Search

CN-121983215-A - Intelligent recording system and method for clinical physical examination data

CN121983215ACN 121983215 ACN121983215 ACN 121983215ACN-121983215-A

Abstract

The invention provides an intelligent recording method of clinical physical examination data, which comprises the steps of S1, user identity authentication and system authority opening, S2, local offline creation of patient files, S3, structuring treatment of nervous system physical sign data, S4, space topology mapping and structuring data package construction of physical sign data, S5, visual rendering of positive physical sign and generation of chevron nerve physical examination vector patterns, S6, association binding of physical sign data and patterns and local offline persistence storage. According to the invention, standardized and structured processing of the physical examination information of the nerve is realized from a data acquisition source, the spatial topological attribute of physical signs is completely reserved, high-fidelity vectorization visual generation and high-efficiency multiplexing of the physical examination map are realized, meanwhile, a full-link patient privacy data safety protection system is constructed, the cognitive load of clinical operation is reduced, the medical information circulation efficiency and the data safety are improved, and standardized and high-reliability data and technical support is provided for clinical diagnosis and treatment, disease course tracking and medical scientific research.

Inventors

  • WANG GANG
  • YAO XIAOYING
  • Pan Yuanmei

Assignees

  • 上海交通大学医学院附属仁济医院

Dates

Publication Date
20260505
Application Date
20260407

Claims (10)

  1. 1. The intelligent recording method for clinical physical examination data is characterized by comprising the following steps of: s1, acquiring an identity credential input by a user, carrying out local unidirectional encryption processing on the identity credential, and then carrying out consistency comparison with a local prestored authorization ciphertext, wherein the user identity authentication is completed after the comparison is passed, and the operation authority of a system is opened; S2, based on the open system operation authority, acquiring patient basic information input by a user, distributing global unique identifiers for corresponding patients, establishing a main table of patient basic information by taking the global unique identifiers as associated keys, and completing local offline creation of patient files; S3, acquiring physical examination sign data of the nervous system, which are input by a user through a preset standardized interaction component, based on the created patient file, performing strong constraint structural processing on the physical sign data, removing natural language ambiguity content, and generating a standardized physical sign data set with anatomical space attributes; S4, mapping the standardized sign data set into a preset two-dimensional kinematic anatomical topological model, matching anatomical logic nodes corresponding to the sign data through a ray intersection inclusion detection algorithm, generating corresponding sign state subtrees for each matched anatomical logic node, and integrating all sign state subtrees to construct a sign structured data packet with spatial topological association; S5, preprocessing the sign structured data packet, stripping normal baseline data without pathological identification meaning, screening positive sign data with clinical diagnosis value, generating a core skeleton of the herringbone nerve physical examination map based on the positive sign data and a preset canvas space coordinate system, completing visual rendering of the positive sign through a parameterized vector fitting algorithm, and generating a standard herringbone nerve physical examination vector map; S6, carrying out unique association binding on the physical sign structured data package and the corresponding generated herringbone nerve physical examination vector map, and storing the associated and bound data package and map into a physical examination record detail table associated with a patient basic information main table through a global unique identifier corresponding to a patient, so as to finish local offline persistence storage of data.
  2. 2. The method for intelligently recording clinical physical examination data according to claim 1, further comprising S7, acquiring an export instruction initiated by a user, calling a data packet and a map which are correspondingly associated and bound in a query body record detail table based on an application scene of the export instruction, performing differential privacy desensitization processing on the physical sign structured data packet if the data packet is a scientific research sharing scene, generating a compliance issuable data set, and converting the herringbone nerve query body vector map into a format picture which accords with medical specifications if the data packet is a medical record embedding scene, so as to finish corresponding export operation.
  3. 3. The intelligent recording method for clinical physical examination data according to claim 1 is characterized in that in S1, the specific process of carrying out local one-way encryption processing on an identity certificate is that a unique high-entropy random salt value is generated for the identity certificate through a kernel-level cryptography secure pseudo-random number generator, after the salt value and the identity certificate are spliced in sequence, a composite hash ciphertext is generated through a hash-based message identity verification code algorithm by combining a global isolation stored security key, and meanwhile, the local encryption processing adopts a cryptography algorithm with a memory hard characteristic, so that each encryption operation is forced to occupy a cache with a preset capacity, and the anti-cracking capability is improved.
  4. 4. The intelligent recording method of clinical physical examination data according to claim 1, wherein in the step S3, the acquired physical examination physical sign data of the nervous system comprises periodic tremor physical sign data, displacement time series signals acquired by a matched external sensor are synchronously acquired aiming at the periodic tremor physical sign data, after denoising processing is carried out on the displacement time series signals, sine wave curve fitting is completed through a nonlinear least square regression algorithm, amplitude, frequency, phase and baseline offset quantization parameters of tremor physical signs are extracted and included in a standardized physical sign data set, and in the step S4, the extracted quantization parameters are synchronously mapped into physical sign state subtrees of corresponding anatomic logic nodes.
  5. 5. The intelligent recording method of clinical physical examination data according to claim 1 is characterized in that in S5, a Canvas2D absolute layering rendering mechanism is adopted for visual rendering, a core framework layer, a pathological sign layer and a text labeling layer are sequentially rendered according to the sequence from a bottom layer to a top layer, wherein the core framework layer calculates anchor coordinates through a trigonometric function mapping matrix to generate a grogram framework, the pathological sign layer finishes vector drawing of continuous signs through three-time Bezier curve parameterization fitting, and the text labeling layer calculates labeling offset through a two-dimensional bounding box collision detection algorithm to avoid overlapping shielding of labeling contents.
  6. 6. The intelligent recording method of clinical physical examination data according to claim 1, wherein in S5, when generating a chevron-shaped nerve physical examination vector map, a sign structured data packet stored at the bottom layer adopts a standard enumeration key value irrelevant to language, a corresponding local dictionary mapping tree is loaded to complete the rendering of text labels based on a language environment selected by a user in view rendering, when receiving a language switching instruction initiated by the user, redrawing operation is only executed on a text label layer of the map, rendering contents of a core skeleton layer and a pathological sign layer are reserved, and lossless dynamic switching is completed.
  7. 7. The intelligent recording method of clinical physical examination data according to claim 1, wherein in S5, before generating the chevron-shaped nerve examination vector map, the pixel density ratio of the current terminal device is automatically detected, a scaling matrix is injected based on the pixel density ratio, the physical resolution of the rendering buffer is amplified, pixel interpolation distortion in the vector rendering process is avoided, and terminal display devices with different resolutions are adapted.
  8. 8. The intelligent recording method of clinical physical examination data according to claim 1, wherein after the local offline persistent storage is completed, if a physical sign data modification instruction initiated by a user is received, a physical sign structured data packet corresponding to a physical sign record detail table is called, reverse analysis and state rehydration are carried out on the data packet, physical sign parameters are written back into a standardized interaction component, and after the physical sign data modified by the user is obtained, the steps of structuring, topology mapping and map rendering are repeatedly executed, so that the responsive redrawing of a map and the updating storage of the data are completed.
  9. 9. The intelligent recording method of clinical physical examination data according to claim 1 is characterized in that, for multiple times of physical examination of the nervous system of the same patient, according to a first-in last-out mode, the data packet and the map after each time of generated association binding are sequentially stored in a physical examination record detail table of the corresponding patient to construct a disease course serialization view with a time axis, when a disease course comparison instruction initiated by a user is received, the corresponding physical examination record is called according to a time sequence, and the chevron-shaped physical examination vector map of the corresponding time node is synchronously rendered to complete visual comparison of physical sign evolution tracks.
  10. 10. An intelligent recording system for clinical physical examination data, characterized in that the intelligent recording method for clinical physical examination data according to any one of claims 2 to 9 is applied, the intelligent recording system for clinical physical examination data comprising: The user identity verification module is used for verifying the identity credentials input by the user, and if the user identity credentials pass the verification, the user identity authentication is completed, and the system operation authority is opened; The patient data core management module is used for completing local offline creation of a patient file and carrying out unique association binding on the physical sign structured data packet and the corresponding generated herringbone nerve examination vector map; The system comprises a structured sign input and quantitative analysis module, a mapping and intersection module, a structural sign input and quantitative analysis module, a structural sign analysis module and a structural sign analysis module, wherein the structured sign input and quantitative analysis module is used for inputting and preprocessing the physical examination sign data of a nervous system of a user to generate a standardized sign data set with anatomical spatial attributes; the visual sign check body rendering engine is used for screening positive sign data with clinical diagnosis value from the sign structured data packet, and generating a standard herringbone nerve check body vector map through visual rendering of the positive sign; The multi-language dynamic switching module is used for executing redrawing operation on the text labeling layer of the map only when receiving a language switching instruction initiated by a user, reserving rendering contents of the core framework layer and the pathological sign layer and completing lossless dynamic switching; The intelligent workbench and high-speed retrieval module is used for constructing a lightweight inverted index tree corresponding to all patient files, wherein when receiving retrieval keywords input by a user, the intelligent workbench and high-speed retrieval module is used for positioning the corresponding patient files and the contents of the query record list in constant or logarithmic time complexity by combining precompiled regular expressions through a matching algorithm; and the offline long graph splicing and exporting module is used for acquiring an export instruction initiated by a user, and calling and exporting the data packet and the map which are correspondingly associated and bound in the query body record detail table based on the application scene of the export instruction.

Description

Intelligent recording system and method for clinical physical examination data Technical Field The invention relates to the technical field of digital medical treatment and medical information intelligent processing, in particular to an intelligent recording system and method for clinical physical examination data. Background In neurology and related neuroscience clinical practice, physical examination of the nervous system is a core basic link of disease diagnosis, disease course monitoring and prognosis evaluation, but the existing physical examination recording technology and electronic scheme have a plurality of technical defects which are difficult to solve: The data unstructured problem stands out. The existing clinical physical records are stored in an electronic medical record system in a natural language text form, a large number of abbreviations, synonyms and grammatical ambiguity exist, a computer cannot realize accurate analysis and standardized query, and a compliance feature extraction basis is difficult to provide for a machine learning model, so that massive clinical real world data cannot be converted into structural data assets capable of being used for the second time. Spatial topology information is severely missing. The nerve sign has extremely strong anatomical space attribute, and the traditional linear text recording mode converts three-dimensional space sign distribution into one-dimensional character string storage, so that irreversible information loss occurs in the data storage link in the space mapping relation of the sign, the cognitive load of a clinician is increased, and information transmission errors are easy to occur in cross-department consultation and course handover. The standardization and multiplexing of the visual record are insufficient. The traditional hand-drawn herringbone physical examination map relies on personal operation habits of doctors and lacks a unified visual sign and standardized presentation system, and the scanned map is a raster image, so that the scanned map has large file size, vector lossless scaling cannot be realized, secondary structured editing cannot be performed, and the circulation efficiency of medical information and the subsequent scientific research data mining work are severely restricted. There is a short-board for data security and privacy compliance. The traditional identity authentication and password storage mechanism has insufficient safety intensity, is easily subjected to various attacks such as violent cracking, rainbow tables and the like, and is difficult to adapt to the use requirements of non-network or weak-network offline scenes such as isolation wards, basic medical treatment and the like. Disclosure of Invention The invention provides an intelligent recording system and method for clinical physical examination data, which realize standardized and structured processing of nerve physical examination information from a data acquisition source, completely reserve spatial topological properties of physical signs, realize high-fidelity vectorization visual generation and efficient multiplexing of physical examination maps, simultaneously construct a full-link patient privacy data safety protection system, finally reduce the cognitive load of clinical operation, improve the medical information circulation efficiency and the data safety, and provide a standardized and high-reliability data base and technical support for clinical diagnosis and treatment, disease course tracking and medical scientific research. In order to achieve the above purpose, the invention adopts the following technical scheme: an intelligent recording method for clinical physical examination data comprises the following steps: s1, acquiring an identity credential input by a user, carrying out local unidirectional encryption processing on the identity credential, and then carrying out consistency comparison with a local prestored authorization ciphertext, wherein the user identity authentication is completed after the comparison is passed, and the operation authority of a system is opened; S2, based on the open system operation authority, acquiring patient basic information input by a user, distributing a global unique identifier for a corresponding patient, establishing a patient basic information main table taking the global unique identifier as an association key, and completing local offline creation of a patient file; S3, acquiring physical examination sign data of the nervous system, which are input by a user through a preset standardized interaction component, based on the created patient file, performing strong constraint structural processing on the physical sign data, removing natural language ambiguity content, and generating a standardized physical sign data set with anatomical space attributes; S4, mapping the standardized sign data set into a preset two-dimensional kinematic anatomical topological model, matching anatomical logic nodes co