en) | 2021-05-28 | 2021-05-28 | Medical image marking method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113223674B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010109351A1 (en) * | 2009-03-26 | 2010-09-30 | Koninklijke Philips Electronics N.V. | A system that automatically retrieves report templates based on diagnostic information |
| CN109460756A (en) * | 2018-11-09 | 2019-03-12 | 天津新开心生活科技有限公司 | Medical image processing method, apparatus, electronic equipment and computer-readable medium |
| WO2019160557A1 (en) * | 2018-02-16 | 2019-08-22 | Google Llc | Automated extraction of structured labels from medical text using deep convolutional networks and use thereof to train a computer vision model |
| CN111640480A (en) * | 2020-05-21 | 2020-09-08 | 上海联影智能医疗科技有限公司 | Medical report generation method, computer device, and storage medium |
| CN112712879A (en) * | 2021-01-18 | 2021-04-27 | 腾讯科技（深圳）有限公司 | Information extraction method, device, equipment and storage medium for medical image report |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6417972B2 (en) * | 2015-01-28 | 2018-11-07 | カシオ計算機株式会社 | Medical image classification system, medical image classification device, medical image classification method and program |
-
2021
- 2021-05-28 CN CN202110597868.4A patent/CN113223674B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2010109351A1 (en) * | 2009-03-26 | 2010-09-30 | Koninklijke Philips Electronics N.V. | A system that automatically retrieves report templates based on diagnostic information |
| WO2019160557A1 (en) * | 2018-02-16 | 2019-08-22 | Google Llc | Automated extraction of structured labels from medical text using deep convolutional networks and use thereof to train a computer vision model |
| CN111727478A (en) * | 2018-02-16 | 2020-09-29 | 谷歌有限责任公司 | Automatic extraction of structured labels from medical text using deep convolutional networks and use thereof for training computer vision models |
| CN109460756A (en) * | 2018-11-09 | 2019-03-12 | 天津新开心生活科技有限公司 | Medical image processing method, apparatus, electronic equipment and computer-readable medium |
| CN111640480A (en) * | 2020-05-21 | 2020-09-08 | 上海联影智能医疗科技有限公司 | Medical report generation method, computer device, and storage medium |
| CN112712879A (en) * | 2021-01-18 | 2021-04-27 | 腾讯科技（深圳）有限公司 | Information extraction method, device, equipment and storage medium for medical image report |
Non-Patent Citations (3)
| Title |
|---|
| Clinical Integration of Picture Archiving and Communication Systems with Pathology and Hospital Information System in Oncology;Duncan, LD，等;AMERICAN SURGEON;第76卷(第09期);第982-986页 * |
| 基于内容的医学图像检索;王春燕，等;医疗卫生装备(第05期);第33-35页 * |
| 跨模态多标签生物医学图像分类建模识别;于玉海，等;中国图象图形学报;第23卷(第06期);第917-927页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113223674A (en) | 2021-08-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Azizi et al. | Robust and data-efficient generalization of self-supervised machine learning for diagnostic imaging | |
| JP6595193B2 (en) | Interpretation report creation device and interpretation report creation system | |
| CN102612696B (en) | Medical information system with report validator and report augmenter | |
| EP2888686B1 (en) | Automatic detection and retrieval of prior annotations relevant for an imaging study for efficient viewing and reporting | |
| US8380013B2 (en) | Case image search apparatus, method and computer-readable recording medium | |
| US11232853B2 (en) | System and method for creating, querying, and displaying a MIBA master file | |
| US10210310B2 (en) | Picture archiving system with text-image linking based on text recognition | |
| JP2019169049A (en) | Medical image specification device, method, and program | |
| US10878564B2 (en) | Systems and methods for processing 3D anatomical volumes based on localization of 2D slices thereof | |
| CN115830017B (en) | Tumor detection system, method, equipment and medium based on image-text multi-mode fusion | |
| US10803986B2 (en) | Automatic layout apparatus, automatic layout method, and automatic layout program | |
| CN111000571A (en) | Method for predicting risk of post-traumatic psychological disease, related device and storage device | |
| US20220366151A1 (en) | Document creation support apparatus, method, and program | |
| CN112530550A (en) | Image report generation method and device, computer equipment and storage medium | |
| US20130275124A1 (en) | Generation of pictorial reporting diagrams of lesions in anatomical structures | |
| CN113223674B (en) | Medical image marking method, device, equipment and storage medium | |
| US20230005580A1 (en) | Document creation support apparatus, method, and program | |
| US11978274B2 (en) | Document creation support apparatus, document creation support method, and document creation support program | |
| Kawa et al. | Radiological atlas for patient specific model generation | |
| CN114241261A (en) | Image processing-based skin disease identification method, device, equipment and storage medium | |
| Xu et al. | Development and characterization of a chest CT atlas | |
| JP2021175454A (en) | Medical image processing apparatus, method and program | |
| Ray et al. | An automatic method for complete brain matter segmentation from multislice CT scan | |
| US20240119750A1 (en) | Method of generating language feature extraction model, information processing apparatus, information processing method, and program | |
| US20220391599A1 (en) | Information saving apparatus, method, and program and analysis record generation apparatus, method, and program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |