Search

CN-121983241-A - Multi-dimensional classification-based home rehabilitation guidance method and system for apoplexy

CN121983241ACN 121983241 ACN121983241 ACN 121983241ACN-121983241-A

Abstract

The invention relates to the field of medical information processing, in particular to a home rehabilitation guidance method and system for stroke based on multidimensional classification. The method comprises the steps of obtaining basic conditions of a stroke patient, classifying disease features, nerve function defect positions, severity and complications to form classification results, performing patient group mapping based on the classification results to establish corresponding key monitoring related indexes, performing change acquisition and change analysis according to the key monitoring related indexes to generate a staged rehabilitation proposal, and performing abnormality judgment and on-line guidance around the staged rehabilitation proposal to form adjustment proposal and write back. The invention takes the classification result as the organization basis penetrating the whole flow, so that a consistent closed loop processing link is formed by monitoring, analyzing, guiding and writing back, and the continuity, traceability and cooperative stability of the scheme generation and adjustment process in the home rehabilitation guidance process of the apoplexy are improved.

Inventors

  • ZHAO JING

Assignees

  • 上海市闵行区中心医院

Dates

Publication Date
20260505
Application Date
20260212

Claims (10)

  1. 1. A home rehabilitation guidance method for apoplexy based on multidimensional classification is characterized by comprising the following steps: S100, acquiring basic conditions of a patient suffering from apoplexy, and performing classification treatment on disease characteristics, nerve function defect parts, severity and complications to obtain classification results; S200, performing mapping processing on 10 types of patient groups based on the classification result to generate 10 types of patient groups; s300, performing key monitoring related index establishment processing based on the 10-class patient groups to generate key monitoring related indexes; S400, based on the key monitoring related indexes, performing change acquisition and change analysis processing to generate a staged rehabilitation proposal; s500, based on the staged rehabilitation proposal, carrying out abnormality judgment and online guiding processing, generating an adjustment proposal and writing back.
  2. 2. The method of claim 1, wherein performing a disease signature, neurological deficit, severity, complication classification procedure comprises: The disease feature classification processing comprises enumeration normalization of the stroke type, time axis normalization of the attack time to generate a disease course duration field, and output of classification labels comprising the stroke type, the attack time, recurrence information and current recovery stage description; The method comprises the steps of sorting and processing the nerve function defect parts, wherein the sorting and processing the nerve function defect parts comprises the steps of carrying out structural analysis on language, movement, cognition and psychological four types of nerve function defect conditions, extracting key phrases and carrying out conflict resolution, and mapping final phrases into corresponding defect part sorting labels; the severity classification processing comprises source priority judgment and merging processing on severity descriptions of doctor terminals and mobile terminals, and outputting severity gears or marks to be confirmed; A syndrome classification process comprising structured splitting of a syndrome description to generate a syndrome entry comprising an entry source and a validation state, and performing a correlation verification of past medication information and the syndrome entry.
  3. 3. The method of claim 1, wherein performing a class 10 patient group mapping process comprises: Performing field disassembly, standardization and consistency check on the classification result, wherein the consistency check comprises source consistency check and logic consistency check; Loading a preset mapping rule set from a storage module, wherein the mapping rule set is maintained in a version number mode, and each rule entry forms a matching condition by a combination mode of four enumerated values of a disease characteristic classification label, a nerve function defect part classification label, a severity gear and a complications classification label; performing rule matching by adopting a coarse-fine two-stage matching link, wherein the two-stage matching link comprises a first-stage matching step, a second-stage matching step and a third-stage matching step, wherein the first-stage matching step is used for forming a combination key based on a nerve function defect part classification label and a severity gear in the classification result and screening a candidate rule item set in the mapping rule set; and when the classification result has a mark to be confirmed, attaching a mapping confidence state field to the 10-class patient group output by the mapping processing.
  4. 4. The method of claim 1, wherein generating the group of 10 types of patients comprises: Generating 10 types of patient groups based on the final rule entry, wherein the 10 types of patient groups comprise patient group identifications, mapping confidence state fields and associated index fields, and the associated index fields are used for recording version numbers of the classification results and version numbers of the mapping rule sets according to which the classification results are generated; And writing the generated 10 types of patient groups and the associated index field into a patient group file of a storage module to form a patient group file index.
  5. 5. The method of claim 1, wherein the process of performing the key-monitor related indicator creation process comprises: resolving group identifications of the 10-class patient groups from the patient group archive, mapping confidence state fields and the association index fields; Loading a heavy point monitoring related index establishment rule set from a storage module for matching based on the group identification, wherein the key monitoring related index establishment rule set is maintained in a version number mode; When the matching is successful, generating a key monitoring related index structure according to the matched rule item, wherein the key monitoring related index structure comprises an index field list and an acquisition frequency configuration, the index field list is composed of a subset of healthy data and training data fields corresponding to language, movement, cognition and psychological nerve function defect parts, and the acquisition frequency configuration comprises an acquisition frequency switching rule for switching between a rest state and a movement or rehabilitation training state based on a training time period mark; and writing the generated key monitoring related index structure and the rule version number according to the key monitoring related index structure into a monitoring configuration file of the storage module.
  6. 6. The method of claim 1, wherein the process of performing the change acquisition and change analysis process comprises: The data transmission unit performs acquisition scheduling according to the acquisition frequency configuration in the key monitoring related indexes, switches the acquisition frequency according to the training time period marks, performs field existence check and field value consistency check on the reported health data and training data, generates a change acquisition record comprising an acquisition time stamp, a data source, a missing measurement mark and a rechecking mark, and is organized by a storage module by taking an acquisition window as a unit; The data processing unit reads the change acquisition record associated with the key monitoring related index from the storage module, executes a combined processing link of time sequence comparison, intra-window comparison and inter-window comparison, generates a change analysis record containing a change direction label, a fluctuation label and a mutation label, and executes isolation processing on data with a missing measurement mark or a recheck mark.
  7. 7. The method of claim 1, wherein generating the key monitoring related indicator comprises: The method comprises the steps of obtaining a patient group record index, carrying out phase switching judgment according to a preset phase switching rule item set based on the change analysis record and the patient group record index to obtain a phase mark, matching and assembling rehabilitation nursing, rehabilitation training and psychological nursing recommended items from a recommended template library maintained by a storage module based on the phase mark and the patient group record index to form a phase recommended item set, registering a recommended version record containing the patient group record index, the key monitoring related index version number, the phase mark and a phase switching basis field for the phase recommended item set, and generating a phase rehabilitation scheme recommendation.
  8. 8. The method of claim 1, wherein the process of performing anomaly determination and online guidance processing comprises: The data processing unit is used for executing advice consistency verification and execution feedback verification according to the advice version record, the change analysis record and the change acquisition record which are related to the staged rehabilitation proposal and the advice version record, and according to the abnormality judgment conditions which are loaded from a storage module and are bound with the patient group archive index, wherein the advice consistency verification is used for verifying the correspondence between the staged rehabilitation proposal advice and the generation basis thereof, the execution feedback verification is used for verifying the consistency of the training record submitted by a mobile terminal and the rehabilitation training advice item in the staged rehabilitation proposal; The online guidance processing comprises the steps that an online guidance module generates an online guidance task based on the abnormal state and pushes the online guidance task to a doctor terminal, and receives structural adjustment suggestion input which is returned by the doctor terminal and comprises an adjustment content field, an application stage field, an application patient group field and a review signing field.
  9. 9. The method of claim 1, wherein generating the adjustment suggestion and writing back comprises: Performing difference merging processing on the adjustment suggestion input and the staged rehabilitation proposal to generate an adjustment suggestion record, wherein the adjustment suggestion record comprises an adjustment suggestion version number, a write-back state field and the association index field which are in father-son association with the suggestion version record number; and writing the adjustment suggestion record into a storage module, updating a processing state field of the abnormal state, and synchronously writing the adjustment suggestion record back to a home management system and a mobile terminal through a data transmission unit.
  10. 10. The home rehabilitation guidance system for the apoplexy is characterized by comprising a data acquisition module, a data processing module, a remote monitoring module, an online guidance module and a storage module, wherein the modules are sequentially connected and used for realizing the method according to any one of claims 1-9.

Description

Multi-dimensional classification-based home rehabilitation guidance method and system for apoplexy Technical Field The invention relates to the field of medical information processing, in particular to a home rehabilitation guidance method and system for stroke based on multidimensional classification. Background In the field of medical information processing, the existing scheme generally processes the basic condition acquisition and key monitoring related index acquisition of a stroke patient, and combines classification parting, scheme recommendation and remote follow-up visit to form closed-loop management, so that the classification result is difficult to form a main index penetrating the whole flow, the processing caliber is difficult to trace back and review in different stages, the on-line guidance is difficult to write back and agree, and the like. The existing method is used for classifying the basic condition of a stroke patient by depending on static typing or experience rules, the referential relation between the classification result and the follow-up key monitoring related index and the staged rehabilitation proposal lacks unified constraint, the mapping caliber inconsistency between key monitoring related index establishment processing and ten kinds of patient group mapping processing is easy to occur under the home rehabilitation guidance of the stroke, the input boundary between the staged rehabilitation proposal and the abnormality judgment and on-line guidance processing is unclear, and the stable realization of adjustment proposal and write-back is difficult to meet. Aiming at classification results, ten kinds of patient groups, key monitoring related indexes, and joint processing of staged rehabilitation scheme suggestions, the prior art generally has short plates in links such as solidification and version registration of mapping rules, abnormal judgment and consistent maintenance of structured records and write-back states of on-line guidance processing, and is difficult to form continuous link consistent flows of acquisition, classification processing, mapping processing, change acquisition and change analysis processing, abnormal judgment and on-line guidance processing and write-back in application scenes of home rehabilitation guidance, so that review between a cross-stage scheme and on-line guidance is difficult to align according to difficult alignment, adjustment suggestions after write-back are difficult to establish traceable relations with existing staged rehabilitation scheme suggestions, and further influences such as unstable flow connection, insufficient record auditability and the like are brought in the process of home rehabilitation service implementation and medical collaborative treatment. Disclosure of Invention In order to solve the technical problems, the invention provides a home rehabilitation guidance method for apoplexy based on multidimensional classification, which comprises the following steps: S100, acquiring basic conditions of a patient suffering from apoplexy, and performing classification treatment on disease characteristics, nerve function defect parts, severity and complications to obtain classification results; S200, performing mapping processing on 10 types of patient groups based on the classification result to generate 10 types of patient groups; s300, performing key monitoring related index establishment processing based on the 10-class patient groups to generate key monitoring related indexes; S400, based on the key monitoring related indexes, performing change acquisition and change analysis processing to generate a staged rehabilitation proposal; s500, based on the staged rehabilitation proposal, carrying out abnormality judgment and online guiding processing, generating an adjustment proposal and writing back. Further, the process of classifying and treating disease features, nerve function defect sites, severity and complications comprises the following steps: The disease feature classification processing comprises enumeration normalization of the stroke type, time axis normalization of the attack time to generate a disease course duration field, and output of classification labels comprising the stroke type, the attack time, recurrence information and current recovery stage description; The method comprises the steps of sorting and processing the nerve function defect parts, wherein the sorting and processing the nerve function defect parts comprises the steps of carrying out structural analysis on language, movement, cognition and psychological four types of nerve function defect conditions, extracting key phrases and carrying out conflict resolution, and mapping final phrases into corresponding defect part sorting labels; the severity classification processing comprises source priority judgment and merging processing on severity descriptions of doctor terminals and mobile terminals, and outputting severity gears or marks to be confirme