Image classfication for diagnosis of neurological disorders-Deep Dive

The deep learning revolution has generated much interest in its use in medicine. Deep learning belongs to a family of machine learning tools and directly learn from the data. By contrast standard machine learning tools require an extra feature extraction step. It can be used in supervised or unsupervised methods. Several different approaches exist for deep learning and are available in data science platforms. An area of interest is the use of supervised deep learning to classify different disorders that the clinicians face. In earlier works, we have explored this issue using probabilistic classification of disease by voxel location on MRI scans for stroke and white matter hyperintensity. This project required a user extraction of the data by segmenting stroke and white matter hyperintensity. In this project, the aim is to apply deep learning methods and compare them with task specific machine learning tools to classify neurological disorders (stroke, multiple sclerosis and white matter hyperintensity) from medical images.