A Stroke Detection System Based on Cincinnati Prehospital Stroke Scale

Abstract

Stroke, also known as cerebrovascular events, is mainly caused by the obstruction of blood flow in the brain, which leads to an inability to supply oxygen to the brain. By 2013, stroke had become the second most common cause of death in Taiwan (accounting for about 12% of all deaths). This study used image processing technology and speech recognition, following the Cincinnati Prehospital Stroke Scale, to determine whether or not the user had a stroke. The Cincinnati Prehospital Stroke Scale has three indicators, including facial droop, arm drift, and speech. Patients with 1 of these 3 findings have a 72% probability of having had an ischemic stroke. If all satisfies, the probability of the stroke is more than 85%. In addition, we developed a mobile APP based on this method to detect whether or not the user had a stroke, and hope to reduce stroke hazards.