Enhanced Computer-Aided Scoring System for Stroke CT Images Using a Priori Knowledge and Case-Based Reasoning

Stroke is a medical emergency and can cause permanent neurological damage, complications, and even death. It is the leading cause of adult disability in Taiwan right now. Recently, IV tPA has emerged as an effective treatment if less than 1/3 MCA (Middle Cerebral Artery) territory is involved. Unfortunately, human reading which suffers from intraobserver and interobserver discrepancy is the norm. The purpose of this study proposed herein is twofold: to develop a computer-aided image scoring system that reduces interobserver as well as intraobserver discrepancy in interpreting non-contrast CT brain images of acute stroke patients, and to push the scoring accuracy of such a system beyond that achievable by expert physicians using unaided eyes. The proposed study is built on the success of the preceding pilot study which showed promising results. We have proposed to take advantage of a priori knowledge embedded in the medical records, and case based reasoning to enhance the scoring accuracy. We plan to retrieve 300 acute stroke cases retrospectively selected from patients with acute stroke symptoms who came to Chang Gung Hospital for treatment over the years for the proposed study.

Project ID:PB10107-1747
External Project ID:NSC101-2221-E182-058