System and Method for Providing Clinical Support Based on Patient-Specific Medical Data

The medical assistance system integrates patient-specific data across multiple systems to detect incomplete states and provide actionable items, ensuring safe and accurate medical data management without altering existing EMR or OCS structures.

KR103046558B1Active Publication Date: 2026-11-10배광욱
View PDF 9 Cites 0 Cited by

Patent Information

Application Number
KR1020260145305
Authority / Receiving Office
KR · KR
Patent Type
Patents
Current Assignee / Owner
Priority Date
2026-06-04
Filing Date
2026-08-04
Publication Date
2026-11-10
Estimated Expiration
2046-08-04

AI Technical Summary

Technical Problem

Existing electronic medical records (EMR) and order communication systems (OCS) fail to integrate patient-specific medical data effectively, leading to incomplete management of test results and follow-up actions, and lack the ability to automatically detect discrepancies and omissions without altering their structure.

Method used

A medical assistance system that acquires patient-specific data through multiple paths, calculates the reliability of patient identification information, generates a patient-specific medical context, and determines incomplete states to provide actionable items to a separate user interface for medical staff confirmation.

Benefits of technology

The system reduces the risk of data miscombination, automatically detects incomplete states, and ensures safe integration of medical data without altering existing systems, preventing erroneous recordation until confirmed by medical staff.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 112026095332857-PAT00001_ABST
    Figure 112026095332857-PAT00001_ABST
Patent Text Reader

Abstract

The present invention relates to a medical assistance system and method, and more specifically, to a medical assistance system and method that operates in conjunction with existing medical information systems, such as electronic medical records (EMR) or order delivery systems (OCS), without modifying them, and performs patient-specific tracking management and medical assistance by acquiring and integrating patient-specific medical data. More specifically, the medical assistance system and method of the present invention determine whether patient medical data acquired through different paths are combined, detect an incomplete state based on the correspondence between a medical act and its result, and calculate a subsequent medical act.
Need to check novelty before this filing date? Find Prior Art

Description

Technology Field

[0001] The present invention relates to a medical assistance system and method, and more specifically, to a medical assistance system and method that operates in conjunction with existing medical information systems, such as electronic medical records (EMR) or order delivery systems (OCS), without modifying them, and performs patient-specific tracking management and medical assistance by acquiring and integrating patient-specific medical data.

[0002] More specifically, the medical assistance system and method of the present invention determine whether patient medical data acquired through different paths are combined, detect an incomplete state based on the correspondence between a medical act and its result, and calculate a subsequent medical act. Background Technology

[0004] Electronic Medical Records (EMR) and Order Communication Systems (OCS) typically structure data based on medical staff's prescriptions and orders; consequently, the data is linked based on what was prescribed or ordered, rather than what results were generated. Under this structure, there are many cases where the status of whether results corresponding to ordered tests were actually generated, received, and verified, as well as whether follow-up actions were implemented, are not managed separately.

[0005] Furthermore, most EMR / OCS systems have limitations in that they fail to reflect the varying examination patterns and risk levels of individual patients by applying standardized screen sequences and input structures regardless of patient characteristics. In addition, even if corresponding pathology test results are received in a separate system after a biopsy is performed during an endoscopy, there is no function to automatically match the two, requiring manual comparison by medical staff. Moreover, there is a problem in that the system fails to automatically detect requirements for verification and action that arise over time, such as exceeding the retest cycle for glycated hemoglobin (A1c) or the approach of a scheduled follow-up test date. Furthermore, as blood test results, medical imaging results, pathology test results, and health check-up results are managed separately across different systems or screens, there is also the issue of difficulty in forming a single integrated clinical context.

[0006] Accordingly, there is a growing need for technology capable of integrating data by reflecting patient-specific test histories and results without altering the structure of existing medical information systems, automatically determining the relationship between medical procedures dispersed across different systems and their corresponding outcomes, and actively detecting omissions in results or delays in follow-up actions. Prior art literature

[0007] Korean Registered Patent No. 10-2827972 The problem to be solved

[0008] The present invention aims to provide a medical assistance system and method capable of safely determining whether patient medical data acquired through different channels is combined without modifying existing electronic medical records or prescription delivery systems.

[0009] The present invention also aims to automatically determine the relationship between a medical procedure and its corresponding result, and to actively detect incomplete states such as omission of results or delay in follow-up measures.

[0010] The present invention also aims to prevent the omission of result verification that may occur during the treatment process by calculating a subsequent medical procedure based on a detected incomplete state and providing it after verification by medical staff.

[0012] Other objects and advantages of this application will become more apparent from the following detailed description, together with the appended claims and drawings. Anything not described in this specification is omitted, as it can be sufficiently recognized and inferred by those skilled in the art of this application or a similar art field. means of solving the problem

[0014] One embodiment of the present invention relates to a patient-specific medical data-based medical assistance system comprising one or more processors, wherein

[0015] An acquisition unit that acquires patient identification information and patient medical data through two or more different paths among screen-based recognition, file synchronization, and test result database linkage without changing the database structure, network configuration, and software of the existing medical information system;

[0016] A data linkage unit that calculates the degree of agreement or reliability between patient identification information obtained through the above-mentioned different paths, and withholds the combination, display, or tracking management reflection of the patient medical data if the degree of agreement or reliability is below a reference value;

[0017] A context generation unit that generates a patient-specific medical context using medical outcome data as reference data from patient medical data allowed to be combined by the data linkage unit above;

[0018] An analysis unit that determines a correspondence between medical act data representing an examination or procedure scheduled or performed on a patient and medical result data generated in response to said medical act, determines one or more incomplete states among non-receipt of result, medical act-result mismatch, or non-reflection of follow-up action based on said correspondence, and calculates one or more medical items requiring verification or action based on said incomplete state and said patient-specific medical context, and the type of subsequent medical act and the standard time for execution corresponding thereto; and

[0019] An output unit that provides the above medical items and the above subsequent medical procedures to a user interface distinct from the existing medical information system, and does not automatically record the information generated by the analysis unit into the existing medical information system until confirmation input from medical staff is received.

[0020] including,

[0021] It can provide a medical assistance system.

[0022] In one embodiment, the acquisition unit can generate current patient information by acquiring one or more of a patient name, registration number, date of birth, gender, date of visit, and date of examination from one or more of the screen, active window, reception queue, examination room queue, reservation information, and DICOM worklist of the existing medical information system.

[0023] In one embodiment, the acquisition unit may extract patient identification information or medical result data by applying optical character recognition to a screen capture image, an ID card image, or a handwritten document image as the screen-based recognition.

[0024] In one embodiment, the acquisition unit may acquire inspection results expressed in different formats by converting them into a common data structure using a file conversion method corresponding to a plurality of entrusted inspection companies or inspection equipment.

[0025] In one embodiment, the acquisition unit may generate a single patient reception queue by integrating two or more of ID card OCR, handwritten document OCR, direct input, reception list, and reservation information.

[0026] In one embodiment, the acquisition unit or data linkage unit may sort the single patient reception queue based on whether one or more of the reliability of each input source, the time of input, and the patient identification information match.

[0027] In one embodiment, the data linkage unit may calculate the degree of agreement or reliability by comparing two or more of the patient name, registration number, date of birth, and test date obtained from different paths.

[0028] In one embodiment, the data linkage unit may request user confirmation by displaying a list of candidate patients and discrepancies, without performing automatic linkage, if one or more of the following exist: people with the same name, registration number mismatch, optical character recognition failure, examination date mismatch, or discrepancy between reservation information and reception information.

[0029] In one embodiment, the data linkage unit may indicate that an item whose degree of agreement or reliability is less than the reference value needs to be checked.

[0030] In one embodiment, the medical result data may include one or more of blood test results, pathology test results, medical imaging results, endoscopy results, ultrasound results, bone density test results, fundus examination results, and health checkup results.

[0031] In one embodiment, the context generation unit may analyze the numerical value, reference range, increase or decrease, or interpretation findings included in the medical result data to structure it into one or more of normal or abnormal, whether follow-up is required, and candidate diagnosis names.

[0032] In one embodiment, the context generation unit may generate different patient-specific medical contexts for the same medical outcome data based on one or more of the patient's medical condition, examination history, and concomitant drugs.

[0033] In one embodiment, the analysis unit may determine the correspondence between the medical procedure data and the medical result data using two or more of the examination date, specimen site, lesion location, and whether a biopsy was performed.

[0034] In one embodiment, the medical procedure data includes information on the performance of a tissue biopsy performed during an endoscopic examination, the medical result data includes pathology test results, and the analysis unit,

[0035] In the case where the pathology test result corresponding to the above tissue biopsy information is not received, a pathology result mismatch state; and

[0036] Matching indeterminate state when pathology results exist but the corresponding endoscopic examination information is not definitively determined;

[0037] It may be a decision made by distinguishing between them.

[0038] In one embodiment, the analysis unit may select a candidate for a chronic disease based on the patient's disease information or a candidate diagnosis, and determine whether the recommended cycle has been exceeded by comparing the final execution date and recommended cycle of the essential follow-up test item corresponding to the selected chronic disease with the current time, thereby determining the incomplete state.

[0039] In one embodiment, the non-reception of results includes a state in which an examination or procedure has been performed but the corresponding medical result data has not been received, the medical act-result mismatch includes a reading mismatch state in which medical result data exists but is not linked to the corresponding medical act data, and the non-reflection of follow-up measures may include a state of waiting for physician confirmation before the calculated follow-up medical act is reflected in the tracking management list after confirmation by medical staff.

[0040] In one embodiment, the analysis unit may update at least one of the incomplete state, the medical item, or the reference point for the performance of the subsequent medical act when new medical act data or medical result data is acquired.

[0041] In one embodiment, the analysis unit may calculate a change trend by comparing current medical result data with past medical result data and reflect this in the verification priority of the medical item.

[0042] In one embodiment, the subsequent medical procedure may include one or more of follow-up examinations, re-examinations, pathology tracking, vaccinations, and outpatient follow-up visits.

[0043] In one embodiment, the analysis unit may calculate the priority of the subsequent medical procedure by considering the recommended date for the follow-up examination, the examination cycle, the previous examination history, and whether the doctor has confirmed it.

[0044] In one embodiment, the output unit may reflect the subsequent medical act in the patient-specific tracking management list only when an approval, modification, hold, or exclusion input is received from the medical staff.

[0045] In one embodiment, the user interface may include a browser screen, a pop-up screen, or an overlay screen that is displayed separately from the screen of the existing medical information system.

[0046] In one embodiment, the output unit summarizes one or more of the refined result, the follow-up test candidate, the abnormal finding candidate, and the representative image in which the match of patient identification information is confirmed, and displays them in the same test row of the user interface.

[0047] Detailed information corresponding to the selected item among the above identical inspection rows is displayed through either a detailed area extended from the above identical inspection row or a separate confirmation screen, and

[0048] Unrefined data or the entire original image may be separated to be processed in the above-mentioned detail area or a separate confirmation screen.

[0049] In one embodiment of the present invention, a patient-specific medical data-based medical assistance method performed by one or more processors,

[0050] A step of acquiring patient identification information and patient medical data through two or more different paths among screen-based recognition, file synchronization, and test result database linkage, without changing the database structure, network configuration, and software of the existing medical information system;

[0051] A step of calculating the degree of agreement or reliability between patient identification information obtained through the above-mentioned different paths, and withholding the combination, display, or tracking management reflection of the patient medical data if the degree of agreement or reliability is below a reference value;

[0052] A step of generating a patient-specific medical context linked to the patient's disease information, prescription history, or procedure history, using medical result data including test results, interpretation results, or pathology results actually generated independently of the existence of order information for tests or prescriptions from the aforementioned patient medical data for which combination is permitted, as reference data;

[0053] A step of determining a correspondence between medical act data representing an examination or procedure scheduled or performed on a patient and medical result data generated in response to said medical act, determining one or more incomplete states among non-receipt of result, medical act-result mismatch, and non-reflection of follow-up measures based on said correspondence, and calculating one or more medical items requiring verification or action based on said incomplete state and said patient-specific medical context, and the type of subsequent medical act and the standard time for performance corresponding thereto; and

[0054] A step of providing the above medical items and the above subsequent medical acts to a user interface distinct from the existing electronic medical record or prescription delivery system, wherein the information generated in the above calculation step is not automatically recorded in the existing medical information system until confirmation input from medical staff is received.

[0055] It can provide a medical assistance method including Effects of the invention

[0057] According to one embodiment of the present invention, the risk of miscombination caused by identical names or discrepancies in registration numbers can be reduced by determining whether to combine patient medical data based on the degree of consistency or reliability between patient identification information obtained through different paths, without changing the database structure, network configuration, and software of existing medical information systems such as electronic medical records or prescription delivery systems.

[0058] In addition, the present invention can automatically determine the correspondence between medical procedure data and medical results to distinguish between non-receipt or non-connection of examination records, results, images such as DICOM, interpretations, and pathology results, and can reduce the possibility of missing result verification and follow-up actions by updating the status upon data arrival or confirmation by medical staff. At this time, each incomplete state is managed together with the source data, observation time, and release condition, and is updated upon the reception of new data or confirmation by medical staff, thereby allowing a normal standby state due to linkage such as night DICOM to be distinguished from an unmatched state that has not yet been completed even though the linkage window has been closed.

[0059] In addition, medical items requiring verification or action and subsequent medical procedures can be calculated based on the detected incomplete state and the patient-specific medical context. At this time, by treating the representative image and interpretation / pathology results, for which the patient identification information matches, as the same examination case and providing them, the mixing of results from different patients or different examinations can be prevented.

[0060] Furthermore, by not automatically recording the generated information in the existing medical information system until confirmation input from medical staff is received, it is possible to prevent the system's judgment from being reflected in medical records before it is finally confirmed, and to provide safe medical support based on a human-in-the-loop structure. Brief explanation of the drawing

[0062] FIG. 1 is a block diagram illustrating the overall configuration of a medical assistance system according to one embodiment of the present invention. FIG. 2 is a structural diagram illustrating the creation and alignment structure of a single patient reception queue according to one embodiment of the present invention. FIG. 3 is a flowchart illustrating a method for determining the degree of matching of patient identification information and determining whether to combine it according to an embodiment of the present invention. FIG. 4 is a flowchart illustrating a method for determining a pathology-endoscopy correspondence and an incomplete state according to an embodiment of the present invention. FIG. 5 is a flowchart illustrating the procedure for transferring inspection result status values ​​and reflecting them after confirmation of intent according to one embodiment of the present invention. FIG. 6 is a representative flowchart illustrating the detailed flow of a medical assistance method according to one embodiment of the present invention. FIG. 7 is a block diagram illustrating the specific configuration of a medical assistance system according to one embodiment of the present invention. FIG. 8a is a drawing illustrating an actual usage screen of a medical assistance user interface (Sidecar) according to an embodiment of the present invention. FIG. 8b and FIG. 8c are drawings illustrating actual usage screens according to different arrangements of the user interface according to an embodiment of the present invention. FIGS. 9a and 9b are drawings illustrating a display screen showing the result of determining the pathology-endoscopy correspondence relationship according to an embodiment of the present invention. FIG. 9c is a drawing illustrating a display screen showing the status of the examination, DICOM, interpretation, and pathology stages according to an embodiment of the present invention. FIG. 10 is a diagram illustrating the display content for a patient whose colonoscopy tracking is imminent on a Sidecar screen according to one embodiment of the present invention. FIG. 11 is a diagram illustrating the display content for tracking glycated hemoglobin (A1c) of a diabetic patient on a Sidecar screen according to one embodiment of the present invention. FIG. 12 is a diagram illustrating the display content regarding the progression of liver enzyme levels in a patient with chronic liver disease on a Sidecar screen according to one embodiment of the present invention. FIG. 13 is a diagram illustrating the display content for managing multiple indicators of a hemodialysis patient on a Sidecar screen according to one embodiment of the present invention. Specific details for implementing the invention

[0063] Embodiments of the present invention are described in detail with reference to the attached drawings so that those skilled in the art can easily implement the invention. The present invention may be embodied in various different forms and is not limited to the embodiments described herein. In the drawings, parts unrelated to the explanation have been omitted to clearly explain the invention, and the same reference numerals are used for identical or similar components throughout the specification. Furthermore, specific descriptions of widely known prior art are omitted.

[0064] Throughout the specification, when a configuration is described as "including" another configuration, this means that, unless specifically stated otherwise, it does not exclude other configurations but may include additional configurations. Furthermore, terms such as "part" and "module" are not limited to implementations in specific hardware but should be understood to refer to functional units implemented by software, hardware, or a combination thereof.

[0066] The meanings of key terms, expressions, and abbreviations used in this specification are as follows. However, the definitions below are intended to aid the understanding of those skilled in the art to which the present invention pertains, and the specific technical meaning of each term should be interpreted according to the context or description in each text.

[0067] The term "patient medical data" as used herein may include, but is not limited to, patient identification information, medical act data regarding examinations, procedures, prescriptions, etc. scheduled or performed on the patient, and medical outcome data such as test results, interpretation results, and pathology results generated therefrom.

[0068] The term "patient medical data" as used herein encompasses all information regarding a patient scattered across multiple systems, namely patient identification information, medical procedure data, and medical outcome data. The term "medical procedure data" as used herein refers to data indicating what was decided to be done or what was done to the patient, including but not limited to test orders, procedure reservations, and execution records. The term "medical outcome data" as used herein refers to data indicating what result was obtained when such an action was actually completed, including but not limited to test results, interpretation results, and pathology results. Conventional electronic medical records (EMR) and order delivery systems (OCS) primarily configure screens and data based on medical procedure data (what was ordered), so in many cases, whether medical outcome data corresponding to that order was actually generated or received is not managed separately. The present invention is distinguished from conventional technology in that it distinguishes these two types of data and reconstructs the context based on medical outcome data.

[0069] The term "patient-specific medical context" as used herein may refer to a set of data formed by combining patient medical data acquired from different systems or acquisition paths based on the same patient and organizing them according to temporal or semantic relationships. Specifically, diagnosis (what disease the patient is suffering from) may be stored in an electronic medical record (EMR), prescription history in an order of prescription (OCS), blood test results in a laboratory database, pathology results in a pathology system, images (DICOM) in a picture archiving and transmission system (PACS), endoscopic findings in an endoscopy program, and past examination results in a health checkup system. In the present invention, "patient-specific medical context" may refer to a data structure that organizes the matters to be verified regarding this patient by grouping this information into one based on a single patient in chronological or semantic order.

[0070] The term "medical information system" used in this institution includes, but is not limited to, electronic medical records (EMR), order communication systems (OCS), picture archiving and transmission systems (PACS), radiology information systems (RIS), and DICOM (standardized image management systems), and is used as a general term for information systems within a hospital that store, manage, or transmit information regarding patient treatment, prescription, examination, or images. The related abbreviations are as follows: "EMR (Electronic Medical Record)" refers to a hospital information system that electronically manages patient medical records; "OCS (Order Communication System)" refers to a system that transmits orders from medical staff regarding examinations, procedures, medication, etc., to relevant departments; "PACS (Picture Archiving and Communication System)" refers to a system that stores and transmits medical images such as X-rays, CT scans, and MRIs; "RIS (Radiology Information System)" refers to a system that manages the scheduling and interpretation of imaging examinations; and "DICOM (Digital Imaging and Communications in Medicine)" refers to an international standard format for storing and transmitting medical images and related information.

[0071] The term "incomplete state" as used herein refers to a condition in which the result or follow-up action corresponding to a medical act scheduled or performed on a patient has not yet been completed. For example, this may include, but is not limited to, cases where a test has been performed but the result has not yet arrived (non-receipt of result); cases where the result has arrived but is not linked to the record of the medical act that caused it (mismatch between medical act and result); and cases where the result has been confirmed but subsequent action has not been taken (non-reflection of follow-up action).

[0072] The term "determination of correspondence" as used in this institution may refer to determining whether medical procedure data and medical outcome data relate to the same examination, procedure, specimen, or treatment event. For example, it may refer to determining whether a tissue biopsy performed on a specific part of the colon on a specific date (medical procedure data) and a subsequent pathology reading regarding that specific part received from the pathology system (medical outcome data) relate to the same specimen by comparing the examination date, specimen site, etc.

[0073] The term "OCR (Optical Character Recognition)" as used herein refers to a technology that recognizes characters contained in an image and converts them into character data, and can be used to acquire patient identification information or patient medical data from identification cards, documents, reception screens, handwritten reception notes, or screens of medical information systems.

[0074] As used herein, the term "Sidecar" may refer to an auxiliary user interface provided in the form of a separate monitor, screen, browser, window, or overlay without replacing an existing medical information system or altering its database structure or software, but is not limited to its name or display form.

[0075] The term "state requiring verification" as used here refers to a condition where the consistency or reliability of data does not meet standards, thereby requiring user confirmation or approval. In this case, the structure in which the system does not automatically reach a conclusion but entrusts the final judgment and approval to a human (medical staff) is also referred to here as a "human-in-the-loop" (a circular structure involving human intervention). In other words, the term "human-in-the-loop" as used here may refer to a structure in which the results of a judgment made by the system are reflected after undergoing user confirmation, modification, or approval.

[0076] As used herein, the term "refining" refers to a process of extracting or structuring numerical values, diagnostic findings, or other meaningful information from original data, and "unrefined data" may refer to the original data prior to such processing.

[0077] The terms "read-only capture" or "non-invasive integration" as used herein may refer to a method of acquiring data from an existing medical information system without recording data in the system or altering the system's database structure, network configuration, or software.

[0078] The term "reception queue" as used in this institution refers to a data set that manages patient reception information in a specific order or under certain conditions, and "single patient reception queue" may refer to a data set that integrates reception information acquired from different acquisition channels based on the same patient.

[0080] FIG. 1 is a block diagram schematically illustrating the overall configuration of a medical assistance system (100) according to an embodiment of the present invention. The medical assistance system (100) includes an acquisition unit (110), a data linkage unit (120), a context generation unit (130), an analysis unit (140), and an output unit (150). Each of the above components may be implemented together on a computing device including one or more processors and memory, or may be implemented distributed across two or more physical devices as needed.

[0081] The medical assistance system (100) can communicate with the existing medical information system (160). In addition to the Electronic Medical Record (EMR) and Order Communication System (OCS), the medical assistance system (100) can communicate with the following: a medical image management system including a Picture Archiving and Communication System (PACS) and a Radiology Information System (RIS), a pathology information system, an electrocardiogram management system (ECG management system), a laboratory information system and commissioned testing agency result files, a health checkup system, a hospital website, and a patient portal. The medical assistance system (100) is distinguished from conventional linkage methods that require direct modification of the internal structure of the existing system in that it accesses necessary data in a non-invasive manner, such as through screen-based recognition, file synchronization, or linkage with the test result database, without changing the database structure, network configuration, or software of the existing medical information system (160). In FIG. 1, the arrow pointing from the existing medical information system (160) to the acquisition unit (110) represents the data acquisition relationship, and the dotted arrow pointing from the output unit (150) to the existing medical information system (160) represents the relationship that is manually reflected after medical staff verification.

[0082] The acquisition unit (110) acquires patient identification information and patient medical data through two or more different paths, and the data linkage unit (120) calculates the degree of agreement or reliability between the acquired patient identification information to determine whether to combine them. The context generation unit (130) generates a medical context for each patient based on medical result data from the data that is allowed to combine, and the analysis unit (140) determines the correspondence between the medical procedure data and the medical result data to determine an incomplete state and calculates a subsequent medical procedure. The output unit (150) provides the above results to a separate user interface, but may not automatically record them in the existing medical information system (160) until there is confirmation input from the medical staff.

[0083] In one embodiment, the acquisition unit (110) acquires patient identification information and patient medical data through two or more different paths among screen-based recognition, file synchronization, and test result database linkage without changing the database structure, network configuration, and software of the existing medical information system (160). Data acquired independently from multiple paths is compared with each other in the data linkage unit (120) described later, thereby reducing the risk of mismatching that may occur when using only a single path.

[0084] In one embodiment, the test result database linkage may include a method of querying structured data by directly accessing the database via SQL through a database administrator account granted by the provider of the existing medical information system, and in this case, the acquisition unit (110) can acquire patient identification information or medical result data with higher accuracy and processing speed compared to screen-based recognition.

[0085] In one embodiment, the acquisition unit (110) generates current patient information by extracting one or more items such as patient name, registration number, date of birth, gender, date of visit, and date of examination from one or more of the execution screen, active window, outpatient reception queue, examination room queue, reservation information, and DICOM worklist of the existing medical information system (160). Screen-based recognition can be performed by combining string pattern matching for screen capture images, screen layout coordinate-based field recognition, or optical character recognition (OCR), without directly accessing or modifying the internal database of the existing medical information system (160).

[0086] In one embodiment, the acquisition unit (110) may acquire image examination results including CT or MRI from an image storage and transmission system (PACS) or pulmonary function test results from a pulmonary function test system and include them in the patient-specific medical data.

[0087] In one embodiment, the acquisition unit (110) applies optical character recognition to a screen capture image, an ID card image, or a handwritten document image as a specific means of screen-based recognition to extract patient identification information or medical result data (710 in FIG. 7). For example, the patient's name, date of birth, registration number, or contact information can be extracted by OCR from an ID card or a handwritten reception note image captured through a camera connected to a reception computer, and this can be combined with direct patient input through a separate touch input device. The extracted string undergoes a consistency verification with a predefined field format, and if the verification fails, the corresponding field is marked as "recognition failed" and reflected in the reliability calculation of the data linkage unit (120).

[0088] In one embodiment, the acquisition unit (110) acquires inspection results expressed in different formats by converting them into a common data structure using a file conversion method corresponding to a plurality of entrusted inspection companies or inspection equipment. For example, result files in CSV, HL7, or unique formats different for each inspection equipment, DICOM format files generated by imaging equipment, and PDF format reading reports provided by entrusted inspection agencies are each collected through the acquisition unit (110) and can be converted into an internal data structure having common fields for inspection type, measurement item, numerical value, unit, and standard range.

[0089] In one embodiment, the acquisition unit (110) creates a single patient reception queue by integrating two or more of ID card OCR, handwritten document OCR, direct input, reception list, and reservation information (Fig. 2, 210). For each integrated item, the type of input source, the time of input, and the OCR reliability are stored together. Subsequently The acquisition part (110) isThe above queue is sorted based on whether one or more of the reliability of the input source, the time of input, and the patient identification information match (220). For elderly patients, patients without accompanying documents, patients who failed OCR recognition, or foreign patients who do not follow the domestic registration number system, an alternative processing procedure involving manual verification or direct input by a reception staff member may be provided. In this case, the acquisition unit (110) may mark the item as "waiting for manual verification" and exclude it from the automatic combination target until verification is completed.

[0091] In one embodiment, the data linkage unit (120) calculates the degree of agreement or reliability between patient identification information acquired by the acquisition unit (110) through different paths, and if the degree of agreement or reliability is less than a reference value, it withholds the combination, display, or tracking management of patient medical data (Fig. 3). This is to prevent in advance the risk of mismatching and subsequent medical errors that may occur by hastily combining data acquired from different systems as belonging to the same patient.

[0092] In one embodiment, the data linkage unit (120) compares two or more of the patient name, registration number, date of birth, and test date obtained from different paths using string similarity comparison, exact match comparison, or date proximity comparison methods, and calculates an overall match or reliability score by assigning weights to the comparison results for each item (320). For example, a high weight may be assigned if the registration number matches completely, and a relatively low weight may be assigned if only the date of birth matches and the patient name meets the string similarity criteria.

[0093] In one embodiment, if there are cases such as identical names, discrepancies in registration numbers, optical character recognition failures, discrepancies in examination dates, or discrepancies between reservation information and reception information, the data linkage unit (120) does not perform automatic linkage and requests user confirmation by displaying a list of candidate patients with possible matches and discrepancies in the user interface (step 340 after the “less than” branch in step 330 of FIG. 3). For example, if two patients with the same name have different dates of birth, the data linkage unit (120) may display both candidates and withhold automatic linkage of the corresponding examination results until it receives a selection input from a medical staff member or reception staff member.

[0094] In one embodiment, the data linkage unit (120) may display items with a calculated match or reliability below a reference value as "needs verification" (step 340 of FIG. 3). Items displayed as "needs verification" may be separately displayed on the user interface by the output unit (150) (see 802 in FIG. 8a - labeled as "waiting for patient recognition"), and until a verification input from a medical staff is received, they are excluded from subsequent processing by the context generation unit (130) and the analysis unit (140), or even if processed, they are not treated as confirmed results.

[0095] In one embodiment, the acquisition unit (110) sorts the single patient reception queue based on whether one or more of the reliability of the input source, the time of input, and the patient identification information match (step 220 of FIG. 2). For example, items that have been matched with the registration number and compared with the reservation information are placed at the top of the queue, and items with low OCR recognition reliability or those requiring manual verification are separately distinguished and placed at the bottom or highlighted at the top with a "verification needed" indication. However, in another embodiment, the process may be carried out in the data linkage unit (120).

[0097] In one embodiment, the context generation unit (130) can generate a patient-specific medical context from patient medical data that is allowed to be combined by the data linkage unit (120), using medical result data as reference data. In one specific embodiment, the context generation unit (130) generates a patient-specific medical context linked to the patient's disease information, prescription history, or procedure history, using medical result data that includes one or more of actually generated test results, reading results, or pathology results as reference data, independently of whether order information for tests or prescriptions exists, from patient medical data that is allowed to be combined by the data linkage unit (120).

[0098] In one embodiment, the medical result data may include one or more of blood test results, pathology test results, medical imaging results, endoscopy results, ultrasound results, bone density test results, fundus examination results, and health checkup results.

[0099] In one embodiment, the context generation unit (130) analyzes the numerical value, reference range, increase / decrease status, or interpretation findings included in the medical result data and structures it into one or more of normal or abnormal, whether follow-up is required, and candidate diagnosis names. For example, for blood test results, the measured value, unit, reference range, direction of increase / decrease compared to the previous test, and whether re-examination is required are extracted, and for bone density test results, the T-score for each measurement site is recognized and structured into osteopenia, osteoporosis, or whether follow-up examination is required. In the case of a test type in which the interpretation text is included in the image itself, such as a fundus examination, the context generation unit (130) does not treat the original image as a separate image interpretation target, but extracts the interpretation result text accompanying the image and reflects it in the test history and medical summary.

[0100] In one embodiment, the context generation unit (130) generates different patient-specific medical contexts for the same medical outcome data based on one or more of the patient's medical condition, test history, and concomitant medications. For example, even with the same creatinine test result, the context generation unit (130) can generate different tracking management items and medical summary phrases for a patient who has diabetes and hypertension as underlying conditions and is taking a diuretic in combination, and for a patient who has no specific underlying conditions.

[0101] In one embodiment, the context generation unit (130) may evaluate the medical context for each patient according to the treatment guidelines and generate one or more of the following: drug review information for reviewing interactions or contraindications between concomitant drugs, necessary test guidance information for guiding test items recommended in the guidelines but not yet performed, and today's check information for organizing items required to be checked during the treatment on the day.

[0103] In one embodiment, the analysis unit (140) determines the correspondence between medical act data indicating an examination or procedure scheduled or performed on a patient and medical result data generated in response to said medical act, and determines one or more incomplete states among result not received, medical act-result mismatch, or follow-up action not reflected based on said correspondence, and calculates one or more medical items requiring verification or action based on said incomplete state and patient-specific medical context, and the type of follow-up medical act and the time of execution corresponding thereto.

[0104] In one embodiment, the analysis unit (140) determines the correspondence between medical procedure data and medical result data using two or more of the examination date, specimen site, lesion location, and whether a biopsy was performed (Fig. 4, 410). The medical procedure data and medical result data to be determined do not necessarily have to be generated in the same system.

[0105] In one embodiment, medical procedure data includes information on tissue biopsy performed during an endoscopic examination, and medical result data includes pathology test results. The analysis unit (140) compares the specimen site information with the pathology diagnosis information to determine the correspondence (410 in FIG. 4), and as a result, distinguishes between the following two states. If the pathology test result corresponding to the tissue biopsy information is not received, it is determined to be in a "pathology result mismatch" state (420 in FIG. 4), and if the pathology test result exists but is not linked to the corresponding endoscopic examination information, it is determined to be in a "reading mismatch" state (430 in FIG. 4). If both are confirmed, it is determined to be in a "matching completed" state, and whether tracking is necessary is calculated (440 in FIG. 4).

[0106] In one embodiment, the analysis unit (140) selects at least one candidate for a chronic disease, including but not limited to diabetes, hypertension, hyperlipidemia, chronic hepatitis B, etc., based on the patient's disease information or candidate diagnosis name, and compares the final execution date and recommended cycle of the essential follow-up test item corresponding to the selected chronic disease with the current time to determine whether the recommended cycle has been exceeded, thereby determining that the state is incomplete.

[0107] In one embodiment, the non-reception of results includes a state in which an examination or procedure has been performed but the corresponding medical result data has not been received, the medical act-result mismatch includes a reading mismatch state in which medical result data exists but is not linked to the corresponding medical act data, and the non-reflection of follow-up measures includes a state awaiting physician confirmation before the calculated follow-up medical act is reflected in the tracking management list after confirmation by medical staff.

[0108] In one embodiment, the test result item progresses from a result non-receipt state (510) through result reception (520), unpurified state (530), and purification completed state (540) to a waiting state for physician confirmation (550), and can transition to a physician confirmation completed state (560) and a tracking management reflection state (570) after confirmation by medical staff (Fig. 5). This state transition exemplifies the entire processing process from the time the test result item is acquired until it is finally reflected in tracking management. Among these, the result non-receipt state (510) and the waiting state for physician confirmation (550) correspond to the result non-receipt and follow-up action non-reflection described above, respectively, and the reading Mismatching The state corresponds to the embodiment described in FIG. 4. When a new result is obtained, the above state and the time point may be updated.

[0110] In one embodiment, when new medical procedure data or medical result data is acquired, the analysis unit (140) updates at least one of the incomplete state, the medical item, or the reference point for the performance of the subsequent medical procedure.

[0111] In one embodiment, the analysis unit (140) calculates a change trend by comparing current medical result data with past medical result data and reflects this in the verification priority of the corresponding medical item.

[0112] In one embodiment, the subsequent medical actions calculated by the analysis unit (140) include one or more of follow-up examinations, re-examinations, pathology tracking, vaccinations, and outpatient follow-up visits, and the calculated subsequent medical actions may be displayed as items requiring tracking in the area displaying the medical summary (804 in FIG. 8a).

[0113] In one embodiment, the analysis unit (140) calculates the priority of subsequent medical procedures by considering the recommended date for the follow-up examination, the examination cycle, the history of previous examinations, and whether the doctor has confirmed the results.

[0115] In one embodiment, the output unit (150) provides the medical items and subsequent medical procedures calculated by the analysis unit (140) to a user interface that is distinct from the existing medical information system (160), and does not automatically record the information generated by the analysis unit (140) in the existing medical information system (160) until a confirmation input from a medical professional is received (see the dotted arrow pointing from the output unit (150) in FIG. 1 to the existing medical information system (160).

[0116] In one embodiment, the output unit (150) reflects the corresponding subsequent medical act in the patient-specific tracking management list only when an approval, modification, hold, or exclusion input is received from the medical staff (steps 550 to 570 of FIG. 5).

[0117] In one embodiment, the user interface generated by the output unit (150) includes a browser screen, a pop-up screen, or an overlay screen ("Sidecar") that is displayed separately from the screen of the existing medical information system (160). FIG. 8 illustrates an example of actual use of such a Sidecar screen (800), and this part will be explained in more detail in the following embodiments.

[0118] In one embodiment, the user terminal may include a medical staff terminal used by medical staff and a patient terminal used by a patient. A medical assistance user interface (Sidecar) generated by the output unit (150) may be displayed on the medical staff terminal, and a patient screen accessible through the hospital homepage and patient portal may be displayed on the patient terminal. Each terminal may be implemented as a known computing device, such as a desktop, laptop, tablet, or smartphone, comprising a processor, memory, a display, and a communication module.

[0119] In one embodiment, the output unit (150) displays the purified result, tracking test candidate, and abnormal finding candidate on the user interface, and distinguishes the unpurified data or original image so that it is processed on a separate confirmation screen.

[0121] Referring to FIG. 6, a patient-specific medical data-based treatment assistance method according to one embodiment of the present invention may be performed by one or more processors. The treatment assistance method may include: a step (610) of acquiring patient identification information and patient medical data through two or more different paths without changing the database structure, network configuration, and software of an existing medical information system; a step (620) of calculating the degree of agreement or reliability between the patient identification information acquired through the different paths and determining whether to combine the patient medical data based thereon; a step (630) of generating a patient-specific medical context using medical result data as reference data from the patient medical data for which combination is permitted; a step (640) of determining the correspondence between medical act data and the medical result data, determining an incomplete state based on the correspondence, and calculating medical items requiring verification or action and subsequent medical acts based on the incomplete state and the patient-specific medical context; and a step (650) of providing the medical items and subsequent medical acts to a user interface that is distinct from the existing electronic medical record or prescription delivery system.

[0122] In the step (610) of acquiring patient identification information and patient medical data, the patient identification information and patient medical data may be acquired through two or more different paths among screen-based recognition, file synchronization, and test result database linkage. The patient medical data may include, but is not limited to, medical action data regarding tests, procedures, or prescriptions scheduled or performed on the patient, and medical result data such as test results, reading results, or pathology results generated in response to said medical action.

[0123] In the step (620) of determining whether to combine, two or more of the patient name, registration number, date of birth, date of examination, or other patient identification information obtained through different paths may be compared. If the calculated agreement or reliability is greater than or equal to a predetermined threshold value, the combination of the patient medical data may be permitted, and if it is less than the threshold value, the combination, display, or tracking management reflection may be withheld. In this case, the data may be displayed as requiring verification and may not be processed as confirmed patient medical data until a user's verification input is received.

[0124] In the step (630) of generating a patient-specific medical context, the medical result data actually generated among the patient medical data allowed to be combined may be used as reference data. The said medical result data may be configured as a patient-specific medical context by linking it with the patient's disease information, prescription history, test history, or procedure history, independently of whether or not order information for tests or prescriptions exists. Accordingly, information distributed across different medical information systems or acquisition paths may be linked temporally or semantically based on the same patient.

[0125] In the step (640) of calculating medical items and subsequent medical acts, it may be determined whether medical act data representing an examination or procedure scheduled or performed on a patient and medical result data generated in correspondence therewith relate to the same examination, procedure, specimen, or medical treatment event. Based on the result of the above determination, one or more incomplete states may be determined, such as a result non-receipt state where medical result data is not received, a medical act-result mismatch state where medical result data and medical act data are not linked, or a follow-up action non-reflection state where confirmation or follow-up action based on the medical result is not reflected.

[0126] Additionally, in the above step (640), one or more medical items requiring verification or action are identified based on the determined incomplete state and the patient-specific medical context, and the type of subsequent medical action and the timing of its execution, such as follow-up examination, re-examination, pathology tracking, vaccination, or outpatient follow-up visit corresponding to each medical item, may be calculated. The subsequent medical action and the timing of execution may be calculated by further considering one or more of the patient's medical condition, past examination history, trends in changes to medical result data, examination cycle, or confirmation by medical staff.

[0127] In the step (650) of providing to the user interface, the medical item, incomplete status, and subsequent medical act may be provided on a browser screen, pop-up screen, overlay screen, or separate terminal screen that is displayed separately from the screen of the existing electronic medical record or prescription delivery system. The information calculated in the step (640) may not be automatically recorded in the existing electronic medical record or prescription delivery system until a confirmation input from the medical staff is received, and the information may be reflected in the patient-specific tracking management list when an approval, modification, hold, or exclusion input is received from the medical staff.

[0128] As illustrated by the dotted line in FIG. 7, when new patient medical data is acquired, at least some of the steps from determining whether to combine (620) to providing to the user interface (650) may be performed again. Accordingly, at least one of the consistency or reliability of patient identification information, patient-specific medical context, incomplete status, medical item, subsequent medical act, or the reference point for its performance may be updated to reflect the newly acquired data.

[0129] Meanwhile, the order of execution of each step illustrated in FIG. 7 is merely an example for convenience of explanation, and within the scope of not departing from the technical concept of the invention, some steps may be performed in parallel, two or more steps may be integrated and performed, or some steps may be repeated.

[0131] In one embodiment, the medical assistance system (100) can be implemented with four functional layers as shown in FIG. 7.

[0132] The input layer (Layer A, 710) includes an ID card shooting recognition module using a camera, a shooting recognition module for a handwritten receipt note, and a dedicated touch input terminal, and implements screen-based recognition and direct input functions of the acquisition unit (110).

[0133] The patient queue server layer (Layer B, 720) implements the functions of the acquisition unit (110) and the data linkage unit (120) by including an integrated patient queue engine that maintains a single patient reception queue arranged in chronological order, records the input method and reliability of each item, and manages the reception status.

[0134] The AI ​​workflow engine layer (Layer C, 730) implements the functions of the context generation unit (130) and the analysis unit (140) by including a module that collects screen recognition data, a module that combines examination, diagnosis, medication, and vaccination history through local inquiry to construct a patient-specific context, and a module that generates a clinical summary including tracking of abnormal results and vaccination recommendations.

[0135] The AI ​​assistance screen layer (Layer D, 740) in the examination room implements the function of the output unit (150) by displaying patient summaries, test trends, detection and tracking notifications of missing explanations, but not automatically recording them in the existing electronic medical record.

[0136] The existing EMR system (750) captures data only in a read-only manner and does not modify its internal structure. The four layers may be implemented together within a single computing device, or they may be implemented separately considering processing performance, network status, inspection type, or image data capacity.

[0138] Hereinafter, specific embodiments actually implemented according to the medical assistance system (100) and the method described in FIGS. 1 to 7 are described. The specific embodiments below are not intended to imply that each component must be implemented in the form shown below, but are intended to provide an example in which a person of ordinary knowledge can implement the technical concept of the claim in an actual clinical environment.

[0140] 1. Specific Example 1: Example of implementation of a system for actual use (Fig. 8)

[0142] FIG. 8a illustrates the basic framework of a user interface (Sidecar, 800) generated by the output unit (150), including, but not limited to, an area (802) displaying patient recognition status and patient basic information, an area (804) displaying medical summary and items requiring tracking, an area (806) displaying test history, an area (808) displaying drug review information, an area (810) displaying surgical history, an area (812) displaying cautionary drugs, an area (814) displaying blood test history by date, an area (816) summarizing so that it can be copied to a medical chart after verification by medical staff, an area (818) displaying test results, an area (820) displaying National Health Insurance Service checkup subjects, an area (822) displaying KDCA vaccination history, and an area (824) displaying a procedure confirmation certificate. FIG. 8a illustrates the state before a patient is selected, and a message indicating that the relevant information has not yet been displayed is displayed in each area.

[0143] The arrangement, size, and display order of each area illustrated in FIG. 8a are exemplary and may be changed arbitrarily according to the user's settings or selections. In one embodiment, each area may be moved to a different location on the screen or incorporated into another area by the user's drag-and-drop input, and such arrangement changes may be saved as settings of individual users (medical staff) and maintained upon subsequent access. Therefore, even if areas 814, 820, and 822 are displayed as being located inside area 804, this may not be because 814, 820, and 822 are fixed sub-components of 804, but rather as a result of the user moving and arranging the corresponding areas to adjacent locations of area 804 according to their convenience.

[0144] FIGS. 8b and FIGS. 8c illustrate actual usage examples in which the arrangement is adjusted as above by the user for different patients.

[0145] Referring to Fig. 8b, the 802b area displays basic patient information (registration number masked, gender / age, date of birth), and the 804b area displays the number of problems, a diagnosis category badge (e.g., "hyperlipidemia"), and a finding headline under the heading "Check Today." When "View Evidence" is expanded, a detailed table consisting of the current problem, currently identified medication, guideline recommendations, safety / exceptions, and evidence is displayed. Below this, a secondary warning message (e.g., "Need to check for deterioration of renal function"), a list of major disease candidate badges, and a highlight bar indicating items requiring follow-up are displayed. Subsequently, the 814b area (displaying recent blood test results compared to previous values), the 820b area (eligible for National Health Insurance Service check-ups), and the 822b area (vaccination) are displayed adjacent to the 804b area. In area 806b, a list of test results is displayed, in area 808b, a list of medication reviews is displayed, in area 810b, a surgical history is displayed, and in area 812b, cautionary drugs are displayed.

[0146] Referring to FIG. 8c, areas of the same type as those in FIG. 8b are displayed, but the display order and frame configuration differ from FIG. 8b as a result of arrangement adjustments by the user. Specifically, an upper frame (804c-1) containing today's check, secondary warning phrases, and major disease candidates, and a lower frame (804c-2) displaying items requiring tracking are placed as separate frames, and area 814c is moved to be located between the two frames. Areas 820c and 822c are placed adjacent to the bottom of area 804c-2, and surgical history and cautionary medications are displayed in areas 810c and 812c, respectively, as in FIG. 8b.

[0147] Meanwhile, in the area 806c of FIG. 8c, a “pathology result mismatch” status badge is displayed for items where tissue biopsy information has been confirmed but the corresponding pathology test result has not yet been received, and a specific example thereof is described in detail in FIG. 9 below. In contrast, in the area 806b of FIG. 8b, there are no such items with incomplete matching, and all displayed test items are in a state where matching has been completed or confirmed as normal findings.

[0148] As such, FIGS. 8b and 8c demonstrate that display areas of the same type (804, 814, 820, 822, etc.) can be rearranged into different positions, order, and frame structures according to the user's drag-and-drop operation. This supports the idea that even if each display area is depicted as being located inside another display area, this does not imply a fixed inclusion relationship between the areas, and that the arrangement of each display area does not limit the technical features of the present invention.

[0150] 2. Example 2: Actual usage screen of the result of determining the pathology-endoscopy correspondence (Fig. 9a)

[0152] FIGS. 9a and 9b are examples of screens in which the results of determining the pathology-endoscopic correspondence calculated by the analysis unit (140) are actually displayed, and correspond to examples of detailed contents of the area (806) displaying the examination history of FIG. 8.

[0153] Referring to FIG. 9a, a list of examination history (900) is displayed at the top of the screen in the order of most recent examinations, and each item consists of text (902) including the examination date, examination name, and interpretation / diagnostic findings. A badge (904) indicating the interpretation or diagnostic findings is attached to each item, for example, a badge such as "Gastritis" may be displayed for the gastroscopy item. A representative image (906) corresponding to the examination is displayed to the right of each item, and if there are multiple images for a single examination, multiple representative images (906) may be displayed together. An item selected by the user is given a selection mark (908) that distinguishes it from other items, and when the selection mark (908) is selected, detailed information corresponding to it is displayed in the selection item detail area (910).

[0154] The above-mentioned selection item detail area (910) includes an area (912) displaying the examination status, an area (914) displaying the readings, an area (916) displaying abnormal findings and corresponding examination images together, and an area (918) displaying follow-up action candidates calculated by the system.

[0155] FIG. 9a illustrates a case where the matching of tissue biopsy information and pathology test results is completed. In the area (912) displaying the test status, a phrase such as "Matching completed — tissue biopsy information and pathology test results are linked" may be displayed. In the area (914) displaying the reading content, the corresponding pathology test result is displayed in text along with the reading findings, and if the matching is completed, the actual pathology diagnosis name, such as "Pathology: Chronic gastritis with intestinal metaplasia," may be displayed. In the area (916) displaying the abnormal findings and the test image together, the location of the lesion is displayed on the test image, and the corresponding reading text is presented together, so that the reading text and the image are combined into a single context. As the reading text, test image, pathology result, and status value are combined and presented on a single screen in this way, medical staff can immediately recognize whether the test, image, reading, and pathology have been received and whether they are linked with the corresponding test information.

[0156] Meanwhile, although gastroscopy is illustrated as an example in Fig. 9a, this is merely illustrative and can be applied in the same way to other types of examinations, such as upper abdominal ultrasound and colonoscopy.

[0157] FIG. 9b illustrates a detailed selection area (910) for cases where tissue biopsy information has been confirmed but the corresponding pathology test result has not yet been received, and represents a case where the item selected (908) in FIG. 9a is in an incomplete matching state. In this case, a “Pathology result not matched” badge (904b) is attached to the item, and a phrase such as “Pathology result not matched — tissue biopsy information has been confirmed but the corresponding pathology test result has not yet been received” may be displayed in the area (912b) indicating the test status. A phrase such as “Pathology: Waiting for result” may be displayed in the area (914b) indicating that the pathology test result has not yet been received. At this time, a notification message regarding the delay in receiving pathology results may be displayed in the area (918b) displaying the system action candidate, and a message indicating that confirmation by medical staff is required may be presented, such as, "Notification of non-receipt of pathology results ― Please check after 3 business days."

[0159] In another specific example, for instance, if an image examination order exists but the scheduled night DICOM linkage window of the medical institution has not closed, the analysis unit (140) determines a DICOM linkage waiting state, and if a corresponding Study is not confirmed even after the linkage window closes, it may be updated to a DICOM mismatched state. If a Study is confirmed but a confirmed reading is not received, a reading mismatched state may be determined. Regarding endoscopic examinations, whether a tissue biopsy is performed is checked using the prescription code or prescription name of the structured prescription list for the relevant day, and if information on performing a tissue biopsy exists but there is no pathology result, a pathology result mismatched state may be determined. If multiple pathology results or multiple examination rows conflict and a single corresponding relationship cannot be determined, automatic linkage may be suspended and a matching ambiguous state may be displayed.

[0160] FIG. 9c is a diagram illustrating how the status of each stage described above is actually displayed on the user interface. Each item consists of an examination date, an examination name, and a status badge, and the basis for determining the corresponding status is noted next to the status badge. For example, DICOM linkage waiting (920) may be displayed when an image examination order exists but the night DICOM linkage window has not yet closed; DICOM mismatch (922) may be displayed when a corresponding Study is not confirmed even after the linkage window has closed; interpretation mismatch (924) may be displayed when a Study is confirmed but a confirmed interpretation report is not received; pathology result mismatch (926) may be displayed when tissue biopsy information is confirmed but a corresponding pathology test result is not received; and matching ambiguity (928) may be displayed when multiple pathology results or examination rows conflict and a single corresponding relationship is not confirmed, but is not limited thereto.

[0161] Each state may be linked to a source data source, the time of the first observation, the time of the last re-query, and a release condition; it transitions to matching completion upon the receipt of new results or confirmation by medical staff, and the state transition record may be preserved even if the corresponding warning is released in the Today Check or Examination Timeline of the user interface.

[0163] 3. Specific Example 3: Pathology-endoscopic matching and follow-up endoscopy recommendation Sidecar indication after colon polyp resection (Fig. 10)

[0165] When a tracking item (e.g., "Colonoscopy pathology: 2025-09-03 Confirmation of colon adenoma / LGD (LGD 2, EMR 1) · 1st year notification") displayed in the 804c-2 area of ​​FIG. 8c is selected, a detailed screen as shown in FIG. 10 may be displayed. However, this is merely an example, and the information shown in FIG. 10 does not necessarily have to be displayed on a separate screen from 804c-2; it may also be displayed in an expanded form within the 804c-2 area itself, or provided together by being included in the summary text displayed in 804c-2. That is, the information displayed by each area shown in FIG. 10 may be provided in any of the following ways: on a separate screen, in an expanded sub-area, or in an abbreviated form within the summary text, and the display form does not limit the technical features of the present invention.

[0167] During a colonoscopy, resection is performed on two polyps measuring 1.2 cm and 0.8 cm in size in the sigmoid colon, and information regarding the corresponding biopsy is stored as medical procedure data. Subsequently, when a pathology test result is obtained from the pathology system and interpreted as tubular adenoma with no dysplasia, the analysis unit (140) matches the pathology test result with the corresponding endoscopy history based on the examination date and lesion location according to the procedure described above in the present invention (matching completed). After the matching is completed, the analysis unit (140) compares the recommended follow-up period corresponding to the histological characteristics of the tubular adenoma with the date of the most recent endoscopy to determine whether the follow-up period has been exceeded.

[0168] FIG. 10 is a diagram illustrating how the above judgment results are actually displayed for a specific patient through the Sidecar screen. In the patient identification area (1010), the patient name and registration number are displayed in an unidentified manner. In the examination history area (1012), the date of the most recent colonoscopy and the elapsed period are displayed, and in the findings summary area (1014), the location, number, and size of the resected polyps are displayed, and in the pathology results area (1016), the pathology test results are displayed.

[0169] In the follow-up recommendation area (1018), a follow-up colonoscopy recommendation calculated by the analysis unit (140) and a warning indication regarding whether the reference time has been exceeded are provided. In the system action candidate area (1020), a follow-up endoscopy appointment notification text message sending candidate calculated by the analysis unit (140) is displayed along with the phrase “Doctor verification required” attached next to it. The “Doctor verification required” indication may indicate that the item is an unconfirmed candidate that is not reflected in the patient-specific follow-up management list until it undergoes approval, modification, suspension, or exclusion input by the medical staff. If only information on tissue biopsy execution exists and the corresponding pathology test results have not yet been received, the analysis unit (140) displays this as a “Pathology result mismatch” state, and in this case, the pathology result area (1016) of FIG. 10 may display the status phrase “Pathology result mismatch” instead of the pathology test results.

[0171] 4. Example 4: Detection of exceeding the glycated hemoglobin (A1c) retest cycle in diabetic patients and Sidecar indication (Fig. 11)

[0173] As explained in Example 3, regarding the following Examples 4 to 6, the information shown in FIGS. 11 to 13 is not limited to being displayed on a separate screen from the area indicating the need for tracking in FIG. 8, but may also be displayed in an expanded manner within the area itself or provided together by being included in a summary text.

[0175] Similar to FIG. 10, if an item related to a glycated hemoglobin retest is displayed in the area indicating the need for tracking in FIG. 8b or FIG. 8c, selecting it may display a detailed screen as shown in FIG. 11. That is, for a patient with underlying conditions of type 2 diabetes, hypertension, and dyslipidemia, if the recommended retest cycle has elapsed since the recent glycated hemoglobin (A1c) measurement date but the retest has not been performed, the analysis unit (140) determines this as an incomplete state and displays the "A1c retest needed" item at the top of the verification priority list, as shown in FIG. 11.

[0176] FIG. 11 is a diagram illustrating how the information calculated as above is displayed through the Sidecar screen. In the patient identification area (1110), the patient name and registration number are displayed in an anonymized manner. In the chronic disease display area (1112), the patient's underlying diseases, such as diabetes, hypertension, and dyslipidemia, are displayed. In the test trend area (1114), the recent A1c measurement date and measurement value are displayed, and in the status display area (1116), a warning message indicating that the measurement cycle has been exceeded and an immediate re-examination is recommended is displayed.

[0177] In the parallel test value area (1118), the fasting blood glucose level measured on the day and its direction of increase are displayed together, thereby providing the analysis unit (140) with parallel indicators that support the need for an A1c re-examination. In the missing items area (1120), the analysis unit (140) displays the remaining items that have exceeded the recommended cycle among the essential follow-up test items corresponding to the patient's disease information, namely microalbuminuria and fundus examination. In the system action candidate area (1122), the recommendation to add an A1c order and the creation of a fundus examination recall candidate are displayed along with the phrase "Physician confirmation required," and are reflected in the patient-specific follow-up management list only if there is an approval input from the medical staff.

[0179] 5. Specific Example 5: Sidecar display reflecting the progression and monitoring priority of liver enzyme levels in patients with chronic liver disease (Fig. 12)

[0181] In the same way, for a patient diagnosed with chronic hepatitis B and liver fibrosis (F2), if the AST and ALT levels measured over the last three times show a continuously rising trend, the analysis unit (140) calculates the increasing trend compared to the previous test and adjusts the priority of verification of the item upward, as illustrated in FIG. 12.

[0182] FIG. 12 is a diagram illustrating how the information calculated as above is displayed through the Sidecar screen. In the patient identification area (1210), the patient name and registration number are displayed in an unidentified manner. In the chronic disease display area (1212), chronic hepatitis B and liver fibrosis are displayed. In the test trend area (1214), AST and ALT values ​​measured at three different points in time are displayed in a time series, and a continuous rise is indicated next to them. In the parallel indicator area (1216), AFP values ​​are displayed along with a judgment that they are within the normal range and have no significant change, thereby displaying the rising trend of AST and ALT and the stable AFP in contrast.

[0183] In the imaging examination history area (1218), recent abdominal ultrasound findings are displayed, and in the follow-up recommendation area (1220), recommendations for abdominal ultrasound follow-up examinations calculated by the analysis unit (140) and warning indications regarding the arrival of the reference time are provided. In the system action candidate area (1222), candidates for abdominal ultrasound appointment guidance and the creation of a liver specialist consultation review memo are displayed along with the phrase "Doctor verification required." For the numerical changes at three points in time displayed in the examination trend area (1214), the analysis unit (140) calculates the trend of change by comparing current medical result data with past medical result data and reflects this in the verification priority.

[0185] 6. Specific Example 6: Sidecar display reflecting integrated management of multiple indicators and verification priority for hemodialysis patients (Fig. 13)

[0186] In the same way, for a patient undergoing hemodialysis (3 times a week, Monday, Wednesday, Friday) for chronic renal failure, when multiple clinical indicators such as dialysis adequacy, anemia, iron metabolism, bone metabolism, nutrition, and vascular access status need to be managed together, the analysis unit (140) calculates whether each indicator is satisfied relative to the target value and raises the priority of verification for the items that need verification among them, as illustrated in FIG. 13.

[0187] FIG. 13 is a diagram illustrating how the information calculated as above is displayed through the Sidecar screen. In the patient area (1310), the patient name, registration number (de-identified display), and the day of the week for hemodialysis are displayed. In the causative disease / dialysis history area (1312), the disease causing the dialysis and the elapsed time since the start of dialysis are displayed.

[0188] In the dialysis adequacy area (1314), Kt / V and URR values ​​calculated based on the average of the last three times are displayed along with target values, and a judgment regarding whether the target has been met is noted. In the anemia management area (1316), hemoglobin levels are displayed along with the target range and the direction of deviation, along with the dosage and administration cycle of the hematopoietic agent currently being administered. In the iron indicator area (1318), ferritin and transferrin saturation (TSAT) levels are displayed along with the direction of deviation relative to the target value. In the bone metabolism area (1320), calcium, phosphorus, parathyroid hormone (iPTH), and vitamin D levels are displayed along with the direction of deviation from their respective normal ranges. In the nutritional assessment area (1322), albumin levels and changes in dry body weight over the recent period are displayed. In the vascular access / infection area (1324), the presence or absence of abnormalities in vascular access, such as arteriovenous fistulas or artificial blood vessels, CRP levels, and a history of fever are displayed.

[0189] In the tracking arrival area (1326), tracking items for which a reference time has arrived are displayed, such as the elapsed period of regular follow-up examinations like echocardiography and confirmation of hepatitis B antibody titers, calculated by the analysis unit (140). In the system action candidate area (1328), action candidates calculated by the analysis unit (140) by synthesizing multiple indicators are displayed. For example, by considering the iron deficiency findings calculated in the iron indicator area (1318) and the reduced response to hematopoietic agents calculated in the anemia management area (1316) together, an action candidate based on the relationship between items that cannot be derived from a single indicator alone, such as "Prioritize iron deficiency correction before increasing the hematopoietic agent dosage," may be displayed along with the phrase "Physician confirmation required." The above system action candidates are not established solely by the values ​​of individual test items, but are calculated based on the relationship between multiple items, such as the increase in phosphorus levels in the bone metabolism area (1320).

[0191] 7. Specific Example 7: Single reception queue processing for elderly patients without identification (Figs. 2 and 3)

[0193] When an elderly patient who does not have an identification card visits, the receptionist records the name, date of birth, and contact information based on the patient's statement in a handwritten registration notebook, and an image of this is recognized through the acquisition unit (110). However, if part of the date of birth is not recognized due to a decrease in the handwriting recognition rate, the acquisition unit (110) creates the item in a "waiting for handwritten verification" state (210 in FIG. 2) and excludes it from the automatic combination target until it is supplemented through direct input by the receptionist. Once the supplementation is completed, the data linkage unit (120) compares it with the same patient information registered in the EMR registration list through a queue alignment procedure (220 in FIG. 2), and if it is greater than or equal to a threshold value, it normally transmits it to the match determination procedure of FIG. 3 (320 to 330 in FIG. 3).

[0195] 8. Example 8: Combination withheld due to discrepancy in patient identification information (Fig. 3)

[0197] When a test is performed on one of two patients with the same name, it may be confirmed that the registration number on the reception screen obtained through path A (screen-based recognition) and the registration number listed in the test equipment output file obtained through path B (file synchronization) are different from each other. In this case, the data linkage unit (120) detects that the name matches but the registration number does not match, and since the score calculated by the data linkage unit (120) is below the reference value (step 330 of FIG. 3, “less than”), the data linkage unit (120) marks the test result as needing verification and withholds automatic combination and tracking management reflection (step 340 of FIG. 3). When the reception staff or medical staff confirm and select the correct patient, the data linkage unit (120) confirms the combination according to the verification result.

[0199] The embodiments described above are merely illustrative of the technical concept of the present invention, and those skilled in the art will be able to make various modifications and variations within the scope of the essential characteristics of the present invention without departing from its nature. Accordingly, the scope of protection of the present invention should be determined by the following claims and should not be limited by the above embodiments.

Claims

Claim 1 A patient-specific medical data-based clinical assistance system comprising one or more processors, comprising: an acquisition unit that acquires patient identification information and patient medical data through two or more different paths among screen-based recognition, file synchronization, and test result database linkage without changing the database structure, network configuration, and software of an existing medical information system; a data linkage unit that calculates the degree of agreement or reliability between the patient identification information acquired through the said different paths, and withholds the combination, display, or tracking management reflection of said patient medical data if the degree of agreement or reliability is less than a reference value; a context generation unit that generates a patient-specific medical context from the patient medical data allowed to be combined by the data linkage unit, using medical result data as reference data; a determination of the correspondence between medical act data representing an examination or procedure scheduled or performed on a patient and said medical result data generated in response to said medical act, and based on said correspondence, determines one or more incomplete states among non-receipt of results, medical act-result mismatch, or non-reflection of follow-up measures, and based on said incomplete state and said patient-specific medical context, one or more medical items requiring verification or action, and the type and performance of the corresponding follow-up medical act. A medical assistance system comprising: an analysis unit for calculating a reference point; and an output unit that provides the medical item and the subsequent medical act to a user interface distinct from the existing medical information system, and does not automatically record the information generated by the analysis unit into the existing medical information system until confirmation input from a medical professional is received. Claim 2 A medical assistance system according to claim 1, wherein the acquisition unit generates current patient information by acquiring one or more of a patient name, registration number, date of birth, gender, date of visit, and date of examination from one or more of the screen, active window, reception queue, examination room queue, reservation information, and DICOM worklist of the existing medical information system. Claim 3 A medical assistance system according to claim 1, wherein the acquisition unit extracts patient identification information or medical result data by applying optical character recognition to a screen capture image, an ID card image, or a handwritten document image as the screen-based recognition. Claim 4 A medical assistance system according to claim 1, wherein the acquisition unit integrates two or more of ID card OCR, handwritten document OCR, direct input, reception list, and reservation information to generate a single patient reception queue. Claim 5 A medical assistance system according to claim 4, wherein the data linkage unit or acquisition unit sorts the single patient reception queue based on whether one or more of the reliability of input source, input time, and patient identification information match. Claim 6 A medical assistance system according to claim 1, wherein the data linkage unit calculates the degree of agreement or reliability by comparing two or more of the patient name, registration number, date of birth, and examination date obtained from different paths. Claim 7 A medical assistance system according to claim 1, characterized in that the medical result data includes one or more of blood test results, pathology test results, medical imaging results, endoscopy results, ultrasound results, bone density test results, fundus examination results, and health checkup results. Claim 8 A medical assistance system according to claim 1, wherein the context generation unit analyzes numerical values, reference ranges, whether an increase or decrease, or interpretation findings included in the medical result data to structure them into one or more of normal or abnormal, whether follow-up is required, and candidate diagnosis names. Claim 9 A medical assistance system according to claim 1, wherein the context generation unit generates different patient-specific medical contexts for the same medical outcome data based on one or more of the patient's medical condition, examination history, and concomitant drugs. Claim 10 A medical assistance system according to claim 1, characterized in that the analysis unit determines the correspondence between the medical procedure data and the medical result data using two or more of the examination date, specimen site, lesion location, and whether a biopsy was performed. Claim 11 A medical assistance system according to claim 10, wherein the medical procedure data includes information on tissue biopsy performed during an endoscopic examination, the medical result data includes pathology test results, and the analysis unit distinguishes and determines from each other a pathology result mismatch state when a pathology test result corresponding to the tissue biopsy information is not received; and a matching ambiguity state when a pathology test result exists but the corresponding endoscopic examination information is not definitively determined. Claim 12 A medical assistance system according to claim 1, wherein the analysis unit selects a candidate for a chronic disease based on the patient's disease information or candidate diagnosis, and determines whether the recommended cycle has been exceeded by comparing the final execution date and recommended cycle of the essential follow-up test item corresponding to the selected chronic disease with the current time, thereby determining the incomplete state. Claim 13 A medical assistance system according to claim 1, wherein the analysis unit updates at least one of the incomplete state, the medical item, or the reference point for the performance of the subsequent medical act when new medical act data or medical result data is acquired. Claim 14 A medical assistance system according to claim 1, characterized in that the subsequent medical act includes one or more of follow-up examinations, re-examinations, pathology tracking, vaccinations, and outpatient follow-up visits. Claim 15 A medical assistance system according to claim 1, characterized in that the output unit reflects the subsequent medical act in the patient-specific tracking management list only when an approval, modification, suspension, or exclusion input is received from the medical staff. Claim 16 A medical assistance system according to claim 1, characterized in that the user interface includes a browser screen, a pop-up screen, or an overlay screen that is displayed separately from the screen of the existing medical information system. Claim 17 A medical assistance system according to claim 1, wherein the output unit summarizes one or more of the results of purification, candidates for follow-up examinations, candidates for abnormal findings, and representative images in which the match of patient identification information is confirmed, and displays them in the same examination row of the user interface, and displays detailed information corresponding to the selected item in the same examination row through either a detailed area extended from the same examination row or a separate confirmation screen, and distinguishes unpurified data or the entire original image so that it is processed in the detailed area or the separate confirmation screen. Claim 18 A method for assisting medical care based on patient-specific medical data performed by one or more processors, comprising: a step of acquiring patient identification information and patient medical data through two or more different paths among screen-based recognition, file synchronization, and test result database linkage without changing the database structure, network configuration, and software of an existing medical information system; a step of calculating the degree of agreement or reliability between the patient identification information acquired through the said different paths, and suspending the combination, display, or tracking management reflection of said patient medical data if the degree of agreement or reliability is less than a reference value; a step of generating a patient-specific medical context linked to the patient's disease information, prescription history, or procedure history, using medical result data including test results, interpretation results, or pathology results actually generated independently of the existence of order information for tests or prescriptions from said patient medical data for which combination is permitted as reference data; a step of determining the correspondence between medical act data indicating a test or procedure scheduled or performed on the patient and said medical result data generated in response to said medical act, and determining one or more incomplete states among non-reception of results, mismatch between medical act and result, and non-reflection of follow-up measures based on said correspondence; and A method for assisting medical treatment, comprising: a step of calculating one or more medical items requiring verification or action based on an incomplete state and the medical context for each patient, and the type of subsequent medical act corresponding thereto and the standard time for performance; and a step of providing the medical items and the subsequent medical act to a user interface distinct from the existing medical information system, wherein the information generated in the calculation step is not automatically recorded in the existing medical information system until a verification input from a medical professional is received.

Citation Information

Patent Citations

  • Method and apparatus of context-based patient similarity

    JP2018060529A

  • Electronic medical information managing system

    KR1020130126047A

  • Method of Normalization for Combination of Clinical Data from Different Electronic Healthcare Databases and System thereof

    KR1020170019739A

  • Method and system for generating interpretable predictive result for patient

    KR1020250006766A

  • System and Method for Peoviding of Diagnosis Result Analysis Information

    KR1020250066003A