Soap-based digital healthcare system

Abstract

The present invention relates to a SOAP-based digital healthcare system capable of automatically collecting and classifying patient symptom information and examination findings using speech recognition, medical natural language processing, vision recognition AI, and field-specific specialized analysis modules in a non-face-to-face medical consultation environment between a patient and medical staff, and generating diagnosis and treatment plans based thereon. More specifically, the invention relates to a SOAP-based digital healthcare system that supports non-face-to-face medical consultation between a patient and medical staff, comprising: a patient interface module for collecting subjective symptom information (Subject) from a patient; a medical staff interface module for inputting objective examination content (Object) by medical staff; a speech recognition engine for converting real-time voice conversation between a patient and medical staff into text; a medical natural language processing engine for extracting medically important keywords and context from the converted text; a conversation record generation module for generating a conversation record by recording the entire conversation content between the patient and medical staff in chronological order; and a SOAP diagnosis engine for generating an assessment and treatment plan by comprehensively analyzing the subjective symptom information, objective examination content, and conversation record. The present invention relates to a SOAP-based digital healthcare system comprising a document generation module that processes the extracted information and conversation records together with the subjective symptom information and objective examination content in a sequential process of Subject, Object, Assessment, and Plan according to the SOAP format to automatically classify them by item and generate a draft medical document, wherein the SOAP diagnostic engine is configured to fuse the analysis results of the medical natural language processing engine and the results of the field-specific AI analysis module in real time and automatically reflect them in each SOAP item.

Description

Technology Field

[0001] The present invention relates to a SOAP (Subject, Object, Assessment, Plan)-based digital healthcare system, and more specifically, to a SOAP-based digital healthcare system capable of automatically collecting and classifying patient symptom information and examination findings using speech recognition, medical natural language processing, vision recognition AI, and field-specific specialized analysis modules in a non-face-to-face medical consultation environment between a patient and medical staff, and generating diagnosis and treatment plans based thereon. Background Technology

[0002] Due to recent advancements in Information and Communication Technology (ICT) and the increasing demand for remote medical services, non-face-to-face digital healthcare systems that enable medical consultations even when patients and medical staff are not physically present in the same space are garnering attention. In particular, the need for remote medical consultation and diagnostic technologies is being further intensified by socio-medical demands, such as the entry into an aging society, the growing need for chronic disease management, and the prevention of infectious disease spread.

[0003] In the medical field, medical records are traditionally written based on the SOAP (Subject, Object, Assessment, Plan) format. This format plays a role in enhancing the consistency and accuracy of medical care by systematically recording the patient's subjective symptoms, objective examination findings, medical evaluations, and treatment plans. However, existing telemedicine systems lack features to automate or intelligently support the creation of such SOAP-based records, resulting in the inconvenience of medical staff having to manually organize and input information.

[0004] Furthermore, while high-resolution imaging and analysis technologies are required in fields where visual observation is crucial, such as dermatology, existing telemedicine platforms suffer from reduced diagnostic accuracy due to issues like degraded image quality, a lack of analysis capabilities, and limitations in tracking and managing lesions. Additionally, in the case of neurology outpatient care, the reliability of diagnoses in a remote environment is limited because the standardized collection of neurological symptoms and test results, as well as AI-based differential diagnosis support functions, are not sufficiently implemented.

[0005] In particular, existing telemedicine systems were limited to simple video call functions, which resulted in significantly reduced diagnostic accuracy in specialties where visual diagnosis is crucial, such as dermatology, ophthalmology, and otolaryngology. Furthermore, there was a lack of advanced visual diagnostic support features based on Vision AI, including real-time lesion analysis, automatic image capture, and cross-viewpoint comparative analysis.

[0006] In this way, existing non-face-to-face healthcare systems,

[0007] - Lack of support for automatic generation and intelligent classification of SOAP-based medical records

[0008] - Insufficient capabilities for photographing, analyzing, and tracking changes in skin lesions

[0009] - Lack of support for differential diagnosis and treatment planning based on neurological examination data

[0010] - Lack of monitoring functions for the safety and ethics of the treatment process

[0011] It has problems such as those mentioned above, and thus, technological development to improve it is required. Prior art literature

[0012] (0001) Korean Registered Patent No. 10-2646876 The problem to be solved

[0013] The technical problem that the present invention aims to solve is to provide a SOAP-based digital healthcare system that utilizes voice recognition technology and vision AI technology to drastically reduce the time required for medical staff to manually write chart records during the treatment process, thereby enabling medical staff to focus more on patient care.

[0014] Another technical problem that the present invention aims to solve is to provide a SOAP-based digital healthcare system capable of efficiently collecting and classifying patients' subjective symptom information and medical staff's objective examination content in a non-face-to-face medical environment, and automatically generating diagnosis and treatment plans based thereon.

[0015] Another technical problem that the present invention aims to solve is to provide a SOAP-based digital healthcare system that can reduce the record-keeping burden on medical staff and ensure consistency in medical treatment by automatically analyzing conversations between patients and medical staff using speech recognition and medical natural language processing technologies, and classifying and storing them by item in SOAP format.

[0016] Another technical problem that the present invention aims to solve is to provide a SOAP-based digital healthcare system that can drastically reduce the unnecessary chart recording time required during medical treatment by automatically recording information generated during conversations with patients and examinations by medical staff through the application of voice recognition-based AI and image recognition-based AI (vision AI) to the medical treatment process.

[0017] Another technical problem that the present invention aims to solve is to provide a SOAP-based digital healthcare system capable of precisely analyzing the color, size, shape, and texture of skin lesions through a high-resolution vision recognition camera and specialized AI analysis modules in various fields such as dermatology, and enabling multispectral image-based malignancy determination, treatment priority calculation, and tracking analysis of lesion changes.

[0018] Another technical problem that the present invention aims to solve is to provide a SOAP-based digital healthcare system capable of enhancing remote support functions for neurology outpatient care by collecting neurological symptoms and test data in a standardized format, calculating diagnosis probabilities by applying an AI-based differential diagnosis algorithm, and automatically generating treatment plans and follow-up schedules.

[0019] Another technical problem that the present invention aims to solve is to provide a SOAP-based digital healthcare system that utilizes behavioral monitoring AI to detect and record abnormal contact, non-medical acts, and ethical violations during the treatment process, and provides automatic warning and video storage functions to ensure the safety of both patients and medical staff.

[0020] Another technical problem that the present invention aims to solve is to provide a SOAP-based digital healthcare system that enables the establishment of personalized treatment and prevention strategies by integrating and managing patient-specific medical data and supporting long-term follow-up and analysis of treatment responses by disease.

[0021] The above system is configured to be scalable for application in various medical fields, including not only neurology and dermatology but also internal medicine, surgery, psychiatry, rehabilitation medicine, otolaryngology, and ophthalmology, and can operate as a general-purpose SOAP-based digital healthcare platform by applying specialized AI models and clinical guidelines tailored to the characteristics of each medical field.

[0022] The technical problems to be solved by the present invention are not limited to those mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art to which the present invention belongs from the description below. means of solving the problem

[0023] A SOAP-based digital healthcare system according to an embodiment of the present invention for achieving the above technical objectives comprises: a patient interface module that collects subjective symptom information (Subject) from a patient; a medical staff interface module that allows medical staff to input objective examination content (Object); a speech recognition engine that converts real-time voice conversation between a patient and medical staff into text; a medical natural language processing engine that extracts medically important keywords and context from the converted text; a conversation record generation module that generates a conversation record by recording the entire conversation content between the patient and medical staff in chronological order; and a SOAP diagnostic engine that generates an assessment and a treatment plan by comprehensively analyzing the subjective symptom information, objective examination content, and conversation record. It includes a document generation module that processes the extracted information and conversation records together with the subjective symptom information and objective examination content in a sequential process of Subject, Object, Assessment, and Plan according to the SOAP format to automatically classify them by item and generate a draft medical document, and the SOAP diagnostic engine is characterized by being configured to automatically reflect the analysis results of the medical natural language processing engine and the results of the field-specific AI analysis module in real time in each SOAP item.

[0024] In addition, it further includes a collaborative editing interface that allows medical staff to review and modify SOAP document drafts generated by the document generation module in real time; a prescription support module that automatically suggests drugs and dosages based on confirmed diagnoses; a safety verification module that verifies prescription safety by reviewing drug interactions, contraindications, and side effect risks in real time; a review and analysis module that generates final medical documents by reflecting modifications made through the collaborative editing interface in real time; and a history management module that integrates all data generated throughout the entire treatment process to build and manage a comprehensive medical history for each patient, wherein the history management module is configured to automatically update a long-term tracking database in conjunction with a field-specific specialized analysis module.

[0025] In addition, it includes a high-resolution vision recognition camera for photographing skin lesions; and a dermatology-specialized AI analysis module that analyzes skin images acquired from the vision recognition camera to extract color, size, shape, and texture features of the skin lesions, wherein the SOAP diagnostic engine integrates the analysis results of the dermatology-specialized AI analysis module with subjective symptom information and objective examination content to generate an evaluation and treatment plan for dermatological diseases, and the dermatology-specialized AI analysis module is configured to verify the reliability of patient statements regarding lesions by fusion analysis of skin images and conversation record data.

[0026] In addition, the vision recognition camera includes a multispectral imaging sensor, and the dermatology specialized AI analysis module is characterized by automatically identifying skin diseases including malignant tumors, benign tumors, and inflammatory diseases through a deep learning-based skin lesion classification algorithm, calculating a malignancy score, treatment priority, and the need for additional examination for each lesion, tracking and analyzing the pattern of change of the lesion by comparing it with past skin images of the same patient, and being configured so that the data on the pattern of change is automatically reflected in the SOAP evaluation stage.

[0027] In addition, it further includes a behavioral monitoring AI module that monitors the real-time video consultation process between a patient and medical staff to detect exceeding the scope of physical contact with the patient, exceeding consultation time limits, and filming for non-medical purposes; a safety management module that automatically generates a warning alert and separately stores the video of the relevant section when the behavioral monitoring AI module detects suspected sexual harassment or ethical violations among the detected behaviors; a smart shooting module that automatically captures important moments during the consultation process to create snapshots and automatically recognizes the timing of skin lesion filming to save high-quality images; and a dermatology-specific treatment planning module that automatically generates personalized treatment plans and lifestyle improvement measures according to dermatology treatment guidelines, wherein the safety management module is configured to automatically adjust warning trigger criteria by referring to a database of medical ethics compliance rules.

[0028] In addition, it includes a neurology-specialized patient interface module that systematically collects neurological symptom information from patients; a neurology specialist medical staff interface module that allows medical staff to input neurological examination findings in a standardized form; and a neurology-specialized AI diagnostic module configured to analyze the neurological symptom information and examination findings to generate a list of differential diagnoses for neurological diseases and calculate the probability for each diagnosis, wherein the neurology-specialized AI diagnostic module is configured to improve diagnostic accuracy based on past records by linking with a patient-specific long-term tracking database.

[0029] In addition, the above-described neurology specialized AI diagnostic module classifies neurological symptoms resulting from central and peripheral nervous system dysfunction by category, and calculates a three-stage diagnostic probability of probable, possible, and unlikely by comprehensively analyzing the frequency of occurrence, duration, aggravating / relief factors, and associated symptoms of each symptom; it includes a differential diagnosis algorithm for vascular, degenerative, paroxysmal, and headache-related neurological diseases; the results of the differential diagnosis are automatically reflected in the SOAP evaluation stage, and the priority is configured to be adjusted in the treatment planning stage.

[0030] In addition, it further includes a neurology treatment planning AI module that derives additional test recommendations and treatment plans based on the aforementioned differential diagnosis list; a treatment strategy formulation module that establishes a comprehensive treatment strategy including drug treatment, non-drug treatment, rehabilitation treatment, and lifestyle improvement measures according to the confirmed diagnosis; a follow-up management module that automatically proposes follow-up schedules and evaluation indicators for monitoring treatment response; and a patient education module that automatically generates and provides patient education materials and management guidelines specialized for each neurological disease, wherein the treatment strategy formulation module is configured to automatically reflect treatment priorities and resource allocation plans based on individual patient risk profiles.

[0031] The above embodiments of the present invention are merely some of the preferred embodiments of the present invention, and various embodiments reflecting the technical features of the present invention can be derived and understood by those skilled in the art based on the detailed description of the present invention to be described below. Effects of the invention

[0032] The present invention described above has the following effects.

[0033] First, the present invention automates SOAP diagnostic recording by simultaneously collecting patient voice and visual data, thereby increasing the work efficiency of medical staff and shortening patient consultation time. Furthermore, it is not limited to neurology or dermatology but is applicable to various medical specialties such as internal medicine, surgery, psychiatry, and rehabilitation medicine, and can be expanded into a general-purpose digital medical support platform by training voice AI and vision AI models tailored to the characteristics of each specialty.

[0034] In addition, the present invention provides advanced vision AI technology that is differentiated from conventional simple video medical services, thereby offering precise visual diagnostic support functions that are difficult for medical staff to perform directly, such as real-time automatic lesion detection, quantitative analysis based on multispectral images, automatic alignment with past images and measurement of changes, and 3D shape analysis of lesions. This enables achieving diagnostic accuracy equivalent to or greater than that of face-to-face medical services even in a remote medical environment, which can be considered an innovative differentiating factor compared to existing remote medical systems.

[0035] Furthermore, the present invention recognizes and records conversations between patients and medical staff in real time and automatically classifies them according to SOAP format to generate draft medical documents, thereby significantly reducing the burden of record-keeping on medical staff and ensuring consistency and standardization of medical records, which is advantageous for improving the quality of medical care and preserving legal and medical evidence.

[0036] Furthermore, the present invention facilitates information sharing and clinical cooperation among medical staff through collaborative editing and real-time review functions, reduces the risk of drug misuse and side effects through automatic prescription suggestions and safety verification functions, and enables the establishment of long-term treatment and prevention strategies by managing comprehensive medical history for each patient.

[0037] In addition, the present invention can improve the accuracy of diagnosing skin diseases even in a remote environment through high-resolution imaging of skin lesions and AI-based analysis, and enables medical professionals to make more objective judgments by quantitatively analyzing various characteristics of the lesions, such as size, color, shape, and texture.

[0038] In addition, the present invention can automatically classify malignant and benign tumors and inflammatory diseases by utilizing multispectral imaging and deep learning-based analysis, accurately determine whether the lesion has progressed through comparative analysis with past images, and quantitatively evaluate the treatment effect.

[0039] In addition, the present invention can ensure medical safety by detecting unethical acts, such as sexual harassment and filming for non-medical purposes, in real time and preserving evidence through behavioral monitoring AI, and can automatically secure optimal timing and image quality when filming skin lesions through a smart filming function.

[0040] In addition, the present invention can increase patient satisfaction by providing a personalized dermatological treatment plan.

[0041] In addition, the present invention can improve the objectivity and reliability of differential diagnosis by collecting and analyzing symptoms and test results in a standardized manner in neurological treatment, and can reduce omissions in diagnosis by enabling data-based analysis of neurological examinations even in a remote environment.

[0042] In addition, the present invention can simultaneously improve diagnostic efficiency and accuracy by classifying neurological symptoms by category and providing probability-based differential diagnosis, and enables differentiated analysis of vascular, degenerative, paroxysmal, and headache-related diseases, thereby assisting in setting treatment directions.

[0043] In addition, the present invention enables the establishment of optimal treatment strategies that reflect individual patient characteristics and the latest medical guidelines through an AI-based treatment plan derivation function, allows for the systematic management of treatment progress by automatically suggesting follow-up schedules and evaluation indicators, and enhances patients' understanding of the disease and self-management abilities by automatically providing patient education materials.

[0044] The effects obtainable from the present invention are not limited to those mentioned above, and other unmentioned effects will be clearly understood by those skilled in the art from the description below. Brief explanation of the drawing

[0045] FIG. 1 is an overall conceptual diagram of the configuration according to an embodiment of the present invention, FIG. 2 is a conceptual diagram of a SOAP process according to an embodiment of the present invention, FIG. 3 is an example diagram of a collaborative editing interface screen according to an embodiment of the present invention, FIG. 4 is a conceptual diagram of a vision recognition system configuration according to an embodiment of the present invention, FIG. 5 is a conceptual diagram of a skin lesion analysis process according to an embodiment of the present invention, FIG. 6 is a conceptual diagram of behavior monitoring and safety management according to an embodiment of the present invention, FIG. 7 is an example diagram of a neurology specialized interface according to an embodiment of the present invention, FIG. 8 is an exemplary diagram of a neurology AI diagnostic system according to an embodiment of the present invention. Specific details for implementing the invention

[0046] Hereinafter, some embodiments of the present invention will be described in detail with reference to exemplary drawings. It should be noted that in assigning reference numerals to the components of each drawing, the same components are given the same reference numeral whenever possible, even if they are shown in different drawings. Furthermore, in describing the embodiments of the present invention, if it is determined that a detailed description of related known components or functions would hinder understanding of the embodiments of the present invention, such detailed description is omitted.

[0047] In addition, terms such as first, second, A, B, (a), (b), etc., may be used when describing the components of the embodiments of the present invention. These terms are intended merely to distinguish the components from other components, and the essence, order, or sequence of the components is not limited by these terms. Where it is stated that a component is "connected," "combined," or "joined" to another component, it should be understood that the component may be directly connected or joined to the other component, but that components may also be "connected," "combined," or "joined" between each component.

[0048] FIG. 1 is an overall configuration concept diagram according to an embodiment of the present invention, FIG. 2 is a SOAP process concept diagram according to an embodiment of the present invention, FIG. 3 is an example diagram of a collaborative editing interface screen according to an embodiment of the present invention, FIG. 4 is a vision recognition system configuration concept diagram according to an embodiment of the present invention, FIG. 5 is a skin lesion analysis process concept diagram according to an embodiment of the present invention, FIG. 6 is a behavior monitoring and safety management concept diagram according to an embodiment of the present invention, FIG. 7 is an example diagram of a neurology specialized interface according to an embodiment of the present invention, and FIG. 8 is an example diagram of a neurology AI diagnosis system according to an embodiment of the present invention.

[0049] A SOAP-based digital healthcare system according to one embodiment of the present invention, in a SOAP-based digital healthcare system that supports non-face-to-face medical consultation between a patient and a medical staff, comprises: a patient interface module (1) that collects subjective symptom information (Subject) from a patient; a medical staff interface module (2) that allows a medical staff to input objective examination content (Object); a speech recognition engine (3) that converts real-time voice conversation between a patient and a medical staff into text; a medical natural language processing engine (4) that extracts medically important keywords and context from the converted text; a conversation record generation module (5) that generates a conversation record by recording the entire conversation content between a patient and a medical staff in chronological order; and a SOAP diagnostic engine (6) that generates an assessment and a treatment plan by comprehensively analyzing the subjective symptom information, objective examination content, and conversation record. It includes a document generation module (7) that processes the extracted information and conversation records together with the subjective symptom information and objective examination content in a sequential process of Subject, Object, Assessment, and Plan according to the SOAP format to automatically classify them by item and generate a draft of a medical document, and the SOAP diagnostic engine (6) can be configured to automatically reflect the analysis results of the medical natural language processing engine and the results of the field-specific AI analysis module in real time in each SOAP item.

[0050] That is, referring to FIG. 1, a SOAP-based digital healthcare system (hereinafter also referred to as the "system") can operate by including a patient interface module (1), a medical staff interface module (2), a voice recognition engine (3), a medical natural language processing engine (4), a conversation history generation module (5), a SOAP diagnostic engine (6), and a document generation module (7).

[0051] The patient interface module (1) can provide structured questions, free-form input, voice message upload, and photo and video upload functions to collect subjective symptom information from the patient, such as symptom description, onset time, aggravating and alleviating factors, pain scale, etc.

[0052] The patient interface module (1) can secure user authentication and consent records for the collected inputs and automatically assign metadata, such as shooting time, device information, location information, optional tags, etc., to ensure the reliability of the medical record.

[0053] The medical staff interface module (2) provides a screen configuration that allows the examiner to input objective examination findings (visual and palpation findings, test results, anthropometric values, etc.) into a standardized template and supports the selection of diagnostic codes or standard terms, such as SNOMED, ​​ICD, etc., thereby enabling the structuring of data.

[0054] The medical staff interface module (2) provides functions for attaching image and video files, linking measurement values ​​from external devices such as blood pressure monitors and blood glucose meters, and adding immediate annotations, thereby allowing objective examination evidence to be easily linked to electronic documents.

[0055] The voice recognition engine (3) can convert real-time voice conversations between patients and medical staff into text in a streaming manner and improve the recognition rate of specialized terms by using an adaptive voice recognition model that reflects a dictionary of specialized medical terms.

[0056] The speech recognition engine (3) can track the original audio location of each utterance by combining speaker diarization and timestamps, and can use the confidence score assigned to each sentence or token for subsequent processing.

[0057] The medical natural language processing engine (4) can perform named entity recognition, processing of negative expressions (e.g., "none"), extraction of time information, extraction of numerical values ​​and test values, extraction of relationships, and normalization based on semantic roles (annotation) by integrating text from the speech recognition engine (3) and text directly input by patients and medical staff.

[0058] The medical natural language processing engine (4) can provide a preprocessing and postprocessing pipeline that combines rule-based processing and machine learning-based classifiers to interpret context-dependent information, such as the severity of symptoms and the duration of symptoms, and maps the extracted information to the Subject, Object, Assessment, and Plan items of SOAP.

[0059] The conversation record generation module (5) can store all collected conversation texts, timestamps, and speaker information in chronological order so that they are searchable and retain an audit trail, and can link and store the original audio and attached media associated with the conversation segments.

[0060] The conversation record generation module (5) can manage conversation records in a way that satisfies legal and ethical requirements by generating storage encryption and access logs according to personal information anonymization and access control policies.

[0061] The SOAP diagnostic engine (6) can generate Assessment and Plan candidates by fusing the structured output of the medical natural language processing engine (4) and the results of field-specific AI analysis modules, such as a dermatology specialized AI analysis module (14) and a neurology specialized AI diagnostic module (21), in real time at the feature level and decision level.

[0062] The SOAP diagnostic engine (6) can derive a judgment using a weighted sum or ensemble technique by reflecting the reliability score of each source during fusion and can provide the explainability of the diagnosis candidate by including medical evidence (dialogue evidence, image evidence, history evidence, etc.).

[0063] The SOAP diagnostic engine (6) can automatically map the generated diagnostic candidates to standard codes and sort them based on priority to present candidates for treatment plans such as drugs, tests, and follow-up schedules.

[0064] The document generation module (7) can automatically construct a draft medical document based on the Subject, Object, Assessment, and Plan items output by the SOAP diagnostic engine (6) and provide it in a form that is easy for medical staff to review by including the corresponding evidence (dialogue segment ID, image file name, AI analysis result ID, etc.) and reliability information as annotations in each sentence.

[0065] The document generation module (7) can convert the generated draft into a standardized electronic medical record (EHR) format or structured data such as XML / JSON and output it so that it is easy to link with external systems and extract data.

[0066] The system can be configured to utilize an event-based message queue or API gateway structure for real-time data consistency between modules so that the text generated by the speech recognition engine (3), the analysis results of the medical natural language processing engine (4), the judgment of the SOAP diagnostic engine (6), and the draft of the document generation module (7) are consistently synchronized.

[0067] The system can ensure safety by highlighting items with low reliability in the automatic processing results and requiring manual verification by medical staff in the document generation module (7).

[0068] The aforementioned modules can continuously improve performance during operation through periodic updates of the learning model, supplementation of rules, and a log-based performance monitoring system.

[0069] Through this configuration, the system can improve the consistency and efficiency of medical care by integratively analyzing the patient's subjective symptoms, the medical staff's objective opinions, and conversation records during the non-face-to-face medical consultation process and automatically generating a draft of the medical document in SOAP format.

[0070] According to one embodiment of the present invention, the patient interface module (1) and the medical staff interface module (2) can collect the patient's voice and facial expressions, skin condition, scenes of motor function tests, and treatment process in real time during treatment.

[0071] The collected voice data is analyzed by a voice recognition engine (3) and a medical natural language processing engine (4), so that subjective symptoms stated by the patient and questions asked by medical staff can be automatically classified.

[0072] Simultaneously, the Vision AI module analyzes visual information, such as the patient's facial expressions, skin lesions, and movement patterns captured by the camera during medical examination, to automatically extract key disease-related findings from the corresponding image data. Going beyond simple image storage, this enables real-time analysis of lesion size, color tone changes, and movement patterns within the images, and automatically reflects these findings in the 'O (Objective)' field of SOAP.

[0073] The data generated in this way is structured into a diagnostic record by the SOAP diagnostic engine (6), and is finalized through the document creation module (7) and the collaborative editing interface (8) so that medical staff can review and modify it, and the entire treatment process can be clearly understood through the collected data.

[0074] In addition, according to one embodiment of the present invention, the system further includes a collaborative editing interface (8) that allows medical staff to review and modify a draft SOAP document generated by a document generation module (7) in real time; a prescription support module (9) that automatically suggests drugs and dosages based on confirmed diagnoses; a safety verification module (10) that verifies prescription safety by reviewing drug interactions, contraindications, and side effect risks in real time; a review and analysis module (11) that generates a final medical document by reflecting the content modified through the collaborative editing interface in real time; and a history management module (12) that integrates all data generated throughout the entire treatment process to build and manage a comprehensive medical history for each patient, wherein the history management module (12) may be configured to automatically update a long-term tracking database by linking with a field-specific specialized analysis module.

[0075] That is, referring further to FIGS. 2 and FIGS. 3, it can operate by additionally including a collaborative editing interface (8), a prescription support module (9), a safety verification module (10), a review and review module (11), and a history management module (12).

[0076] The collaborative editing interface (8) provides a real-time collaborative editing environment in which multiple medical staff can simultaneously view and edit the SOAP document draft generated by the document creation module (7), and can separate display and modification rights according to roles such as attending physician, co-patient, and pharmacy team through editing rights and role-based access control.

[0077] The collaborative editing interface (8) provides an editing history (audit trail), a change suggestion function, inline comments, and a discussion thread so that the basis for the modification is clearly recorded, and for each modification, the original text conversation segment ID and the image / AI analysis result ID can be automatically linked and displayed.

[0078] The collaborative editing interface (8) can generate automatic highlights and verification requests for specific items in the document, such as diagnosis candidates in Assessment and prescriptions in Plan, thereby inducing manual review by medical staff for items with low reliability transmitted from the document creation module (7).

[0079] The prescription support module (9) can automatically suggest indications, usage, and dosage based on confirmed or candidate diagnoses, and calculate individualized dosage by querying the patient's age, weight, kidney function, comorbidities, and past drug history information from the history management module (12).

[0080] The prescription support module (9) can derive a safe prescription by linking with a standard drug code system (RxNorm, etc.), referring to the latest approval information and clinical guidelines provided by pharmaceutical companies, and applying dosage adjustment rules for special groups such as children, renal impairment, and hepatic impairment.

[0081] The safety verification module (10) can check for drug interactions, contraindications, allergy history, duplicate prescriptions, and anthropometric-based dosage overdose in real time for the drugs and dosages proposed by the prescription support module (9), and generate immediate warnings according to the risk level.

[0082] The safety verification module (10) periodically synchronizes external drug databases, country-specific contraindications and precautions lists, and pharmaceutical company safety notices (DB) to perform up-to-date verification, and when a warning occurs, provides the supporting literature or guideline clauses together to present the basis for the medical staff's judgment.

[0083] The review and review module (11) can convert the final document draft approved in the collaborative editing interface (8) into an electronic medical record standard (FHIR, HL7 CDA, etc.) format, synchronize it with the electronic medical record system, and finalize and lock the final document through an electronic signature or approval process.

[0084] The review and review module (11) can automatically record the changer, time of change, and reason for change for all changes prior to approval, and, if necessary, store a snapshot of that point in time in accordance with legal requirements to manage evidence preservation.

[0085] The history management module (12) can integrate and store conversation records, images, AI analysis results, prescription records, and review logs generated throughout the system into a comprehensive history for each patient and automatically update the long-term tracking database in conjunction with the specialized analysis module for each field.

[0086] The history management module (12) can continuously improve the decision quality of the SOAP diagnostic engine (6) and prescription support module (9) by supporting follow-up notifications based on long-term data, automatic provision of comparison of previous lesions upon re-examination, analysis of cumulative drug exposure history, and calculation of patient-specific risk profiles.

[0087] The system can process data exchange between the collaborative editing interface (8), prescription support module (9), safety verification module (10), review and review module (11) and history management module (12) using standardized APIs and event-based message queues, and can manage data integrity and synchronization issues that occur during interoperability based on logs and transactions.

[0088] The system can ensure personal information protection and regulatory compliance by applying access control, encryption, audit log generation, and patient consent-based data utilization policies to the storage and transmission of sensitive medical information.

[0089] Through the aforementioned configuration, the system according to one embodiment of the present invention can simultaneously improve the efficiency of medical treatment and patient safety by supporting rapid and safe document confirmation and prescription verification through collaboration among medical staff.

[0090] Additionally, according to one embodiment of the present invention, the system includes a high-resolution vision recognition camera (13) for photographing skin lesions; and a dermatology-specialized AI analysis module (14) that analyzes skin images obtained from the vision recognition camera (13) to extract color, size, shape, and texture features of the skin lesions. The SOAP diagnostic engine (6) integrates the analysis results of the dermatology-specialized AI analysis module (14) with the subjective symptom information and objective examination content to generate an evaluation and treatment plan for dermatological diseases. The dermatology-specialized AI analysis module (14) may be configured to verify the reliability of patient statements regarding lesions by fusion analysis of skin images and conversation record data.

[0091] That is, referring further to FIG. 4, the system can perform functions specialized for dermatological treatment by including a vision recognition camera (13) and a dermatology specialized AI analysis module (14).

[0092] This vision AI-based analysis system is a core technology that is fundamentally different from existing simple video consultation systems. It quantifies subtle changes in lesions that are difficult for medical staff to detect with the naked eye, providing objective diagnostic evidence. It also enables medical staff with relatively little clinical experience to achieve diagnostic accuracy at the level of a specialist, thereby enabling the standardization of medical services and the equalization of quality.

[0093] The vision recognition camera (13) is a high-resolution medical imaging device that includes an optical zoom, an LED ring light, and a polarizing filter to minimize light reflection and shadows on the skin surface so as to be able to photograph skin lesions in detail in a medical setting.

[0094] The vision recognition camera (13) can be configured as a fixed or portable type and can be connected to the medical staff interface module (2) or the patient interface module (1) via wired or wireless connection to enable real-time shooting even during remote medical treatment.

[0095] The camera can support various shooting modes, and for example, by selectively activating standard RGB shooting, near-infrared shooting, and blue light shooting modes, it is possible to differentiate and photograph pigmentation, vascular structures, and sebum distribution of lesions.

[0096] The dermatology specialized AI analysis module (14) can perform image normalization such as color uniformization, noise removal, and contrast enhancement by preprocessing skin image data input from the vision recognition camera (13).

[0097] The analysis module (14) can utilize a boundary detection algorithm and a segmentation network (such as U-Net) to automatically detect skin lesion areas, and can extract key features such as the color histogram, size, shape, boundary irregularity, and surface texture of the detected lesion area. This feature data can be used as input values ​​for an AI model for lesion classification, and the model is trained based on large-scale skin lesion training data to automatically determine whether the lesion is benign or malignant, the disease category, and the stage of progression.

[0098] The dermatology specialized AI analysis module (14) can quantitatively analyze changes in lesions and, for example, calculate the diameter change rate, color variation index, irregularity index, etc., and use them for follow-up observation.

[0099] The SOAP diagnostic engine (6) can automatically generate a comprehensive assessment and treatment plan for dermatological diseases by integrating the analysis results of the dermatology specialized AI analysis module (14) with subjective symptom information collected from the patient interface module (1) and objective examination content entered into the medical staff interface module (2). In this process, the SOAP diagnostic engine (6) can be linked with a collaborative editing interface (8) to assign a reliability score to the AI ​​analysis results and to allow results with low reliability to be finalized after review by medical staff.

[0100] The system can store the captured original image and the analysis result image in the patient history management module (12) to enable comparison analysis with the past during subsequent treatment.

[0101] The vision recognition camera (13) and the dermatology specialized AI analysis module (14) may include high-efficiency image compression and progressive resolution transmission functions during low-bandwidth transmission so that they can operate stably even in a remote environment.

[0102] Through this configuration, the system enables precise imaging of skin lesions and AI-based analysis even in a non-face-to-face environment, thereby improving the accuracy and efficiency of dermatological treatment.

[0103] In addition, according to one embodiment of the present invention, the vision recognition camera (13) includes a multispectral imaging sensor (uncoded), and the dermatology specialized AI analysis module (14) can be configured to automatically identify skin diseases including malignant tumors, benign tumors, and inflammatory diseases through a deep learning-based skin lesion classification algorithm, calculate a malignancy score, treatment priority, and the need for additional examination for each lesion, track and analyze the pattern of change of the lesion by comparing it with the skin image of the same patient in the past, and automatically reflect the data on the pattern of change in the SOAP evaluation stage.

[0104] That is, referring further to FIG. 5, the entire process from multispectral image collection to deep learning-based classification, tracking and comparison with past images, and reflection of the results in the SOAP evaluation stage can be implemented by the system centering on the vision recognition camera (13) and the dermatology specialized AI analysis module (14).

[0105] The vision recognition camera (13) includes a multispectral imaging sensor and can simultaneously collect images of multiple wavelengths, such as near-infrared and blue light as well as visible light bands.

[0106] The vision recognition camera (13) is equipped with an optical zoom, a polarizing filter, uniform lighting (such as a ring LED of variable intensity), and a RAW image acquisition function, so that it can acquire a precise image with minimal reflection and shadow effects on the skin surface.

[0107] The vision recognition camera (13) can correct for differences in color and optical conditions between shots by recording a color correction chart and a radiometric correction value together during shooting.

[0108] The collected multispectral data can be normalized into a spectral data cube through preprocessing steps such as band registration, radiometric correction, noise removal, and contrast enhancement in a dermatology specialized AI analysis module (14).

[0109] The dermatology specialized AI analysis module (14) can generate an accurate boundary mask of the lesion by applying a high-performance segmentation network to automatically detect the lesion area to the preprocessed data, and based on the segmentation results, the dermatology specialized AI analysis module (14) can calculate morphological features such as the area, perimeter, asymmetry index, and boundary irregularity index of the lesion.

[0110] The dermatology specialized AI analysis module (14) can quantitatively extract spectral characteristics, color distribution (e.g., CIE series index), texture characteristics (e.g., GLCM, wavelet-based characteristics) and vascular patterns (signals in the near-infrared band) for each wavelength band to analyze the spectral, texture, and vascular characteristics of the lesion.

[0111] The dermatology specialized AI analysis module (14) can distinguish between malignant tumors, benign tumors, and inflammatory diseases by applying a structure including a deep learning-based classifier, such as a convolutional neural network-based ensemble model and an attention mechanism, which takes extracted multi-domain features as input, and calculate a probability-based malignancy score (e.g., 0 to 1 scale) for each lesion.

[0112] The dermatology specialized AI analysis module (14) can provide explainability for clinical decision-making by outputting the necessity of examination (e.g., recommendation for biopsy), treatment priority (e.g., recommendation for immediate action / short-term follow-up / long-term follow-up) and the basis for recommendation (key feature values, related spectral features, reference similar cases) in addition to the malignancy score.

[0113] The dermatology specialized AI analysis module (14) can classify the generated malignancy score and recommendation based on confidence and threshold criteria, and set a flag to indicate high-confidence cases as automatic recommendations and low-confidence cases to request additional review by medical staff.

[0114] To compare with stored skin images of the same patient from the past, the dermatology specialized AI analysis module (14) can perform alignment of the same area using an image registration algorithm and process it so that direct comparison between time points is possible by applying deformable registration and size and angle normalization.

[0115] In change tracking, patterns of change over time can be quantified by calculating quantitative indicators such as the rate of change in area, color change (e.g., CIEDE2000-based distance), changes in boundary irregularities, and changes in texture indices.

[0116] The dermatology specialized AI analysis module (14) can determine the meaning of the change, such as significant growth or an increase in the likelihood of malignancy due to color change, by performing statistical significance testing and trend analysis on the above change indicators.

[0117] The dermatology specialized AI analysis module (14) can generate the malignancy score, priority, need for additional examination, and change pattern data into a structured data packet (JSON, etc.) and provide it to the SOAP diagnostic engine (6).

[0118] The SOAP diagnostic engine (6) can automatically reflect the change pattern data and malignancy score received from the dermatology specialized AI analysis module (14) into the SOAP Assessment item and record them as the basis for diagnosis.

[0119] The SOAP diagnostic engine (6) can automatically include recommendations in the Plan item based on the change pattern data, and, for example, if the malignancy score exceeds a specific threshold or the change indicator shows statistical significance, it can generate prescription and test recommendations including recommendations for biopsy or immediate visit.

[0120] All original images, analysis results, matching metadata, and change indicators are stored for each patient in the history management module (12) and can be used for future comparison and tracking.

[0121] The system can maintain a labeled verification dataset and cross-validation procedure to continuously verify the classification performance of the dermatology specialized AI analysis module (14), monitor performance indicators such as sensitivity, specificity, and AUC, and periodically perform model retraining and threshold reset.

[0122] The system is designed to assist medical staff in their judgment and can be linked to allow medical staff to review and modify the basis of the recommendation through a collaborative editing interface (8) even when the recommendation of the dermatology specialized AI analysis module (14) is automatically applied.

[0123] To protect personal information and comply with regulations, explicit patient consent is recorded prior to imaging, stored images and analysis data are encrypted, and access rights and audit logs can be managed.

[0124] Through this configuration, the system can automatically calculate the malignancy and priority of lesions through multispectral image-based quantitative feature extraction, deep learning classification, and sophisticated time-change analysis, and automatically reflect the results in the SOAP evaluation stage, thereby improving the accuracy and tracking efficiency of dermatological telemedicine.

[0125] In addition, according to one embodiment of the present invention, the system further includes: a behavior monitoring AI module (15) that monitors the real-time video consultation process between a patient and medical staff to detect exceeding the patient's physical contact range, exceeding the consultation time limit, and filming for non-medical purposes; a safety management module (16) that automatically generates a warning notification and separately stores the video of the relevant section when the behavior monitoring AI module (15) detects suspected sexual harassment or ethical violations among the detected behaviors; a smart shooting module (17) that automatically captures important moments during the consultation process to generate snapshots and automatically recognizes the time of skin lesion filming to save high-quality images; and a dermatology-specific treatment plan module (18) that automatically generates personalized treatment plans and lifestyle improvement measures according to dermatology treatment guidelines, wherein the safety management module (16) may be configured to automatically adjust the warning generation criteria by referring to a database of medical ethics compliance rules.

[0126] That is, referring further to FIG. 6, the system can support patient safety, ethical compliance, and the creation of personalized treatment plans during the treatment process by including a behavioral monitoring AI module (15), a safety management module (16), a smart imaging module (17), and a dermatology-specific treatment plan module (18).

[0127] The behavior monitoring AI module (15) can monitor the patient's physical contact range, examination time, and filming activities during the real-time video consultation between the patient and the medical staff. The module can detect abnormal behavior or filming activities outside the purpose of treatment by utilizing camera images, sensor data, location tracking information, etc.

[0128] The behavior monitoring AI module (15) can identify access to specific parts of the patient's body, excessive contact during examination, and repetitive suspicious movements through a machine learning-based behavior recognition algorithm, and determine whether there is suspected sexual harassment or a violation of medical ethics by comparing it with ethical regulations.

[0129] The safety management module (16) automatically generates a warning notification when a suspicious behavior detected by the behavior monitoring AI module (15) occurs, and saves the video of the relevant section separately so that it can be used for future review and as evidence.

[0130] The smart shooting module (17) can automatically capture important moments during the treatment process to create snapshots and automatically recognize the time of shooting skin lesions to save high-resolution images. The smart shooting module (17) can automatically adjust the shooting angle, lighting, and focus conditions to ensure optimal image quality.

[0131] The dermatology-specialized treatment planning module (18) can automatically generate individual treatment plans and lifestyle improvement measures based on the diagnosis and evaluation results provided by the SOAP diagnostic engine (6) and the dermatology-specialized AI analysis module (14). The dermatology-specialized treatment planning module (18) can derive customized recommendations by comprehensively considering information such as lesion type, malignancy score, patient age, underlying diseases, and past medical history.

[0132] The smart shooting module (17) and the dermatology specialized treatment planning module (18) can be linked with the patient history management module (12) to support continuous treatment tracking and plan adjustment by comparing and referencing past images, treatment plans, and diagnostic results.

[0133] The behavior monitoring AI module (15) and the safety management module (16) can control ethical risks that may occur during treatment in real time and provide immediate notifications to medical staff, thereby ensuring patient safety and compliance with medical ethics.

[0134] Through this configuration, the system can monitor safe interactions between patients and medical staff even in a non-face-to-face medical environment, and improve the accuracy and reliability of dermatological care by capturing high-quality images and providing personalized treatment plans.

[0135] In addition, according to one embodiment of the present invention, the system includes a neurology-specialized patient interface module (19) for systematically collecting neurological symptom information from a patient; a neurology-specialized medical staff interface module (20) for medical staff to input neurological examination findings in a standardized form; and a neurology-specialized AI diagnosis module (21) configured to analyze the neurological symptom information and examination findings to generate a differential diagnosis list for neurological diseases and calculate a probability for each diagnosis, wherein the neurology-specialized AI diagnosis module (21) may be configured to improve diagnostic accuracy based on past records by linking with a patient-specific long-term tracking database.

[0136] That is, referring further to FIG. 7, the system can perform SOAP-based digital healthcare functions specialized for neurology outpatient care, including a neurology-specialized patient interface module (19), a neurology specialist medical staff interface module (20), and a neurology-specialized AI diagnostic module (21).

[0137] The neurology-specific patient interface module (19) can systematically collect information on neurological symptoms from the patient, and, for example, can receive input through structured question forms and checklists regarding motor function abnormalities, sensory changes, headaches, dizziness, seizure symptoms, etc. The neurology-specific patient interface module (19) can provide various interfaces, such as voice input, touchscreen input, and mobile app input, to enable the patient to conveniently provide information.

[0138] The neurology specialist medical staff interface module (20) can input neurological examination findings in a standardized form and can support recording results of muscle strength tests, reflex tests, sensory tests, balance and coordination tests as standardized data. The medical staff interface may include real-time data input, voice command recognition, and automatic recording of test results.

[0139] The neurology specialized AI diagnostic module (21) can comprehensively analyze patient symptom and test data collected from the neurology specialized patient interface module (19) and the neurology specialized medical staff interface module (20). The neurology specialized AI diagnostic module (21) can quantify the frequency of occurrence, duration, aggravating and alleviating factors, and accompanying symptoms of each symptom, and learn patterns to generate a list of differential diagnoses for neurological diseases.

[0140] The neurology specialized AI diagnostic module (21) calculates the probability of each diagnosis based on the analysis results, and, for example, applies a three-stage probability system of probable, possible, and unlikely so that medical staff can refer to it when making a diagnosis.

[0141] The neurology specialized AI diagnostic module (21) classifies symptoms caused by abnormalities in the central nervous system and peripheral nervous system by category and may include differential diagnosis algorithms for various disease groups such as vascular, degenerative, paroxysmal, and headache-related neurological diseases. Through this, the patient's neurological condition can be systematically evaluated and evidence necessary for establishing follow-up examinations or treatment plans can be provided.

[0142] The neurology specialized AI diagnostic module (21) can provide analysis results and probability information to the SOAP diagnostic engine (6) to automatically reflect in the Assessment item and provide base data for deriving the treatment plan and additional tests in the Plan item.

[0143] The system stores data collected from the neurology specialized patient interface module (19) and the neurology specialist medical staff interface module (20) in the history management module (12), allowing for comparison and reference with past data during future medical treatment, and supporting continuous follow-up observation and evaluation of treatment response.

[0144] Through this configuration, the system systematically collects and analyzes the symptoms and test findings of neurological patients and provides probability-based differential diagnosis, thereby enabling accurate and reliable diagnostic support in remote or face-to-face environments.

[0145] In addition, according to one embodiment of the present invention, a neurology specialized AI diagnostic module (21) classifies neurological symptoms caused by central nervous system and peripheral nervous system dysfunction by category, and calculates a three-stage diagnostic probability of probable, possible, and unlikely by comprehensively analyzing the frequency of occurrence, duration, aggravating / relief factors, and accompanying symptoms of each symptom, and includes a differential diagnosis algorithm for vascular, degenerative, paroxysmal, and headache-related neurological diseases, and the differential diagnosis results are automatically reflected in the SOAP evaluation stage and configured to adjust the priority in the treatment planning stage.

[0146] That is, referring further to FIG. 8, the system can specify the differential diagnosis function of the neurology specialized AI diagnosis module (21) to systematically classify neurological symptoms caused by central nervous system and peripheral nervous system dysfunction, and comprehensively analyze the frequency of occurrence, duration, aggravating and alleviating factors, and accompanying symptoms of each symptom.

[0147] The neurology specialized AI diagnostic module (21) can calculate the probability value of each neurological symptom based on input patient symptoms and test data, and can support medical staff in making intuitive diagnostic judgments by applying a three-stage diagnostic probability system of probable, possible, and unlikely.

[0148] The AI ​​diagnostic module (21) specialized in neurology can classify the patient's symptoms into abnormalities in the central nervous system (brain, spinal cord) and abnormalities in the peripheral nervous system (peripheral nerves, muscles), and can quantitatively evaluate abnormalities in motor function, sensory function, reflexes, balance and coordination function, etc. within the respective categories.

[0149] The neurology specialized AI diagnostic module (21) may include differential diagnosis algorithms for major neurological disease groups such as vascular disease, degenerative disease, epileptic disease, and headache disease, and can compare and analyze the patient's symptom patterns and test results by applying standardized judgment rules and machine learning models for each disease group.

[0150] The neurology specialized AI diagnostic module (21) can improve the reliability of differential diagnosis by referencing past similar patient data for each disease group to correct the probability calculation and by integrating the correlation between the frequency and duration of symptom occurrence, factors for aggravation and relief, and associated symptoms.

[0151] The neurology specialized AI diagnostic module (21) transmits the analysis results to the SOAP diagnostic engine (6) to be reflected in the Assessment item and can be used as a basis for recommending additional tests or setting the initial treatment direction in the Plan item.

[0152] The neurology specialized AI diagnostic module (21) can display probability values, major symptoms, accompanying symptoms, and evaluations by disease group on the interface so that the diagnosis results and analysis basis can be visualized and easily understood and reviewed by medical staff.

[0153] Through this configuration, the system can systematically classify and analyze neurological symptoms and provide probability-based differential diagnosis, thereby enabling reliable diagnostic support and the establishment of follow-up treatment plans even in a remote medical care environment.

[0154] In addition, according to one embodiment of the present invention, the system further includes a neurology treatment plan AI module (22) that derives additional test recommendations and treatment plans based on the differential diagnosis list; a treatment strategy establishment module (23) that establishes a comprehensive treatment strategy including drug treatment, non-drug treatment, rehabilitation treatment, and lifestyle improvement measures according to the confirmed diagnosis name; a tracking management module (24) that automatically proposes a tracking schedule and evaluation indicators for monitoring treatment response; and a patient education module (25) that automatically generates and provides patient education materials and management guidelines specialized for each neurology disease, wherein the treatment strategy establishment module (23) may be configured to automatically reflect treatment priorities and resource allocation plans based on a patient-specific risk profile.

[0155] That is, referring to FIG. 8, the system can perform the functions of establishing a comprehensive treatment plan and patient management for neurological outpatient care, including a neurological treatment plan AI module (22), a treatment strategy establishment module (23), a follow-up management module (24), and a patient education module (25).

[0156] The neurology treatment plan AI module (22) can derive additional test recommendations and a treatment plan tailored to individual patients based on the differential diagnosis list and diagnosis probability generated as described above. The module can generate personalized recommendations by comprehensively considering the confirmed diagnosis name, symptom severity, patient age, underlying disease, drug allergy, and past treatment history.

[0157] The treatment strategy establishment module (23) can establish a comprehensive treatment strategy including drug treatment, non-drug treatment, rehabilitation treatment, and lifestyle improvement measures based on recommendations provided by the neurology treatment plan AI module (22). The treatment strategy establishment module (23) can automatically reflect the priority, start time, duration, and whether combination therapy is applied for each treatment item in the plan, and can support medical staff in modifying it through the collaborative editing interface (8) if necessary.

[0158] The tracking management module (24) can monitor the treatment response by reflecting the established treatment strategy in the patient-specific follow-up schedule and evaluation indicators, automatically generate regular re-examinations, timing of image and test data collection, and symptom re-evaluation schedules, and can reflect the progress evaluation and updates of the Plan items based on the progress of treatment.

[0159] The patient education module (25) can automatically generate patient-specific educational materials and management guidelines according to the confirmed diagnosis and treatment plan, and includes disease-specific symptom management methods, rehabilitation exercises, medication methods, and recommendations for lifestyle improvement, and can provide them in the form of digital documents, videos, and notifications to support the patient in effectively understanding and carrying out the treatment plan.

[0160] The neurology treatment planning AI module (22), treatment strategy establishment module (23), tracking management module (24), and patient education module (25) are linked with the history management module (12) to integrate and analyze past treatment history, patient response, and test results, and to reflect the Plan items in the latest state.

[0161] The aforementioned modules can be linked with the SOAP diagnostic engine (6) to automatically document the differential diagnosis produced in the Assessment stage and the treatment strategy and follow-up management content derived in the Plan stage, and support medical staff in reviewing and modifying them through the collaborative editing interface (8).

[0162] Through this configuration, the system can improve the accuracy of medical care and the effectiveness of treatment by providing integrated support for neurology outpatients, ranging from establishing customized treatment plans based on diagnostic results to tracking treatment progress and patient education.

[0163] An example of the usage state according to the present invention described above is as follows.

[0164] When a patient accesses the system, they can first input subjective symptom information (Subject) through the patient interface module (1). The patient can provide current symptoms, onset time, symptom severity, past medical history, etc., using a mobile app, tablet, web interface, etc. During this process, the system can automatically check for missing input data and, if necessary, provide supplementary questions.

[0165] At the same time, medical staff can check the patient's basic information and existing medical records through the medical staff interface module (2) and input objective examination content (Object) in real time if necessary. Medical staff can record the results of questioning, visual inspection, palpation, auscultation, etc. in a standardized format, and the input data can be transmitted to the SOAP diagnostic engine (6) in real time.

[0166] Real-time voice conversations between the patient and medical staff are converted into text by a voice recognition engine (3). Key keywords and contextual information are extracted from the converted text through a medical natural language processing engine (4), and the patient's symptoms, examination details, and conversation records are recorded in chronological order by a conversation record generation module (5). This record is used as a basis for comprehensively analyzing the entire medical process and for the SOAP diagnostic engine (6) to perform assessment and generate a treatment plan.

[0167] The SOAP diagnostic engine (6) generates an evaluation and treatment plan by synthesizing the patient's subjective symptom information, medical staff input, and conversation records, and the generated content is automatically classified into SOAP format by the document generation module (7) to create a draft of the medical document. The document generation module (7) organizes each item in the order of Subject, Object, Assessment, and Plan, and can be configured in a form that medical staff can easily review.

[0168] When a draft of a medical document is created, medical staff can review and modify the content in real time through a collaborative editing interface (8). During the modification process, the prescription support module (9) automatically suggests drugs and dosages based on the confirmed diagnosis, and the safety verification module (10) can verify the safety of the prescription by reviewing drug interactions, contraindications, and risks of side effects in real time.

[0169] When medical staff finally review the document and reflect the modifications, the review and review module (11) can integrate them to generate a final medical document. Data generated throughout the entire process is stored in the history management module (12) and managed as a comprehensive medical history for each patient, which can then be compared and referenced with past records during subsequent medical treatment.

[0170] In a dermatology treatment setting, a high-resolution vision recognition camera (13) can be used to photograph the patient's skin lesions. Features such as color, size, shape, and texture are extracted from the captured image by a dermatology-specialized AI analysis module (14), and the SOAP diagnostic engine (6) can integrate this with subjective symptoms and objective examination information to generate an evaluation and treatment plan.

[0171] Ethical violations such as exceeding the scope of contact with the patient's body during the treatment process, exceeding the standard time for the examination, or filming for non-medical purposes can be detected in real time by the behavior monitoring AI module (15), and when detected, the safety management module (16) can immediately generate a warning notification and save the video of the relevant section separately.

[0172] Important moments during treatment or times when skin lesions are photographed can be automatically captured by the smart shooting module (17) and saved as high-quality images, and the dermatology specialized treatment planning module (18) can automatically generate a patient-customized treatment plan and lifestyle improvement plan based on the results of the SOAP diagnostic engine (6) and the dermatology specialized AI analysis module (14).

[0173] Through this step-by-step process, the system safely and accurately supports telemedicine between patients and medical staff, and can improve the efficiency and reliability of medical care by automating medical documents, verifying prescription safety, and generating dermatology-specific diagnoses and treatment plans.

[0174] In addition, another example of a usage state according to the present invention is described as follows.

[0175] When a patient accesses the system, they can first input information regarding neurological symptoms through a neurology-specific patient interface module (19). The patient can provide various neurological symptoms, such as motor dysfunction, sensory changes, headaches, dizziness, and seizure symptoms, through a structured checklist and question form, and can provide information in various ways, such as voice input, touchscreen input, and mobile app input.

[0176] Medical staff can check the patient's symptoms and existing medical records through the neurology specialist medical staff interface module (20) and input neurological examination findings in a standardized form. Medical staff can input results of muscle strength tests, reflex tests, sensory tests, and balance and coordination tests, and can utilize voice command recognition or real-time data recording functions if necessary.

[0177] Real-time voice conversation between a patient and medical staff is converted into text by a voice recognition engine (3), and key keywords and contextual information are extracted through a medical natural language processing engine (4). A conversation record generation module (5) can record the conversation content in chronological order and provide it to a SOAP diagnosis engine (6), and the SOAP diagnosis engine (6) can generate a differential diagnosis list and diagnosis probability for neurological diseases by comprehensively analyzing the patient's symptoms, test findings, and conversation records.

[0178] The neurology specialized AI diagnostic module (21) analyzes input symptoms and test results to classify neurological symptoms caused by central nervous system and peripheral nervous system dysfunction into categories, and can calculate three-stage diagnostic probabilities of probable, possible, and unlikely by comprehensively analyzing the frequency of symptom occurrence, duration, aggravating and alleviating factors, and accompanying symptoms. The module can apply differential diagnosis algorithms for major neurological disease groups such as vascular, degenerative, paroxysmal, and headache diseases to provide highly reliable information that medical staff can refer to for diagnostic judgment.

[0179] The analysis results are reflected in the SOAP diagnostic engine (6), automatically recorded in the Assessment item, and used as basic data for establishing a treatment plan in the Plan item. The neurology treatment plan AI module (22) can derive additional test recommendations and personalized treatment plans based on the differential diagnosis list and confirmed diagnosis name, and can be designed to comprehensively reflect the patient's age, underlying diseases, drug allergies, past treatment history, etc.

[0180] The treatment strategy establishment module (23) can establish comprehensive treatment strategies, such as drug treatment, non-drug treatment, rehabilitation treatment, and lifestyle improvement measures, based on recommendations generated by the neurology treatment plan AI module (22). The module can automatically set the priority, start time, duration, and whether to apply combination therapy for each treatment item, and can support medical staff in modifying them through the collaborative editing interface (8).

[0181] The tracking management module (24) can monitor the treatment response by generating a patient tracking schedule and evaluation indicators according to the treatment strategy, and can automatically provide guidance on regular re-examinations, symptom re-evaluations, and the timing of test data collection. If necessary, the Plan item can be updated to reflect the latest status.

[0182] The patient education module (25) can automatically generate patient-specific educational materials and management guidelines based on the confirmed diagnosis and established treatment strategy. The module includes symptom management, rehabilitation exercises, medication instructions, and recommendations for lifestyle improvement, and can be provided in the form of digital documents, videos, and notifications to support the patient in understanding and carrying out the treatment plan.

[0183] Data generated during all treatment processes is integrated and stored in the history management module (12), allowing for comparison and reference with past records during future treatment, and can be utilized for continuous follow-up observation and evaluation of treatment effects.

[0184] Through this step-by-step process, the system provides integrated support for neurological patient symptom collection, differential diagnosis, establishment of customized treatment plans, tracking of treatment progress, and patient education, thereby enabling accurate and reliable diagnosis and treatment management even in a telemedicine environment.

[0185] In addition, according to an example of use for the multi-AI collaborative diagnosis of the present invention, during the treatment process, a patient interface module (1) collects voice data, and at the same time, a vision recognition camera (13) photographs a skin lesion, and a behavior monitoring AI module (15) can analyze changes in the patient's motor ability.

[0186] The data collected in this way is processed by the medical natural language processing engine (4), the dermatology specialized AI analysis module (14), and the neurology specialized AI diagnosis module (21), respectively, and then integrated into the SOAP diagnosis engine (6).

[0187] Through this, the patient's subjective symptoms, objective skin findings, and neurological examination results are analyzed simultaneously and cross-verified, which can enhance the reliability of the final SOAP record.

[0188] In addition, according to an example of use regarding the integration of safety and ethics monitoring of the present invention, patient statements and image data input through the patient interface module (1) during the treatment process can be monitored in real time through the safety verification module (10) and the safety management module (16).

[0189] The safety verification module (10) can detect and warn of a lack of evidence or ethical issues (e.g., concerns about the misuse of specific drugs) regarding the diagnosis and treatment plan within the AI-generated SOAP document, and the safety management module (16) can detect emergency situations (e.g., decreased consciousness, abnormal behavior) in the patient's image in real time and send an immediate notification to the medical staff interface module (2).

[0190] In addition, according to an example of use for patient-customized tracking management of the present invention, patient records generated by the SOAP diagnostic engine (6) are stored in the tracking management module (24) and the history management module (12), and can be provided to the patient as customized educational content through the patient education module (25).

[0191] For example, in the case of dermatology patients, the trend of change in the lesion is visually displayed by comparing previous images with current photos of the lesion, and in the case of neurology patients, motor function test data is provided in the form of a graph over time, allowing patients to monitor their progress themselves.

[0192] In addition, these tracking results can be automatically reflected in the 'Plan' item within the SOAP record and linked with the treatment strategy establishment module (23).

[0193] In addition, according to an example of use regarding multidisciplinary collaboration support of the present invention, if a patient has a complex disease, SOAP records can be automatically separated and organized by department.

[0194] For example, in cases where skin lesions and neurological symptoms are present simultaneously, a dermatology-specialized AI analysis module (14) and a neurology-specialized AI diagnosis module (21) can each generate a SOAP record and integrate them in a collaborative editing interface (8) to provide an integrated SOAP document for multidisciplinary collaboration.

[0195] In the foregoing, although all components constituting an embodiment of the present invention have been described as being combined or operating in combination, the present invention is not necessarily limited to such embodiments. That is, within the scope of the purpose of the present invention, all components may be selectively combined in one or more ways to operate. Furthermore, terms such as "include," "constitute," or "have" described above, unless specifically stated otherwise, mean that the relevant component may be inherent; thus, they should be interpreted as allowing for the inclusion of additional components rather than excluding other components. All terms, including technical or scientific terms, have the same meaning as generally understood by those skilled in the art to which the present invention pertains, unless otherwise defined. Terms commonly used, such as those defined in advance, should be interpreted in accordance with their meaning in the context of the relevant technology and, unless explicitly defined in the present invention, should not be interpreted in an ideal or overly formal sense.

[0196] The foregoing description is merely an illustrative explanation of the technical concept of the present invention, and those skilled in the art to which the present invention pertains will be able to make various modifications and variations within the scope of the essential characteristics of the present invention. Accordingly, the embodiments disclosed in this invention are intended to explain, not limit, the technical concept of the present invention, and the scope of the technical concept of the present invention is not limited by such embodiments. The scope of protection of the present invention shall be interpreted by the claims below, and all technical concepts within an equivalent scope shall be interpreted as being included within the scope of rights of the present invention. Explanation of the symbols

[0197] 1: Patient Interface Module 2: Medical Staff Interface Module 3: Speech recognition engine 4: Medical natural language processing engine 5: Conversation log generation module 6: SOAP diagnostic engine 7: Document Creation Module 8: Collaborative Editing Interface 9: Prescription Support Module 10: Safety Verification Module 11: Review and Review Module 12: History Management Module 13: Vision recognition camera 14: Dermatology specialized AI analysis module 15: Behavior Monitoring AI Module 16: Safety Management Module 17: Smart Shooting Module 18: Dermatology-Specialized Treatment Planning Module 19: Neurology-Specialized Patient Interface Module 20: Neurology Specialist Interface Module 21: Neurology Specialized AI Diagnosis Module 22: Neurology Treatment Planning AI Module 23: Treatment Strategy Formulation Module 24: Tracking Management Module 25: Patient Education Module

Claims

Claim 1 In a SOAP-based digital healthcare system supporting non-face-to-face medical consultation between a patient and medical staff, the system comprises: a patient interface module that collects subjective symptom information (Subject) from a patient; a medical staff interface module that allows medical staff to input objective examination details (Object); a speech recognition engine that converts real-time voice conversations between the patient and medical staff into text; a medical natural language processing engine that extracts medically important keywords and context from the converted text; a conversation record generation module that generates a conversation record by recording the entire conversation content between the patient and medical staff in chronological order; and a SOAP diagnostic engine that generates an assessment and treatment plan by comprehensively analyzing the subjective symptom information, objective examination details, and conversation record. and a document generation module that processes the extracted information and conversation records together with the subjective symptom information and objective examination content in a sequential process of Subject, Object, Assessment, and Plan according to the SOAP format to automatically classify them by item and generate a draft medical document; a high-resolution vision recognition camera for photographing skin lesions; a dermatology-specialized AI analysis module that analyzes skin images acquired from the vision recognition camera to extract color, size, shape, and texture features of skin lesions; a neurology-specialized patient interface module that systematically collects neurological symptom information from patients; and a neurology-specialized medical staff interface module that allows medical staff to input neurological examination findings in a standardized format.It includes a neurology-specialized AI diagnostic module configured to generate a list of differential diagnoses for neurological diseases by analyzing the neurological symptom information and examination findings and to calculate the probability for each diagnosis; the SOAP diagnostic engine is configured to automatically reflect the analysis results of the medical natural language processing engine, the dermatology-specialized AI analysis module, and the neurology-specialized AI diagnostic module in real-time at the feature and decision levels in each SOAP item; the SOAP diagnostic engine is configured to derive a judgment using a weighted sum or ensemble technique by reflecting the reliability scores of each source during fusion and to provide the explainability of the diagnosis candidate by indicating the medical evidence together; the SOAP diagnostic engine integrates the analysis results of the dermatology-specialized AI analysis module with the subjective symptom information and objective examination content to generate an evaluation and treatment plan for dermatological diseases; the dermatology-specialized AI analysis module is configured to verify the reliability of patient statements regarding lesions by fusion analysis of skin images and conversation record data, tracks and analyzes the pattern of change of lesions by comparing with past skin images of the same patient, and is configured so that the data on the pattern of change is automatically reflected in the SOAP evaluation stage. A SOAP-based digital healthcare system characterized by a neurology-specialized AI diagnostic module that classifies neurological symptoms resulting from central and peripheral nervous system dysfunction by category, calculates three-stage diagnostic probabilities (probable, possible, unlikely) by comprehensively analyzing the frequency of occurrence, duration, aggravating / relief factors, and associated symptoms of each symptom, includes differential diagnosis algorithms for vascular, degenerative, paroxysmal, and headache-related neurological diseases, and is configured to improve diagnostic accuracy based on past records by linking with a patient-specific long-term tracking database. Claim 2 A SOAP-based digital healthcare system according to claim 1, further comprising: a collaborative editing interface that allows medical staff to review and modify a draft SOAP document generated by the document generation module in real time; a prescription support module that automatically suggests drugs and dosages based on confirmed diagnoses; a safety verification module that verifies prescription safety by reviewing drug interactions, contraindications, and side effect risks in real time; a review and analysis module that generates a final medical document by reflecting the content modified through the collaborative editing interface in real time; and a history management module that integrates all data generated throughout the entire treatment process to build and manage a comprehensive medical history for each patient, wherein the history management module is configured to automatically update a long-term tracking database in conjunction with a field-specific specialized analysis module. Claim 3 delete Claim 4 A SOAP-based digital healthcare system characterized in that, in claim 1 or 2, the vision recognition camera includes a multispectral imaging sensor, and the dermatology specialized AI analysis module automatically identifies skin diseases including malignant tumors, benign tumors, and inflammatory diseases through a deep learning-based skin lesion classification algorithm, calculates a malignancy score, treatment priority, and the need for additional examination for each lesion, tracks and analyzes the pattern of change of the lesion by comparing it with past skin images of the same patient, and is configured so that the data on the pattern of change is automatically reflected in the SOAP evaluation stage. Claim 5 A SOAP-based digital healthcare system according to claim 1 or 2, further comprising: a behavioral monitoring AI module that monitors the real-time video consultation process between a patient and medical staff to detect exceeding the scope of physical contact with the patient, exceeding the consultation time limit, and filming for non-medical purposes; a safety management module that automatically generates a warning notification and separately stores the video of the relevant section when the behavioral monitoring AI module detects suspected sexual harassment or ethical violations among the detected behaviors; a smart shooting module that automatically captures important moments during the consultation process to generate snapshots and automatically recognizes the time of skin lesion filming to save high-quality images; and a dermatology-specific treatment plan module that automatically generates personalized treatment plans and lifestyle improvement measures according to dermatology treatment guidelines, wherein the safety management module is configured to automatically adjust the warning generation criteria by referring to a database of medical ethics compliance rules. Claim 6 delete Claim 7 delete Claim 8 A SOAP-based digital healthcare system according to claim 1 or 2, further comprising: a neurology treatment plan AI module that derives additional test recommendations and treatment plans based on the differential diagnosis list; a treatment strategy establishment module that establishes a comprehensive treatment strategy including drug treatment, non-drug treatment, rehabilitation treatment, and lifestyle improvement measures according to the confirmed diagnosis; a follow-up management module that automatically proposes follow-up schedules and evaluation indicators for monitoring treatment response; and a patient education module that automatically generates and provides patient education materials and management guidelines specialized for each neurological disease, wherein the treatment strategy establishment module is configured to automatically reflect treatment priorities and resource allocation plans based on patient-specific risk profiles.

Images