A data processing method based on inspection check results

Abstract

Embodiments of the present application relate to a kind of data processing methods based on inspection examination result, comprising: receiving initial inspection examination report data of hospital laboratory information system LIS, image filing and communication system PACS, electronic medical record system EMR;Initial inspection examination report data includes at least one of text, value, image;Image includes at least one of electrocardiogram and pathological section;Text, value, image in initial inspection examination report data are parsed, extract key medical entity, and key medical entity is mapped to standardization field, and target inspection examination report data is generated according to standardization field, and target inspection examination report data is structured inspection examination report data;Target inspection examination report and pre-set rule set are matched, and first quantity rules are determined from rule set;According to the priority of first quantity rules, determine target rule;The result information corresponding to target rule is sent to doctor terminal.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a data processing method based on inspection and testing results. Background Technology

[0002] Current clinical laboratory testing systems, such as Hospital Information Systems (LIS), Picture Archiving and Communication Systems (PACS), and Electronic Medical Records (EMR), only provide the function of displaying test results. Doctors need to manually analyze the data for diagnosis, which has the following problems:

[0003] 1. Inefficiency: The testing and diagnostic processes are separated, requiring doctors to switch between multiple platforms;

[0004] 2. Risk of misjudgment: Manual analysis relies on experience, and complex cases are prone to oversights;

[0005] 3. Insufficient data utilization: Unstructured inspection reports, such as text and images, are difficult for automated systems to parse.

[0006] While existing CDSS systems can provide diagnostic suggestions, their rules are fixed and they rely on structured input, making them unable to adapt to the needs of multi-format test data and dynamic medical scenarios. Summary of the Invention

[0007] The purpose of this invention is to address the shortcomings of existing technologies by providing a data processing method based on inspection and testing results, thereby solving the problems existing in the prior art.

[0008] To achieve the above objectives, the present invention provides a data processing method based on inspection results, comprising:

[0009] Receive initial laboratory test report data from the Hospital Laboratory Information System (LIS), the Picture Archiving and Communication System (PACS), and the Electronic Medical Record System (EMR); the initial laboratory test report data includes at least one of text, numerical values, and images; the images include at least one of electrocardiograms and pathological slides;

[0010] The text, values, and images in the initial test and examination report data are parsed to extract key medical entities, and the key medical entities are mapped to standardized fields. Target test and examination report data is generated based on the standardized fields. The target test and examination report data is structured test and examination report data.

[0011] The target inspection report is matched with a preset rule set, and a first number of rules are determined from the rule set;

[0012] Determine the target rule based on the priority of the first number of rules;

[0013] The result information corresponding to the target rule is sent to the doctor's terminal.

[0014] In one possible implementation, the step of parsing the text, numerical values, and images in the initial test and examination report data, extracting key medical entities, mapping the key medical entities to standardized fields, and generating target test and examination report data based on the standardized fields specifically includes:

[0015] Cluster the initial inspection and testing report data and extract common structural features; the common structural features include inspection and testing items.

[0016] When new inspection report data is detected, the text is preprocessed to remove redundant characters and segment semantic units.

[0017] The system determines whether the parsed semantic unit exists in the parsing template library. If it does not exist in the parsing template library, entity recognition is performed on the semantic unit based on a model in the medical field to identify key entities. The key entities include test and examination items, numerical values ​​and units, and descriptive text.

[0018] Regular expression matching and standard conversion are performed on the inspection and testing items, numerical values ​​and units, and descriptive text to generate structured first inspection and testing report data;

[0019] Text descriptions are extracted from examination images using OCR technology, and image feature processing is performed using a convolutional neural network to generate structured numerical indicators. The image feature processing includes identifying abnormal cell morphology from pathological sections and converting abnormal cell morphology into structured numerical indicators based on predefined standardized fields.

[0020] Based on the structured first inspection report data and structured numerical indicators, the target inspection report data is obtained.

[0021] In one possible implementation, calculating the intersection-union ratio of each rule in the first number of rules specifically includes:

[0022] The crossover ratio (IoU) of each rule is determined using the formula: IoU = number of matching rules / total number of rules related to the disease.

[0023] In one possible implementation, the method further includes:

[0024] The weighted intersection-union ratio (IoU1) of each rule is calculated using the formula IoU1 = Σ(matching rule weight) / Σ(weight of all associated rules).

[0025] In one possible implementation, the matching of the target inspection report with the first rule set, specifically including the matching of a first number of rules from the first rule set:

[0026] Calculate the overlap ratio of any two rules. When the overlap ratio is greater than the overlap ratio threshold, the rule is assigned to the first rule set.

[0027] When any two rules point to the same disease, the two rules are assigned a higher similarity weight and the rules are assigned to the second rule set.

[0028] The probability of statistical rules being triggered simultaneously is used to determine the rules whose probability is greater than a preset probability, and these rules are then assigned to the third rule set.

[0029] According to the clustering algorithm, the first rule set, the second rule set, and the third rule set are represented as feature vectors, and clustering is performed to obtain multiple clusters;

[0030] When the target inspection report triggers any cluster, calculate the intersection-union ratio of multiple rules within that cluster;

[0031] When a target inspection report triggers multiple clusters, the crossover and union ratio of each cluster is calculated based on the cluster priority, the target cluster is determined, and then the crossover and union ratio of multiple rules within the target cluster is determined.

[0032] The rule that determines the intersection-union ratio (IU) to be greater than the IU threshold is one of the first number of rules.

[0033] In one possible implementation, the clusters include anemia clusters, infection clusters, metabolic disease clusters, and respiratory disease clusters; the conditions for the anemia cluster include HGB and MCV; the conditions for the infection cluster include white blood cells, CRP, and body temperature; the metabolic disease cluster includes blood glucose, lactate, and pH; and the respiratory disease cluster includes white blood cells, body temperature, and pH.

[0034] In one possible implementation, the method further includes:

[0035] Receive diagnostic results sent from the doctor's terminal;

[0036] Based on the diagnostic results, the preset rule set is updated.

[0037] In one possible implementation, the method further includes:

[0038] Based on the time information, obtain the target inspection and examination report data prior to the current time; the target inspection and examination report data includes the name of the abnormal indicator, the value of the abnormal indicator, and the diagnosis result of the abnormal indicator;

[0039] Based on the target inspection report data, determine the corresponding rules;

[0040] Update the cluster weights according to the rules.

[0041] In one possible implementation, the method further includes:

[0042] Based on the current geographical location information, obtain the target inspection report data within the geographical location information range;

[0043] Based on the target inspection report data, determine the corresponding rules;

[0044] Update the cluster weights according to the rules.

[0045] In one possible implementation, the method further includes:

[0046] Obtain the patient ID;

[0047] Based on the patient ID, retrieve the patient's medical history corresponding to the patient ID; the medical history includes major illnesses, chronic diseases, infectious diseases, surgical history, and allergy history;

[0048] Matching is performed based on the aforementioned major diseases, chronic diseases, infectious diseases, surgical history, and allergy history, and a preset set of rules;

[0049] When a rule is matched, the target test report data of the patient ID is matched again with the matched rule;

[0050] If no rule is matched, the target test report data of the patient ID is matched with the rule set.

[0051] By applying the data processing method based on test results provided in this embodiment of the invention, initial test report data from a hospital laboratory information system (LIS), a picture archiving and communication system (PACS), and an electronic medical record system (EMR) are received. The initial test report data includes at least one of text, numerical values, and images; the images include at least one of electrocardiograms and pathological slides. The text, numerical values, and images in the initial test report data are parsed to extract key medical entities, which are then mapped to standardized fields. Target test report data, which is structured, is generated based on these standardized fields. The target test report data is matched with a preset rule set to determine a first number of rules. A target rule is determined based on the priority of the first number of rules. The result information corresponding to the target rule is sent to the doctor's terminal. Thus, by structuring the test report data, the data is unified, and based on a clustering algorithm, automatic diagnostic results are generated based on the test results, reducing reliance on manual analysis, improving user experience, and avoiding oversights in complex cases. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of a data processing method based on inspection results provided in an embodiment of the present invention;

[0053] Figure 2 for Figure 1 The specific content of step 120;

[0054] Figure 3 for Figure 1 The specific content of step 130. Detailed Implementation

[0055] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0056] Figure 1 This is a schematic flowchart of a data processing method based on inspection results provided in an embodiment of the present invention. The following is in conjunction with... Figure 1 The technical solution of the present invention will be described with reference to specific embodiments. Figure 1 As shown, this application includes the following steps:

[0057] Step 110: Receive initial laboratory test report data from the Hospital Laboratory Information System (LIS), the Picture Archiving and Communication System (PACS), and the Electronic Medical Record System (EMR); the initial laboratory test report data includes at least one of text, numerical values, and images; the images include at least one of electrocardiograms and pathological slides.

[0058] Specifically, this application can directly connect to LIS, PACS, and EMR systems to automatically obtain test and examination report data from each system. The text in the report data can be various test reports, and the values ​​can be the values ​​corresponding to the test reports or images. The images can be electrocardiograms, electroencephalograms, MRI results, etc. This application does not limit these.

[0059] Step 120: Parse the text, values, and images in the initial test and examination report data, extract key medical entities, map the key medical entities to standardized fields, and generate target test and examination report data based on the standardized fields. The target test and examination report data is structured test and examination report data.

[0060] Specifically, step 120 includes the following steps:

[0061] Step 1201: Preprocess the text by removing redundant characters and segmenting semantic units;

[0062] Specifically, irrelevant symbols and extra spaces are removed using regular expressions. Semantic units are then obtained by segmenting the text according to paragraphs or punctuation marks.

[0063] Step 1202: Determine whether the parsed semantic unit exists in the parsing template library. If it does not exist in the parsing template library, perform entity recognition on the semantic unit based on the medical domain model to identify key entities. Key entities include test and examination items, numerical values ​​and units, and descriptive text.

[0064] The semantic units are matched with regular rules in the template library. If no match is found, an NLP model is used to identify unmapped inspection items.

[0065] The tests and examinations include, but are not limited to, complete blood counts and biochemical indicators. The numerical values ​​are the test results, the units are the units of measurement, and descriptive text includes terms such as "increased" or "decreased."

[0066] Step 1203: Perform regular expression matching and standard conversion on the inspection and testing items, numerical values ​​and units, and descriptive text to generate structured first inspection and testing report data;

[0067] Specifically, this application can preprocess various initial test reports, including converting each report into a feature vector. For example, it can detect the presence of each item by encoding (1 for presence, 0 for absence), extracting item location features (e.g., if the report has a fixed paragraph structure, such as "complete blood count" or "biochemical indicators"), recording the location of the items, and using a clustering algorithm to divide the report into K classes based on the feature vectors (e.g., K=5, corresponding to complete blood count, blood gas analysis, imaging reports, etc.). It then analyzes the structural similarity between reports using a dendrogram. For each report class, it identifies frequently occurring test items, such as "HGB, WBC, PLT" in the complete blood count class, and subsequently generates a common structure template. This template structure includes template type, test items, numerical format, paragraph position, etc.

[0068] Step 1204: Extract text descriptions from examination images using OCR technology, and perform image feature processing using a convolutional neural network to generate structured numerical indicators; image feature processing includes identifying abnormal cell morphology from pathological sections and converting abnormal cell morphology into structured numerical indicators based on predefined standardized fields.

[0069] Specifically, a custom model is trained to adapt to the font of the medical report, and the pre-trained model is used for transfer learning to label medical record slice images, such as normal, cancerous, and inflammatory, and the model output is mapped to standardized indicators, such as the percentage of abnormal cells being 30%.

[0070] Step 1205: Obtain the target inspection report data based on the structured first inspection report data and the structured numerical indicators.

[0071] Furthermore, this application may also include, after step 120 and before step 130: obtaining the patient ID;

[0072] Based on the patient ID, retrieve the patient's past medical history, which includes major illnesses, chronic diseases, infectious diseases, surgical history, and allergy history.

[0073] Matching is performed based on major illnesses, chronic diseases, infectious diseases, surgical history, and allergy history, and a pre-defined set of rules;

[0074] When a rule is matched, the target test report data of the patient ID is matched again with the matched rule;

[0075] If no rule is found, the target test report data of the patient ID is matched with the rule set.

[0076] Specifically, the patient ID is obtained from the target test report, and then the patient's medical history is retrieved and matched with the rule base. When a match is successful, the matched rule is used as one of the first number of rules in step 130.

[0077] Step 130: Match the target inspection report with the preset rule set, and determine the first number of rules from the rule set;

[0078] Step 140: Determine the target rule based on the priority of the first number of rules;

[0079] Specifically, step 130 includes the following steps:

[0080] Step 1301: Calculate the overlap ratio of any two rules. When the overlap ratio is greater than the overlap ratio threshold, the rule is assigned to the first rule set.

[0081] Specifically, calculate the overlap ratio between the two rules in the condition section (such as inspection indicators). For example:

[0082] The conditions for rule A are "HGB < 12 and MCV < 80";

[0083] The conditions for rule B are "HGB < 12 and Ferritin < 15";

[0084] The overlap index for the two is HGB, and the overlap ratio is 1 / 2.

[0085] The degree of overlap can be calculated using the following formula:

[0086] Suppose the condition fields for rule A and rule B are C respectively. A and C B The overlap ratio is:

[0087]

[0088] If Overlap(A, B) is greater than the overlap ratio threshold, for example, if the overlap ratio threshold is 1 / 2, then rule A and rule B can be classified into the same rule set, which can be labeled as the first rule set. Step 1302: When any two rules point to the same disease, assign higher similarity weights to the two rules and classify the rules into the second rule set;

[0089] Specifically, if two rules point to the same type of disease, such as the same target disease field, like anemia, then they are assigned a higher similarity weight and are grouped into the same rule set.

[0090] One approach is to merge rules using a weighted overlap ratio, such as by using Overlap(A, B) * weight.

[0091] Step 1303: Calculate the probability of rules being triggered simultaneously, identify rules with a probability greater than the preset probability, and classify the rules into the third rule set;

[0092] Specifically, the probability of statistical rules being triggered simultaneously in practical applications, such as the frequency of rule A and rule B being triggered simultaneously in historical data, rules with high co-occurrence frequency are more similar, and rules with a probability greater than the probability threshold are grouped into the same rule set.

[0093] The trigger probability can be calculated using the following formula: P(A∩B) = Number of reports that trigger both A and B simultaneously / Total number of reports.

[0094] Step 1304: According to the clustering algorithm, the first rule set, the second rule set, and the third rule set are represented as feature vectors, and clustering is performed to obtain multiple clusters;

[0095] Specifically, applying K-means or hierarchical clustering algorithms to group rules and representing them as feature vectors may include the following features: the encoding of each field, such as test items and threshold ranges; the target disease label, which is a unique hot encoding; condition indicators, disease categories, and historical trigger counts;

[0096] Clustering algorithms, such as K-means and hierarchical clustering, are used to group the rule set into multiple clusters, each representing a class of related rules. For example, cluster 1 contains rules related to anemia, with conditions involving HGB, MCV, and Ferritin; cluster 2 contains rules related to infection, with conditions involving white blood cells, CRP, and body temperature; and cluster 3 contains rules related to metabolic diseases, with conditions involving blood glucose, lactate, and pH. The distance between clusters can be measured using weighted Euclidean distance or cosine similarity, combined with overlap ratio and disease label weighting. The number of clusters can be determined by optimizing the number of clusters using the elbow method or silhouette coefficient.

[0097] Step 1305: When the target inspection report triggers any cluster, calculate the intersection-union ratio of multiple rules within the cluster;

[0098] When a target inspection report triggers a cluster, it is necessary to calculate the crossover ratio (CRO) of the rules within the cluster. The specific steps include trigger rule extraction, CRO calculation, and CRO threshold filtering.

[0099] Trigger rule extraction involves matching all triggered rules within a cluster based on the content of the target test report, such as the test results. For example, rule A: Hemoglobin <12g / dL → triggers anemia; rule B: Mean corpuscular volume <80fL → triggers iron deficiency anemia. The intersection-over-union (IoU) ratio of the rule condition fields is calculated. If the condition fields for rules A and B are CA and CB respectively, then the IoU ratio is... For example, if both rule A and rule B check hemoglobin and red blood cell volume, then the intersection is 2, the union is 2, and IoU = 1.

[0100] If the IoU of a rule within a cluster is greater than a threshold, such as a threshold of 0.8, then these rules are considered to be highly overlapping, and only one of them, such as the rule with the highest priority, is retained.

[0101] Step 1306: When the target inspection report triggers multiple clusters, calculate the crossover and union ratio of each cluster according to the cluster priority, determine the target cluster, and then determine the crossover and union ratio of multiple rules within the target cluster.

[0102] Specifically, when a patient's target test report data triggers multiple clusters, the clusters need to be sorted by priority, such as disease severity, rule trigger frequency, etc., and the IoU of each cluster needs to be calculated in turn.

[0103] Different clusters can have priorities, which are represented by cluster weights. A higher weight indicates a higher priority. Clusters can have preset priorities and dynamic priorities. The preset priority is acute disease clusters > chronic disease clusters; the dynamic priority is determined based on the trigger frequency of acetic acid in historical reports. The highest priority cluster is selected, and its cross-over-union ratio (CUI) is calculated. If the CUI does not meet a threshold, the next highest priority clusters are checked sequentially. Within the target cluster, the rule with the highest CUI is output. This rule may be a single rule or a combination of overlapping rules.

[0104] For example, in one instance, cluster 1 is an anemia-related cluster, which includes rule A (hemoglobin <12g / dL) and rule B (hemoglobin <12g / dL and ferritin <30ng / mL).

[0105] Since the condition fields completely overlap, the intersection-union ratio (IoU(A,B)) of rules A and B is 1. If the threshold is 0.9, then only rule B will be retained because rule B is more stringent.

[0106] Cluster 2 is a diabetes-related cluster, which includes rule C (fasting blood glucose >126 mg / dL) and rule D (glycated hemoglobin >6.5%). There is no overlap between rules C and D, and the crossover ratio IoU(C,D) = 0, so both are retained.

[0107] If the report triggers both cluster 1 and cluster 2, and cluster 1 has a higher priority, then rule B will be output in the end, because the IoU meets the threshold.

[0108] Rules within the same cluster, due to their high degree of condition overlap, have a higher intersection-union ratio (IU) that better reflects the core characteristics of the disease. Therefore, rules within each cluster also have priorities; for example, rules can be categorized into multiple priority levels (e.g., urgent > high > medium > low). The priorities of different rules within the same cluster can also be represented by weights, with higher weights indicating higher priorities. Rules with higher priorities are processed first.

[0109] For example, in one cluster, there are rule A and rule B. Rule A has a priority of "urgent": "PaO2 < 60 mmHg → hypoxemia"; rule B has a priority of "medium": "blood glucose > 7.0 mmol / L → risk of diabetes". The processing order is to check rule A first, then check rule B.

[0110] Each rule is assigned a weight, such as based on the strength of medical evidence or historical accuracy, and rules with higher weights are given priority in case of conflict.

[0111] In another example, rule 1, weight 0.9: "White blood cells > 12 × 10" 9 / L AND body temperature > 38℃ → bacterial infection; Rule 2 (weight 0.7): "CRP > 10mg / L → inflammatory response".

[0112] If both match, the diagnostic recommendation from Rule 1 should be adopted first.

[0113] When logical conflicts exist, logical verification is performed on rules pointing to different diseases, and the conflict is adjudicated by combining context or additional data.

[0114] For example, rule A: "cough + fever → influenza", rule B: "cough + swollen lymph nodes → tuberculosis". Solution: If the patient has both "fever" and "swollen lymph nodes", further examination of chest X-ray or laboratory indicators, such as PPD test, is required for diagnosis.

[0115] When calculating the crossover ratio (IoU), the IoU of each rule can be determined using the formula IoU = number of matching rules / total number of rules related to the disease.

[0116] Alternatively, the intersection-union ratio (IoU) of each rule can be determined using the formula IoU = number of matching rules / total number of rules related to the disease.

[0117] In another example, the intra-cluster rule intersection and union is determined as follows:

[0118] Define the rule sets within each cluster: Cluster 0: Contains reports 1 and 4, involving items {HGB, WBC, PLT}. Cluster 1: Contains reports 2 and 5, involving items {WBC, PLT}. Cluster 2: Contains report 3, involving items {HGB, PLT}.

[0119] Calculate the intra-cluster regular intersection-union ratio: cluster 0, report 1 and report 4:

[0120] Rule 1: {HGB, WBC}, Rule 4: {HGB, WBC, PLT}, Intersection: {HGB, WBC} Union: {HGB, WBC, PLT}.

[0121] IoU** = 2 (number of intersection items) / 3 (number of union items) ≈ 0.67;

[0122] Cluster 1 (Reports 2 and 5):

[0123] Rule 2: {WBC, PLT}, Rule 5: {PLT}, intersection: {PLT}, union: {WBC, PLT}, IoU** = 1 / 2 = 0.5.

[0124] Cluster 2, only report 3: single rule, no intersection-union ratio calculation.

[0125] Step 1307: Determine that the rule whose crossover-union ratio is greater than the crossover-union ratio threshold is one of the first number of rules.

[0126] Specifically, if rules for multiple clusters are triggered, the global intersection-union ratio is calculated by combining the cluster priorities or weights.

[0127] For example, the weight of the emergency disease cluster is higher than that of the general disease cluster.

[0128] In one example, the input data is "HGB":11.0,"MCV":75,"Ferritin":10,"White blood cells":15×10 9 / L;

[0129] After the rules are matched, the rules for cluster 1 anemia are triggered: HGB<12, MCV<80, Ferritin<15; and the rules for cluster 2 infection are triggered at the same time: white blood cells>12.

[0130] Cross-union ratio (CUN) calculation: Intra-cluster calculation: The regular CUN of the anemia cluster is 3 / 3 = 1.0; Cross-cluster weighted calculation: The CUN of the infection cluster is 1 / 1 = 1.0, but due to its lower priority, the overall weight is 0.7. The recommended results are: "Primary Diagnosis": "Iron Deficiency Anemia", "Secondary Diagnosis": "Bacterial Infection", "Confidence Level": [0.95, 0.85]

[0131] In this application, before performing rule matching through a clustering algorithm, it may also include filtering invalid rules.

[0132] Specifically, invalid rules are excluded based on contraindications. First, the contraindications for each rule are checked (such as age restrictions, gender restrictions, specific medical history). If the patient's test data meets the contraindications for a certain rule, then that rule is directly excluded.

[0133] For example, the rule: "If the age is <18 years old, exclude prostate cancer-related diagnoses" → if the patient's age is 15 years old, then all prostate cancer rules are invalid.

[0134] Step 150: Send the result information corresponding to the target rule to the doctor's terminal.

[0135] Specifically, the results information corresponding to the rules, including the diagnosis results and the confidence level of each diagnosis result, can be sent to the doctor's terminal.

[0136] Furthermore, this application also includes the following steps:

[0137] Receive diagnostic results sent from the doctor's terminal;

[0138] Based on the diagnostic results, the preset rule set is updated.

[0139] Furthermore, this application may also include the following steps:

[0140] Based on the time information, retrieve the target inspection and testing report data prior to the current time; the target inspection and testing report data includes the name of the abnormal indicator, the value of the abnormal indicator, and the diagnosis result of the abnormal indicator;

[0141] Based on the data in the target inspection report, determine the corresponding rules;

[0142] Update the cluster weights according to the rules.

[0143] Specifically, the priority or weight of rules will be dynamically adjusted based on epidemiological data or the latest medical guidelines.

[0144] For example, during flu season, rules related to respiratory symptoms are given higher priority. Temporary rules take effect: New temporary rules are added to address public health emergencies.

[0145] Furthermore, this application may also include:

[0146] Based on the current geographical location information, obtain the target inspection and testing report data within the geographical location information range;

[0147] Based on the data in the target inspection report, determine the corresponding rules;

[0148] Update the cluster weights according to the rules.

[0149] Specifically, for example, if a certain disease suddenly occurs in a certain area, and the current geographical location information happens to be within this area, then within this area, the rule corresponding to the disease might be a body temperature greater than 37 degrees Celsius. If this rule is not in the rule base, a temporary rule can be added to update the rule base. If this rule is in the rule base, the weight of the cluster containing the rule "body temperature greater than 37 degrees Celsius" can be increased, so that the cluster containing this rule can be matched more quickly. The weight of the cluster and the weight of the rule within the cluster can also be increased, so that the rule within the cluster can be matched more quickly.

[0150] By applying the data processing method based on test results provided in this embodiment of the invention, initial test report data from LIS, PACS, and EMR are received. The initial test report data includes at least one of text, numerical values, and images; the images include at least one of electrocardiograms and pathological slides. The text, numerical values, and images in the initial test report data are parsed to extract key medical entities, which are then mapped to standardized fields. Target test report data, which is structured, is generated based on these standardized fields. The target test report data is matched with a preset rule set to determine a first number of rules. A target rule is determined based on the priority of the first number of rules. The result information corresponding to the target rule is sent to the doctor's terminal. Thus, by structuring the test report data, the data is unified, and based on a clustering algorithm, automatic diagnostic results are generated based on the test results, reducing reliance on manual analysis, improving user experience, and avoiding oversights in complex cases.

[0151] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0152] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented in hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0153] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A data processing method based on inspection and testing results, characterized in that, The method includes: Receive initial laboratory test report data from the Hospital Laboratory Information System (LIS), the Picture Archiving and Communication System (PACS), and the Electronic Medical Record System (EMR); the initial laboratory test report data includes at least one of text, numerical values, and images; the images include at least one of electrocardiograms and pathological slides; The text, values, and images in the initial test and examination report data are parsed to extract key medical entities, and the key medical entities are mapped to standardized fields. Target test and examination report data is generated based on the standardized fields. The target test and examination report data is structured test and examination report data. The target inspection report is matched with a preset rule set, and a first number of rules are determined from the rule set; Determine the target rule based on the priority of the first number of rules; Send the result information corresponding to the target rule to the doctor's terminal; The step of matching the target inspection report with the first rule set, specifically including the first number of rules matched from the first rule set: Calculate the overlap ratio of any two rules. When the overlap ratio is greater than the overlap ratio threshold, the rule is assigned to the first rule set. When any two rules point to the same disease, the two rules are assigned a higher similarity weight and the rules are assigned to the second rule set. The probability of statistical rules being triggered simultaneously is used to determine the rules whose probability is greater than a preset probability, and these rules are then assigned to the third rule set. According to the clustering algorithm, the first rule set, the second rule set, and the third rule set are represented as feature vectors, and clustering is performed to obtain multiple clusters; When the target inspection report triggers any cluster, calculate the intersection-union ratio of multiple rules within that cluster; When a target inspection report triggers multiple clusters, the crossover and union ratio of each cluster is calculated based on the cluster priority, the target cluster is determined, and then the crossover and union ratio of multiple rules within the target cluster is determined. The rule that determines the intersection-union ratio (IU) to be greater than the IU threshold is one of the first number of rules.

2. The method according to claim 1, characterized in that, The process of parsing the text, values, and images in the initial test and examination report data, extracting key medical entities, mapping these key medical entities to standardized fields, and generating target test and examination report data based on the standardized fields specifically includes: Cluster the initial inspection and testing report data and extract common structural features; the common structural features include inspection and testing items. When new inspection report data is detected, the text is preprocessed to remove redundant characters and segment semantic units. The system determines whether the parsed semantic unit exists in the parsing template library. If it does not exist in the parsing template library, entity recognition is performed on the semantic unit based on a model in the medical field to identify key entities. The key entities include test and examination items, numerical values ​​and units, and descriptive text. Regular expression matching and standard conversion are performed on the inspection and testing items, numerical values ​​and units, and descriptive text to generate structured first inspection and testing report data; Text descriptions are extracted from examination images using OCR technology, and image feature processing is performed using a convolutional neural network to generate structured numerical indicators. The image feature processing includes identifying abnormal cell morphology from pathological sections and converting abnormal cell morphology into structured numerical indicators based on predefined standardized fields. Based on the structured first inspection report data and structured numerical indicators, the target inspection report data is obtained.

3. The method according to claim 1, characterized in that, The calculation of the intersection-union ratio of multiple rules within the cluster specifically includes: The intersection-over-union ratio (IoU) for each rule is determined using the formula IoU = number of matching rules / total number of rules related to the disease.

4. The method according to claim 1, characterized in that, The calculation of the intersection-union ratio of multiple rules within the cluster includes: The weighted intersection-union ratio (IoU1) of each rule is calculated using the formula IoU1 = Σ (weight of matching rule) / Σ (weight of all associated rules).

5. The method according to claim 1, characterized in that, The clusters include anemia clusters, infection clusters, metabolic disease clusters, and respiratory disease clusters; the anemia cluster includes HGB and MCV; the infection cluster includes white blood cells, CRP, and body temperature; the metabolic disease cluster includes blood glucose, lactate, and pH; and the respiratory disease cluster includes white blood cells, body temperature, and pH.

6. The method according to claim 1, characterized in that, The method is followed by: Receive diagnostic results sent from the doctor's terminal; Based on the diagnostic results, the preset rule set is updated.

7. The method according to claim 1, characterized in that, The method further includes: Based on the time information, obtain the target inspection and examination report data prior to the current time; the target inspection and examination report data includes the name of the abnormal indicator, the value of the abnormal indicator, and the diagnosis result of the abnormal indicator; Based on the target inspection report data, determine the corresponding rules; Update the cluster weights according to the rules.

8. The method according to claim 1, characterized in that, The method further includes: Based on the current geographical location information, obtain the target inspection report data within the geographical location information range; Based on the target inspection report data, determine the corresponding rules; Update the cluster weights according to the rules.

9. The method according to claim 1, characterized in that, The method further includes: Obtain the patient ID; Based on the patient ID, retrieve the patient's medical history corresponding to the patient ID; the medical history includes major illnesses, chronic diseases, infectious diseases, surgical history, and allergy history; Matching is performed based on the aforementioned major diseases, chronic diseases, infectious diseases, surgical history, and allergy history, and a preset set of rules; When a rule is matched, the target test report data of the patient ID is matched again with the matched rule; If no rule is matched, the target test report data of the patient ID is matched with the rule set.

Images