Tumor patient test laboratory sheet interpretation system and method

Abstract

The invention discloses a tumor patient test laboratory sheet interpretation system and method, and relates to the field of medical informatization, and the system comprises an information verification module which is used for the identity verification and classification of a patient and a doctor; the laboratory test report analysis module is used for carrying out structured analysis on the test laboratory test report through natural language processing and a computer vision technology, and verifying the qualification of a detection unit based on a fuzzy matching algorithm; the data storage module is used for integrating knowledge base data related to tumor diagnosis and treatment; and the index analysis module is used for performing tumor type identification, index abnormal item analysis and triage matching through a cascade model of a support vector machine, an isolated forest and a decision tree. According to the technical scheme, automatic interpretation of the test laboratory sheet of the tumor patient and matching of professional doctors can be achieved, and the method has high application value.

Description

Technical Field

[0001] The present invention relates to the field of medical informationization, and in particular to a system and method for interpreting test reports of tumor patients. Background Art

[0002] In the field of tumor diagnosis and treatment, accurate interpretation of test reports and efficient doctor-patient matching are key links in improving the quality of clinical decision-making. Traditional test report analysis mainly relies on manual interpretation, which has problems such as low efficiency, strong subjectivity and high misdiagnosis rate. In recent years, artificial intelligence technology has been gradually applied to medical data processing. Some studies have attempted to achieve automated analysis of test report indicators through a single algorithm, but its generalization ability is limited and it is difficult to cope with the complex multi-indicator correlation and dynamic change characteristics of tumor patients. In addition, existing methods mostly focus on data analysis itself, lack deep integration with clinical resources, and cannot form a closed-loop service.

[0003] Traditional methods usually process text and image data independently, and fail to fully explore the semantic relationship between the two. For example, the numerical indicators and imaging features in the test report are not analyzed collaboratively, resulting in a one-sided interpretation of the results; the system based on the black box model lacks a transparent display of the decision logic, and it is difficult to meet the requirements of clinicians for the credibility of the results; at the same time, the static model is difficult to adapt to the differences in data distribution in different hospitals and requires frequent retraining. Existing matching methods mostly rely on keyword searches or simple rules, and do not consider the dynamic update of doctors' professional labels, the needs of multidisciplinary collaboration, and the personalized disease characteristics of patients, resulting in low matching efficiency and accuracy. Therefore, based on the above-mentioned difficulties, the present invention proposes a system and method for interpreting test reports for tumor patients. Summary of the invention

[0004] Technical Purpose In order to solve the above problems, the purpose of the present invention is to provide a system and method for interpreting test reports of tumor patients, aiming to solve the problems of low efficiency, high misdiagnosis rate and inaccurate matching of medical resources in manual interpretation of test reports of tumor patients. By constructing a system with multimodal data fusion and intelligent algorithm collaboration, automatic and accurate analysis of test indicators, multi-dimensional disease assessment and dynamic matching of cross-disciplinary doctors can be achieved, thereby improving the accuracy, efficiency and utilization of medical resources in diagnosis and treatment decisions, while strengthening data privacy protection and ethical compliance.

[0005] Technical Solution In order to achieve the above-mentioned objectives, the present invention provides a system and method for interpreting test reports of tumor patients. The system and method combine natural language processing and computer vision technology to realize the joint analysis and structured extraction of text indicators and medical images, and use support vector machine, isolation forest and decision tree algorithms to complete tumor type classification, anomaly detection and triage matching respectively, and introduce transfer learning and dynamic feature selection mechanisms to enhance the generalization of the model; in addition, homomorphic encryption is used to protect patient privacy, and knowledge graphs and incremental learning are combined to dynamically update doctor labels to achieve multidisciplinary precise recommendations based on disease complexity and professional ability, and integrate generative adversarial network data amplification, automated ethical review and visual anomaly interpretation functions to ensure system compliance and clinical interpretability.

[0006] In a first aspect, the present invention provides a system for interpreting test reports of tumor patients, comprising: Information verification module, used for identity verification and classification of patients and doctors; The test report parsing module is used to perform structured parsing of test reports through natural language processing and computer vision technology, and to verify the qualifications of the testing unit based on the fuzzy matching algorithm; Data storage module, used to integrate knowledge base data related to tumor diagnosis and treatment, including tumor markers, treatment plans and clinical guidelines; The indicator analysis module is used to identify tumor types, analyze indicator abnormalities, and perform triage matching through a cascade model of support vector machines, isolation forests, and decision trees; Feature fusion module, which is used to realize the temporal fusion of cross-modal features based on dynamic weight allocation and long short-term memory network; The weight optimization module is used to quantify the differences in data distribution across institutions through KL divergence and realize dynamic aggregation of federated models based on meta-learning.

[0007] Furthermore, the qualification of the testing unit verified by the test report analysis module specifically includes: (1) Establish a standardized name database for CNAS-certified laboratories, including full names, abbreviations, and historical name change records; (2) Use OCR technology to extract the name of the testing agency from the test order image, and calculate its similarity with the database using the Levenshtein distance algorithm; (3) When the similarity is ≥85%, it will be automatically marked as a certification body. Otherwise, the manual review process will be triggered and the review results will be recorded to optimize the database; (4) Generate risk warnings for testing agencies that have not passed the certification and restrict their subsequent analysis authority of test reports.

[0008] Furthermore, the test report analysis module corrects the tilt and blur of the test report image by preprocessing the image, extracts key indicator data from the image using OCR technology, verifies the integrity of the data through a rule engine, and compares and analyzes the data with standard thresholds in the medical knowledge base.

[0009] Furthermore, the test report parsing module is based on the text positioning subunit of the convolutional neural network, which is used to extract text areas from the test report image, use the CRNN model to recognize the names and values ​​of the test items in the extracted area, and store the recognition results in a standardized manner according to a predetermined format.

[0010] Furthermore, the predetermined format includes: test item-value-unit-reference range.

[0011] Furthermore, the data storage module stores tumor diagnosis and treatment data in a structured manner in the form of entity-relationship-attribute, and uses crawler technology to capture the latest clinical research literature in real time to automatically correct outdated rules in the knowledge base.

[0012] Furthermore, the indicator analysis module realizes the classification of tumor types based on kernel function optimization, calculates the anomaly score according to the path length, and dynamically adjusts the threshold to adapt to the anomaly detection requirements of different tumor types, and constructs a triage rule tree based on the principle of maximizing information gain, supporting multi-label classification to match interdisciplinary medical teams.

[0013] Furthermore, the kernel function uses a Gaussian kernel, and the hyperparameters are determined through grid search and cross validation.

[0014] Furthermore, the anomaly detection process of the isolation forest model includes: Automatically adjust the weights of input features based on tumor type; Generate visual reports to illustrate the contribution of abnormal indicators and potential associated diseases.

[0015] By optimizing the model's adaptability to local data distribution through transfer learning and introducing a dynamic feature selection mechanism to improve the sensitivity and specificity of anomaly detection, the system can explore potential pathological associations from multi-dimensional indicators and generate interdisciplinary diagnosis and treatment recommendations. At the same time, it enhances doctors' trust in the algorithm output through visual explanations, significantly improving the efficiency and scientificity of diagnosis and treatment of complex cases.

[0016] Furthermore, a comprehensive doctor score is generated based on dimensions such as doctor's title, patient evaluation, and scientific research results.

[0017] Furthermore, a multidisciplinary expert recommendation list is generated based on the output results of the indicator analysis module and the doctor's professional labels, the doctor's practice status change information is synchronized through an incremental learning algorithm, and the patient's sensitive data is desensitized using homomorphic encryption technology.

[0018] Furthermore, it also includes a user interaction module for online doctor-patient consultations and report interpretation. By converting the patient's voice into structured text and triggering knowledge base retrieval through keyword matching, the trend changes of the patient's historical test data are displayed in the form of dynamic charts and abnormal indicators are indicated. It supports seamless switching between iOS, Android and Web, and the interface layout is adaptively adjusted according to the device screen size.

[0019] Accurate recommendations are achieved by combining doctors' professional capabilities to dynamically update labels, privacy-preserving computing, and multidisciplinary collaborative needs analysis. The fit between matching results and patients' conditions is significantly improved, and the dynamic scheduling efficiency of medical resources is optimized. At the same time, by ensuring the compliance of the recommendation logic, a personalized, full-cycle diagnosis and treatment service closed loop is built for patients.

[0020] Furthermore, the feature fusion module extracts text features from the test report through the BERT model, extracts medical image features through ResNet-50, and introduces the inter-modality covariance matrix to quantify the correlation between features:

[0021] In the formula, It is the dynamic fusion weight of text and image features; is the feature vector extracted from the test report text; is the feature vector extracted from the test image; and is the learnable linear transformation parameter; is the cross-modal interaction matrix; is the matrix trace operation.

[0022] The covariance matrix and attention mechanism are used to capture the deep semantic associations between text and images, and the temporal dependencies are modeled through long short-term memory networks to accurately characterize the dynamic evolution of tumor disease. The system's ability to parse complex multimodal data is significantly enhanced, and the interpretability of clinical decisions is greatly improved. At the same time, the computational load of redundant features is reduced, providing an efficient solution for real-time processing of high-dimensional medical data.

[0023] Furthermore, the weight optimization module dynamically adjusts the weights according to the difference between local and global data distribution:

[0024] In the formula, For the The weight of each participant in the federation aggregation; For the Probability distribution of local data of each participant; is the estimated distribution of global data; For local distribution With global distribution The Kullback-Leibler divergence between is the adjustment factor.

[0025] By quantifying the differences in data distribution across institutions through KL divergence, the federation aggregation weights are dynamically adjusted, and meta-learning is combined to quickly adapt to new institution data in small sample scenarios. The generalization of the model in cross-hospital collaboration is significantly improved, and the risk of privacy leakage in the data sharing process is greatly reduced. At the same time, the deployment cycle of the new institution model is shortened, providing a safe and efficient distributed learning paradigm for multi-center medical collaboration.

[0026] Furthermore, the basic data module of the system is deployed on the hospital's local server to reduce latency, and the core algorithm module is deployed on the cloud, achieving distributed computing and elastic resource expansion through an API interface.

[0027] In a second aspect, the present invention further provides a method for interpreting a test report of a tumor patient, the method being based on the system described in the first aspect, comprising: Complete data entry and preliminary integration based on test orders and medical images; Extract text and image features and achieve dynamic fusion through attention mechanism; Complete tumor classification, anomaly detection and triage recommendation generation based on collaborative analysis of SVM, isolation forest and decision tree; Doctor matching based on disease characteristics is achieved through knowledge graph and privacy-preserving computing; Based on the dynamic contribution weight aggregation model, rapid adaptation of new organizations is achieved through meta-learning; Complete closed-loop management of diagnosis and treatment and optimization of patient services based on visual reports and online consultation portal.

[0028] In a third aspect, the present invention also provides a computer device, comprising a management platform and a memory, wherein the management platform is connected to the memory, the memory is used to store computer programs, and the management platform is used to execute the computer programs stored in the memory, so that the computer device executes at least one step of the aforementioned method for interpreting test reports of tumor patients.

[0029] In a fourth aspect, the present invention further provides a computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a management platform, at least one step of the aforementioned method for interpreting test reports of tumor patients is implemented.

[0030] The present invention integrates support vector machines, isolation forests and decision tree algorithms to build a full-chain analysis system from tumor classification, abnormal detection to triage matching; multi-dimensional label modeling based on knowledge graphs and incremental learning breaks through the static and one-sided nature of traditional rule matching; dynamic weight allocation and cross-modal time series modeling solve the feature fragmentation problem caused by independent analysis of traditional multimodal data; dynamic contribution evaluation and meta-learning adaptation mechanisms are proposed for the heterogeneity of medical data distribution and privacy protection needs. The system and method have a wide range of application scenarios and can realize the automatic interpretation of tumor patient test reports and the matching of professional doctors.

[0031] Beneficial Effects By implementing the above-mentioned system and method for interpreting test reports of tumor patients provided by the present invention, the following technical effects are achieved: (1) By integrating support vector machines, isolation forests, and decision tree algorithms, a full-chain analysis system from tumor classification, anomaly detection to triage matching is constructed. Transfer learning is used to optimize the model's adaptability to local data distribution, and a dynamic feature selection mechanism is introduced to improve the sensitivity and specificity of anomaly detection. The system can mine potential pathological associations from multi-dimensional indicators and generate interdisciplinary diagnosis and treatment recommendations. At the same time, visual explanations are used to enhance doctors' trust in the algorithm output, significantly improving the efficiency and scientificity of diagnosis and treatment of complex cases.

[0032] (2) Multi-dimensional label modeling based on knowledge graph and incremental learning breaks through the static and one-sided nature of traditional rule matching. Accurate recommendations are achieved by combining doctors’ professional capabilities to dynamically update labels, privacy-preserving computing, and multidisciplinary collaborative needs analysis. The matching results are significantly more consistent with the patient’s condition, and the dynamic scheduling efficiency of medical resources is optimized. At the same time, by ensuring the compliance of the recommendation logic, a personalized, full-cycle diagnosis and treatment service closed loop is built for patients.

[0033] (3) Solve the problem of feature fragmentation caused by independent analysis of traditional multimodal data through dynamic weight allocation and cross-modal time series modeling. The covariance matrix and attention mechanism are used to capture the deep semantic association between text and images, and the long short-term memory network is used to model the time series dependency to accurately describe the dynamic evolution of tumor disease. The system's ability to parse complex multimodal data is significantly enhanced, the interpretability of clinical decisions is greatly improved, and the computational load of redundant features is reduced, providing an efficient solution for real-time processing of high-dimensional medical data.

[0034] (4) A dynamic contribution evaluation and meta-learning adaptation mechanism is proposed to address the heterogeneity of medical data distribution and privacy protection needs. The federation aggregation weights are dynamically adjusted by quantifying the differences in data distribution across institutions through KL divergence, and meta-learning is combined to quickly adapt new institution data in small sample scenarios. The generalization of the model in cross-hospital collaboration is significantly improved, and the risk of privacy leakage in the data sharing process is greatly reduced. At the same time, the deployment cycle of the new institution model is shortened, providing a safe and efficient distributed learning paradigm for multi-center medical collaboration. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to make the above-mentioned tumor patient test report interpretation system and method of the present invention more obvious and easy to understand, the drawings required for use in the specific implementation of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without paying creative labor.

[0036] Figure 1 A flow chart showing the method of interpreting test reports of tumor patients; Figure 2 A schematic diagram showing the interface of the system for interpreting test reports of tumor patients. DETAILED DESCRIPTION

[0037] Embodiment 1: A system and method for interpreting test results of tumor patients is provided. The method flow is as follows: Figure 1 As shown, it includes an information verification module, a test report parsing module, a data storage module and an indicator analysis module, which are described in detail as follows.

[0038] The information verification module is used for identity verification and classification of patients and doctors.

[0039] The test report parsing module is used to perform structured parsing of test reports through natural language processing and computer vision technology, and to verify whether the qualifications of the testing unit have passed the medical laboratory quality and capability accreditation based on the fuzzy matching algorithm.

[0040] The qualification of the testing unit verified by the test report analysis module specifically includes: (1) Establish a standardized name database for CNAS-certified laboratories, including full names, abbreviations, and historical name change records; (2) Use OCR technology to extract the name of the testing agency from the test order image, and calculate its similarity with the database using the Levenshtein distance algorithm; (3) When the similarity is ≥85%, it will be automatically marked as a certification body. Otherwise, the manual review process will be triggered and the review results will be recorded to optimize the database; (4) Generate risk warnings for testing agencies that have not passed the certification and restrict their subsequent analysis authority of test reports.

[0041] The test report analysis module corrects the tilt and blur of the test report image by preprocessing the image, extracts key indicator data in the image using OCR technology, verifies the integrity of the data through a rule engine, and compares and analyzes it with standard thresholds in the medical knowledge base.

[0042] The test report parsing module is based on the text positioning subunit of the convolutional neural network, which is used to extract the text area from the test report image, use the CRNN model to identify the test item name and value in the extracted area, and store the recognition result in a standardized manner according to a predetermined format.

[0043] The predetermined format includes: test item-value-unit-reference range.

[0044] The data storage module is used to integrate knowledge base data related to tumor diagnosis and treatment, including tumor markers, treatment plans and clinical guidelines.

[0045] The data storage module stores tumor diagnosis and treatment data in a structured manner in the form of entity-relationship-attribute, and uses crawler technology to capture the latest clinical research literature in real time to automatically correct outdated rules in the knowledge base.

[0046] The indicator analysis module is used to identify tumor types, analyze indicator abnormalities, and perform triage matching through a cascade model of support vector machines, isolation forests, and decision trees.

[0047] The indicator analysis module realizes the classification of tumor types based on kernel function optimization, calculates the anomaly score according to the path length, and dynamically adjusts the threshold to adapt to the anomaly detection requirements of different tumor types, and constructs a triage rule tree based on the principle of maximizing information gain, supporting multi-label classification to match a cross-disciplinary doctor team.

[0048] The kernel function uses a Gaussian kernel, and the hyperparameters are determined through grid search and cross validation.

[0049] The anomaly detection process of the isolation forest model includes: Automatically adjust the weights of input features based on tumor type; Generate visual reports to illustrate the contribution of abnormal indicators and potential associated diseases.

[0050] Generate a comprehensive doctor score based on dimensions such as doctor's title, patient evaluation, and scientific research results.

[0051] A multidisciplinary expert recommendation list is generated based on the output results of the indicator analysis module and the doctor's professional labels, the doctor's practice status change information is synchronized through an incremental learning algorithm, and the patient's sensitive data is desensitized using homomorphic encryption technology.

[0052] It also includes a user interaction module for online doctor-patient consultations and report interpretation. It converts patient voice into structured text and triggers knowledge base retrieval through keyword matching. It displays trend changes in patients' historical test data in the form of dynamic charts and indicates abnormal indicators. It supports seamless switching between iOS, Android and Web, and the interface layout is adaptively adjusted according to the device screen size.

[0053] The system interface is as follows Figure 2 As shown, the specific steps for tumor patients to interpret test reports and match doctors in the system include: Step 1: The patient registers and logs in using his / her mobile phone number. The system collects data such as the basic information of the cancer patient, specific pathological type, and treatment progress.

[0054] Step 2: The doctor registers and logs in using his / her mobile phone number. The system collects the doctor’s basic personal information, medical qualification certificate certification, level classification and other data.

[0055] Step 3: Identify and classify the contents of the test report, and determine whether the test unit meets the certification requirements; form a preliminary diagnosis and treatment opinion on the test report by retrieving knowledge data related to medical testing, and match the patient with a doctor in a related field.

[0056] Step 4: The patient starts the online consultation process. The doctor accepts the patient's questions and reviews the patient's information. According to the patient's condition, the doctor fills out the standardized receipt record form for the online consultation and provides the patient with a breakpoint chart of the patient's medical history results.

[0057] Step 5: The patient pays the doctor's consultation fee online. The doctor can withdraw the consultation fee stored in the system through the withdrawal system. After the consultation process is completed, the system records the process and saves the consultation results.

[0058] Embodiment 2: On the basis of the aforementioned embodiment, a feature fusion module is added, specifically an adaptive multimodal feature fusion mechanism, which is used to capture the complementarity of cross-modal features and the law of disease evolution through dynamic weight allocation and time series modeling, thereby enhancing the system's ability to analyze complex tumor data.

[0059] Extracting text features from test sheets using the BERT model , and extract medical image features through ResNet-50 .

[0060] Construct a two-stream network, calculate the dynamic weights of text and image, and introduce the inter-modal covariance matrix Quantify the correlation between features:

[0061] In the formula, It is the dynamic fusion weight of text and image features; is the feature vector extracted from the test report text; is the feature vector extracted from the test image; and are learnable linear transformation parameters used to map text and image features into the same space; is the cross-modal interaction matrix, which is used to quantify the correlation between text and image features; is the matrix trace operation, which is used to quantify the interaction strength of the covariance matrix.

[0062] Verification shows that under the condition of obtaining an average error similar to that of the above embodiment, on the TCGA dataset, the accuracy of tumor type recognition is increased from 89.2% to 93.5%, the feature dimension is reduced by about 30%, the reasoning speed is increased by about 25%, and the weight visualization shows that the text contribution accounts for 60%-80%, which is consistent with the clinical dependency logic. The results show that this mechanism can more comprehensively integrate text and image features, capture the dynamic evolution of tumor indicators, and significantly improve the fit between the disease classification results and the actual clinical needs. At the same time, the efficiency optimization of the feature fusion process reduces redundant calculations, the system response speed is significantly accelerated, and the weight visualization function enhances the doctor's trust in the algorithm decision logic, providing reliable technical support for accurate diagnosis and treatment.

[0063] Embodiment 3: On the basis of the aforementioned embodiment, a weight optimization module is further added, which is specifically a dynamic contribution evaluation and meta-learning adaptation mechanism, which is used to quantify data distribution differences through KL divergence and quickly fine-tune based on small samples to achieve accurate aggregation and personalized deployment of cross-institutional models.

[0064] Each hospital trains the SVM model based on local data and outputs parameters .

[0065] Dynamically adjust weights based on differences in local and global data distribution:

[0066] In the formula, For the The weight of each participant in the federation aggregation; For the Probability distribution of local data of each participant; is the estimated distribution of global data; For local distribution With global distribution The Kullback-Leibler divergence between , used to quantify the difference in distribution; It is a regulating factor, a hyperparameter used to control the influence of KL divergence on the contribution weight.

[0067] The global model is updated as:

[0068] In the formula, For the Model parameters trained locally by each participant; are the global model parameters after federation aggregation.

[0069] When a new institution joins, the MAML algorithm is fine-tuned on 10% of the local data to generate a personalized model.

[0070] For example, there are 5 hospitals (H1-H5), the data distribution is highly heterogeneous, the local model SVM is RBF kernel, and the federation rounds are 50 rounds.

[0071] Assume that the local data distribution of hospitals H1-H5 With global distribution The KL divergences are:

[0072] Hypothesized Moderating Factors , then the contribution weight is:

[0073] Calculated:

[0074] The global model parameters are:

[0075] Newly added hospital Using 10% local data, after fine-tuning with MAML, the local accuracy increased from 68% to 85%.

[0076] The effect of the weight optimization module is shown in Table 1.

[0077] Table 1. Summary of module effects

[0078] According to the experimental table, Hetero-FL reduces the impact of distribution differences through dynamic weights, and the standard deviation of cross-hospital test accuracy is reduced from 12.3% to 4.7%, indicating that the stability of the model under different data distributions is significantly enhanced; the contribution weight optimizes the parameter aggregation direction, and the number of rounds required to achieve a 90% benchmark accuracy is reduced by 40%; the KL divergence penalty term reduces the exposure of sensitive data, and the probability of leakage is reduced to 0.8%; the meta-learning mechanism reduces the fine-tuning time of new institutions from 12 hours to 2 hours, and the adaptation efficiency is improved by 83.3%. The framework shows stronger adaptability in cross-institutional collaboration, and the performance stability of the model under different data characteristics is significantly improved. At the same time, the cycle of model convergence and new institution access is greatly shortened. In addition, the strengthening of the privacy protection mechanism significantly reduces the risk of sensitive data leakage, provides systematic protection for the safe sharing and collaborative optimization of medical data, and promotes the balanced and intelligent development of tumor diagnosis and treatment resources.

[0079] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable non-transient storage media containing computer-usable program code.

[0080] The present invention can provide computer program instructions to a management platform of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the management platform of the computer or other programmable data processing device produce a device for implementing the system.

[0081] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a product including an instruction device that implements the functions of the system.

[0082] These computer program instructions may also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, whereby the instructions executed on the computer or other programmable device provide steps for implementing the functions of the system.

Claims

1. A system for interpreting test reports of tumor patients, characterized in that: include: Information verification module, used for identity verification and classification of patients and doctors; The test report parsing module is used to perform structured parsing of test reports through natural language processing and computer vision technology, and to verify the qualifications of the testing unit based on the fuzzy matching algorithm; Data storage module, used to integrate knowledge base data related to tumor diagnosis and treatment; The indicator analysis module is used to identify tumor types, analyze indicator abnormalities, and perform triage matching through a cascade model of support vector machines, isolation forests, and decision trees; Feature fusion module, which is used to realize the temporal fusion of cross-modal features based on dynamic weight allocation and long short-term memory network; The weight optimization module is used to quantify the differences in data distribution across institutions through KL divergence and realize dynamic aggregation of federated models based on meta-learning.

2. The system according to claim 1, characterized in that: The qualification of the testing unit verified by the test report analysis module specifically includes: (1) Establish a CNAS-certified laboratory directory database; (2) Extract the name field of the testing agency from the test report image; (3) Perform fuzzy matching between the extracted name and the database and return the matching score; (4) When the matching degree is ≥85%, it is determined to be a certification body, otherwise the manual review process is triggered.

3. The system according to claim 1, characterized in that: The indicator analysis module realizes the classification of tumor types based on kernel function optimization, dynamically adjusts the threshold according to the path length to adapt to the abnormality detection requirements of different tumor types, and constructs a triage rule tree to match the interdisciplinary doctor team based on the principle of maximizing information gain.

4. The system according to claim 1, characterized in that: A multidisciplinary expert recommendation list is generated based on the output results of the indicator analysis module and the doctor's professional labels, the doctor's practice status change information is synchronized through an incremental learning algorithm, and the patient's sensitive data is desensitized using homomorphic encryption technology.

5. The system according to claim 1, characterized in that: It also includes a user interaction module for online doctor-patient consultations and report interpretation. By converting the patient's voice into structured text and triggering knowledge base retrieval through keyword matching, it can display the trend changes of the patient's historical test data in the form of dynamic charts and indicate abnormal indicators.

6. The system according to claim 1, characterized in that: The feature fusion module extracts text features from the test report through the BERT model, extracts medical image features through ResNet-50, and introduces the inter-modality covariance matrix to quantify the correlation between features: In the formula, is the dynamic fusion weight of text and image features; is the feature vector extracted from the test report text; is the feature vector extracted from the test image; and is the learnable linear transformation parameter; is the cross-modal interaction matrix; is the matrix trace operation.

7. The system according to claim 1, characterized in that: The weight optimization module dynamically adjusts the weights according to the difference between local and global data distribution: In the formula, For the The weight of each participant in the federation aggregation; For the Probability distribution of local data of each participant; is the estimated distribution of global data; For local distribution With global distribution The Kullback-Leibler divergence between is the adjustment factor.

8. A method for interpreting a tumor patient's test report, characterized in that: The method is implemented based on the system according to any one of claims 1 to 7: The method comprises: Complete data entry and preliminary integration based on test orders and medical images; Extract text and image features and realize dynamic fusion through attention mechanism; Complete tumor classification, anomaly detection and triage recommendation generation based on collaborative analysis of SVM, isolation forest and decision tree; Doctor matching based on disease characteristics is achieved through knowledge graph and privacy-preserving computing; Based on the dynamic contribution weight aggregation model, rapid adaptation of new organizations is achieved through meta-learning; Complete closed-loop management of diagnosis and treatment and optimization of patient services based on visual reports and online consultation portal.

9. A computer device, comprising a management platform and a memory, wherein the management platform is connected to the memory, and the memory is used to store a computer program, characterized in that: The management platform is used to execute the computer program stored in the memory, so that the computer device performs at least one step in the method of claim 8.

10. A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, characterized in that: The computer program implements at least one step of the method of claim 8 when executed.

Images