Poster
23
0
0
全文
(2) MIST2002 Proceeding. A study of the PC simulation on Drugs Management for Chronic Disease Patient Data Card JIUN-TZE a. LAI a、CHING-MING LIN b、CHING-MING CHAO c. Department of ES National Cheng Kung University、 b Department of ES National Cheng Kung University、 c Department of ES National Cheng Kung University a b c [email protected] [email protected] [email protected]. Abstract. may occur without doctor’s notice, thus may result in patients’ overdose and the medical resources waste. Such phenomenon demonstrates that the medical resource waste may also cause the patient’s health burden.. The key point of this study is on the design of the 「Chronic Disease Patient Data Card 」, and the simulation of procedure how a doctor care a patient, the computer checking the patient repetitive drug doses, and to correct the prescription , checking sever drugs interaction . To record the statistics of repetitive drugs doses and sever drugs interaction for later medicinal consultations. According to the three years drugs consumption sample data of 192 high blood pressure patients group collected from NHI, by undertaking through the proposed PC simulation schema, each HBP patient can save about 5428 $ per year. Take about one million HBP patients as estimated by NHI’s statistics, this could be up to a 5.5 billion $ saving per year. So that the Chronic Disease Patient Data Card will bring benefit and improve the service offered by Bureau of National Health Insurance. It can also save the resource for business continuity.. 2.Conclusion The treatment data of some patients provided by the NHI was drawn in this study to attribute to the personal drugs record database, design the “Chronic Disease Patient Data Card”,and simulate real-time diagnosis. From early 1999 to early 2000, there were 22064 items of drugs being prescribed to 43 diabetes patients, the total sum of repetitive drugs treatment number is 1358, the days of repetitive drugs doses is 71926, preliminarily the drugs expenditure waste amounts 836620 dollars. While from early 1999 to early 2000, there were 88277 items of drugs being prescribed to 192 high blood pressure patients, the total sum of repetitive drugs treatment number is 5484, the days of repetitive drugs doses is 269961, preliminarily the drugs expenditure waste amounts 3126792 dollars. Using this Chronic Disease Patient Data Card computer simulated diagnosis program, take the case of high blood pressure, each patient can save 5428 dollars for the NHI each year. If there were about 1 million of high blood pressure patients, it would save nearly 5.428 billion dollars each year. That would be a huge sum of budget. Though the price analyzed in the statistics includes only the drugs item price, while the partial fee paid by the patients were not included in the analysis, it demonstrates clearly the effect that the Chronic Disease Patient Data Card have on the saving of the medical appliances.. 1.Repetitive Drugs Doses Check Result This study simulates under the condition that the chronic disease patients using the Chronic Disease Patient Data Card. The Chronic Disease Patient Data Card defaults 2Kbyte capacity, there might be some existing treatment data that had not been stored into the card, as a result, the amount the patients actually waste may be higher than the numbers of the repetitive drugs doses. Therefore, the data may be closer to the attributed statistics and findings if upgrading the capacity of the Chronic Disease Patient Data Card to more than 4Kbyte or simplifying the recording format of the data card. The original recording column 158byte is changed to 90 byte. Applying this recording format, 2Kbyte capacity of Chronic Disease Patient Data Card could record as 1.5 times as much drugs doses prescription. Adopting the PC simulation of this study, the check of repetitive drugs doses and the drugs interaction would match the results of the attributed drugs record analysis. When simulating the interrogation, we found that some of the doctors did not enquire about the condition of the patient’s on-going drugs, while the patients were undergoing other diseases. The ingredient of the drugs repetition in the prescription s2.
(3) MIST2002 Proceeding. A Study on Implementation of Drugs Management Database for Chronic Disease Patient Data Card CHING-MING CHAO a、JIUN-TZE a. LAI b、WEI-I LO c. Department of ES National Cheng Kung University 、b Department of ES National Cheng Kung University 、c Department of ES National Cheng Kung University a [email protected] b [email protected] c [email protected]. Abstract. 1 Research Result. This thesis studies the chronic patient’s medical. 1.1 Diabetes patient group statistics. administration and implementation of the drugs. According to the inquiring programming ,we take a. management database for the proposed chronic. diabetes patient for example，who’s ID number is. disease patient data card. Starting from the. 「DS05650678」.There are 6 data，the totally wasted. post-clustering algorithm, chronic disease patient’s. money is 611.4333 $. After calculation, there were. records has been implemented a feasible archives.. 22,064 items of drugs being prescribed to 43 diabetes. Then the patient's drug records, such as precious and. patients, and spent 4,680,860 $ on it. Among it, the. taboo medicine prescribed at every medical clinic. diabetes. and. 5,8018medicine drugs and wasted 965,315$.. various. hospital. departments,. can. be. patients. group. totally. repeated. summarized and analyzed for personal chronically. Base on the preliminary estimates, the threes years. medical. Regularly. from1999 to Feb, 2002, each diabetes patient. clustering and analyzing the chronic disease drug. repeated 134.906 medicine drugs and wasted. records archives, both the doctors and the patients. 22,449.186$. On the whole，The diabetes group has. can take it for consultation.. 26.29% repetitive medicine drug prescriptions, and. According to drugs consumption sample data of 43. has wasted 20.62% of the total medicine cost. diabetes patient group and 192 high blood pressure. 1.2 HBP patients group statistics. patient group collected from NHI in the three years.. According to the inquiring programming ,we take a. By undertaking through the checking of repetitive. HBP patient for example ， who’s ID number is. medicine drugs and serious drugs interaction, each. 「DS99104185」. There are 5 data，the totally wasted. diabetes patient repeated 135 medicine drugs and. money is 17.3333 $. After calculation, there were. wasted 22,449$, HBP patient repeated 120 medicine. 88277 items of drugs being prescribed to 192 HPB. drugs and wasted 21,813$. On the whole, the. patients, and spent 18,968,006$ on it. Further, the. diabetes group has 26.29% repetitive medicine drug. HPB patients group totally repeated 22,978medicine. prescriptions, and has wasted 20.62% of the total. drugs and wasted 4,189,013$.. medicine cost. The HBP group has 26.03% repetitive. Base on the preliminary estimates, the threes years. medicine drug prescriptions, and has wasted 22.08%. from1999 to Feb, 2002, each HBP patient repead. of the total medicine cost. As for the serious drugs. 119.68 medicine drugs and wasted 21,813.78$. On. interaction searching no serious interaction of 2,666. the whole，The HBP group has 26.03% repetitive. medical prescriptions for 43 diabetes patient group. medicine drug prescriptions, and has wasted 22.08%. but 7 serious interaction of 12,556 medical. of the total medicine cost.. contraindicated. archives.. prescriptions for 192 HBP patient group. s3.
(4) MIST2002 Proceeding. A study on security mechanism of HPC for Chronic Disease Patient Data bCard a c CHUN-LIN YEH 、YI-JIAN CHEN 、CHING-MING CHAO. a. Department of ES National Cheng Kung University、bDepartment of ES National Cheng Kung University c Department of ES National Cheng Kung University a [email protected]、 [email protected]、 [email protected]. 1.Abstract. 3-2 Authentication Process (1) The authentication process between SAM card and HPC: (by three-pass authentication) (2) The authentication process between SAM card and HC: (by three-pass authentication) 3-3 the Message Authentication Code (MAC) of HPC This thesis uses the HPC as the “Card Key” and the Chronic Disease Patient Data Card as the “Card Database”. Thus patient prescriptions are stored as secure electronic data with a digital signature of the physician. This establishes the undeniable signed prescription in case of mediation contradiction.. This thesis is a part of the feasibility study on the implementation of the Chronic Disease Patient Data Card System (PDCS) for NHI. The main object is to simulate the secure application for the Health Professional Card (HPC) on PDCS. A further study on the authentication process between PDC and HPC has been carried out. Here, a simulation of the Security Access Module (SAM) Card’s operation is proposed for this particular verification and data signature processes related to this PDC system. Those designed procedures and protocols not only able to guarantee the patient data integrity and authenticity, also enable to prove the undeniable signed prescription in case of mediation contradiction. This will provide the NHI’s ongoing Health Care Card (HCC) system a solid secure infrastructure . This thesis is focusing on the planning and implementation of HPC system by a series of security operations. They include the card issuing, secure authorization, and medical prescription digital signature operations. Extension of this research toward electronic health records application can enhance the secure protection of the electronic health information. And the proposed Public Key infrastructure (PKI) will provide the secure mechanism to achieve the secure and trustful Web Health Service (WHS) scenario.. 4.Conclusion and Future： In near future, there will issue 23,000 thousand National Health Insurance Card in Taiwan. The system will be achieved by most passive card-readers using PC/SC frame and less separate active card-readers. Furthermore, an extra planning to issue thirty thousand Health Professional Card,which are designed by Java language with embedding system. Hence we save AP, DES algorithms and a RSA cipher system into the card system. Thus not only enhance the security of system management, but also accomplish digital signature using a RSA cipher system. This result lets HPC can be executed onto the management of secure access and make more perfection to carry out National Health Insurance Card in Taiwan.. 2.Research Motive This thesis establishes a secure structure and objectives base on SAM card for the feasibility study on the implementation of the Chronic Disease Patient Data Card System (PDCS) for NHI. By using authentication between SAM card and HPC, we can ensure patient’s right. As for PDC, we can also use SAM card to ensuring the legitimate right of patient. Because we hope that Chronic Patient Data Card can apply to current National Health Insurance Card system, we plan the structure of Chronic Patient Data Card base on current National Health Insurance Card system.. 3.Main Thesis 3-1 Issue Systems There are SAM card and Health Professional Card (HPC) included Doctor Card, Medicine Card and Patient Data Card in this thesis. (1) Issue the SAM card (2) Issue the HPC (3) Issue the HC s4.
(5) MIST2002 Proceeding. A Web-based National Health Insurance Database Application Health Knowledge Information Discovery System (Health KIDS) Po-Huang Chiang a*; Der-Ming Lioub; Shu-I Chiuc; Yi-Ling Houc; Yuan An Liuc a. b. Division of Health Policy Research, National Health Research Institutes, Taipei 11 4, Taiwan, R.O.C. Dept. of Health Informatics and Decision Making, National Yang-Ming University, Taipei 112, Taiwan, R.O.C c Dept. of Research Resources, National Health Research Institutes, Taipei 114, Taiwan, R.O.C.. Abstract. National Health Insurance Database provided by the Bureau of National Health Insurance, Department of Health in Taiwan and managed by National Health Research Institutes (NHRI). The discussion begins with the history and initial rationale of NHIRD, which began 3 years ago, and continues with a description of the development of making available and useful of health insurance claim data upon which NHIRD is based. It summarized the development of Health KIDS which use the Sagent solution to extract, transform and organize the insurance data to bring the information to decision makers for review and analysis.. To improve our understanding of factors and mechanisms that affect our health industry is a fundamental necessity for planning appropriate health policy and preventive interventions in Taiwan. National Health Insurance database, with at least 21 million outpatient and inpatient records every month, provide a unique opportunity to examine the health issues to the society as a whole. This project describes the planning and development of the Health Knowledge Information Discovery System (Health KIDS) in use of National Health Insurance Research Database (NHIRD), which is based on data from the. Health Informatics Education Janelle Craig BAppSc(MRA) MComma and Angelika Lange, MA (Inf. Sc.), GradCert (CDM), Dipl.Psychb a. Lecturer, School of Health Information Management, Faculty of Health Sciences, The University of Sydney [email protected] b Lecturer, School of Health Information Management, Faculty of Health Sciences, The University of Sydney [email protected] Correspondence to: Janelle Craig, Lecturer, School of Health Information Management, Faculty of Health Sciences, The University of Sydney, PO Box 170, LIDCOMBE NSW 1825 AUSTRALIA [email protected]. ABSTRACT In 2001, the School of Health Information Management at the University of Sydney introduced the first coursework Masters program in health informatics to be offered by an Australian university, the Master of Health Science (Health Informatics).. This paper details the background to the development of the program, and describes the content and structure of the program. Methods to ensure the currency and relevance of the program are also addressed, these focusing on adherence to the IMIA’s Health and Medical Informatics Education Working Party’s recommendations, criteria established by the Australian Universities Teaching Committee to assess best practice in postgraduate coursework programs, and through extensive industry consultation and participation.. The paper concludes by considering the future directions of the program and its graduates.. s5.
(6) MIST2002 Proceeding. Applications of Dictaphone and Transcription on the Internet Tsair-Wei Chien a*, Hung-Jung Lin a, On- Yick Liang a a. Chimei Foundation Hospital , No.901 Junghua Rd., Yungkang City, Tainan, Taiwan 710, R.O.C. *The corresponding author: [email protected]. Background and objectives: The Global Budget. verification by email and pager from time to time. A. Payment System has been manipulated for hospitals. total of 237 physicians from 30 sampled big hospitals. st. since July 1 , 2002 in Taiwan. The readable medical. in Taiwan participated in this dictation survey in June,. records are requisite for the oncoming DRGs. 2002.. reimbursement soon adopted from current U.S. where. Results： The digital transcription, “what you say is. has about 6.5 billion transcription markets. The. what you soon get”, has saved time from days to. ICD_9 coding today is with plentiful problems.. hours and got the decipherable charts. The turnaround. Seeking solutions as the LifeCodeR engine other than. time was improved, shorter from analog to digital.. employing professional manual coding is a good way. The dictated notes in faithfulness, expressiveness,. to profit hospitals. Dictating and transcribing the. and elegance were assessed, more excellent than the. records are advantageous for decreasing the paper. typed ones. The analyses are from medical centers. work time, improving the record legibility and quality,. and regional hospitals: About 46% agreed with. and allowing performers to have more time for. talking better than typing, in contrast to 21%. patients. Attributed to the participants’ workload,. disagreeing; Up to 50% subjects considered dictated. writing charts may increase inaccuracies which may. notes helpful; 53%, dictation beneficial to chart. be minimized by dictating which saves much more. quality. Around 73% typed less than 30 words per. time than what is anticipated. We hereby share the. minute (WPM), 23 WPM averagely, which revealed. experience of setting up a web-hosted dictation via a. the difficulty on typing. According to half of the. phone system for medical applications in our. testees, to do the manual work for the digitalized. hospital.. notes was difficult enough.. Methods: The hardware requirement for Dictaphone. Conclusions: The state-of-the-art systems over the. consists of three IBM eServers, AP and transcription. phone are easy for performance: recording on a. database, 64-phoneline ports, sound cards, and. digital dictation system, transcribing through the. Philips LFH-6277 series including PC headsets, foot. internet, transmitting a signature, printing at the site,. pedals, etc. Many PHP web-based programs were. and eventually storing an EMR. Owing to the DRGs. developed in conjunction with MySQL Databases. payment coming, to dictate and transcribe the clear. under the Linux operation system. The C language. and concise medical notes are conducive. The. was applied to the control mode of dictating records. dictation system which comprises the point-N-click. which can be backward/forward, and paused, etc.. on workstation and portable handheld digital devices. There were 5 transcriptionists available for several. is. types of dictated reports. After patient’s PIN, work. “one-stop-shopping” in the nearest future. Further. type, and the narrative free text (English or/and. research would be required on the speech recognition. Chinese) via phone. for medicine.. lines were input, the dictated. notes would be sent to the practitioners for s6. to. complete. the. EMRs. efficiently ─.
(7) MIST2002 Proceeding. An attitude investigation of medical users toward search engines Shu-sheng Liaw a, Hsiu-Mei Huang b, Wen-chih Chenc Computer Center, China Medical College, 91 Shiuesh Rd., Taichung, 404, Taiwan, E-mail: [email protected], Phone: 886-422053366 Ext. 1710a Department of Information Management, National Taichung Institute of Technology, 129, Sec. 3, Saming Rd., Taichung, 404, Taiwan, E-mail: [email protected], Phone: 886-422211181 Ext. 2620b Department of Pharmacy, China Medical College, 91 Shiuesh Rd., Taichung, 404, Taiwan, E-mail:[email protected]. Abstract. total of 114 responses, 80 were male and 34 were female students. The data for this study was gathered by a questionnaire survey. The questionnaire survey included three major components: (a) demographic information, (b) computer and Internet experience, and (c) perceptions of search engines.. Many search engines have been developed to help users look for relevant health or medical information on the Web. In fact, there is just too much information on the Internet. The current study is to investigate medical students’ perceptions toward Internet search engines as a tool for information and knowledge retrieval. Regarding search types of search engines, search engines derive from three basic paradigms: directory-based services, query-based, and filter-based search engines. Directory-based services, such as Yahoo! (http://www.yahoo.com) for general purpose or MedWeb (http://www.medweb.emory.edu/MedWeb), Medical Matrix (http://www.medmatrix.org) for medical purpose. Query-based search engines, such as Google (http://www.google.com) for general purpose or med411 (http://www.med411.com/) and HON (http://www.hon.ch) for medical purpose. Filter-based search engines, such as OMNI (http://omni.ac.uk), MEDBOT (http://medworld.stanford.edu/medbot), and HONselect (http://www.hon.ch/HONselect) for medical purpose. Although more and more search engines have emerged on the Web, major problems (such as precision rate, recall rate, and retrieval speed) with information retrieval still remain. In other words, users still are not satisfied the quality of search engines. The research hypotheses combine four different perspectives: individual computer experience, quality of information systems, Technology acceptance Model (TAM) and motivation. The study was conducted in a medical college in central. Taiwan with a sample of 120 medical students. A. Experience with. Quality of search. Internet response. Experience using. word processing. engines. time. operating systems. 0.29**. 0.31**. 0.35**. 0.26**. 0.32**. Perceived enjoyment of. Perceived easy to use. search engines. search engines. 0.58**. 0.31**. 0.26**. Perceived usefulness of search engines 0.67** Intention to use search engines. Figure 1:The results of regression analyses. **p<0.01. The results provide supports for all hypotheses that were tested (Figure 1). Additionally, the results also confirm theories, such as TAM and motivation. Moreover, individual computer-related experience and perceptions are key factors to influence users’ intention to use search engines. Further, we suggest that retrieval information from search engines should contain five parts: the Web page’s URL, a title, percentage of relevance, a brief description, and a group of keywords intended to describe the Web page’s contents.. s7.
(8) MIST2002 Proceeding. Simplified TCP/IP Communication Procedure for Information Retrieval I-Cheng Shen, M. S. Young Department of Electrical Engineering National Cheng Kung University, Tainan 70101 Taiwan, ROC. [email protected]. Abstract. The communication procedure in this study, network packet manager handles these receiving the network frames. ARP request, ICMP Echo request, and ARP reply would be replied back immediately. TCP frame would be delivered to TCP connection manager using 4 functions to handle SYN, FIN, ACK, and normal TCP frame. TCP connection starts by 3-way SYN handshake and ends with 3-way FIN handshake after transmission. When TCP connection manager receives SYN and FIN packet, it would immediately reply SYN-ACK and FIN-ACK respectively, and waiting for an ACK. Lately, normal TCP frame is used to receive data.. Internet communication is an important communication method. TCP/IP is the standard of Internet communication procedure and protocol [1], but is complex. The communication of TCP/IP obeys a standard procedure to establish connection and deliver data. These protocols are usually implemented in the PC OS such as Windows, Linux, and BSD. This study designs a simple procedure of TCP/IP to communicate with a standard TCP/IP protocol. This method generates the same network packets, and it can communicate well with a standard TCP/IP communication in Windows.. 1. Introduction. 3. Result and Discussion. Till year 2002, there are more than three thousand RFCs (The Request for Comments) published. The most important protocol to communicate on the internet is TCP/IP. In 7 layers of OSI (Open System Interconnect) structure, TCP/IP presents the service in both network and transmission layers. The completely TCP/IP communication procedures and protocols are complex. General speaking, PC’s OS is suitable to implement completely TCP/IP protocol with a lot of memory and the fast running time, but using the microcontroller to implement these is difficult. Information retrieval in local area network (LAN) usually needs one host-to-host connection, one port-to-port, less requirement for flow control, and ability to be the server. So it’s better to reduce TCP/IP procedure to simplify information retrieval communication process. We decided to design a simplified TCP/IP communication procedure will be easier to implement using a microcontroller and spend less CPU time to handle with CPU time. It will communicate well with a standard TCP/IP protocols.. Table I the features of communication protocols. Network communication protocols Features Standard This design Support ports 65535 1 Support hosts 256*4 1 Transmission WAN LAN Range Flow control Yes Less capacity Reliability Yes Yes Client-Server Both Server ability Connection Yes Simulation establishment Connection Yes Simulation termination Date Packet Yes Yes acknowledgement Table I lists the feature difference between this design and a standard TCP/IP protocol. This design simplifies the connection establishment, connection termination, and flow control. This design will communicate well with completely TCP/IP protocol and be useful to retrieve information through the network when we only use an 8-bit microcontroller. It is also useful for local area network information retrieval. In the future, we will strengthen both flow control to enhance ports and hosts capacity at the same time.. . Method This study designs a new flow from the RFCs (Reference for comments). The RFC793 is the standard of TCP protocol. The features of TCP Protocol are both client and server abilities, and have 65535 ports available for connection, flow control, connection establishment, termination, and long distance transmission [2]. These features of this design are for server ability, one host-to-host connection, one port-to-port connection, less capacity for flow control, connection establishment and termination, local area network transmission. In this study, we use two Windows-based PCs to simulate the communication flow. Microsoft Network Monitor analyzes the communication flow and network packets between the two hosts in the network. TCP standard defines 11 states in its’state machine. This design split those states to be 4 stand-alone state machines.. 4. Reference [1] Lorcan Dempsey, Libraries, ”Networks and OSI : A Review, with a Report on North American Developments”, (Westport, CT: Meckler in association with the UK O? ce for Library Networking, 1992). [2] J. Postel, ”Transmission Control Protocol”, RFC793, Sep 01, 1981. [3] J. Postel,”Internet Protocol”, RFC791 Sep 01, 1981.. s8.
(9) MIST2002 Proceeding. The Development of PDA – Based Mobile Expert Meeting Support Systems a. 柯柏光、 b 李慶琪、 c 劉嘉良、 d 蕭芳瑩、 e 張博論. a、c、d、e a. 國立陽明大學衛生資訊與決策研究所 b 國立清華大學科技管理研究所 [email protected]; b [email protected]; c [email protected]; d [email protected]; e [email protected]. Abstract. Telemedicine(遠距醫療). Since the medical staffs are always on the run, it is never an easy task to have group meetings with them to make decisions. However, as the development of technology, the traditional form of meeting can be transformed into a mobile meeting by utilizing new technologies, and thus solves the problem of having a group meeting with the medical staffs.. To achieve wireless network transmission and to efficiently apply it on the voting mechanism in a project meeting, in this research, the database server 、the NGT Meeting server and the IIS server are constructed on the same host, and the wireless transmission is achieved via TCP/IP protocol. To ensure the consistency of the database and the integration of the system, a Client-Server structure is adopted. During the process of the meeting, frequent message transmission between the chairman and the anticipator can be predicted. This system uses Winsock transmission methods precisely designed for PDA database transportation, Remote data access, or replication methods.. The mobility and low cost of PDA in combination with the wireless network environment can enhance the flow of information. Therefore, this study proposes a mobile meeting system compatible with the wireless environment by using Nominal Group Technique and PDA. In such system, the medical staffs will be able to have a group meeting over the wireless network via PDA.. Insuring the smoothness of the process, we enlisted the “Normal Group Technique” to use as an accessory to the meeting process. Through the declaration, interpretation, and evaluation of the opinions, the anticipators can reach an agreement through this system and produce a consistent outcome.. Key Word PDA(個人數位助理) Nominal Group Technique(名義群組技術) Wireless(無線) Mobile Information System(行動資訊系統). 公開金鑰密碼系統應用於院際間電子病歷傳遞之研究 潘妍伶 a、陳正鎔 b、劉原祺 a. c. 國防管理學院國防資訊所研究生 [email protected] b 萬能技術學院資管系 [email protected] c 大華技術學院國貿系 [email protected]. 摘要. 題。在公開金鑰密碼系統【1-2】中，一份文件若 要傳遞多個接收者，需要分別對接收者進行加密手 續，時間的耗費反而降低了網際網路的便利性；若 採用傳統式密碼系統【3-5】 ，雖然可以利用一次加 密手續傳遞多個接收者，但由於擁有相同密碼的人 過多，增加了洩密的風險。因此，本論文提出一套 以公開金鑰密碼系統為基礎，以一次加密手續，達 到不同醫院可接收到不同或相同的電子病歷之研 究。 關鍵字：公開金鑰密碼系統、傳統密碼系統、電子 病歷、資訊安全. 電子簽章法業於去（90）年完成立法，該法雖源自 於電子商務之推動，但對醫療界的影響，並不亞於 商業應用，隨著該法於九十一年四月一日公告實 行，國內醫療院所推行已久之電子病歷也將取得合 法性。電子病歷之優點除了降低營運成本及人事費 用外，最重要的是有助於醫學研究及統計，經由院 際間病歷交流，醫師可充分掌握病患的病史，掌握 醫療的第一時間並提升醫療品質。然而病歷交流雖 可提高醫療效率及醫療品質，但是在網際網路上傳 遞個人醫療資料卻面臨了安全性及保密性之問 s9.
(10) MIST2002 Proceeding. Building a Medical Multimedia Teaching Warehouse 簡文山 a. a,b,c. ,張博論 b,徐建業 d,簡嘉惠 e,萬序恬 e,林恩德 e,李友專 b. d. c. 臺北醫學大學醫務管理學系, 國立陽明大學衛生資訊與決策研究所, 國立陽明大學公共衛生研究所 d 台北醫學大學醫學資訊研究所 e 台北醫學大學資訊服務中心 [email protected]. 一、中文摘要. 者介面是一律採用瀏覽器介面，後端是採用個別獨. 本研究主要是建立及架構醫學多媒體教學共. 立的資料庫元件，使用者可透過資料上傳維護介. 享的 warehouse，讓學習者在最短的時間內在網路. 面，將醫學教育的多媒體資料上傳於系統之各種資. 上取其所要的醫學多媒體教學資源。利用現有高頻. 料庫之中，我們可以將這些資料庫的元件再利用多. 網路的特色和 windows 2000 的影音伺服器的新功. 媒體資料分類索引系統予以有系統的分類，分門別. 能，使得學生不需集中視聽教室，僅利用一般的網. 類後此系統可以提供使用類別及連結等之關鍵字. 路電腦上網至本研究計劃的多媒體倉儲中心，點取. 搜尋。. 同步廣播或 video on demand，即可上課或作非同步. 四、結果. 的網路教學，達到低成本、高效益，真正無空間和 時間限制的多媒體網路醫學教育。本研究也進一步. 此研究成功的架設了此一醫學多媒體資料倉. 提供研究成果及經驗來推展各醫學教學科目的教. 庫，提供了醫學教育者一個線上教育資料的交換及. 學教材網路化，同時培養新一代的網路醫學多媒體. 儲存的空間。. 傳播人才。. 1.. 對於此研究開發之系統的使用結果，我們也 進行了問卷調查，針對使用此系統輔助上課的. 二、緣由與目的. 57 位學生進行調查，回收有效問卷 52 份，其. 醫學教育屬於描述性之教學，大部份都是圖. 調查結果是 92%的使用者此系統對於他們的. 片和影像之呈現。醫學教育如停留在刻板不動之課. 學習認為有幫助或有些幫助，64%的使用者認. 本和一堆的文字表達，無法讓同學能整體貫通。一. 為完全不可能或些許不可能完全取代傳統之. 套生動之多媒體醫學教材才能達到教學目標. 面對面課程，88%的使用者覺得很值得或可以 推廣至其他的課程。. 三、方法 五、討論. 本研究之主要目標是於現有的頻寬架設一個 醫學多媒體教育倉庫(Medical Multimedia Teaching. 由於影像的資料太過龐大，雖然利用了先進的網路. Warehouse)，主要利用 Windows 2000 伺服器及專. 技術來減少資料的傳輸量，但是 Internet 的使用者. 業影片、語音及時廣播伺服器，架設以醫學專業為. 日益增多，網路頻寬永遠不敷所需，產生了影像延. 主的教育傳播網，配合醫學教育的需求，訂定不同. 遲之現象，因此較不適合網際網路使用。. 規格模式的環境，讓將來各界可以輕易取得和加入 本系統之行列。選擇 Windows 2000 伺服器及其專 業影片語音及時廣播伺服器的主要原因是因為它 可讓使用者和提供者不需增加額外的裝置，更不需 昂貴的配備，尤其讓學生更可在其家裡使用基本配 備上網學習。 本系統乃是採用三層式的系統架構前端使用 s10.
(11) MIST2002 Proceeding. 以 Grid 技術建構邦聯式醫院電子病歷分享環境 Using Grid Computing for a Federated Electronic Medical Record Sharing Environment among Hospitals 簡文山 a,b,c, 李友專 d, 蔣以仁 d, 朱賢良 e, 張博論 b, 林恩德 d a. 臺北醫學大學醫務管理學系, b 國立陽明大學衛生資訊與決策研究所, c 國立陽明大學公共衛生研究所 d 台北醫學大學醫學資訊研究所, e 國家高速電腦中心 [email protected]. 一、中文摘要. 四、預期效益. 本文提出在各醫院設立一個資源分享之伺服. 使用 Grid 技術主要提供系統二大功用:. 器，各醫院可將確認後並提供查詢之病患資料存放. Computational Grid 及 Data Grid 功能。我們預期此. 於伺服器的 Public Area Folder，系統即可利用 Grid. 系統將有下列之能力: 分散式資料收集能力. 技術串聯各醫院之伺服器，形成一個單一檔案資料. (Distributed data collection capabilities)、自動化數位. 夾或資料庫，再利用 Web Server 作為一 Middle. 圖書館能力(Digital Library capabilities)、強大檔案. Ware，讓使用者可以利用瀏覽器來存取檔案資料或. 存取控管能力(Persistent Archive capabilities)、分散. 下達操作指令，例如 SQL 查詢等等。最後提出其. 式的資料查詢能力(Distributed data query. 預期效益並加以討論。. capabilities) 同時依此模式建立的聯邦式醫院電子病歷分享環. 二、緣由與目的. 境也將會有下列的優點:. 在醫學上之電子病歷資料量龐大，往往也需. 1.. 位置的透通性(Location transparency). 要大量的儲存空間及搜尋計算能力，以全國性之病. 2.. 提高可靠度及可利用率(Improved reliability. 歷資料來說，皆是以 Tera Bytes 來計算，且各醫院. and availability). 之資料皆分散在各地，為了配合各資料來源分散、. 3.. 單一權限控管(administrative). 能夠整合異質平台、儲存空間資源整合且能快速進. 4.. 資料容錯(Fault tolerance). 行分散式計算之目的，採用 Grid 技術來建構一個. 5.. 加速運算(Accelerate Computing). 聯邦式的醫院電子病歷分享環境，不失為一個可行. 五、討論. 的方式。. 集合眾多小型電腦之資源共享，來達到大型電腦之. 三、方法. 能力，是未來的趨勢，由於 Grid 技術尚在發展當. 本計畫預計在各醫院設立一類似 Storage. 中，但所幸已有一些組織努力推展中，如. Resource Broker[8]之伺服器，各醫院可將確認後並. GGF(Global Grid Forum)[10]等，各學術領域也開始. 提供查詢之病患資料存放於伺服器的 Public Area. 運用此技術來達到研究資源共享的目的，以醫院電. Folder，系統即可利用 Grid 技術串聯各醫院之伺服. 子病歷的 Append Only 特性來說， 更適合利用此. 器，形成一個單一檔案資料夾或資料庫，再利用. 一技術來達到資料共享及資源整合的目的。. Web Server 作為一 Middle Ware，讓使用者可以利 用瀏覽器來存取檔案資料或下達操作指令，例如 SQL 查詢等等，且利用 SSL 機制來達到安全性的 要求。. s11.
(12) MIST2002 Proceeding. 以網路服務為基礎之智慧型整合式糖尿病健康管理支援系統 Advanced Intelligent Web-Service-based Integrated Diabetes Health Management Support Systems 張國航 a、唐千惠 b、沈鈺翔 c、張博論. d. abcd a. [email protected]. b. 國立陽明大學衛生資訊與決策研究所 [email protected] c [email protected]. d. [email protected]. Abstract The key factor of successfully controlling diabetics and improving patient’s health status requires a seamless cooperation between health providers, patients, and his family. However, it is very difficult to make a complete caring plan in reality. The involvement of different health providers and visiting behavior of residents prevents the integration of patient medical data and self- education plans from different sources. The objective of this study is to provide a facilitate. A new integrated framework was proposed as a solution to Diabetics Care. With the new Web service. information. model,. client-server. techniques. We developed an integrated diabetes. management and processing system as a distributed. health management support system using the. application model. It provides a strong and stable. techniques of web service、PDA、SMS and IC card.. message exchange platform for our framework.. We will provide a demonstrating prototype of the. Mobile devices such as PDA, cellular phones were. system.. used as "message shuttle" between everyone who is. l. concerned. Moreover, the personal IC cards, which. Web Service、Diabetes Care Management、Decision Support System、Personal Digital Agent(PDA) 、IC Card. it. extends. the. existing. would be available national wide in the late 2002, were adopted for patient identification and essential information. By joining all advantages of emerging technologies, our framework would provide a complete supporting system for diabetics care.. After merging techniques we have discussed, we designed. the. Web-Service-based. “Advanced Integrated. Intelligent. Diabetes. Health. Management Support Systems” (illustrated below). Four sub-systems are included:. s12. infrastructure. using. advanced. IT. Key words:.
(13) MIST2002 Proceeding. 以網路服務暨 HL7 為基礎之聯合網路掛號智慧代理人系統之開發 The Development of Web Services and HL7 Based Universal Web Registry Intelligent Agents 廖玉里 a、張博論. b. a b a. 國立陽明大學衛生資訊與決策研究所 [email protected] b [email protected]. 摘要. feasibility of developing smarter web sites.. 本研究是以微軟.Net C#為開發工具，建置一套聯合. system responded to user’s requests for hospital. 網路掛號智慧代理系統網站與模擬二套醫院Web. admission on an intelligent agent, which was. Services系統，運用智慧型代理人的原理與技術，. developed using Simple Multi-Attribute Rating. 集結大台北地區醫療院所的預約門診表於聯合網. Technique (SMART). The system would recommend. 路掛號系統網站上，配合交談式對話的服務介面應. the users which doctors to see based on their. 用於網際網路上，依據事實與關係建置症狀科別知. symptoms and performance, and would make an. 識 庫 ， 利 用 簡 單 多 屬 性 評 比 技 術 (Simple. appointment on behalf of patients.. Multi-Attribute Rating Technique )之推論模式，提出. processed all information based on the HL7 standard. 推論就醫選擇資訊的結果供民眾就診參考，並將推. in an XML format. The registry was then done by. 論的結果轉換成HL7-XML的訊息交換格式，利用. the system to connect to the other web-services-based. 醫院所提供Web Services網路服務系統，直接代理. registry system provided by participant hospitals.. 民眾預約門診掛號、取消預約門診掛號、查詢目前. The system was run successfully and pointed the. 預約門診掛號人數等服務。. great potentials of developing more intelligent web. 未來在實施應用聯合網路掛號智慧代理系統上，除. sites.. 必須提升科別症狀知識庫的代表性與檢討修正簡. To make the results of this study to be more practical,. 單多屬性評比推論技術之外，應結合防火牆或公立. the system would need to be capable of catching. 認證機構，提供一個資訊交換安全的網路環境，並. user’s preferences and providing more accurate. 且找出院際間在預約掛號上的HL7 共同交換資訊. recommendation for seeing doctors.. 標準，突破.Net的技術，與其他家Web Services系統. issues would also need to be well taken care of to. 結合，真正達到網際網路上跨平台、跨語言、應用. ensure the secure transmission of user ’s personal. 系統與應用系統整合性的網路醫療服務。. health. According to the studies, the contents of the. information. on. the. web. The. The system. The security. and. appropriateness of appointment scheduling.. the Web. web sites in health care as well as the way we used. services appeared to be a good technique for the. those. for. hospitals to provide services on the web and the HL7. improvement. People are expecting not only more. could then guarantee the suitable exchange of. valuable information online, but also more intelligent. information between hospitals.. behaviors of web sites.. Key Words(關鍵字):. have. attracted. people’s. attention. Web Services 、 Intelligent Agents 、 HL7 、 Web Registry(網路掛號)、SMART. This research was to use the Microsoft .Net C# to. develop. the. web-services-and. -HL7-based. universal hospital web registry intelligent agents, taking this as an example, to technically test the s13.
(14) MIST2002 Proceeding. 台灣地區符合醫院評鑑等級醫院之網站內容普查報告 The Website Census for The Accredited Hospitals In Taiwan a. b. 蕭芳瑩. ab. 輔仁大學公共衛生學系 a [email protected]. c. 魏中仁. 張博論. c. 國立陽明大學衛生資訊與決策所 b [email protected] c [email protected]. 中英文關鍵字 Census 普查；Website 網站； internet 網際網路；Hospital Accreditation 醫院評 鑑. 整體來看醫學中心具有的網站總服務項目數 量最多，分開各項目醫院簡介、醫療相關服務與其 他資訊服務皆是醫學中心提供較多的服務，而地區. 英文摘要. 醫院在各部分的服務數量皆為最低。在權屬部分， 財團法. The websites have been used as an useful channel for service in many industries. Though some studies. 人醫院所提供的總服務項目，以及其他各細項都較. had been done on the websites developed by hospitals. 公立醫院及私立醫院網站項目服務的數量多，而私. in health care, there has been no related census data. 立醫院在各部分的服務數量皆為最低。整體服務數. available in Taiwan. This study analyzed 491. 量以及診療相關部分以台北區的服務數量為最豐. accredited hospitals on their websites. It was found. 富，東區其次，醫院簡介和其他資訊方面則以高屏. that only 30% had set up their websites. Among. 地區為最低。網路掛號的數量來看，醫學中心以經. these which had not, most of them were district or. 超過九成，區域醫院也有一半的設置比例，楊雅惠. private hospitals. Regarding. the contents of the. （2000）研究指出區域醫院只有 29.5%的比例，表. websites, the majority provided information , instead. 示網路的熱潮帶動醫院網路掛號的興盛，而其他研. of practical services such as consultation .. 究中亦指出，醫院開發網路系統一直在各醫院持續 進行中，可預期的是未來將會有更高的設置比例。. 中文摘要. 研究發現醫療領域中最接近民眾的地區醫院及民 間醫療院所，較不重視網路的發展，與現今商業的 網路開發有相當的距離，非常讓我們意外的是，民 間機構應該會更努力投入這個市場，但實際上則為 政府機構腳步上較為快速。而在服務項目上，大多 是提供 參考性質的服務，而互動性質的服務較不 普及，內容設計上有相當的提升空間。另外一點值 得我們憂心的是，城市與鄉間原本醫療資源及有差 距，作者認為應該是可利用網站服務來拉近其距 離，但網路上北部醫院網站服務的確較為豐富，因 此可能更是把原本醫療資源不均的鴻溝拉大。. 網站建置已成為提供服務的重要管道，在醫療 領域國內雖若干研究對於部分醫院網站內容分 析，但尚無一普及的完整分析的資料。本研究便是 針對國內 419 家符合醫院評鑑之醫院網站，以線上 瀏覽方式進行深度內容分析。研究發現，目前國內 僅有三成的醫院有架設網站。在評鑑別部分，醫學 中心大多提供 17(100%)家全部都有設置網站，區 域醫院 55 家(93.2%)，地區醫院共 415 家只有 21.9% 設置網站。地區別方面東區分局 15 家醫院就有 11 家設有網站，百分比達 73.3%，其次台北分局 43.7%，而其他分局大約設有網站比例為 20%-30%。權屬別方面，其設立網站以公立醫院所 佔比例最高 66 家有(85.7%)，財團法人 37 家 (77.1%)，私立醫院 60 家(16.4%)。 s14.
(15) MIST2002 Proceeding. 多媒體電腦輔助教學應用於臨床護理教學之初探研究 -以某地區教學醫院為例 莊美如 a、廖小瑞. b. 行政院衛生署嘉義醫院. [email protected] [email protected]. 摘要. 上述資料顯示護理人員最需要的是工作上立即可. 臨床護理教學是護理教育中最重要的一部份，但在. 用及能使病人立刻受益的課程。而在職教育實施方. 一般中小型醫院中，因受限於人力及成本的考量，. 面，護理人員認為最困難的是講師的聘請，其次是. 其知識來源主要是以師徒式的方式講授或派至醫. 經費，再其次則是工作人員的調派。對於多媒體電. 學中心受訓及參加研討會為主，此種方式除增加醫. 腦輔助教學之認知方面，知道什麼是多媒體電腦輔. 院成本外，也常造成人力調度之困難，因此尋找一. 助教學只佔受訪者的 22%，且其中對電腦輔助教學. 套有效的教材及教學方法是很重要的。. 的真正含意並不十分了解。. 隨著資訊科技的進步，傳統課堂講授的學習方式已 轉變為結合多媒體、網路等新形態教學，護理教育. 結論與建議. 當然也不例外，過去電腦輔助教學在機械、化工、. 由上述結果可知，護理人員所需之課程多屬基本護 理課程，其中以技能領域為最多，而技能領域的學 習則有賴課程發展的技巧才能達成，而這些課程也 多符合以多媒體電腦輔助教學來呈現，如以電腦課 程軟體呈現，不但教學品質固定且護理人員隨時都 可練習，且不受時間、空間的限制，惟目前此類軟 體在國內較為不足，而一般護理人員對於電腦輔助 教學之概念也較為生疏，因此本研究在最後嘗試實 作一個「中風病人的復健」護理教學課程的雛形系 統，以提供日後發展之參考。. 特教等均發現有顯著的成效，在護理上亦有學者進 行多媒體於護理教學的實驗，但國內醫院在臨床護 理教學上則很少使用此種方法。因此本研究試圖先 以半結構式問卷調查護理人員在職教育中最需要 之課程，並實際製作一個臨床護理教學系統的雛形 (prototype)以供參考。. 研究方法 本研究採半結構式問卷調查方法，以某地區教學醫 院內之護理人員進行資料收集。研究對象以門診、 病房之護理長及護理人員為主，共收集六十四份問 卷資料。. 研究結果 研究結果顯示受訪者認為最有用的課程前三名依 序為常見疾病之處理及護理、儀器操作方法、書寫 案例分析及個案報告的方法，而上述課程受訪者認 為最合適的教學方法依序為演講教學式、反覆練習 式、問題解決式。另本研究以護理人員分級制度所 設計之課程請受訪者自行排列優先次序，結果前三 名為形象塑造及禮儀認識、常見疾病的護理問題及 處理、身體評估。 s15.
(16) MIST2002 Proceeding. 行動化臨床資訊系統－以應用於加護病房為例 Mobile Clinical Information System Applied to an Intensive Care Unit 張芳聆 a. （一）. a. ，劉德明. a. 陽明大學衛生資訊與決策研究所. 摘要. 本系統的設計，主要是希望在提高病歷資料行動效. 加護病房之病患通常為重症需特殊照護者，因此病. 率的前提下，為加護病房醫護人員，設計一個電子. 床邊都會需要接上各式各樣的監測儀器，用來維持. 化的前端病歷輸入介面。. 病患的生命及監測病患的生命徵象，而病患的生命. 因此在系統架構的設計上，運用了嵌入式資料庫及. 參數必須由護理師定時記錄，以作為醫生臨床診斷. 非即時同步（Asynchronous Integration）的技術，. 時的重要參考。. 將嵌入式資料庫建置在個人數位助理端，使用者可. 隨著資訊技術的進步，病歷資料逐漸電子化，護理. 透過網路進行同步，先將資料儲存在嵌入式資料庫. 人員在床邊以紙本方式記錄病患生命參數後，必須. 內，避免因網路環境因素而影響系統的運作。本系. 再回護理站後，將抄寫的資料輸入病患照護系統，. 統架構，如 Figure 1 所示。. 造成醫護人員更多的文書工作量，尤其對於工作較 Nework LAN. 忙碌的 ICU 及急診室中的醫護人員，形成更大的 負擔。基於以上原因，我們希望在提高病歷資料行. 病歷報表. Bedside Monitor. MS Module. Oracle 8. Webtogo. 動效率的前提下，為加護病房醫護人員，設計一個. JDBC. 電子化的前端病歷輸入介面，協助加護病房護理人. Database. Mobile CIS Server. Alert System. Access Point. 員照護重症病患，維護病患病歷資料的完整與正確. Mobile Patient Record. 性，以提升醫療照護的品質。. Figure 1 系統架構圖. 本研究在開發以個人數位助理為基礎的行動化臨 床資訊系統。將個人數位助理配合無線網路技術， 架構出行動化臨床資訊系統，提高護理人員記錄病. （四）. 患生理資料時的行動化性，減少因病歷電子化後造. 本系統功能主要可分為二大部分，一為病患資料管. 成護理人員文書工作量增加的問題，並提供高品. 理功能，另一為病患生命參數資料的輸入功能。病. 質、高效率的醫療照護。. 患資料管理功能，提供了病患轉入、病患轉出、基. （二）. 本資料輸入等功能；而病患生命參數輸入功能方. 研究方法. 系統展示. 本系統以 Client / Server 方式架構，Client 端以. 面，提供了每日診斷及一般資料、生理資料、呼吸. Microsoft Windows CE Toolkit for Visual C++開發. 檢驗值、生化檢驗值、特殊用藥給藥滴數、監視器. 應用程式，採用 Oracle 9iLite 的嵌入式資料庫；. 資料的輸入等功能。系統主要畫面展示如 Figure. Server 端則包含 Oracle 8.0.5 Database Server 建構. 2、Figure 3：. 的臨床資訊系統資料庫，及 Oracle 9iLite 控制同步 機制的 Mobile Server。本系統功能主要可分為二大 部分，病患資料管理及病患生命參數資料輸入功 能。. （三）. 系統設計 s16.
(17) MIST2002 Proceeding 如下圖 Figure 4 所示。. Figure 4 系統同步畫面圖. （五）. 結論. 在病歷電子化後，病患的生命參數須由護理師在特 Figure 2 登入畫面. 定的時間到病床邊記錄，再以人工的方式將病患的 生命參數輸入電腦，這樣的處理方式，不但浪費人 力、時間，而且容易因重複輸入而產生錯誤，影響 資料的一致性及正確性。因此本系統的開發，主要 的使用者為加護病房的護理人員，讓護理人員可利 用 PDA，直接在病床邊進行護理記錄，再透過 Oracle 9iLite 提供的同步機制，將有更改或新增的 資料與後端的臨床資料庫（Clinical Database） ，進 行資料的同步。增加了病歷記錄的行動性，減少了 護理人員約 55％的文書工作時間，有助於醫療品 質的提升。 本系統的設計上，尚有安全上的加強及多媒體的加 入等方面，可繼續加強，因此建議此系統未來可加 強安全性上的考量、結合醫學影像資料、結合語音 辨識設備及整合醫院 HIS，將臨床資訊系統與醫院 既有資訊系統結整合，提高醫療資料使用的效率， 減少護理人員的文書處理時間，並提升醫療照護品 質。. Figure 3 系統主畫面. 當護理人員完成當天護理記錄後，可按下介面上 「同步」的按鈕，啟動同步的功能，則 MoCIS 應 用程式會向 Mobile Server 要求進行同步，Mobile Server 會先判別進行同步的使用者，並將曾經異動 過及新增的資料，透過網際網路送回後端的臨床資 料庫（Clinical Database）進行資料同步。本系統的 同步功能，必須先在 SERVER 端啟動 Mobile Server －Webtogo 與 MGP，才能開始提供同步的服務， s17.
(18) MIST2002 Proceeding. 運用 HL7 介面導入隨身醫囑之研究 李賢輝 a、陳澤生 a. b. 財團法人台灣基督長老教會新樓醫院資訊室、b 國立成功大學工程科學系 a [email protected] b [email protected]. 摘要 由於病患的病歷資料繁多，醫院為了醫師診斷的需. 目前使用架構雖然說大致上是一個穩定的架. 求而必須運用很多人力來記錄以及整理這些資. 構。但是，這樣一個系統是必須建置另一個資料庫. 料。隨著資訊科技的發展，醫院的作業、資料處理. 再透過 ODBC 來連結後端現行的資料庫主機，依. 及病歷都朝向電子化發展。此外，為了解決行動性. 預先設定的處理邏輯進行兩個資料庫的同步處. 的不足，許多的單位紛紛投入開發應用於醫師查房. 理。對於資料庫的安全性與一致性及額外的成本與. 之個人數位助理（Personal Digital Assistant，PDA）. 工作負擔，是需特別考慮的；並且透過 ODBC 來. 醫療輔助系統，協助醫師管理查房所需的資料，用. 聯結後端資料庫主機，系統運作的效能與安全控. 以節省人力資源、減少人為疏失，並掌握時效，以. 管，都會受到些影響。. 提昇醫療照護的品質。. 其次，醫療資訊交換標準第七層(HL7)的發展. 另一方面，醫療資訊交換標準第七層(HL7)的. 越來越受到重視，運用 HL7 標準與其他現有醫療. 發佈與運用技術的逐漸成熟，讓醫療系統間資訊交. 系統的結合也正逐漸發展中。. 換有了一致的標準。. 本文的目的是希望能將 HL7 訊息交換標準之運作 模式，導入個人數位助理輔助醫囑系統中，在目前 新樓醫院資訊系統以 IBM AS/400 為資料庫主機的 架構下，建立一套利用 HL7 訊息來傳遞臨床資訊 的隨身醫囑系統。. 目前已有多家廠商投入個人數位助理醫療輔 助系統的開發工作，也有多家醫療院所，正紛紛投 入相關領域的研究或是推行導入。常見的個人數位 助理醫療輔助醫囑的系統，多為三層式架構： 第一層、前端使用者介面層：為個人數位助理並搭 配微型資料庫，為醫師實際操作及記錄診斷、開立 醫囑之工具。 第二層、中繼端邏輯及資料交換層：主要工作為建 立與個人數位助理之連結以及與後端資料儲存層 之同步機制。 第三層、後端資料儲存層：為資料實際儲存之實 體。 同步機制在運作時，依據已事先設定之資料同 步時所需進行的動作、資料表的對應關係及資料處 理的動作與資料衝突時處理原則 ，透過 ODBC( Open DataBase Connectivity )功能與後端的 關聯資料庫主機進行同步。. s18.
(19) MIST2002 Proceeding. 網際資料管理系統在整合型臨床醫學研究上的應用 The Application of Web-based Data Management System in Integrated Clinical Research 江淑瓊 ac，何橈通 ab，詹文泰 c，胡啟民 b，郭英調 ab，陳曾基 b，劉媛芝 ac，李中原 c，唐大鈿 ac 台北榮民總醫院生物醫學統計小組 a、教學研究部 b、資訊室 c [email protected] 之粗估盛行率：高血壓(36.4%)、心臟病(21.3%)、 糖尿病(14.1%) 。 醫學研究資料管理有其基本的共同需求，但對系統 功能的要求，隨著不同的研究計畫本質而有輕重之 分。臨床試驗的研究計畫，對「線上即時輸入的功 能」與網路「回應速度」就有很高的要求；但對於 像社區醫學，以訪查追蹤為主的流行病學研究， 「查 詢功能」的完備就很重要。對於整合型臨床醫學研 究，網際研究資料管理系統是個極為有力的工具， 再左以完善的行政管理制度，彈性運作，確可滿足 臨床醫學研究團隊：醫護藥技之專家、生物統計學 家、流行病學家等不同的需求，達成整合一體，發 揮全方位的成效。. 摘要 本研究主要在探討網際網路研究資料管理系統，應 用在多中心臨床試驗及整合型社區醫學研究上的 功效。藉由網際網路的即時性、無空間阻隔性和資 料品質上的正確有效性及使用者介面的友善度，期 能達到無紙化的信任；進而整合各研究子計畫之共 同相關資料，使之正確有效地提供即時資訊及安全 合適地的進行資料交流。 利用 ASP(Active Server Pages)做為 Web 伺服器端 之程式語言，接收 Client 端以 Browser 與 JavaScript. 關鍵字：Web-Based Data Management System， Clinical Trial，Community Medicine. 回傳之資料，聯結 Microsoft Access 資料庫做最終 之存取運作所建立的『研究資料管理系統』，在網 際網路的平台上展現極佳的使用友善度，對資訊的 即時傳播，達到沒有空間障礙。簡單的應用卻帶給 研究人員相當大的便利。 利用多中心臨床試驗資料測試線上輸入功能，並進 行「雙重輸入」(Double Key In )比對，發現線上輸 入之資料正確性很高，錯誤率僅為 1.73%。用專屬 伺服器鎖定 IP 位址，以硬體設備來加強資料使用 之安全性管控，也達到很好的效果。藉由分離資料 庫的設計達成各研究中心資料隱密性的保證。對臨 床試驗計劃經由網際資料管理系統，進而省略個案 報告表(Case Report Form) 提供先趨性可靠度的探 討。 預防保健是建構「全民健康」的保證，更是二十一 世紀衛生政策的重點。台北榮總「石牌社區醫學服 務研究計畫」希望透過有系統的追蹤，建立一個理 想的社區醫學示範，是一涵蓋 14 個子計劃的整合 型研究。彙整各子計劃收集之資料，篩選共同研究 項目，透過網際資料管理系統的各種查詢功能，達 到資訊即時交流，資源共享，節省訪查追蹤的人 力。就彙整之 3487 筆資料分析，結果發現石牌社 區 不 論 受 訪 者 本 人 或 其 親 屬( 父 母 子 女 兄 弟 姊 妹)，都以患有高血壓之比例為最高。受訪者罹病 s19.
(20) MIST2002 Proceeding. 糖尿病患者人口結構及檢驗項目組合之探討 謝建成 a、葉秀真. b. 佛光人文社會學院 資訊學研究所. 摘要. 料。 糖尿病人口叢集實驗時的考量變數為性別、地區和 年齡，而作檢驗項目組合關聯規則實驗時的考量變 數為檢驗項目。. 根據我國國民十大死因統計，民國 84 年至 90 年， 糖尿病皆占第五位。又根據民國 86 年行政院衛生 署之「國民營養健康狀況變遷調查」 ，發現 45 歲以 上民眾罹患糖尿病盛行率為 11%，糖尿病對國人的 健康及社會的負擔影響甚巨。故如何有效的預防糖 尿病，正是我國各醫療主管單位所要突破的地方。 本研究即是希望利用資料探勘技術，找出個案醫院 糖尿病患者的人口結構以及糖尿病患者(診斷前三 碼為 250)的檢驗項目之間的關聯性，以作為「預防 勝於治療」醫療參考。 關鍵字 糖尿病、叢集分析、關聯規則分析。. 肆、資料探勘結果 糖尿病人口叢集的資料探勘結果：以男性較女性偏 高的患病率，佔顯著比例的年齡層是以 65 至 73 歲居多，但地區的差異性不大，表示糖尿病罹病率 與年齡、性別關係比較顯著，但與居住的地區別關 係不顯著。 其檢驗組合關聯性的結果：在糖尿病患者的檢驗項 目中最具有關聯性的是以信賴度 100%的腦血管核 子醫學檢查組合，原因是糖尿病患者容易併發大、 小血管疾病，前者有心肌梗塞、中風等，後者有視 網膜病變、腎病變等併發症，所以糖尿病患者常須 監控其腦血管疾病。次之的是 B 型肝炎與 C 型肝 炎的檢查組合，因有 C 型肝炎者(檢查項為 Anti-HCV)大多會再檢查是否有 B 型肝炎(檢查項 為 HBsAg)，此組合的關聯信賴度是 92.68%，但有 B 型肝炎未必會再檢查 C 型肝炎，所以關聯信賴度 略低是 88.85%。 此次檢驗組合關聯性的結果，經與臨床新陳代謝科 醫師的訪談發現，實驗結果是合理的。. 壹、研究動機 糖尿病是種高發生高費用之疾病，每年全國之醫療 費用 11%花在糖尿病治療上。且罹患糖尿病的比率 更有逐年上升的趨勢，對國人的健康及社會的負擔 影響甚巨。因此，糖尿病的防治工作，為目前醫療 主管單位首要的健康議題。台灣糖尿病人口約佔總 人口的 4%，即約 90 萬人，其中就診的只有 60 幾 萬人，大部份的人都未被診斷出來，因此本研究是 就宜蘭地區個案醫院糖尿病人口資料，利用 IBM Intelligence Miner 之叢集分析(Clustering Analysis) 及關聯規則分析(Association Rule Analysis)分析工 具，探究糖尿病人口結構及檢驗項目的關聯性，以 達預防治療之效。 貳、研究方法 目前我國針對糖尿病患者所做過的研究，大多針對 盛行率(以各種不同之診斷標準表示)依年齡、性別 來比較，或糖尿病患者之死亡率、生活型態、運動 情形、家族史分析、慢性病併發症和用藥統計等， 而以上的研究資料多數是以統計的方法完成。 本研究則是利用資料探勘技術之叢集分析及關聯 規則分析技術，應用在糖尿病患者的資料上，以建 構其人口結構及檢驗項目組合之分析結果。. 伍、研究限制 上述研究祇能考量到交易量，而無法將病患之資料 量（Information Size）列入考量，其主要因素為每 位病患之病情不同；反之，醫師對病情不同的患者 亦會有不同之診察與治療，在醫令方面開立上，如 藥品、檢驗項目等就會不相同。因而在病患資料 中，亦有存在不同差異性。. 陸、結論與未來研究方向 此研究僅就人口結構及檢驗項目的關聯作實驗，未 來可再朝向醫師用藥項目的關聯性，或混合式的找 出用藥與檢驗項目之間的關聯性做探討。而糖尿病 患者之人口特質資料，更可由多層面去探討，並考 慮不同的資料探勘技術研究。. 參、資料說明 資料來源是宜蘭某一醫院 90 年下半年的門診資 料，其中糖尿病患者定義為診斷前三碼為 250 的患 者，兩個實驗的統計檔： 一. 糖尿病患者人口檔：共約 18000 筆實驗資料； 二. 糖尿病患者檢驗明細檔：約 20000 筆實驗資 s20.
(21) MIST2002 Proceeding. 應用軟式計算技術於醫療診斷之研究— 以肝病診斷為例 林錦昌. 駱至中. 王鄭慈. 佛光人文社會學院 資訊學研究所. 摘要. 類神經網路模式，其中：鹼性磷酸脢(ALP)、麩草 酸轉胺脢(GOT)、麩丙酮酸轉胺脢(GPT)、轉磷脢(γ -GT)、總膽紅素(T-Bil)、直接膽紅素(D-Bil)、血清 總蛋白(Total Protein)、白蛋白(Albumin)、球蛋白 (Globulin)、澱粉脢(AMY)為輸入變數，配合以病 患的主次診斷為輸出資料。全部資料共計有 816 筆，其中有肝病患者為 384 筆而無肝病患者為 432 筆。為配合類神經網路的訓練，亦將其分成訓練資 料集(408 筆，50%)、測試資料集(204 筆，25%)、 驗證資料集(204 筆，25%)等三群。除本研究提出 之整合式模糊類神經網路模式外，也以相同資料測 試其他常用的倒傳遞類神經網路、半徑基底函數類 神經網路及 C4.5 決策樹分類模式等單一的軟式計 算技術，以比較整合與單一之間的效能差異，而其 結果之彙整如表 1 所示。. 肝病是台灣民眾罹患率極高的一項重大疾病，也因 此常被稱之為我們的「國病」。對於肝病的診斷， 由於初期病症不明顯，醫師們大都是依據個人的經 驗法則來進行評估與診斷，所以診斷錯誤的可能性 是存在的。因此，本研究嘗試結合軟式計算 (Soft Computing)領域中模糊理論 (Fuzzy Theory)、類神 經網路 (Artificial Neural Network)、主成份分析 (Principal Component Analysis)等理論，將它們應用 於醫療肝病診斷及預測，期望得到一個良好的智慧 型疾病診斷評估模式，以此系統輔助醫師進行肝病 的診斷，進而提昇醫療品質。 關鍵詞： 軟式計算、模糊系統、類神經網路、半 徑基底函數類神經網路、主成份分析。. 本文 肝臟疾病為國人所罹患疾病之首，更甚者﹕ 所有的肝病在患病初期症狀並不明顯，惟當 病情相當嚴重時才會出現明顯症狀，但對大 多數的患者而言已經是為時晚矣。目前肝病 的診斷及預測均由肝功能檢查結果來做為診 斷的主要依據﹔在醫師診斷方面，得仰賴醫 師個人的經驗而缺乏特定且絕對有效的判斷 法則，因人為因素而發生診斷及預測錯誤的 情況時有所聞。若能建立一套智慧型的肝病 診斷預測模型來協助醫師提早發現肝病症， 儘早施予有效治療，將能防治肝病並提昇醫 療品質。為此，在本研究中，我們將模糊理 論、主成份分析和類神經網路中的半徑基底 函數類神經網路等三種軟式計算技術加以整 合應用來建置協助肝病診斷的智慧型決策支 援系統，圖 1 所示即為所提出之整合式智慧 型系統的基本運作模式。 知識庫. 模糊 推理. 主成份 分析. 半徑基底 函數類神 經網路. 表 1、各系統效能比較表 系統模式. 分類正確率. 整合式模糊類神經網路. 83.1%. 倒傳遞類神經網路. 77.0%. 半徑基底函數類神經網路. 76.0%. C4.5 決策樹分類模式. 78.3%. 本研究模式的特色在於：先以模糊推論系統的模糊 規則對輸入值加以模糊化再經由主成份分析，將類 神經網路的輸入維度最佳化，最後再利用半徑基底 函數類神經網路來做為疾病診斷預測。如此將可使 其不受面臨絕對性歸屬的選擇，以提高其診斷正確 率。由表 1 中，我們可明顯看出本研究之整合式模 糊類神經網路模式，藉由模糊系統及主成份分析等 軟式計算技術來彌補單一半徑基底函數類神經網 路的不足，不但在分類正確率上高於倒傳遞類神經 輸出 預測 網路以及其他的單一模式的正確率。此外，其中之 模糊推論系統提供了模型解釋能力，這又比一般的 類神經網路更具使用親和性與透明度。本研究所提 出之系統的優異性由此可見。. 圖 1、整合式模糊類神經網路肝病診斷模式 本研究以東部某一區域醫院肝膽胃腸科病患的檢 驗資料來訓練所提出之智慧型系統- 整合型模糊 s21.
(22) MIST2002 Proceeding. 應用學習路徑與知識管理於健康醫學教育網 Integrating the Learning Pathway and Knowledge base Management into a Web-Based Learning System for Health Education 詹如晴 a、徐建業 a、陳恆順 a. 台北醫學大學醫學資訊研究所. b. b. 台灣大學醫學院醫學系、醫學資訊組、家庭醫學科. [email protected]、 [email protected]、 [email protected]. 摘要. with and follows the theory of health-and-physical. 民眾對健康醫學知識的關注程度，有明顯增加. education course, which is a part of the so called the. 的趨勢，為提升其自我健康照護的知能，擁有免於. Nine-year. 疾病恐懼的知識能力。本研究整合消費者健康資訊. Elementary and Junior High Education in Taiwan.. （CHI – Consumer Health Information）和數位學習. The health education learning system built in this. （E-learning）系統發展的特點，並和教育部中小. study is also one of the six major learning systems. 學健康與體育領域九年一貫課程主軸架構理論相. described in the blueprint of information education. 結合，建立一不受時空限制的健康醫學教育網路。. infrastructure promoted by the Ministry of Education.. 健康醫學學習網是教育部「資訊教育總藍圖」計劃. The content structure of the system is designed by. 中六大學習網之一，內容架構的規劃採納國內近一. more than a hundred experts who are devoted to. 百位中西醫相關醫護背景專家和中小學老師等教. public health works including medical doctors,. 育專家的建議，系統平台以多媒體知識庫為基礎，. nurses, professors, school teachers, and so on. The. 整合視覺化操作介面、學習路徑和 FAQ 自然語言. system platform is designed by using a Multimedia. 知識庫管理系統，本學習網將提供全民健康醫學知. Knowledge base and integrated with visualized user. 識理論、發展健康教育教學輔助資源並成為健康醫. interface, learning pathway and FAQ natural. 學學習入口網站，讓民眾從做中學的過程中，促進. language management. HELS will provide the. 健康和預防疾病。. general medical knowledge for public, give help on. Comprehensive. Curriculum. for. the. the development of the assistant teaching materials. Abstract. and become a portal website of Consumer Health. According to the advance of the medical. Information. Following the “learning by doing”. informatics, health related issues are more concerned. methodology, the public will obtain more health. by the public recently. It is very important that a. knowledge and lead to the promotion of national. person has the capability of increasing health. health and prevention of disease.. knowledge and can avoid the fear of disease. In this study, we have established an Internet Health. 關鍵字. Education Learning System. Health education, E-learning, Consumer Health Information, Learning pathway, Knowledge base. (HELS) which is. based on the ideas of Consumer Health Information (CHI) and E-learning. This study is also combined. s22.
相關文件
The Senior Secondary Health Management and Social Care Curriculum and Assessment Guide incorporates the key recommendations in the Curriculum Development Council's
嚴重精神病 (Severe Mental Illness, SMI) 、一般精神障礙 (Common Mental Disorders, CMD).
Topic 4 - Promotion and Maintenance of Health and Social Care in the Community 4CAspects of risk assessment and
• Centre for Food Safety, Food and Environmental Hygiene Department – Report of study on sodium content in local foods. • Centre for Food Safety, Food and Environment
between the roles of the individuals (private sector) and the public or government in the provision of social care and health services responsibility of the government, e.g.
• The Tolerable Upper Intake level (UL) is the highest nutrient intake value that is likely to pose no risk of adverse health effects for individuals in a given age and gender
Key concepts :personal growth (family roles) , family relationship, family problems, social welfare system, interpersonal relationship, communication among family members,
• The Health Information Technology for Economic and Clinical H ealth (HITECH) Act seeks to improve American health care deli very and patient care through an unprecedented
