Application of Machine Learning Techniques for The Prediction of Decompressive Hemicraniectomy Prognosis in Acute Ischemic Stroke

Stroke is one of the leading causes of death in the world with the number of people suffering from it increasing every year. Ischemic strokes, one of the two main types of stroke, occur when blood clots block brain arteries which leads to infarction eventually leading to brain edema. If not addressed quickly enough it may lead to disability and in worst case scenario may even lead to death. In this thesis we proposed a machine learning based MATLAB tool that aids in speeding up the prognosis of acute ischemic stroke patients. From a set of patient medical data such as patient age, blood pressure reading and infarction volume from first CT scan, we created three prediction models which predict second infarction volume, decision for surgery and treatment time. We also experimented with utilizing the technique of feature reduction and implementing Fuzzy Inference System to consider improving the generated models and combined the best performing models into a MATLAB application.