Health data management platform based on cloud computing

Abstract

The invention relates to a health data management platform based on cloud computing. The method comprises the following steps: carrying out suspicious lesion multilevel analysis on a chest CT image of a patient by adopting an artificial intelligence technology based on deep learning, respectively extracting morphological features and textural features of a suspicious lesion object in the image, and then carrying out dual interaction joint perception on the morphological features and the textural features of the suspicious lesion object; and potential abnormal information of the focus is mined, so that the malignant risk level of the focus is intelligently diagnosed. Therefore, the accuracy and efficiency of medical diagnosis can be effectively improved, auxiliary decision support is provided for doctors, and the workload of medical staff is relieved.

Description

Technical Field

[0001] The present application relates to the technical field of health data management, and in particular, to a health data management platform based on cloud computing. Background Art

[0002] With the continuous advancement of medical technology and the increasing aging of the population, the management and analysis of health data has become an indispensable part of the modern medical system. Traditional medical and health data management methods often rely on the internal system of a single medical institution, which has problems such as data silos, limited access, and low processing efficiency. In addition, there are limitations on computing resources and storage space when processing large-scale data, which not only affects the rational allocation of medical resources, but also restricts cross-institutional and cross-disciplinary collaborative research, thereby hindering the improvement of medical service quality.

[0003] Among many health data, medical imaging data has extremely high clinical value due to its intuitiveness and rich diagnostic information. In particular, chest CT images, as an important means of early screening for respiratory diseases such as lung cancer, the accuracy of their analysis and diagnosis is directly related to the timeliness and effectiveness of patient treatment. However, traditional chest CT image analysis methods mainly rely on the doctor's personal experience and naked eye observation, which has certain subjectivity and limitations. For example, doctors may miss or misdiagnose due to fatigue or distraction, and there are differences in diagnostic standards and experience levels among different doctors, which may also lead to inconsistent diagnostic results; in addition, for complex or blurred images, the difficulty of diagnosis is further increased, and misjudgment is prone to occur.

[0004] Therefore, in order to achieve efficient and accurate health data analysis and management, a cloud computing-based health data management platform is expected Summary of the invention

[0005] This summary is provided to introduce concepts in a brief form that will be described in detail in the detailed description below. This summary is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0006] In a first aspect, the present application provides a health data management platform based on cloud computing, the platform comprising: Patient information management module, medical record module, diagnosis assistance module, treatment plan assistance module, condition tracking module and data analysis and reporting module deployed in the cloud; The patient information management module is responsible for inputting, updating and querying the basic information of patients; The medical record module is used to support medical staff in creating and editing medical records; The diagnosis assistance module is used to provide disease auxiliary diagnosis results based on artificial intelligence algorithms; The treatment plan auxiliary formulation module is used to automatically generate or assist the medical staff in formulating a personalized treatment plan based on the disease auxiliary diagnosis results; The condition tracking module is used to record the changes in the condition during treatment and support regular follow-up reminders; The data analysis and reporting module is used to perform statistical analysis on medical record data and generate various statistical reports.

[0007] Optionally, the diagnosis assistance module includes: a CT image acquisition unit, used to acquire a chest CT image of a diagnosed patient object; a suspected lesion object capture unit, used to extract the suspected lesion area of ​​the chest CT image to obtain a suspected lesion region of interest image; a suspected lesion multi-level feature extraction unit, used to respectively extract the morphological features and texture features of the suspected lesion region of interest image to obtain a suspected lesion morphological feature map and a suspected lesion texture feature map; a joint perception unit, used to jointly perceive the suspected lesion morphological feature map and the suspected lesion texture feature map to obtain a suspected lesion morphology-texture joint significant perception feature map; a disease-assisted diagnosis result generation unit, used to generate the disease-assisted diagnosis result based on the suspected lesion morphology-texture joint significant perception feature map, and the disease-assisted diagnosis result is used to indicate the malignancy risk level of the lesion.

[0008] Optionally, the suspicious lesion object capture unit includes: an image standardization processing subunit, used to perform standardization processing on the chest CT image to obtain a standardized chest CT image; and a region of interest extraction subunit, used to input the standardized chest CT image into a suspicious lesion object capturer based on the Yolo model to obtain the suspicious lesion region of interest image.

[0009] Optionally, the suspected lesion multi-level feature extraction unit includes: a morphological feature extraction subunit, used to input the image of the suspected lesion region of interest into a morphological feature extractor based on a first atrous convolutional neural network model to obtain the suspected lesion morphological feature map; a texture feature extraction subunit, used to input the image of the suspected lesion region of interest into a texture feature extractor based on a second atrous convolutional neural network model to obtain the suspected lesion texture feature map.

[0010] Optionally, the joint perception unit is used to: input the suspicious lesion morphology feature map and the suspicious lesion texture feature map into a joint feature saliency perception attention network to obtain the suspicious lesion morphology-texture joint saliency perception feature map.

[0011] Optionally, the joint perception unit includes: a first feature shape reshaping subunit, used to perform feature shape reshaping on the suspected lesion morphological feature map and the suspected lesion texture feature map to obtain a suspected lesion morphological feature shape reshaping matrix and a suspected lesion texture feature shape reshaping matrix; a bidirectional dependency calculation subunit, used to input the suspected lesion morphological feature shape reshaping matrix and the suspected lesion texture feature shape reshaping matrix into a feature channel-by-channel interactive perception module to obtain a morphology-texture dependency matrix and a texture-morphology dependency matrix; a random inactivation subunit, used to input the morphology-texture dependency matrix and the texture-morphology dependency matrix into a random inactivation module to obtain a pruned morphology-texture dependency matrix and a pruned texture-morphology dependency matrix; a dependency optimization subunit , used to optimize the suspected lesion texture feature shape reshaping matrix and the suspected lesion morphological feature shape reshaping matrix based on the pruned morphology-texture dependency matrix and the pruned texture-morphology dependency matrix to obtain a dependency optimized suspected lesion texture feature matrix and a dependency optimized suspected lesion morphological feature matrix; a second feature shape reshaping subunit, used to perform feature shape reshaping on the dependency optimized suspected lesion morphological feature matrix and the dependency optimized suspected lesion texture feature matrix to obtain an optimized suspected lesion morphological feature map and an optimized suspected lesion texture feature map; a multi-scale joint subunit, used to calculate the weighted sum of the optimized suspected lesion morphological feature map and the optimized suspected lesion texture feature map to obtain the suspected lesion morphology-texture joint significant perceptual feature map.

[0012] Optionally, the bidirectional dependency calculation subunit is used to: calculate the suspected lesion morphological feature shape reshaping matrix multiplied by the transposed matrix of the suspected lesion texture feature shape reshaping matrix to obtain a suspected lesion morphology-texture association representation matrix; divide the suspected lesion morphology-texture association representation matrix by the scale of the suspected lesion texture feature shape reshaping matrix and input it into a softmax function to obtain the morphology-texture dependency matrix; calculate the transposed matrix of the suspected lesion morphological feature shape reshaping matrix multiplied by the suspected lesion texture feature shape reshaping matrix to obtain a suspected lesion texture-morphology association representation matrix; divide the suspected lesion morphology-texture association representation matrix by the scale of the suspected lesion morphological feature shape reshaping matrix and input it into a softmax function to obtain the texture-morphology dependency matrix.

[0013] Optionally, the dependency optimization subunit is used to: perform matrix multiplication of the pruned morphology-texture dependency matrix with the suspected lesion texture feature shape reshaping matrix to obtain the dependency-optimized suspected lesion texture feature matrix; perform matrix multiplication of the pruned texture-morphology dependency matrix with the suspected lesion morphological feature shape reshaping matrix to obtain the dependency-optimized suspected lesion morphological feature matrix.

[0014] Optionally, the disease-assisted diagnosis result generating unit is used to: input the suspicious lesion morphology-texture joint significant perceptual feature map into a classifier-based disease-assisted diagnosis result generator to obtain the disease-assisted diagnosis result.

[0015] The above technical solution uses deep learning-based artificial intelligence technology to perform multi-level analysis of suspicious lesions on the patient's chest CT images, extract the morphological features and texture features of the suspicious lesion objects in the image, and then through dual interactive joint perception of its morphological features and texture features, dig out the potential abnormal information of the lesion, so as to intelligently diagnose the malignant risk level of the lesion. In this way, the accuracy and efficiency of medical diagnosis can be effectively improved, auxiliary decision support can be provided for doctors, and the workload of medical staff can be reduced.

[0016] Other features and advantages of the present application will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The above and other features, advantages and aspects of the embodiments of the present application will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale. In the drawings: Figure 1 The invention is a block diagram of a health data management platform based on cloud computing according to an exemplary embodiment.

[0018] Figure 2 It is a block diagram of the diagnosis assistance module of a health data management platform based on cloud computing according to an exemplary embodiment.

[0019] Figure 3 The figure is a flow chart of a health data management method based on cloud computing according to an exemplary embodiment.

[0020] Figure 4 It is a block diagram of an electronic device according to an exemplary embodiment.

[0021] Figure 5 The present invention is an application scenario diagram of a health data management platform based on cloud computing according to an exemplary embodiment. DETAILED DESCRIPTION

[0022] The embodiments of the present application will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present application are shown in the accompanying drawings, it should be understood that the present application can be implemented in various forms and should not be construed as being limited to the embodiments described herein. Instead, these embodiments are provided to provide a more thorough and complete understanding of the present application. It should be understood that the drawings and embodiments of the present application are only for exemplary purposes and are not intended to limit the scope of protection of the present application.

[0023] It should be understood that the various steps described in the method implementation of the present application can be performed in different orders and / or performed in parallel. In addition, the method implementation may include additional steps and / or omit the steps shown. The scope of the present application is not limited in this respect.

[0024] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0025] It should be noted that the concepts such as "first" and "second" mentioned in this application are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0026] It should be noted that the modifications of "one" and "plurality" mentioned in the present application are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0027] The names of the messages or information exchanged between multiple devices in the embodiments of the present application are only used for illustrative purposes and are not used to limit the scope of these messages or information.

[0028] The specific implementation of the present application is described in detail below with reference to the accompanying drawings.

[0029] Figure 1 FIG. 1 is a block diagram of a health data management platform based on cloud computing according to an exemplary embodiment. Figure 1 As shown, the health data management platform 100 includes: Patient information management module 1, medical record module 2, diagnosis assistance module 3, treatment plan assistance module 4, condition tracking module 5 and data analysis and reporting module 6 deployed in the cloud; The patient information management module 1 is responsible for inputting, updating and querying the basic information of patients; The medical record module 2 is used to support medical staff in creating and editing medical records; The diagnosis auxiliary module 3 is used to provide disease auxiliary diagnosis results based on artificial intelligence algorithms; The treatment plan auxiliary formulation module 4 is used to automatically generate or assist the medical staff in formulating a personalized treatment plan based on the disease auxiliary diagnosis results; The condition tracking module 5 is used to record the changes in the condition during treatment and support regular follow-up reminders; The data analysis and reporting module 6 is used to perform statistical analysis on the medical record data and generate various statistical reports.

[0030] In response to the above technical problems, the technical concept of this application is to use artificial intelligence technology based on deep learning to perform multi-level analysis of suspicious lesions on the patient's chest CT images, extract the morphological features and texture features of the suspicious lesion objects in the images, and then through the dual interactive joint perception of its morphological features and texture features, dig out the potential abnormal information of the lesion, so as to intelligently diagnose the malignant risk level of the lesion. In this way, the accuracy and efficiency of medical diagnosis can be effectively improved, auxiliary decision support can be provided for doctors, and the workload of medical staff can be reduced.

[0031] In one embodiment of the present application, the functional modules of the health data management platform can implement: 1. Data management, image storage: Use cloud storage services (such as AWSS3, Azure Blob Storage) to store CT image data. Data retrieval: Provide efficient image data retrieval function, support query by patient ID, examination date and other conditions.

[0032] 2. Image analysis, automatic detection: Automatically detect suspicious lesions in CT images and mark the location and size of the lesions. Feature extraction: Extract the morphological and texture features of the lesions as the basis for further analysis.

[0033] 3. Risk assessment, classification model: Use the trained classification model to assess the risk of malignancy of the lesion and output the risk level (such as low, medium, high). Decision support: Provide decision support and recommend further examinations or treatment plans.

[0034] 4. Report generation, automatic reporting: Automatically generate diagnostic reports, including image analysis results and risk assessment conclusions. Report review: Radiologists can review and modify automatically generated reports to ensure the accuracy of the final report.

[0035] Figure 2is a block diagram of the diagnosis auxiliary module of a health data management platform based on cloud computing according to an exemplary embodiment, such as Figure 2 As shown, the diagnosis auxiliary module 3 includes: a CT image acquisition unit 101, which is used to acquire a chest CT image of a diagnosed patient object; a suspicious lesion object capture unit 102, which is used to extract the suspicious lesion area of ​​the chest CT image to obtain a suspicious lesion region of interest image; a suspicious lesion multi-level feature extraction unit 103, which is used to extract the morphological features and texture features of the suspicious lesion region of interest image respectively to obtain a suspicious lesion morphological feature map and a suspicious lesion texture feature map; a joint perception unit 104, which is used to perform joint perception on the suspicious lesion morphological feature map and the suspicious lesion texture feature map to obtain a suspicious lesion morphology-texture joint significant perception feature map; a disease-assisted diagnosis result generation unit 105, which is used to generate the disease-assisted diagnosis result based on the suspicious lesion morphology-texture joint significant perception feature map, and the disease-assisted diagnosis result is used to indicate the malignancy risk level of the lesion.

[0036] Specifically, in the diagnosis auxiliary module, first, a chest CT image of the patient being diagnosed is obtained. Chest CT (Computed Tomography) images can clearly and accurately display the internal structure of the chest, including key tissues such as the lungs, heart, and blood vessels. For patients suspected of having lung diseases (such as lung cancer, lung nodules, etc.), chest CT images can show the shape, size, location, and relationship with surrounding tissues of the lesions in detail, which is an important basis for disease assessment and diagnosis.

[0037] In one embodiment of the present application, the suspicious lesion object capture unit includes: an image standardization processing subunit, used to perform standardization processing on the chest CT image to obtain a standardized chest CT image; and a region of interest extraction subunit, used to input the standardized chest CT image into a suspicious lesion object capturer based on the Yolo model to obtain the suspicious lesion region of interest image.

[0038] Considering that chest CT images taken by different CT devices may have differences in brightness, contrast, resolution, etc., which may affect the accuracy of subsequent image analysis and processing. Therefore, in the technical solution of the present application, the chest CT images are standardized to eliminate imaging differences and generate standardized chest CT images, so that image data from different sources are comparable, providing a reliable data basis for accurate analysis and diagnosis of lesions. In an embodiment of the present application, the standardized processing includes grayscale correction, contrast enhancement and noise filtering.

[0039] Next, since the chest CT image contains a large number of normal tissues and structures, directly analyzing the entire image may lead to a waste of computing resources and a reduction in diagnostic accuracy. Therefore, in order to improve the efficiency and accuracy of image analysis, this application further uses a suspicious lesion object capturer based on the Yolo model to detect suspicious lesions in the chest CT image after the standardized processing, and quickly and accurately locates the suspicious lesion area through steps such as grid division, bounding box prediction and confidence scoring, thereby extracting the image of the region of interest of the suspicious lesion.

[0040] In one embodiment of the present application, the suspected lesion multi-level feature extraction unit includes: a morphological feature extraction subunit, used to input the image of the suspected lesion region of interest into a morphological feature extractor based on a first atrous convolutional neural network model to obtain the suspected lesion morphological feature map; a texture feature extraction subunit, used to input the image of the suspected lesion region of interest into a texture feature extractor based on a second atrous convolutional neural network model to obtain the suspected lesion texture feature map.

[0041] Considering that in the process of disease diagnosis, the geometric characteristics (such as size, shape and boundary clarity, etc.) and texture characteristics (such as roughness, uniformity, etc.) of the lesion object are important bases for disease diagnosis. Therefore, in the technical solution of the present application, in order to more comprehensively analyze the lesion characteristics, the first hole convolutional neural network model and the second hole convolutional neural network model with different receptive fields are further used as morphological feature extractors and texture feature extractors to process the image of the region of interest of the suspicious lesion. Among them, the first hole convolutional neural network model has a large receptive field and can fully capture the overall morphological characteristics of the lesion object to obtain a morphological feature map of the suspicious lesion; while the second hole convolutional neural network model has a smaller receptive field, focusing on capturing the subtle structure and tissue characteristics of the lesion object, and generating a texture feature map of the suspicious lesion.

[0042] In one embodiment of the present application, the joint perception unit is used to: input the suspected lesion morphological feature map and the suspected lesion texture feature map into an attention network for joint feature saliency perception to obtain the suspected lesion morphology-texture joint saliency perception feature map.

[0043] Since the morphological features and texture features of suspicious lesion objects are complementary in the diagnosis process, in order to comprehensively utilize the complementary advantages of multi-source information to improve the accuracy of diagnosis, the present application further introduces a joint feature saliency-aware attention network to fuse the suspicious lesion morphological feature map and the suspicious lesion texture feature map, and adopts a two-way attention interaction mechanism to model the dependency between the two, thereby achieving the optimization fusion of multi-level features. Specifically, the network first reshapes the shape of the suspicious lesion morphological feature map and the suspicious lesion texture feature map into a feature matrix so that its dimension and shape are adapted to the input requirements of the attention mechanism. Then, by calculating the two-way dependency between the two sets of feature matrices, the interaction between the features is quantified to generate morphology-texture and texture-morphology dependency matrices. At the same time, in order to reduce the model's sensitivity to noise and prevent overfitting, the two sets of dependency matrices are further subjected to random inactivation processing, and the generalization ability of the model is enhanced by selectively ignoring or suppressing a part of the feature interaction. Furthermore, the two groups of dependency matrices after random dropout processing are used as the guiding basis for feature association optimization, and the two groups of feature matrices after shape reshaping are weighted optimized. Then, they are restored to the original feature map form through shape reshaping and weighted fusion to generate a morphological-texture joint significant perception feature map of suspicious lesions, thereby realizing comprehensive mining of potential abnormal information of lesions.

[0044] In one embodiment of the present application, the joint perception unit includes: a first feature shape reshaping subunit, used to perform feature shape reshaping on the suspected lesion morphological feature map and the suspected lesion texture feature map to obtain a suspected lesion morphological feature shape reshaping matrix and a suspected lesion texture feature shape reshaping matrix; a bidirectional dependency calculation subunit, used to input the suspected lesion morphological feature shape reshaping matrix and the suspected lesion texture feature shape reshaping matrix into a feature channel-by-channel interactive perception module to obtain a morphology-texture dependency matrix and a texture-morphology dependency matrix; a random inactivation subunit, used to input the morphology-texture dependency matrix and the texture-morphology dependency matrix into a random inactivation module to obtain a pruned morphology-texture dependency matrix and a pruned texture-morphology dependency matrix; a dependency optimization ... a first subunit, configured to optimize the suspected lesion texture feature shape reshaping matrix and the suspected lesion morphological feature shape reshaping matrix based on the pruned morphology-texture dependency matrix and the pruned texture-morphology dependency matrix to obtain a dependency-optimized suspected lesion texture feature matrix and a dependency-optimized suspected lesion morphological feature matrix; a second feature shape reshaping subunit, configured to perform feature shape reshaping on the dependency-optimized suspected lesion morphological feature matrix and the dependency-optimized suspected lesion texture feature matrix to obtain an optimized suspected lesion morphological feature map and an optimized suspected lesion texture feature map; and a multi-scale joint subunit, configured to calculate the weighted sum of the optimized suspected lesion morphological feature map and the optimized suspected lesion texture feature map to obtain the suspected lesion morphology-texture joint significant perceptual feature map.

[0045] Further, in one embodiment of the present application, the bidirectional dependency calculation subunit is used to: calculate the suspected lesion morphological feature shape reshaping matrix multiplied by the transposed matrix of the suspected lesion texture feature shape reshaping matrix to obtain a suspected lesion morphology-texture association representation matrix; divide the suspected lesion morphology-texture association representation matrix by the scale of the suspected lesion texture feature shape reshaping matrix and input it into a softmax function to obtain the morphology-texture dependency matrix; calculate the transposed matrix of the suspected lesion morphological feature shape reshaping matrix multiplied by the suspected lesion texture feature shape reshaping matrix to obtain a suspected lesion texture-morphology association representation matrix; divide the suspected lesion morphology-texture association representation matrix by the scale of the suspected lesion morphological feature shape reshaping matrix and input it into a softmax function to obtain the texture-morphology dependency matrix.

[0046] Furthermore, in one embodiment of the present application, the dependency optimization subunit is used to: matrix multiply the pruned morphology-texture dependency matrix with the suspected lesion texture feature shape reshaping matrix to obtain the dependency optimized suspected lesion texture feature matrix; matrix multiply the pruned texture-morphology dependency matrix with the suspected lesion morphological feature shape reshaping matrix to obtain the dependency optimized suspected lesion morphological feature matrix.

[0047] Specifically, the suspicious lesion morphology feature map and the suspicious lesion texture feature map are processed by the following joint perception formula to obtain the suspicious lesion morphology-texture joint significant perception feature map, wherein the joint perception formula is:

[0048] in, A morphological feature diagram representing the suspicious lesion, represents the texture feature map of the suspicious lesion, represents the feature shape reshaping, Represents the shape reshaping matrix of the suspicious lesion morphological features, Represents the shape reshaping matrix of the texture features of suspicious lesions, represents matrix multiplication, represents the transpose of a matrix, represents the scale of the suspicious lesion morphological feature shape reshaping matrix, that is, the width of the suspicious lesion morphological feature shape reshaping matrix multiplied by the height, and the scale of the suspicious lesion morphological feature shape reshaping matrix and the suspicious lesion texture feature shape reshaping matrix are the same, is the normalized exponential function, represents the morphology-texture dependency matrix, represents the texture-morphology dependency matrix, represents random dropout processing, represents the pruned morphology-texture dependency matrix, represents the pruned texture-morphology dependency matrix, Represents dependency relationship to optimize the morphological feature matrix of suspicious lesions, Represents dependency relationship to optimize suspicious lesion texture feature matrix, Indicates optimization of the morphological feature map of suspicious lesions. represents the optimization of the texture feature map of suspicious lesions. and Represent different weight coefficients, Represents the morphology-texture joint significant perceptual feature map of the suspicious lesion.

[0049] Furthermore, the suspected lesion morphology-texture joint significant perceptual feature map is input into a classifier-based disease-assisted diagnosis result generator to obtain a disease-assisted diagnosis result, and the disease-assisted diagnosis result is used to indicate the malignancy risk level of the lesion.

[0050] In one embodiment of the present application, the disease-assisted diagnosis result generating unit is used to: input the suspected lesion morphology-texture joint significant perceptual feature map into a classifier-based disease-assisted diagnosis result generator to obtain the disease-assisted diagnosis result.

[0051] In a preferred example, considering that the suspected lesion morphological feature map and the suspected lesion texture feature map respectively represent the suspected lesion morphological image features and the suspected lesion texture image features determined based on void convolution coding of the image of the region of interest of the suspected lesion, when performing attention-driven feature joint saliency perception, the difference in attention weights caused by the difference in the spatial domain expression dimensions of the image semantic features will introduce image semantic joint perception aggregation offset. Therefore, it is expected to further improve the semantic spatial domain aggregation regression comprehensibility of the suspected lesion morphological-texture joint saliency perception feature map, thereby improving the accuracy of the disease auxiliary diagnosis results obtained by inputting the classifier-based disease auxiliary diagnosis result generator.

[0052] Based on this, in the preferred example, inputting the suspected lesion morphology-texture joint significant perceptual feature map into a classifier-based disease-aided diagnosis result generator to obtain a disease-aided diagnosis result specifically includes: Expanding the suspected lesion morphology-texture joint significant perception feature map into a suspected lesion morphology-texture joint significant perception feature vector; Subtracting the zero norm of the suspected lesion morphology-texture joint significant perception feature vector from the length of the suspected lesion morphology-texture joint significant perception feature vector to obtain a suspected lesion morphology-texture joint significant perception isolated zero value; Calculating the logarithm of the square of the suspected lesion morphology-texture joint significant perceived isolated zero value and the sum of the suspected lesion morphology-texture joint significant perceived isolated zero value with base two to obtain a suspected lesion morphology-texture joint significant perceived superposition information value; Calculate a power function with each eigenvalue of the suspected lesion morphology-texture joint significant perception eigenvector as the base, the difference of the suspected lesion morphology-texture joint significant perception isolated zero value minus one as the exponent, and perform a point multiplication with the suspected lesion morphology-texture joint significant perception superposition information value to obtain a first suspected lesion morphology-texture joint significant perception intermediate vector; After performing a dot multiplication on the suspected lesion morphology-texture joint significant perception feature vector and the difference between the suspected lesion morphology-texture joint significant perception isolated zero value minus one, the dot multiplication is then performed on the reciprocal of the suspected lesion morphology-texture joint significant perception isolated zero value to obtain a second suspected lesion morphology-texture joint significant perception intermediate vector; Calculate an exponential function with a natural constant as the base and each eigenvalue of the second suspicious lesion morphology-texture joint significant perception intermediate vector as the exponent to obtain a third suspicious lesion morphology-texture joint significant perception intermediate vector; performing dot addition of the first suspicious lesion morphology-texture joint significant perception intermediate vector and the third suspicious lesion morphology-texture joint significant perception intermediate vector to obtain an optimized suspicious lesion morphology-texture joint significant perception feature vector; and The optimized suspicious lesion morphology-texture joint significant perception feature vector is input into the classifier-based disease auxiliary diagnosis result generator to obtain the disease auxiliary diagnosis result.

[0053] That is, the morphology-texture joint significant perceptual feature vector of the suspicious lesion is, for example, denoted as The optimization process is expressed as:

[0054] in, represents the morphological-textural joint significant perceptual feature vector of the suspicious lesion, represents the length of the suspected lesion morphology-texture joint significant perception feature vector, represents the zero norm of a vector, Indicates the suspected lesion morphology-texture joint significant perceived isolated zero value, represents the logarithm operation with base 2, represents the exponential operation with base e, represents a power function that calculates the eigenvalues ​​of each position in the eigenvector using the difference of the suspected lesion morphology-texture joint significant perception isolated zero value minus one as an exponent, Indicates point addition, represents dot product, Represents the optimized suspicious lesion morphology-texture joint significant perception feature vector.

[0055] In this way, the high-dimensional feature manifold of the suspected lesion morphology-texture joint significant perceptual feature map is represented by the eigenvalues ​​of the feature set as the vector field of the aggregation dimension, and the overlapping value of the vector field of the suspected lesion morphology-texture joint significant perceptual feature map at the isolated zero position is used as the order information to fix the eigenvalue morphological position of its feature set, and a bias for the reversibility of the feature regression distribution field of the suspected lesion morphology-texture joint significant perceptual feature map is added as a reward to achieve the mapping target tracking of the regression distribution of the suspected lesion morphology-texture joint significant perceptual feature map for the eigenvalue position, so that the feature set of the suspected lesion morphology-texture joint significant perceptual feature map perceives the mapping migration to the aggregate distribution, thereby improving the accuracy of the disease auxiliary diagnosis result obtained by inputting the suspicious lesion morphology-texture joint significant perceptual feature map into the classifier-based disease auxiliary diagnosis result generator by improving the semantic space domain aggregation regression comprehensibility of the suspected lesion morphology-texture joint significant perceptual feature map.

[0056] In summary, the above scheme uses deep learning-based artificial intelligence technology to perform multi-level analysis of suspicious lesions on the chest CT images of patients, extracts the morphological features and texture features of the suspicious lesions in the images, and then mines the potential abnormal information of the lesions through dual interactive joint perception of their morphological features and texture features, thereby intelligently diagnosing the malignant risk level of the lesions. In this way, the accuracy and efficiency of medical diagnosis can be effectively improved, auxiliary decision support can be provided for doctors, and the workload of medical staff can be reduced.

[0057] Figure 3 is a flow chart of a health data management method based on cloud computing according to an exemplary embodiment. Figure 3 As shown, the method includes: Step S201: Use the patient information management module to input, update and query the patient's basic information; Step S202: Use the medical record module to support medical staff in creating and editing medical records; Step S203: using the diagnosis assistance module to provide disease auxiliary diagnosis results based on the artificial intelligence algorithm; Step S204: using a treatment plan auxiliary formulation module to automatically generate or assist the medical staff in formulating a personalized treatment plan according to the disease auxiliary diagnosis result; Step S205: Use the condition tracking module to record the changes in the condition during treatment and support regular follow-up reminders; Step S206: Use the data analysis and reporting module to perform statistical analysis on the medical record data and generate various statistical reports.

[0058] Reference below Figure 4, which shows a schematic diagram of the structure of an electronic device 600 suitable for implementing the embodiment of the present application. The terminal device in the embodiment of the present application may include but is not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 4 The electronic device shown is merely an example and should not bring any limitation to the functions and scope of use of the embodiments of the present application.

[0059] like Figure 4 As shown, the electronic device 600 may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 are also stored. The processing device 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0060] Typically, the following devices may be connected to the I / O interface 605: an input device 606 including, for example, a touch screen, a touch pad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 607 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 608 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 609. The communication device 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. Although Figure 4 The electronic device 600 is shown with various devices, but it should be understood that it is not required to implement or possess all the devices shown. More or fewer devices may be implemented or possessed instead.

[0061] In particular, according to an embodiment of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present application includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication device 609, or installed from the storage device 608, or installed from the ROM 602. When the computer program is executed by the processing device 601, the above-mentioned functions defined in the method of the embodiment of the present application are executed.

[0062] It should be noted that the computer-readable medium mentioned above in the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0063] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), an internet (e.g., the Internet), and a peer-to-peer network (e.g., an ad hoc peer-to-peer network), as well as any currently known or future developed network.

[0064] The computer-readable medium may be included in the electronic device, or may exist independently without being incorporated into the electronic device.

[0065] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including, but not limited to, object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0066] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0067] The modules involved in the embodiments described in this application can be implemented by software or hardware. The name of the module does not limit the module itself in some cases. For example, the test parameter acquisition module can also be described as a "module for acquiring device test parameters corresponding to the target device".

[0068] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0069] In the context of the present application, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0070] Figure 5 FIG. 1 is an application scenario diagram of a health data management platform based on cloud computing according to an exemplary embodiment. Figure 5 As shown, in this application scenario, first, a chest CT image of a patient to be diagnosed is obtained (for example, Figure 5 Then, the acquired chest CT image is input to a server (e.g., Figure 5 In S) shown in , the server is capable of processing the chest CT image based on a health data management algorithm based on cloud computing to generate the disease-assisted diagnosis result, and the disease-assisted diagnosis result is used to indicate the malignancy risk level of the lesion.

[0071] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in this application (but not limited to) by each other to form a technical solution.

[0072] In addition, although each operation is described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or to be performed in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the application. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0073] Although the subject matter has been described in language specific to structural features and / or method logic actions, it should be understood that the subject matter defined is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely exemplary forms of implementation. With respect to the apparatus in the above-mentioned embodiments, the specific manner in which each module performs the operation has been described in detail in the embodiments related to the method, and will not be described in detail here.

Claims

1. A cloud computing-based health data management platform, characterized in that: include: Patient information management module, medical record module, diagnosis assistance module, treatment plan assistance module, condition tracking module and data analysis and reporting module deployed in the cloud; The patient information management module is responsible for inputting, updating and querying the basic information of patients; The medical record module is used to support medical staff in creating and editing medical records; The diagnosis assistance module is used to provide disease auxiliary diagnosis results based on artificial intelligence algorithms; The treatment plan auxiliary formulation module is used to automatically generate or assist the medical staff in formulating a personalized treatment plan based on the disease auxiliary diagnosis results; The condition tracking module is used to record the changes in the condition during treatment and support regular follow-up reminders; The data analysis and reporting module is used to perform statistical analysis on medical record data and generate various statistical reports.

2. The cloud computing-based health data management platform according to claim 1, characterized in that: The diagnostic auxiliary module comprises: A CT image acquisition unit, used to acquire a chest CT image of a patient being diagnosed; A suspicious lesion object capturing unit, used for extracting the suspicious lesion area of ​​the chest CT image to obtain a suspicious lesion region of interest image; A suspicious lesion multi-level feature extraction unit, used to extract the morphological features and texture features of the image of the region of interest of the suspicious lesion to obtain a suspicious lesion morphological feature map and a suspicious lesion texture feature map; A joint perception unit, used for performing joint perception on the suspicious lesion morphology feature map and the suspicious lesion texture feature map to obtain a suspicious lesion morphology-texture joint significant perception feature map; A disease-assisted diagnosis result generating unit is used to generate the disease-assisted diagnosis result based on the suspected lesion morphology-texture joint significant perception feature map, and the disease-assisted diagnosis result is used to indicate the malignancy risk level of the lesion.

3. The cloud computing-based health data management platform according to claim 2, characterized in that: The suspicious lesion object capturing unit comprises: An image standardization processing subunit, used for performing standardization processing on the chest CT image to obtain a standardized chest CT image; The region of interest extraction subunit is used to input the standardized chest CT image into a suspicious lesion object capturer based on the Yolo model to obtain the region of interest image of the suspicious lesion.

4. The cloud computing-based health data management platform according to claim 3, characterized in that: The suspicious lesion multi-level feature extraction unit comprises: A morphological feature extraction subunit, used for inputting the image of the region of interest of the suspicious lesion into a morphological feature extractor based on the first hole convolutional neural network model to obtain the morphological feature map of the suspicious lesion; The texture feature extraction subunit is used to input the image of the region of interest of the suspicious lesion into a texture feature extractor based on the second hole convolutional neural network model to obtain the texture feature map of the suspicious lesion.

5. The cloud computing-based health data management platform according to claim 4, characterized in that: The joint sensing unit is used for: The suspicious lesion morphology feature map and the suspicious lesion texture feature map are input into a joint feature saliency perception attention network to obtain the suspicious lesion morphology-texture joint saliency perception feature map.

6. The cloud computing-based health data management platform according to claim 5, characterized in that: The joint sensing unit comprises: A first characteristic shape reshaping subunit is used to perform characteristic shape reshaping on the suspicious lesion morphology characteristic map and the suspicious lesion texture characteristic map to obtain a suspicious lesion morphology characteristic shape reshaping matrix and a suspicious lesion texture characteristic shape reshaping matrix; A bidirectional dependency calculation subunit, for inputting the suspicious lesion morphological feature shape reshaping matrix and the suspicious lesion texture feature shape reshaping matrix into a feature channel-by-channel interactive perception module to obtain a morphology-texture dependency matrix and a texture-morphology dependency matrix; a random dropout subunit, configured to input the morphology-texture dependency matrix and the texture-morphology dependency matrix into a random dropout module to obtain a pruned morphology-texture dependency matrix and a pruned texture-morphology dependency matrix; A dependency optimization subunit is used to optimize the suspected lesion texture feature shape reshaping matrix and the suspected lesion morphology feature shape reshaping matrix based on the pruned morphology-texture dependency matrix and the pruned texture-morphology dependency matrix to obtain a dependency-optimized suspected lesion texture feature matrix and a dependency-optimized suspected lesion morphology feature matrix; A second characteristic shape reshaping subunit is used to perform characteristic shape reshaping on the dependency-optimized suspicious lesion morphological feature matrix and the dependency-optimized suspicious lesion texture feature matrix to obtain an optimized suspicious lesion morphological feature map and an optimized suspicious lesion texture feature map; The multi-scale joint subunit is used to calculate the weighted sum of the optimized suspicious lesion morphological feature map and the optimized suspicious lesion texture feature map to obtain the suspicious lesion morphological-texture joint significant perceptual feature map.

7. The cloud computing-based health data management platform according to claim 6, characterized in that: The bidirectional dependency calculation subunit is used to: Calculating the transposed matrix of the suspicious lesion morphological feature shape reshaping matrix multiplied by the suspicious lesion texture feature shape reshaping matrix to obtain a suspicious lesion morphology-texture association representation matrix; The suspicious lesion morphology-texture association representation matrix is ​​divided by the scale of the suspicious lesion texture feature shape reshaping matrix and then input into a softmax function to obtain the morphology-texture dependency matrix; Calculating the transposed matrix of the suspicious lesion morphological feature shape reshaping matrix and multiplying it by the suspicious lesion texture feature shape reshaping matrix to obtain a suspicious lesion texture-morphology association representation matrix; The suspicious lesion morphology-texture association representation matrix is ​​divided by the scale of the suspicious lesion morphology feature shape reshaping matrix and then input into the softmax function to obtain the texture-morphology dependency matrix.

8. The cloud computing-based health data management platform according to claim 7, characterized in that: The dependency optimization subunit is used to: Performing matrix multiplication of the pruned morphology-texture dependency matrix and the suspicious lesion texture feature shape reshaping matrix to obtain the dependency-optimized suspicious lesion texture feature matrix; The pruned texture-morphology dependency matrix is ​​matrix-multiplied with the suspicious lesion morphology feature shape reshaping matrix to obtain the dependency-optimized suspicious lesion morphology feature matrix.

9. The cloud computing-based health data management platform according to claim 8, characterized in that: The disease auxiliary diagnosis result generating unit is used for: The suspected lesion morphology-texture joint significant perceptual feature map is input into a classifier-based disease auxiliary diagnosis result generator to obtain the disease auxiliary diagnosis result.

Images