Stroktor-A System to Predict Ischemic Brain Stroke using Learning Techniques

Abstract

According to WHO, over 15 million people suffer from a stroke which causes 5 million deaths and approximately 30% of survivors are facing serious disability. CDC (Centers for Disease Control and Prevention) has identified stroke as the fifth-leading cause of death globally. More than 70% of strokes are first events, hence making primary stroke prevention is particularly an important aspect. With the availability of a system to detect the brain stroke with the early symptoms occurring in patients would lead to early diagnosis and prevent severe consequences. This research work is dedicated to build a system that would detect brain stroke with premature symptoms and generate accurate results using neural networks and Computer Vision. Considering the severity of stroke, it is necessary to immediately consult medical practitioners to prevent the consequences which is a tough task. This state of art focuses on obtaining stroke possibility based on change in facial features as prominent symptoms and providing immediate precautionary measures. Along with this, providing an interactive platform for users to connect with available doctors using video calling for immediate consultation.