Development of a Knowledge-Based Clinical Decision Support System for Community Medicine using PDA

本研究旨在开发一个可移动的、基于知识的社区医学临床决策支持系统，并对所开发的系统进行现场评估。该系统具有客户端-服务器体系结构，并包括以下组件：(1)服务器系统：(A)从存储在医院信息系统上的医疗数据生成诊断知识的知识生成子系统。(B)根据生成的知识推断可能的诊断候选的推理系统。(C)从客户端系统接收查询并返回推理结果的服务器应用程序。2.客户端系统：提供输入患者症状和显示推理结果的界面的客户端应用程序。2004年，我们在岛根大学配备了用于软件开发的服务器系统，并在其上实现了先前开发的知识生成系统和推理系统。然后，我们开发了一个与客户端系统进行通信的服务器应用程序，并在一台Note PC和一台PDA上进行了通信实验。2005年，我们在2台Note PC和8台PDA上实现了客户端系统。之后，我们在一家医院进行了社区医疗岛曼网络成员的现场沟通实验，从而确认了该系统可以在个人手机系统服务的支持领域下与笔记本电脑一起使用。然而，我们也面临着以下问题。(1)目前的PDA软件开发环境适合于构建用户界面，但不足以构建智能客户端系统。因此，我们必须在PDA上使用Web界面。(2)在一些人口稀少的地区，小灵通服务不可用。因此，我们不得不使用其他通信设备。这些结果表明，基于PDA的移动远程医疗系统在当前区域通信环境中的使用存在局限性。

This research aims at developing a mobile, knowledge-based clinical decision support system for community medicine and evaluating the usefulness of the developed system on-site. The system has a client-server architecture and consists of the following components.1.server system : (a)A knowledge generation subsystem that generates diagnostic knowledge from medical data stored on the hospital information systems. (b)An inference system that infers possible candidates of diagnosis based on the generated knowledge. (c)A server application that receives queries from client systems and returns inference results.2.client system : A client application that provides an interface for inputting patient's symptoms and displaying the inference results.In 2004, we have equipped server systems for software development on Shimane University, and implemented the previously developed knowledge generation system and inference systems thereon. We then developed a server application that communicates with client systems, and performed communication experiments with a note PC and a PDA.In 2005, we have implemented the client systems on 2 note PCs and 8 PDAs. After that, we performed an on-site communication experiment in a hospital who is a member of the Shimane network for community medicine.Consequently, we have confirmed that the system can be used with note PCs under the support areas of personal handyphone system services. However, we've also faced the following problems. (1)Current software development environments for PDA are suitable for building the user interface, however, are not sufficient for building intelligent client systems. Therefore, we have to use a Web interface on PDA. (2)In some underpopulated regions, PHS services were not available. Therefore, we had to use other communication devices. These results suggested the limitation of the use of PDA-based mobile telemedicine systems in current regional communication environments.