Title: Intelligent diagnosis system for jaundice based on dynamic uncertain causality graph

Authors: Nan Deng; Shichao Gengy; Shaorui Hao; Qin Zhang; Lanjuan Liz

Addresses: School of Software Engineering, Beijing Institute of Technology, Beijing, China ' School of Communication, Shandong Normal University, Shandong, China ' The First Affiliated Hospital, Zhejiang University, Zhejiang, China ' School of Software Engineering, Beijing Institute of Technology, Beijing, China ' The First Affiliated Hospital, Zhejiang University, Zhejiang, China

Abstract: The healthcare system in China still has some defects such as the imbalance of medical resources. With the development of computer science, medical diagnostics digitisation has become possible. In this paper, a medical diagnosis system for jaundice based on dynamic uncertain causality graph (DUCG) is proposed. After a brief introduction to DUCG, a knowledge base of the DUCG for jaundice diagnosis is built. It can represent the experienced knowledge of human experts explicitly with graphical symbols. During the construction of the knowledge base, how to classify the medical diagnosis data in a structural and standard manner is important. In this paper, we propose to classify these data as four categories: general symptoms, medical signs, results of laboratory tests and results of imaging examinations. The first two form the general clinical information and the last two are further information. Each category can be further classified as sub-categories, so that users are easy to find the right position to fill the clinical information. Based on such designed DUCG medical software, 203 randomly selected jaundice related cases out of 3,985 case records of a hospital are tested. The final diagnostic accuracy of the system reaches 99.01%.

Keywords: uncertainty; causality; probabilistic reasoning; medical diagnosis.

DOI: 10.1504/IJICT.2022.123159

International Journal of Information and Communication Technology, 2022 Vol.20 No.4, pp.439 - 462

Received: 13 Jul 2020
Accepted: 28 Sep 2020
Published online: 01 Jun 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article