Stroke Cognitive Medical Assistant (StrokeCMA)

Abstract

Stroke is the number two killer after heart disease since it is responsible for almost 10% of all deaths worldwide. The main problem with a stroke is a significant delay in treatment that happened mainly due to inappropriate detection of stroke symptoms or inability of a person to perform further necessary actions, and might cause death, permanent disabilities, as well as more expensive treatment and rehabilitation. Nowadays assessment of a stroke is done by human, following widely adopted FAST approach of stroke assessment. Since a human factor become one of the causes of treatment delay, offered solution will try to minimize this factor. Artificial Intelligence, Cognitive Computing, Machine Learning and Data Mining, NLP and other technologies make possible to elaborate a smart solution that enable automated stroke symptoms detection on earlier stages without self-assessment or assistance of another person, solution that in time provides notification to corresponding caregivers (family members, responsible medical worker, etc.) and even able to directly call emergency, explaining the cases and providing all necessary evidences to support further decision making. Thus, the paper presents feasibility study of IBM Watson cognitive computing services and tools to address the issue of automated stroke symptoms detection to elaborate smart supportive tool in the pocket of people under high risk of a stroke attack.