神经急危重症计算机辅助诊断系统的研究

目的探讨CBR技术在神经内科急危重症计算机辅助诊断中的应用及其辅助诊断软件的开发。方法以神经内科学权威教材和临床专家的经验,确定神经内科急危重症疾病的诊断要素及初步确定其诊断权重值,建立起疾病原型库,采用Delphi7和SQLServer2000进行软件系统的开发。结果原创开发了具有独立知识产权的神经内科急危重症病例库及其辅助诊断系统。结论CBR技术可应用于神经内科急危重症疾病的辅助诊疗,具有它独到的优势。

Research on Computerization of Auxiliary Diagnostic System for Neurocritical Care

Objective To discuss the computerization of auxiliary diagnostic system for neurocritical care (CADSNC) based on Case-based Reasoning (CBR) and its possible use in neuro intensive care unit. Methods Diagnostic factors and retrieval algorithm are the focus of the CADSNC,and they determine the speed and accuracy of the whole system. Firstly,based on knowledge of neurological clinic experts and authoritative textbook,diagnostic factors and their weight were established from six aspects,including semiology,anamnesis,check-up,medical imaging,electrophysiology and clinical laboratory medicine. Secondly,the better algorithm was looked for from related retrieval algorithms. Thirdly,case-based database was founded by picking up the electronic medical records (EMR) in hospital information system. Finally,CADSNC was developed with Borland dephi 7 and Microsoft SQLServer 2000. Results Diagnostic factors and their weight were confirmed for each neurological disease.Nearest-neighbor match was selected as the retrieval algorithm of CADSNC. Neurological case-based database was establihed. CADSNC was established and was made an attempt to applied in clinic practice. Conclusions CBR is a successful technique for CADSNC in neuro intensive care unit.