CN114332023A - Pneumothorax automatic diagnosis and crisis early warning method, device, equipment and storage medium - Google Patents
Pneumothorax automatic diagnosis and crisis early warning method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN114332023A CN114332023A CN202111648452.7A CN202111648452A CN114332023A CN 114332023 A CN114332023 A CN 114332023A CN 202111648452 A CN202111648452 A CN 202111648452A CN 114332023 A CN114332023 A CN 114332023A
- Authority
- CN
- China
- Prior art keywords
- lung
- pneumothorax
- image
- region
- volume
- 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.)
- Pending
Links
Images
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The embodiment of the application belongs to the technical field of medical intelligence and artificial intelligence, and relates to an automatic pneumothorax diagnosis and crisis early warning method, which comprises the following steps: acquiring a lung CT image, processing the lung CT image through gray histogram equalization and acquiring a gray value of the lung CT image, optimizing the gray value of the lung CT image based on OSTU, calculating to obtain the volume of a lung parenchyma region, identifying a pneumothorax region in the lung CT image based on a 3D U-net depth convolution network, segmenting the pneumothorax region, calculating the volume of the pneumothorax region, and determining the occupied proportion of pneumothorax according to the volume of the lung parenchyma region and the volume of the pneumothorax region. Accurate identification and segmentation of pneumothorax areas are achieved. The application also provides an automatic pneumothorax diagnosis and crisis early warning device, equipment and a storage medium. In addition, the application also relates to a block chain technology, and pneumothorax diagnosis results can be stored in the block chain. The application completes automatic diagnosis and crisis early warning of pneumothorax.
Description
Technical Field
The application relates to the technical field of medical intelligence and artificial intelligence, in particular to a pneumothorax automatic diagnosis and crisis early warning method, device, equipment and storage medium.
Background
In the aspect of pneumothorax diagnosis, the current algorithm still has the following problems: firstly, the number of accurately labeled pneumothorax lung images is insufficient, the resolution is low, most of image segmentation algorithms, especially segmentation algorithms based on deep learning, need to prepare a large number of accurately labeled training images, however, the labeling of medical images not only depends on the professional knowledge of doctors, but also needs to consider medical ethics and the privacy of patients, so that a small number of special medical image databases can be obtained, and the labeled medical images are affected by the difference of clinical practices of different doctors, and errors may exist in the labeled medical images; secondly, the precision of the segmentation algorithm of the pneumothorax region needs to be improved, the shape of pneumothorax focus is irregular, individual difference is large, small pneumothorax can be tiny and difficult to identify, and the current segmentation algorithm has a high promotion space; thirdly, most pneumothorax diagnosis systems cannot automatically diagnose the critical condition of a patient, a large area of pneumothorax may cause ventilation obstruction, so that the patient may have the phenomena of difficult breathing and the like, and if timely medical intervention is not performed, many dangerous conditions may be caused.
Disclosure of Invention
An object of the embodiments of the present application is to provide a method, an apparatus, a device, and a storage medium for pneumothorax auto-diagnosis and crisis early warning, so as to solve the technical problems of insufficient number of precisely labeled pneumothorax lung images, low resolution, and low accuracy in the prior art.
In order to solve the above technical problems, an embodiment of the present application provides an automatic pneumothorax diagnosis and crisis early warning method, which adopts the following technical solutions: the method comprises the following steps:
acquiring a lung CT image;
processing the lung CT image through gray histogram equalization and acquiring a gray value of the lung CT image;
optimizing the gray value of the lung CT image based on OSTU, and calculating to obtain the volume of a lung parenchymal region;
identifying a pneumothorax region in the lung CT image based on the 3D U-net depth convolution network, segmenting the pneumothorax region and calculating the volume of the pneumothorax region;
determining the occupation proportion of pneumothorax according to the volume of the lung parenchyma area and the volume of the pneumothorax area;
providing crisis early warning when a preset numerical value is reached;
and when the preset numerical value is not reached, providing a maintenance scheme.
Further, the step of processing the lung CT image through gray histogram equalization and acquiring the gray value of the lung CT image includes:
performing relational mapping according to the probability density function of the lung CT images and the pixel intensity of each lung CT image;
and acquiring the gray value of the lung CT image.
Further, the step of optimizing the gray-scale values of the lung CT images based on the OSTU and calculating the volume of the lung parenchymal region includes:
calculating the inter-class variance according to the gray value of the lung CT image;
when the inter-class variance is maximum, carrying out lung parenchymal region segmentation based on the optimal gray value corresponding to the maximum inter-class variance;
the volume of the segmented lung parenchymal region is calculated based on morphological filtering.
Further, the gray value of the lung CT image includes a foreground portion and a background portion, and the step of calculating the inter-class variance according to the gray value of the lung CT image includes:
acquiring the average value of the foreground part and the average value of the background part of the gray value of the lung CT image, the proportion of the pixel number in the foreground part to the total pixel number and the proportion of the pixel number in the background part to the total pixel number;
and calculating the result to be the inter-class variance according to the square of the difference between the average value of the foreground part and the average value of the background part multiplied by the proportion of the pixel number in the foreground part to the total pixel number and the proportion of the pixel number in the background part to the total pixel number.
Further, the step of identifying a pneumothorax region in the lung CT image based on a 3D U-net depth convolution network and segmenting the pneumothorax region and calculating a volume of the pneumothorax region comprises:
constructing a 3D U-net depth convolution network of the lung CT image;
mining key features of the lung CT image according to a 3D U-net deep convolution network, and establishing a network model according to the key features;
identifying a pneumothorax region in the network module and segmenting the pneumothorax region and calculating a pneumothorax region volume.
Further, the step of constructing a 3D U-net depth convolution network of the lung CT image comprises:
establishing a coding path, wherein each layer of the coding path comprises two convolution layers with convolution kernel size of 3 multiplied by 3, a ReLU activation layer, and a pooling layer with convolution kernel size of 2 multiplied by 2 and step length of 2;
establishing a decoding road stiffness, wherein each layer of the decoding road stiffness comprises a 2 multiplied by 2 deconvolution layer with the step length of 2, two convolution layers with the convolution kernel size of 3 multiplied by 3 are followed by a ReLU activation layer;
and constructing a 3D U-net deep convolutional network according to the coding path and the decoding path.
Further, the mining key features of the lung CT image according to the 3D U-net deep convolution network, and the step of establishing a network model from the key features specifically includes:
mining key features of the same resolution in the coding path;
passing the key features into the decoding path and outputting a binary label;
and establishing a network model according to the coding path, the decoding path and the binary label.
In order to solve the above technical problem, an embodiment of the present application further provides an pneumothorax automatic diagnosis and crisis early warning device, the device includes:
the acquisition module acquires a lung CT image;
the image processing module is used for processing the lung CT image through gray histogram equalization and acquiring a gray value of the lung CT image;
the lung parenchyma acquisition module is used for optimizing the gray value of the lung CT image based on OSTU and calculating the volume of a lung parenchyma region;
the pneumothorax region acquisition module is used for identifying a pneumothorax region in the lung CT image based on a 3D U-net deep convolution network, segmenting the pneumothorax region and calculating the volume of the pneumothorax region;
the judging module is used for determining the occupied proportion of pneumothorax according to the volume of the lung parenchyma area and the volume of the pneumothorax area, and providing crisis early warning when a preset value is reached; and when the preset numerical value is not reached, providing a maintenance scheme.
In order to solve the above technical problem, an embodiment of the present application further provides a computer device, which adopts the following technical solutions: the pneumothorax automatic diagnosis and crisis early warning method comprises a memory and a processor, wherein computer readable instructions are stored in the memory, and the steps of the pneumothorax automatic diagnosis and crisis early warning method are realized when the processor executes the computer readable instructions.
In order to solve the above technical problem, an embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solutions: the computer readable storage medium has stored thereon computer readable instructions which, when executed by a processor, implement the steps of the pneumothorax auto-diagnosis and crisis pre-warning method as described above.
Compared with the prior art, this application is through obtaining lung CT image, handles through grey level histogram equalization lung CT image and acquisition the grey scale value of lung CT image will the grey scale value of lung CT image is optimized based on OSTU to calculate the volume that reachs lung parenchyma region, based on 3D U-net degree of depth convolution network discernment pneumothorax region in the lung CT image and cut apart out pneumothorax region and calculate the volume of pneumothorax region, according to lung parenchyma region's volume and the proportion occupied of pneumothorax is confirmed to pneumothorax region's volume, when reaching preset value, provides the crisis early warning, when not reaching preset value, provides the maintenance scheme. Accurate identification and segmentation of pneumothorax areas are achieved, automatic diagnosis, positioning and quantitative analysis of pneumothorax lesions are facilitated, radiologists are prompted to perform priority treatment according to critical degrees of disease conditions, the time of pneumothorax patients waiting for reports in radiology departments is shortened, and medical safety of critical patients is guaranteed.
Drawings
In order to more clearly illustrate the solution of the present application, the drawings needed for describing the embodiments of the present application will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present application, and that other drawings can be obtained by those skilled in the art without inventive effort.
FIG. 1 is an exemplary system architecture diagram in which the present application may be applied;
FIG. 2 is a flow diagram of one embodiment of a pneumothorax auto-diagnosis and crisis warning method;
FIG. 3 is a construction diagram of a 3D U-net deep convolutional network;
FIG. 4 is a flow chart of an embodiment of an automatic pneumothorax diagnosis and crisis early warning method;
FIG. 5 is a schematic structural diagram of an embodiment of an automatic pneumothorax diagnosis and crisis early warning device;
FIG. 6 is a schematic block diagram of one embodiment of a computer device according to the present application.
Detailed Description
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used in the description of the application herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "including" and "having," and any variations thereof, in the description and claims of this application and the description of the above figures are intended to cover non-exclusive inclusions. The terms "first," "second," and the like in the description and claims of this application or in the above-described drawings are used for distinguishing between different objects and not for describing a particular order.
Reference herein to "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the application. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is explicitly and implicitly understood by one skilled in the art that the embodiments described herein can be combined with other embodiments.
In order to make the technical solutions better understood by those skilled in the art, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings.
As shown in fig. 1, the system architecture 100 may include terminal devices 101, 102, 103, a network 104, and a server 105. The network 104 serves as a medium for providing communication links between the terminal devices 101, 102, 103 and the server 105. Network 104 may include various connection types, such as wired, wireless communication links, or fiber optic cables, to name a few.
The user may use the terminal devices 101, 102, 103 to interact with the server 105 via the network 104 to receive or send messages or the like. The terminal devices 101, 102, 103 may have various communication client applications installed thereon, such as a web browser application, a shopping application, a search application, an instant messaging tool, a mailbox client, social platform software, and the like.
The terminal devices 101, 102, 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, e-book readers, MP3 players (Moving Picture Experts Group Audio Layer III, mpeg compression standard Audio Layer 3), MP4 players (Moving Picture Experts Group Audio Layer IV, mpeg compression standard Audio Layer 4), laptop portable computers, desktop computers, and the like.
The server 105 may be a server providing various services, such as a background server providing support for pages displayed on the terminal devices 101, 102, 103.
It should be noted that the information retrieval method based on the voice semantics provided by the embodiment of the present application is generally executed by a server, and accordingly, the information retrieval device based on the voice semantics is generally disposed in the server.
It should be understood that the number of terminal devices, networks, and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for implementation.
With continued reference to fig. 2-4, a flow chart of one embodiment of a pneumothorax auto-diagnosis and crisis warning method according to the present application is shown. The pneumothorax automatic diagnosis and crisis early warning method comprises the following steps:
step S1, acquiring a lung CT image;
on one hand, the lung CT image is obtained by real-time detection of a patient, and on the other hand, the lung CT image can be extracted from a pneumothorax special disease database, wherein the pneumothorax special disease database is established by acquiring a lung CT image of a clinical hospital and marking a pneumothorax area by a doctor to establish the pneumothorax special disease database. Specifically, an ITK-SNAP special labeling software is used for labeling the pneumothorax area for algorithm training and testing. Currently, 61 three-dimensional lung CT images are collected in the pneumothorax disease database, and each three-dimensional lung CT image group at least comprises 150 two-dimensional cross sections.
Step S2, processing the lung CT image through gray histogram equalization and obtaining the gray value of the lung CT image;
specifically, the gray histogram is obtained by counting the frequency of occurrence of all pixels in the lung CT image according to the size of the gray value. The gray histogram is a function of gray levels, which represents the number of pixels in the lung CT image having a certain gray level, reflecting the frequency of occurrence of a certain gray level in the lung CT image.
If the total pixel brightness (gray level) of the lung CT image is regarded as a random variable, the distribution reflects the statistical characteristics of the lung CT image, which can be characterized and described by a Probability Density Function (PDF) and is represented as a gray histogram.
Histogram equalization is an image processing method of adjusting contrast by using an image histogram; the intensity (intensity) of the lung CT image is subjected to certain nonlinear transformation, so that the transformed image histogram is approximately uniformly distributed, and the aims of improving the image contrast and enhancing the image are fulfilled. In this embodiment, histogram equalization employs a non-linear transformation of the form:
assuming that f is an original gray image and g is a histogram equalized gray image, the mapping relationship of each pixel of g and f is as follows:
wherein p isnThe probability density function of a gray-scale image histogram processed by the original lung CT image is equal to the proportion of the pixel number with the intensity of n in the image f to the total pixel number; f. ofi,jIndicating the pixel intensity of the ith row and the jth column in the image f; gi,jIndicating the pixel intensity in the ith row and the jth column in the image g. And performing relational mapping according to the probability density function of the lung CT images and the pixel intensity of each lung CT image to obtain the gray value of the lung CT images.
Step S3, optimizing the gray value of the lung CT image based on OSTU, and calculating to obtain the volume of the lung parenchymal region;
and according to the gray characteristic of the image, dividing the lung CT image into a background and a foreground. The larger the inter-class variance between the background and the foreground is, the larger the difference between the two parts constituting the image is, and the smaller the difference between the two parts is caused when part of the foreground is mistaken for the background or part of the background is mistaken for the foreground. Thus, a segmentation that maximizes the inter-class variance means that the probability of false positives is minimized.
Calculating the inter-class variance according to the gray value of the lung CT image, comprising the following steps: acquiring the average value of the foreground part and the average value of the background part of the gray value of the lung CT image, the proportion of the pixel number in the foreground part to the total pixel number and the proportion of the pixel number in the background part to the total pixel number; and calculating the result to be the inter-class variance according to the square of the difference between the average value of the foreground part and the average value of the background part multiplied by the proportion of the pixel number in the foreground part to the total pixel number and the proportion of the pixel number in the background part to the total pixel number.
When the inter-class variance is maximum, carrying out lung parenchymal region segmentation based on the optimal gray value corresponding to the maximum inter-class variance; the volume of the segmented lung parenchymal region is calculated based on morphological filtering.
Specifically, a gray value t is arbitrarily selected, a classification threshold, that is, a gray level, is set, iteration starts from 0, and according to a threshold segmentation method, the lung CT image can be divided into two parts, namely a foreground part larger than the gray value t and a background part lower than the gray value t. The average value of the foreground part and the background part is IfAnd IbgThe ratio of the number of pixels in the foreground portion to the total number of pixels is denoted as PfThe ratio of the number of pixels in the background portion to the total number of pixels is designated as PbgCalculating the mean value of the gray values of the lung CT image ImAnd the inter-class variance C may be defined as:
Im＝IfPf+IbgPbg
C＝Pf(If-Im)2+Pbg(Ibg-Im)2
will ImThe substitution into C gives:
C＝PfPbg(If-Ibg)2
the optimization goal of the OSTU is to find t to enable the value of C to be maximum, and the lung parenchyma segmentation based on the threshold is realized mainly by traversing the gray values and taking the gray value corresponding to the maximum C as the global threshold of the lung CT image. Followed byAnd removing scattered isolated points by using morphological filtering, namely performing primary image corrosion and image expansion. Calculating the volume of the finally segmented lung parenchymal region, and recording the volume as Vlung。
Step S4, identifying a pneumothorax region in the lung CT image based on the 3D U-net depth convolution network, segmenting the pneumothorax region and calculating the volume of the pneumothorax region;
specifically, a 3D U-net depth convolution network of the lung CT image is constructed firstly, and the construction method of the 3D U-net depth convolution network is as follows: establishing a coding path, wherein each layer of the coding path comprises two convolution layers with convolution kernel size of 3 multiplied by 3, a ReLU activation layer, and a pooling layer with convolution kernel size of 2 multiplied by 2 and step length of 2; establishing a decoding road stiffness, wherein each layer of the decoding road stiffness comprises a 2 multiplied by 2 deconvolution layer with the step length of 2, two convolution layers with the convolution kernel size of 3 multiplied by 3 are followed by a ReLU activation layer; and constructing a 3D U-net deep convolutional network according to the coding path and the decoding path.
Mining key features of the lung CT image according to a 3D U-net deep convolution network, and establishing a network model according to the key features, specifically mining key features with the same resolution in the coding path; passing the key features into the decoding path and outputting a binary label; and establishing a network model according to the coding path, the decoding path and the binary label.
And finally, identifying the pneumothorax area in the network module, segmenting the pneumothorax area and calculating the volume of the pneumothorax area.
It should be noted that 3D U-net depth convolution network is constructed, and the intelligent identification and segmentation of pneumothorax region are realized by mining the key features of lung CT image according to 3D U-net depth convolution network. The 3D U-net deep convolutional network includes an encoding path and a decoding path. In the coding path, each layer includes two convolution layers with convolution kernel size of 3 × 3 × 3, one ReLU active layer, and a pooling layer with convolution kernel size of 2 × 2 × 2 and step size of 2. In the decoding path, each layer contains a 2 × 2 × 2 deconvolution layer with step size 2, followed by two convolution layers with convolution kernel size 3 × 3 × 3, followed by a ReLU activation layer. Key features of the same resolution in the encoding path are passed into the decoding path through the hopping connection to augment the information flow of the 3D U-net deep convolutional network. The final output layer of the 3D U-net deep convolutional network outputs a binary label, and the network model is as shown in FIG. 2. The model adopts an Adam optimizer to carry out parameter optimization, the weight attenuation is 1e-6, the learning rate is 1e-4, the Adam optimizer combines the characteristics of two optimization algorithms of AdaGrad and RMSProp, and the formula is as follows:
where t is the number of iterations, α represents the step size, β1And beta2Is the decay rate, θ represents the parameter of the model, m represents the exponential moving mean, v represents the squared gradient, and e represents a very small parameter that prevents division by 0 in the calculation. The loss function of the function is cross-loss entropy, and the formula is as follows:
p is the true class of the pixel x,
to predict the probability that x belongs to class 1.
The pneumothorax area which can be divided by the network model is calculated to be the volume and is marked as Vleision。
Step S5, determining the proportion of pneumothorax according to the volume of the lung parenchyma area and the volume of the pneumothorax area;
lung parenchymal volume V obtained by S3 and S4lungAnd pneumothorax region volume VleisionCalculating the ratio of quantitative analysis of pneumothorax, Pleision＝Vleision/Vlung. Low, medium and high risk are assigned at less than 30%, 30% -70% and above 70%, giving the physician an indication.
Step S6, providing crisis early warning when a preset value is reached;
in this embodiment, if the preset value is 70%, that is, if it exceeds 70%, the patient is in a high risk state, and a surgical plan, a treatment plan and a subsequent rehabilitation plan are provided according to the disease condition of the patient.
And step S7, when the preset value is not reached, providing a maintenance scheme.
In this embodiment, if the preset value is 70%, that is, if the preset value is less than 70%, the patient is at a medium risk or a low risk, and a maintenance plan is presented according to the lesion condition of the patient, it is needless to say that the pneumothorax can be subdivided if the proportion is less than 70%, for example, the interval is less than 30% and 30% to 70%, the interval is 30% to 70% is medium risk, and the interval is less than 30% is low risk, and a physical maintenance plan and deterioration warning are provided for medium risk, and a physical good detection state is provided for low risk.
The invention verifies the result: the results obtained by the above method were used to randomly draw 12 sets of lung CT images as test set A and the rest as training set B, and the experiment result was 0.84 using DICE coefficient as a measure, and thus the method outputted results with 100% sensitivity to middle and high risk pneumothorax lesions.
The lung CT image is processed through gray level histogram equalization, the lung CT image is obtained, the gray level value of the lung CT image is optimized based on OSTU, the volume of a lung parenchyma region is obtained through calculation, the pneumothorax region in the lung CT image is divided into a pneumothorax region and the volume of the pneumothorax region is calculated based on 3D U-net depth convolution network identification, the proportion of the pneumothorax is determined according to the volume of the lung parenchyma region and the volume of the pneumothorax region, when a preset value is reached, crisis early warning is provided, and when the preset value is not reached, a maintenance scheme is provided. Accurate identification and segmentation of pneumothorax areas are achieved, automatic diagnosis, positioning and quantitative analysis of pneumothorax lesions are facilitated, radiologists are prompted to perform priority treatment according to critical degrees of disease conditions, the time of pneumothorax patients waiting for reports in radiology departments is shortened, and medical safety of critical patients is guaranteed.
It is emphasized that, to further ensure privacy and security of the pneumothorax diagnosis, the pneumothorax diagnosis may also be stored in a node of a block chain.
The block chain referred by the application is a novel application mode of computer technologies such as distributed data storage, point-to-point transmission, a consensus mechanism, an encryption algorithm and the like. A block chain (Blockchain), which is essentially a decentralized database, is a series of data blocks associated by using a cryptographic method, and each data block contains information of a batch of network transactions, so as to verify the validity (anti-counterfeiting) of the information and generate a next block. The blockchain may include a blockchain underlying platform, a platform product service layer, an application service layer, and the like.
The embodiment of the application can acquire and process related data based on an artificial intelligence technology. Among them, Artificial Intelligence (AI) is a theory, method, technique and application system that simulates, extends and expands human Intelligence using a digital computer or a machine controlled by a digital computer, senses the environment, acquires knowledge and uses the knowledge to obtain the best result.
The artificial intelligence infrastructure generally includes technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technologies, operation/interaction systems, mechatronics, and the like. The artificial intelligence software technology mainly comprises a computer vision technology, a robot technology, a biological recognition technology, a voice processing technology, a natural language processing technology, machine learning/deep learning and the like.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware associated with computer readable instructions, which can be stored in a computer readable storage medium, and when executed, the processes of the embodiments of the methods described above can be included. The storage medium may be a non-volatile storage medium such as a magnetic disk, an optical disk, a Read-Only Memory (ROM), or a Random Access Memory (RAM).
It should be understood that, although the steps in the flowcharts of the figures are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and may be performed in other orders unless explicitly stated herein. Moreover, at least a portion of the steps in the flow chart of the figure may include multiple sub-steps or multiple stages, which are not necessarily performed at the same time, but may be performed at different times, which are not necessarily performed in sequence, but may be performed alternately or alternately with other steps or at least a portion of the sub-steps or stages of other steps.
With further reference to fig. 5, as an implementation of the method shown in fig. 2-4, the present application provides an embodiment of an automatic pneumothorax diagnosis and crisis early warning apparatus, which corresponds to the embodiment of the method shown in fig. 2-4, and which can be applied to various electronic devices.
As shown in fig. 5, the pneumothorax auto-diagnosis and crisis pre-warning apparatus 400 of the present embodiment includes: an acquisition module 401, an image processing module 402, a lung parenchyma acquisition module 403, a pneumothorax region acquisition module 404, and a judgment module 405. Wherein:
an acquisition module 401 for acquiring a lung CT image;
an image processing module 402, configured to process the lung CT image through gray histogram equalization and obtain a gray value of the lung CT image;
a lung parenchyma acquisition module 403, which optimizes the gray value of the lung CT image based on the OSTU and calculates the volume of the lung parenchyma region;
a pneumothorax region acquisition module 404, configured to identify a pneumothorax region in the lung CT image based on a 3D U-net deep convolution network, segment the pneumothorax region, and calculate a volume of the pneumothorax region;
the judging module 405 determines the occupied proportion of pneumothorax according to the volume of the lung parenchyma region and the volume of the pneumothorax region, and provides crisis early warning when a preset value is reached; and when the preset numerical value is not reached, providing a maintenance scheme.
The lung CT image is processed through gray level histogram equalization, the lung CT image is obtained, the gray level value of the lung CT image is optimized based on OSTU, the volume of a lung parenchyma region is obtained through calculation, the pneumothorax region in the lung CT image is divided into a pneumothorax region and the volume of the pneumothorax region is calculated based on 3D U-net depth convolution network identification, the proportion of the pneumothorax is determined according to the volume of the lung parenchyma region and the volume of the pneumothorax region, when a preset value is reached, crisis early warning is provided, and when the preset value is not reached, a maintenance scheme is provided. Accurate identification and segmentation of pneumothorax areas are achieved, automatic diagnosis, positioning and quantitative analysis of pneumothorax lesions are facilitated, radiologists are prompted to perform priority treatment according to critical degrees of disease conditions, the time of pneumothorax patients waiting for reports in radiology departments is shortened, and medical safety of critical patients is guaranteed.
In some optional implementations of this embodiment, the image processing module 402 includes:
the mapping submodule is used for carrying out relational mapping according to the probability density function of the lung CT images and the pixel intensity of each lung CT image;
and the acquisition submodule acquires the gray value of the lung CT image.
In some optional implementations of the present embodiment, the lung parenchyma obtaining module 403 includes:
the between-class variance calculation submodule calculates the between-class variance according to the gray value of the lung CT image;
the segmentation submodule is used for carrying out lung parenchymal region segmentation based on the optimal gray value corresponding to the maximum inter-class variance when the inter-class variance is maximum;
the calculate lung parenchymal region submodule calculates a volume of the segmented lung parenchymal region based on the morphological filtering.
The calculate between class variance submodule includes:
the acquisition subunit is used for acquiring the average value of the foreground part and the average value of the background part of the gray value of the lung CT image, the proportion of the pixel number in the foreground part to the total pixel number and the proportion of the pixel number in the background part to the total pixel number;
and the inter-class variance calculating subunit is used for calculating the inter-class variance according to the square of the difference between the average value of the foreground part and the average value of the background part, the ratio of the number of pixels in the foreground part to the total number of pixels and the ratio of the number of pixels in the background part to the total number of pixels.
In some optional implementations of this embodiment, the pneumothorax region acquisition module 404 includes:
a construction sub-module that constructs a 3D U-net depth convolution network of the lung CT image;
establishing a model submodule, excavating key characteristics of the lung CT image according to the 3D U-net depth convolution network, and establishing a network model according to the key characteristics;
and the identification submodule identifies a pneumothorax area in the network module, divides the pneumothorax area and calculates the volume of the pneumothorax area.
The building submodule comprises:
the coding path establishing subunit establishes a coding path, wherein each layer of the coding path comprises two convolution layers with convolution kernel size of 3 multiplied by 3, a ReLU activation layer, and a pooling layer with convolution kernel size of 2 multiplied by 2 and step length of 2;
a decoding path establishing subunit, which establishes a decoding path, wherein each layer of the decoding path comprises a 2 × 2 × 2 deconvolution layer with a step length of 2, and is followed by two convolution layers with convolution kernel sizes of 3 × 3 × 3, and then a ReLU activation layer;
and the network construction subunit is used for constructing the 3D U-net deep convolutional network according to the coding path and the decoding path.
The model building submodule comprises:
the mining subunit is used for mining key features with the same resolution in the coding path;
a transferring subunit, for transferring the key feature to the decoding path and outputting a binary label;
and the model establishing subunit establishes a network model according to the coding path, the decoding path and the binary label.
In order to solve the technical problem, an embodiment of the present application further provides a computer device. Referring to fig. 6, fig. 6 is a block diagram of a basic structure of a computer device according to the present embodiment.
The computer device 4 comprises a memory 41, a processor 42, a network interface 43 communicatively connected to each other via a system bus. It is noted that only computer device 4 having components 41-43 is shown in FIG. 6, but it is understood that not all of the illustrated components are required to be implemented, and that more or fewer components may alternatively be implemented. As will be understood by those skilled in the art, the computer device is a device capable of automatically performing numerical calculation and/or information processing according to a preset or stored instruction, and the hardware includes, but is not limited to, a microprocessor, an Application Specific Integrated Circuit (ASIC), a Programmable Gate Array (FPGA), a Digital Signal Processor (DSP), an embedded device, and the like.
The computer device can be a desktop computer, a notebook, a palm computer, a cloud server and other computing devices. The computer equipment can carry out man-machine interaction with a user through a keyboard, a mouse, a remote controller, a touch panel or voice control equipment and the like.
The memory 41 includes at least one type of readable storage medium including a flash memory, a hard disk, a multimedia card, a card type memory (e.g., SD or DX memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a Read Only Memory (ROM), an Electrically Erasable Programmable Read Only Memory (EEPROM), a Programmable Read Only Memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the memory 41 may be an internal storage unit of the computer device 4, such as a hard disk or a memory of the computer device 6. In other embodiments, the memory 41 may also be an external storage device of the computer device 6, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), and the like, provided on the computer device 4. Of course, the memory 41 may also include both internal and external storage devices of the computer device 4. In this embodiment, the memory 41 is generally used for storing an operating system and various application software installed on the computer device 4, such as computer readable instructions of pneumothorax auto-diagnosis and crisis pre-warning method. Further, the memory 41 may also be used to temporarily store various types of data that have been output or are to be output.
The processor 42 may be a Central Processing Unit (CPU), controller, microcontroller, microprocessor, or other data Processing chip in some embodiments. The processor 42 is typically used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is configured to execute computer readable instructions stored in the memory 41 or process data, such as computer readable instructions for executing the pneumothorax auto-diagnosis and crisis-warning method.
The network interface 43 may comprise a wireless network interface or a wired network interface, and the network interface 43 is generally used for establishing communication connection between the computer device 4 and other electronic devices.
The lung CT image is processed through gray level histogram equalization, the lung CT image is obtained, the gray level value of the lung CT image is optimized based on OSTU, the volume of a lung parenchyma region is obtained through calculation, the pneumothorax region in the lung CT image is divided into a pneumothorax region and the volume of the pneumothorax region is calculated based on 3D U-net depth convolution network identification, the proportion of the pneumothorax is determined according to the volume of the lung parenchyma region and the volume of the pneumothorax region, when a preset value is reached, crisis early warning is provided, and when the preset value is not reached, a maintenance scheme is provided. Accurate identification and segmentation of pneumothorax areas are achieved, automatic diagnosis, positioning and quantitative analysis of pneumothorax lesions are facilitated, radiologists are prompted to perform priority treatment according to critical degrees of disease conditions, the time of pneumothorax patients waiting for reports in radiology departments is shortened, and medical safety of critical patients is guaranteed.
The present application further provides another embodiment, which is to provide a computer-readable storage medium storing computer-readable instructions executable by at least one processor to cause the at least one processor to perform the steps of the pneumothorax auto-diagnosis and crisis-warning method as described above.
The lung CT image is processed through gray level histogram equalization, the lung CT image is obtained, the gray level value of the lung CT image is optimized based on OSTU, the volume of a lung parenchyma region is obtained through calculation, the pneumothorax region in the lung CT image is divided into a pneumothorax region and the volume of the pneumothorax region is calculated based on 3D U-net depth convolution network identification, the proportion of the pneumothorax is determined according to the volume of the lung parenchyma region and the volume of the pneumothorax region, when a preset value is reached, crisis early warning is provided, and when the preset value is not reached, a maintenance scheme is provided. Accurate identification and segmentation of pneumothorax areas are achieved, automatic diagnosis, positioning and quantitative analysis of pneumothorax lesions are facilitated, radiologists are prompted to perform priority treatment according to critical degrees of disease conditions, the time of pneumothorax patients waiting for reports in radiology departments is shortened, and medical safety of critical patients is guaranteed.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the present application may be embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, magnetic disk, optical disk) and includes instructions for enabling a terminal device (such as a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present application.
It is to be understood that the above-described embodiments are merely illustrative of some, but not restrictive, of the broad invention, and that the appended drawings illustrate preferred embodiments of the invention and do not limit the scope of the invention. This application is capable of embodiments in many different forms and is provided for the purpose of enabling a thorough understanding of the disclosure of the application. Although the present application has been described in detail with reference to the foregoing embodiments, it will be apparent to one skilled in the art that the present application may be practiced without modification or with equivalents of some of the features described in the foregoing embodiments. All equivalent structures made by using the contents of the specification and the drawings of the present application are directly or indirectly applied to other related technical fields and are within the protection scope of the present application.
Claims (10)
1. An automatic pneumothorax diagnosis and crisis early warning method is characterized by comprising the following steps:
acquiring a lung CT image;
processing the lung CT image through gray histogram equalization and acquiring a gray value of the lung CT image;
optimizing the gray value of the lung CT image based on OSTU, and calculating to obtain the volume of a lung parenchymal region;
identifying a pneumothorax region in the lung CT image based on the 3D U-net depth convolution network, segmenting the pneumothorax region and calculating the volume of the pneumothorax region;
determining the occupation proportion of pneumothorax according to the volume of the lung parenchyma area and the volume of the pneumothorax area;
providing crisis early warning when a preset numerical value is reached;
and when the preset numerical value is not reached, providing a maintenance scheme.
2. The pneumothorax automatic diagnosis and crisis early warning method according to claim 1, wherein the step of processing the lung CT image by gray histogram equalization and obtaining gray values of the lung CT image comprises:
performing relational mapping according to the probability density function of the lung CT images and the pixel intensity of each lung CT image;
and acquiring the gray value of the lung CT image.
3. The pneumothorax auto-diagnosis and crisis pre-warning method as claimed in claim 1, wherein the step of optimizing the gray level values of the lung CT images based on the OSTU and calculating the volume of the lung parenchymal region comprises:
calculating the inter-class variance according to the gray value of the lung CT image;
when the inter-class variance is maximum, carrying out lung parenchymal region segmentation based on the optimal gray value corresponding to the maximum inter-class variance;
the volume of the segmented lung parenchymal region is calculated based on morphological filtering.
4. The pneumothorax automatic diagnosis and crisis early warning method as claimed in claim 3, wherein the gray value of the lung CT image includes a foreground part and a background part, and the step of calculating the between-class variance according to the gray value of the lung CT image comprises:
acquiring the average value of the foreground part and the average value of the background part of the gray value of the lung CT image, the proportion of the pixel number in the foreground part to the total pixel number and the proportion of the pixel number in the background part to the total pixel number;
and calculating the result to be the inter-class variance according to the square of the difference between the average value of the foreground part and the average value of the background part multiplied by the proportion of the pixel number in the foreground part to the total pixel number and the proportion of the pixel number in the background part to the total pixel number.
5. The pneumothorax automatic diagnosis and crisis early warning method as claimed in claim 1, wherein the step of identifying the pneumothorax region in the lung CT image based on the 3D U-net deep convolution network and segmenting the pneumothorax region and calculating the volume of the pneumothorax region comprises:
constructing a 3D U-net depth convolution network of the lung CT image;
mining key features of the lung CT image according to a 3D U-net deep convolution network, and establishing a network model according to the key features;
identifying a pneumothorax region in the network module and segmenting the pneumothorax region and calculating a pneumothorax region volume.
6. The pneumothorax automatic diagnosis and crisis early warning method as claimed in claim 5, wherein the step of constructing the 3D U-net deep convolution network of the lung CT images comprises:
establishing a coding path, wherein each layer of the coding path comprises two convolution layers with convolution kernel size of 3 multiplied by 3, a ReLU activation layer, and a pooling layer with convolution kernel size of 2 multiplied by 2 and step length of 2;
establishing a decoding road stiffness, wherein each layer of the decoding road stiffness comprises a 2 multiplied by 2 deconvolution layer with the step length of 2, two convolution layers with the convolution kernel size of 3 multiplied by 3 are followed by a ReLU activation layer;
and constructing a 3D U-net deep convolutional network according to the coding path and the decoding path.
7. The pneumothorax automatic diagnosis and crisis early warning method as claimed in claim 6, wherein the step of mining key features of the lung CT image according to 3D U-net deep convolution network and building a network model from the key features specifically comprises:
mining key features of the same resolution in the coding path;
passing the key features into the decoding path and outputting a binary label;
and establishing a network model according to the coding path, the decoding path and the binary label.
8. An automatic pneumothorax diagnosis and crisis early warning device, characterized in that, the device includes:
the acquisition module acquires a lung CT image;
the image processing module is used for processing the lung CT image through gray histogram equalization and acquiring a gray value of the lung CT image;
the lung parenchyma acquisition module is used for optimizing the gray value of the lung CT image based on OSTU and calculating the volume of a lung parenchyma region;
the pneumothorax region acquisition module is used for identifying a pneumothorax region in the lung CT image based on a 3D U-net deep convolution network, segmenting the pneumothorax region and calculating the volume of the pneumothorax region;
the judging module is used for determining the occupied proportion of pneumothorax according to the volume of the lung parenchyma area and the volume of the pneumothorax area, and providing crisis early warning when a preset value is reached; and when the preset numerical value is not reached, providing a maintenance scheme.
9. A computer device comprising a memory having computer readable instructions stored therein and a processor that when executed performs the steps of the pneumothorax auto-diagnosis and crisis pre-warning method of any one of claims 1 to 7.
10. A computer readable storage medium having computer readable instructions stored thereon, which when executed by a processor, implement the steps of the pneumothorax auto-diagnosis and crisis pre-warning method as claimed in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111648452.7A CN114332023A (en) | 2021-12-30 | 2021-12-30 | Pneumothorax automatic diagnosis and crisis early warning method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111648452.7A CN114332023A (en) | 2021-12-30 | 2021-12-30 | Pneumothorax automatic diagnosis and crisis early warning method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114332023A true CN114332023A (en) | 2022-04-12 |
Family
ID=81018415
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111648452.7A Pending CN114332023A (en) | 2021-12-30 | 2021-12-30 | Pneumothorax automatic diagnosis and crisis early warning method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114332023A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102240212A (en) * | 2010-05-14 | 2011-11-16 | Ge医疗***环球技术有限公司 | Method and apparatus for measuring pneumothorax |
| CN109727251A (en) * | 2018-12-29 | 2019-05-07 | 上海联影智能医疗科技有限公司 | The system that lung conditions are divided a kind of quantitatively, method and apparatus |
| CN111047591A (en) * | 2020-03-13 | 2020-04-21 | 北京深睿博联科技有限责任公司 | Focal volume measuring method, system, terminal and storage medium based on deep learning |
| WO2021248187A1 (en) * | 2020-06-09 | 2021-12-16 | Annalise-Ai Pty Ltd | Systems and methods for automated analysis of medical images |
-
2021
- 2021-12-30 CN CN202111648452.7A patent/CN114332023A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102240212A (en) * | 2010-05-14 | 2011-11-16 | Ge医疗***环球技术有限公司 | Method and apparatus for measuring pneumothorax |
| CN109727251A (en) * | 2018-12-29 | 2019-05-07 | 上海联影智能医疗科技有限公司 | The system that lung conditions are divided a kind of quantitatively, method and apparatus |
| CN111047591A (en) * | 2020-03-13 | 2020-04-21 | 北京深睿博联科技有限责任公司 | Focal volume measuring method, system, terminal and storage medium based on deep learning |
| WO2021248187A1 (en) * | 2020-06-09 | 2021-12-16 | Annalise-Ai Pty Ltd | Systems and methods for automated analysis of medical images |
Non-Patent Citations (2)
| Title |
|---|
| 柳杨: "《数字图像物体识别理论详解与实战》", 北京邮电大学出版社, pages: 17 * |
| 王继元: "基于深度学习的正位DR胸片质控体系研究与应用", 《万方学术期刊数据库》, pages 1 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110059697B (en) | Automatic lung nodule segmentation method based on deep learning | |
| CN111369576B (en) | Training method of image segmentation model, image segmentation method, device and equipment | |
| CN111932547B (en) | Method and device for segmenting target object in image, electronic device and storage medium | |
| WO2021135449A1 (en) | Deep reinforcement learning-based data classification method, apparatus, device, and medium | |
| CN114462412B (en) | Entity identification method, entity identification device, electronic equipment and storage medium | |
| Xiao et al. | An automated segmentation method for lung parenchyma image sequences based on fractal geometry and convex hull algorithm | |
| CN113168912A (en) | Determining growth rate of objects in 3D data sets using deep learning | |
| CN113240661A (en) | Deep learning-based lumbar vertebra analysis method, device, equipment and storage medium | |
| US20240161035A1 (en) | Multi-model medical scan analysis system and methods for use therewith | |
| CN115564756A (en) | Medical image focus positioning display method and system | |
| CN109871866B (en) | Model training method, device, equipment and medium for hospital infection prediction | |
| CN114822741A (en) | Processing device, computer equipment and storage medium of patient classification model | |
| CN114706985A (en) | Text classification method and device, electronic equipment and storage medium | |
| CN113822846A (en) | Method, apparatus, device and medium for determining region of interest in medical image | |
| CN114241411B (en) | Counting model processing method and device based on target detection and computer equipment | |
| CN113850260B (en) | Key information extraction method and device, electronic equipment and readable storage medium | |
| CN116521915A (en) | Retrieval method, system, equipment and medium for similar medical images | |
| CN114332023A (en) | Pneumothorax automatic diagnosis and crisis early warning method, device, equipment and storage medium | |
| CN113821641B (en) | Method, device, equipment and storage medium for classifying medicines based on weight distribution | |
| CN115631370A (en) | Identification method and device of MRI (magnetic resonance imaging) sequence category based on convolutional neural network | |
| CN115762721A (en) | Medical image quality control method and system based on computer vision technology | |
| CN114664458A (en) | Patient classification device, computer device and storage medium | |
| CN113780855A (en) | Medical institution supervision method and device, computer equipment and storage medium | |
| CN112086154A (en) | Intelligent pediatric information filing method, device, equipment and storage medium | |
| CN111696637A (en) | Quality detection method and related device for medical record data |
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 |