本發明係關於一種應用於臨床的決策支援系統，尤指一種應用於臨床輔助診斷的模組化醫學檢驗決策支援系統及方法。The invention relates to a decision support system applied in clinical practice, and in particular to a modular medical test decision support system and method applied in clinical auxiliary diagnosis.
目前的醫學檢驗之臨床決策支援系統（Clinical Decision Support System），主要是應用在進行診斷、治療方式等資學相關決策時，給予醫師或相關人員相關建議，而建議的來源通常是根據多年累積的經驗、臨床實驗分析、統計而取得，資料庫相當龐大，必須從歷年累積的大量病患資料中，經由資料探勘、分析等技術而歸納的資訊與知識。The current clinical decision support system for medical examinations is mainly used to give advice to doctors or related personnel when making relevant decisions about diagnosis and treatment, and the source of the recommendations is usually based on years of accumulation. Obtained through experience, clinical experiment analysis, and statistics, the database is quite large. It is necessary to summarize information and knowledge through data exploration, analysis and other technologies from a large number of patient data accumulated over the years.
然而現有技術中的臨床決策支援系統的資料庫過於龐大，儲存了大量陳述性和程序性的知識，在沒有易於使用、擴增、更新和維護的系統結構之下，使得現有技術中的臨床決策支援系統對於醫療及臨床領域知識的使用無法多元化及彈性應用，造成臨床決策不確定性仍然偏高，因此，就現有技術而言，如何同時提升臨床決策支援系統的決策判斷精確度、多元性及使用方便性，確實有待進一步提供解決方案的必要性。However, the database of the clinical decision support system in the prior art is too large, storing a large amount of declarative and procedural knowledge. Without a system structure that is easy to use, expand, update and maintain, it makes the clinical decision in the prior art The use of support systems for medical and clinical knowledge cannot be diversified and applied flexibly, resulting in high uncertainty in clinical decision-making. Therefore, in terms of existing technology, how to improve the accuracy and diversity of decision-making in clinical decision support systems at the same time? And ease of use, it really needs to provide further solutions.
有鑑於上述現有技術之不足，本發明的主要目的在於提供一種應用於臨床輔助診斷的模組化醫學檢驗決策支援系統及方法，其透過由專家建模之模組化技術結合編碼處理，並配合專家資料庫進行資料處理，以提升決策判斷的精確度、多元性以及使用方便性。In view of the above-mentioned shortcomings of the prior art, the main object of the present invention is to provide a modular medical test decision support system and method applied to clinical auxiliary diagnosis, which uses a modular technology modeled by an expert in combination with coding processing, and Expert database performs data processing to improve the accuracy, diversity and convenience of decision making.
為達成上述目的所採取的主要技術手段係令前述應用於臨床輔助診斷的模組化醫學檢驗決策支援系統包括：
一模組編輯單元，係供編輯多數的檢驗項目資訊，以產生一個以上群組化的檢驗模組資訊；
一編碼處理單元，係與該模組編輯單元連結，根據該檢驗模組資訊產生一組以上的特徵碼；
一專家決策經驗資料庫，係與該編碼處理單元連接，且該專家決策經驗資料庫儲存有多數臨床個案資料；
其中，當該專家決策經驗資料庫接收該組特徵碼，則根據該組特徵碼與多數臨床個案資料進行比對，以產生一筆回饋資料。The main technical measures adopted to achieve the above purpose are to make the aforementioned modular medical test decision support system for clinical auxiliary diagnosis include:
A module editing unit is used to edit most inspection item information to generate more than one grouped inspection module information;
An encoding processing unit is connected with the module editing unit and generates more than one set of feature codes based on the information of the inspection module;
An expert decision-making experience database is connected to the coding processing unit, and the expert decision-making experience database stores most clinical case data;
Among them, when the expert decision-making experience database receives the set of feature codes, it is compared with most clinical case data according to the set of feature codes to generate a piece of feedback data.
依上述構造，醫師或專業人員可透過該模組編輯單元編輯該等檢驗項目資訊以產生該群組化的檢驗模組資訊，並由該編碼處理單元根據該檢驗模組資訊產生該組特徵碼，當該專家決策經驗資料庫接收該組特徵碼，則根據該組特徵碼與多數臨床個案資料進行比對，以產生該筆回饋資料，藉此達到提升決策判斷的精確度、多元性以及使用方便性之目的。According to the above structure, a physician or a professional can edit the inspection item information through the module editing unit to generate the grouped inspection module information, and the encoding processing unit generates the set of feature codes according to the inspection module information. When the expert decision-making experience database receives the set of feature codes, the set of feature codes are compared with most clinical case data to generate the feedback data, thereby improving the accuracy, diversity and use of decision-making judgments. The purpose of convenience.
為達成上述目的所採取的主要技術手段係令前述應用於臨床輔助診斷的模組化醫學檢驗決策支援方法，係執行於一醫學檢驗決策支援系統，該方法係包括以下步驟：
執行一建模程序，編輯多數的檢驗項目資訊，以產生一個以上群組化的檢驗模組資訊；
根據該檢驗模組資訊產生一組以上的特徵碼；
執行一專家決策程序，根據該組特徵碼與多數臨床個案資料進行比對，以產生一筆回饋資料。The main technical means adopted to achieve the above purpose is to make the aforementioned modular medical test decision support method applied to clinical auxiliary diagnosis, which is executed in a medical test decision support system, the method includes the following steps:
Run a modeling process to edit most inspection item information to generate more than one grouped inspection module information;
Generate more than one set of feature codes based on the information of the inspection module;
An expert decision-making process is performed, and the majority of clinical case data are compared according to the set of feature codes to generate a feedback data.
依上述步驟，本發明可由醫師或專業人員使用該醫學檢驗決策支援系統，並於該醫學檢驗決策支援系統上執行該建模程序，編輯該等檢驗項目資訊以產生該群組化的檢驗模組資訊，該醫學檢驗決策支援系統再根據該檢驗模組資訊產生該組特徵碼，並執行該專家決策程序，該醫學檢驗決策支援系統根據該組特徵碼與多數臨床個案資料進行比對以產生該筆回饋資料，藉此達到提升決策判斷的精確度、多元性以及使用方便性之目的。According to the above steps, the present invention can use the medical test decision support system by a doctor or a professional, and execute the modeling program on the medical test decision support system, edit the test item information to generate the grouped test module. Information, the medical test decision support system then generates the set of signatures based on the test module information and executes the expert decision process. The medical test decision support system compares the majority of clinical case data with the set of signatures to generate the Pen feedback data to achieve the purpose of improving the accuracy, diversity and convenience of decision-making.
關於本發明應用於臨床輔助診斷的模組化醫學檢驗決策支援系統之較佳實施例，請參閱圖1所示，該模組化醫學檢驗決策支援系統包括一模組編輯單元10、一編碼處理單元20以及一專家決策經驗資料庫30，該模組編輯單元10係與該編碼處理單元20構成連結，該編碼處理單元20係與該專家決策經驗資料庫30構成連結；其中該模組編輯單元10係供醫生或專家使用並編輯多數的檢驗項目資訊，以產生一個以上群組化的檢驗模組資訊，並由該編碼處理單元20根據該檢驗模組資訊產生一組以上的特徵碼，而且於本較佳實施例中，該專家決策經驗資料庫30可預先儲存有多數臨床個案資料、或者隨時藉由內部／外部資料庫更新臨床個案資料；當該專家決策經驗資料庫30接收該組特徵碼，則根據該組特徵碼與多數臨床個案資料進行比對，該等臨床個案資料均另具有特徵碼，藉由該檢驗模組資訊產生的該組特徵碼與多數臨床個案資料的特徵碼進行比對，以產生一筆回饋資料，該筆回饋資料的形式是由百分比所構成。For a preferred embodiment of the modular medical test decision support system of the present invention applied to clinical auxiliary diagnosis, please refer to FIG. 1. The modular medical test decision support system includes a module editing unit 10 and an encoding process. Unit 20 and an expert decision-making experience database 30. The module editing unit 10 is connected to the encoding processing unit 20, and the encoding processing unit 20 is connected to the expert decision-making experience database 30. The module editing unit 10 is used by doctors or experts to edit most test item information to generate more than one grouped test module information, and the coding processing unit 20 generates more than one set of feature codes based on the test module information, and In the preferred embodiment, the expert decision-making experience database 30 can store most clinical case data in advance, or update the clinical case information at any time by internal / external databases; when the expert decision-making experience database 30 receives the set of features Code, it is compared with most clinical case data according to this set of feature codes, and these clinical case data have additional feature codes. The group signature by this test module generates information for comparison with the signature most clinical cases of data to produce a sum in the form of feedback information, the extra feedback data is constituted by percentage.
醫師或專業人員透過該模組編輯單元10編輯該等檢驗項目資訊以產生該群組化的檢驗模組資訊，並由該編碼處理單元20根據該檢驗模組資訊產生該組特徵碼，當該專家決策經驗資料庫30接收該組特徵碼，則根據該組特徵碼與多數臨床個案資料進行比對，以產生該筆回饋資料，藉此提升決策判斷的精確度、多元性以及使用方便性。The doctor or professional edits the inspection item information through the module editing unit 10 to generate the grouped inspection module information, and the encoding processing unit 20 generates the set of feature codes according to the inspection module information. The expert decision experience database 30 receives the set of feature codes, and compares the set of feature codes with most clinical case data to generate the feedback data, thereby improving the accuracy, diversity, and convenience of decision making.
進一步的，為舉例說明本較佳實施例的一應用方式，請參考圖2所示，其中該模組編輯單元10可設置於一電腦裝置100中，該電腦裝置100係供醫師或專家使用及操作，該編碼處理單元20及該專家決策經驗資料庫30可分別設置於一伺服器200中，並且該伺服器200可透過網路與該電腦裝置100構成連結，以交換資料。Further, in order to illustrate an application method of the preferred embodiment, please refer to FIG. 2, in which the module editing unit 10 may be set in a computer device 100 which is used by a physician or an expert and Operation, the encoding processing unit 20 and the expert decision-making experience database 30 can be respectively set in a server 200, and the server 200 can form a connection with the computer device 100 through a network to exchange data.
前述該專家決策經驗資料庫30所儲存或更新的多數臨床個案資料，多數臨床個案資料可包括有各種器官及感染疾病資料等臨床個案資料。
如上表格所示，醫師或專家於該模組編輯單元10執行一建模程序，編輯多數的檢驗項目資訊，多數的檢驗項目資訊包括多種檢驗結果，並於本較佳實施例中以三個檢驗項目資訊設為一群組，並針對該等檢驗項目資訊對應產生該等群組化的檢驗模組資訊；多數的檢驗項目資訊更包括多數的數值及其數值範圍、多數的量化資訊、多數的權重資訊、多數的代碼資訊。
再如上表所示，該編碼處理單元20根據該檢驗模組資訊產生一組以上的特徵碼，於本較佳實施例中，產生該組特徵碼的方式，係根據該檢驗模組資訊取得所對應到多數的數值，並將該等數值分別與該檢驗模組資訊所對應的數值範圍進行比較，以產生多數的量化資訊，再根據該等量化資訊，設定對應的一權重資訊，使得該等群組化的檢驗模組資訊均取得多數的權重資訊，並將該等群組化的檢驗模組資訊之權重資訊進行加總，以分別取得一代碼資訊，最後將該等群組化的檢驗模組資訊的進行最後加總，以產生該組特徵碼。As shown in the above table, the encoding processing unit 20 generates more than one set of feature codes according to the inspection module information. In the preferred embodiment, the manner of generating the set of feature codes is obtained from the inspection module information. Corresponds to the majority of values, and compares these values with the range of values corresponding to the inspection module information to generate a majority of quantified information, and then sets a corresponding weight information according to the quantized information, so that The grouped inspection module information has obtained most of the weight information, and the weighted information of the grouped inspection module information is summed to obtain a code information, and finally the grouped inspection information is obtained. The module information is finally added up to generate the set of feature codes.
當取得該組特徵碼後，由該編碼處理單元20及該專家決策經驗資料庫30執行一專家決策程序，根據該組特徵碼與多數臨床個案資料進行比對，以產生一筆具有百分比形式的回饋資料；進一步的，醫師或專家可根據該筆回饋資訊進行精確、多元的判斷，並提升使用方便性，而且，若出現新案例，則啟動一新增個案程序，將新的臨床個案加入該專家決策經驗資料庫30。After obtaining the set of feature codes, the coding processing unit 20 and the expert decision experience database 30 execute an expert decision-making process, and compare the majority of clinical case data with the set of feature codes to generate a feedback in the form of a percentage Data; further, the physician or expert can make accurate and diversified judgments based on the feedback information, and improve the convenience of use, and if a new case appears, start a new case procedure to add a new clinical case to the expert Decision-making experience database 30.
根據上述較佳實施例及應用方式本發明可進一步歸納出一應用於臨床輔助診斷的模組化醫學檢驗決策支援方法，該方法係執行於前述的該醫學檢驗決策支援系統，該方法包括以下步驟：
由該模組編輯單元10執行一建模程序，編輯多數的檢驗項目資訊，以產生一個以上群組化的檢驗模組資訊(S31)；於本較佳實施例中以三個檢驗項目資訊設為一群組，並針對該等檢驗項目資訊對應產生該等群組化的檢驗模組資訊；
由該編碼處理單元20根據該檢驗模組資訊，以產生一組以上的特徵碼(S32)；
於該編碼處理單元20及該專家決策經驗資料庫30之間執行一專家決策程序，根據該組特徵碼與多數臨床個案資料進行比對(S33)，以產生一筆回饋資料(S34)。According to the above-mentioned preferred embodiments and application methods, the present invention can further summarize a modular medical test decision support method for clinical auxiliary diagnosis. The method is executed in the aforementioned medical test decision support system, and the method includes the following steps. :
The module editing unit 10 executes a modeling program to edit most inspection item information to generate more than one grouped inspection module information (S31); in the preferred embodiment, three inspection item information are set. Is a group, and generates corresponding grouped inspection module information corresponding to the inspection item information;
The encoding processing unit 20 generates more than one set of feature codes according to the inspection module information (S32);
An expert decision-making process is executed between the encoding processing unit 20 and the expert decision-making experience database 30, and a comparison is made with most clinical case data according to the set of feature codes (S33) to generate a piece of feedback data (S34).
於本較佳實施例中，產生該組特徵碼的方式係根據該檢驗模組資訊取得所對應到多數的檢驗結果之數值，並將該等檢驗結果之數值分別與該檢驗模組資訊所對應的檢驗結果之數值範圍進行比較，以產生多數的量化資訊，再根據該等檢驗結果之量化資訊，設定對應的一權重資訊，使得該等群組化的檢驗模組資訊均取得多數的權重資訊，並將該等群組化的檢驗模組資訊之權重資訊進行加總，以分別取得一代碼資訊，最後將該等群組化的檢驗模組資訊的進行最後加總，以產生該組特徵碼In this preferred embodiment, the method of generating the set of signatures is to obtain the corresponding values of the majority of the test results according to the information of the test module, and to respectively correspond the values of the test results with the information of the test module. The numerical range of the test results is compared to generate a majority of quantitative information, and then a corresponding weight information is set according to the quantitative information of the test results, so that the grouped inspection module information all obtains a majority of the weight information And sum up the weight information of the grouped inspection module information to obtain a code information respectively, and finally sum up the grouped inspection module information to generate the group of features code
本發明可由醫師或專業人員使用該醫學檢驗決策支援系統，模擬醫師的疾病診斷模式，不僅進一步把知識型及法則型專家系統和個案庫應用整合於決策支援模式中，可以達到決策支援目標多元化目的，而且經該醫學檢驗決策支援系統的效能評估，對冠狀動脈疾病與腦中風的偵測敏感性分別為100%及99.5%，而對健康成人經由此醫學檢驗決策支援模組偵測出異常陽性反應的亞健康警訊提示率為82.9%，由此證明本發明在臨床醫療的輔助診斷與個人健康管理方面有極大的發展應用空間。The invention can be used by doctors or professionals to use the medical test decision support system to simulate a doctor's disease diagnosis mode. It not only further integrates knowledge-based and rule-based expert systems and case libraries into the decision support mode, and can achieve diversified decision support goals. Purpose, and through the performance evaluation of the medical test decision support system, the detection sensitivity for coronary artery disease and stroke is 100% and 99.5%, respectively, and abnormalities are detected in healthy adults through this medical test decision support module The sub-health warning prompt rate of positive reactions is 82.9%, which proves that the present invention has great development and application space in the auxiliary diagnosis of clinical medicine and personal health management.
10‧‧‧模組編輯單元10‧‧‧Module editing unit
20‧‧‧編碼處理單元20‧‧‧coding processing unit
30‧‧‧專家決策經驗資料庫30‧‧‧Expert database of expert decision-making experience
100‧‧‧電腦裝置100‧‧‧Computer device
200‧‧‧伺服器200‧‧‧Server
圖1 係本發明之較佳實施例的系統架構方塊圖。
圖2 係本發明之較佳實施例的另一系統架構方塊圖。
圖3 係本發明之較佳實施例的模組化醫學檢驗決策支援方法之流程圖。FIG. 1 is a block diagram of a system architecture according to a preferred embodiment of the present invention.
FIG. 2 is a block diagram of another system architecture according to a preferred embodiment of the present invention.
FIG. 3 is a flowchart of a modular medical test decision support method according to a preferred embodiment of the present invention.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW108106062ATWI680469B (en) | 2019-02-22 | 2019-02-22 | Modular medical test decision support system and method applied to clinical auxiliary diagnosis |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW108106062ATWI680469B (en) | 2019-02-22 | 2019-02-22 | Modular medical test decision support system and method applied to clinical auxiliary diagnosis |
| Publication Number | Publication Date |
|---|---|
| TWI680469Btrue TWI680469B (en) | 2019-12-21 |
| TW202032578A TW202032578A (en) | 2020-09-01 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW108106062ATWI680469B (en) | 2019-02-22 | 2019-02-22 | Modular medical test decision support system and method applied to clinical auxiliary diagnosis |
| Country | Link |
|---|---|
| TW (1) | TWI680469B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI781721B (en)* | 2021-08-10 | 2022-10-21 | 展市華科技有限公司 | Methods of diagnosing symptoms based on images |
| TWI810891B (en)* | 2022-04-13 | 2023-08-01 | 何桂芳 | Medical care standard knowledge-based decision support system |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105787252A (en)* | 2016-01-18 | 2016-07-20 | 贡京京 | Medical decision supporting method and system |
| CN106126958A (en)* | 2016-07-06 | 2016-11-16 | 温冬梅 | Method and system for automatic verification of clinical biochemical tests in medical laboratories |
| CN106650292A (en)* | 2017-01-04 | 2017-05-10 | 梁月强 | Personal health record system with process decision support function |
| CN108962394A (en)* | 2018-07-05 | 2018-12-07 | 广东工业大学 | A kind of medical data decision support method and system |
| CN109346169A (en)* | 2018-10-17 | 2019-02-15 | 长沙瀚云信息科技有限公司 | A kind of artificial intelligence assisting in diagnosis and treatment system and its construction method, equipment and storage medium |
| TWM582676U (en)* | 2019-02-22 | 2019-08-21 | 國泰醫療財團法人國泰綜合醫院 | Modularized medical examination decision support system applied to clinical auxiliary diagnosis |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105787252A (en)* | 2016-01-18 | 2016-07-20 | 贡京京 | Medical decision supporting method and system |
| CN106126958A (en)* | 2016-07-06 | 2016-11-16 | 温冬梅 | Method and system for automatic verification of clinical biochemical tests in medical laboratories |
| CN106650292A (en)* | 2017-01-04 | 2017-05-10 | 梁月强 | Personal health record system with process decision support function |
| CN108962394A (en)* | 2018-07-05 | 2018-12-07 | 广东工业大学 | A kind of medical data decision support method and system |
| CN109346169A (en)* | 2018-10-17 | 2019-02-15 | 长沙瀚云信息科技有限公司 | A kind of artificial intelligence assisting in diagnosis and treatment system and its construction method, equipment and storage medium |
| TWM582676U (en)* | 2019-02-22 | 2019-08-21 | 國泰醫療財團法人國泰綜合醫院 | Modularized medical examination decision support system applied to clinical auxiliary diagnosis |
| Publication number | Publication date |
|---|---|
| TW202032578A (en) | 2020-09-01 |
| Publication | Publication Date | Title |
|---|---|---|
| Lüscher et al. | Artificial intelligence in cardiovascular medicine: clinical applications | |
| CN109805898B (en) | Critical death prediction method based on attention mechanism time sequence convolution network algorithm | |
| Wang et al. | TWIN-GPT: digital twins for clinical trials via large language model | |
| EP1941410A2 (en) | Medical-risk stratifying method and system | |
| Yang et al. | An approach to automatic process deviation detection in a time-critical clinical process | |
| Chan et al. | Medtsllm: Leveraging llms for multimodal medical time series analysis | |
| KR20240100366A (en) | Predicting clinical trial facilitator performance using patient claims and historical data | |
| TWI680469B (en) | Modular medical test decision support system and method applied to clinical auxiliary diagnosis | |
| CN120108740B (en) | Method and system for constructing risk assessment model for myocardial infarction patients | |
| Vásquez-Iturralde et al. | Automatic classification of cardiac arrhythmias using deep learning techniques: A systematic review | |
| Hachamovitch et al. | Imaging registries and single-center series | |
| CN118824500B (en) | A neurosurgery nursing management system based on cloud computing | |
| JP7623120B2 (en) | Medical Support System | |
| Nugent et al. | Playing by the numbers: how collecting outcomes data changed my life | |
| US12205687B2 (en) | Pathway information | |
| Baheti | Study of need and framework of expert systems for medical diagnosis | |
| TWM582676U (en) | Modularized medical examination decision support system applied to clinical auxiliary diagnosis | |
| Strobel et al. | Healthcare in the Era of Digital twins: towards a Domain-Specific Taxonomy. | |
| Rani et al. | Quantile loss function empowered machine learning models for predicting carotid arterial blood flow characteristics | |
| KR102536982B1 (en) | Artificial intelligence-based disease prediction method and apparatus using medical questionnaire | |
| Adar et al. | RETRACTED: A user-friendly AI-based clinical decision support system for rapid detection of pandemic diseases: Covid-19 and Monkeypox | |
| TWI755731B (en) | Biomedical data risk prediction system and method | |
| Wang et al. | Cardiovascular disease prediction model based on patient behavior patterns in the context of deep learning: a time-series data analysis perspective | |
| Herman et al. | Human Digital Twin Modeling for Cardiovascular System | |
| Surojudin et al. | Implementation of the Naive bayes algorithm for death due to heart failure using rapid miner |