Machine Learning in medical image application for diagnosis, modeling and treatment guidance

With rapid economic and demographic changes in an aging society, neurodegenerative diseases, cardiac disease and cancers are becoming serious healthcare problems, affecting social, economic and medical systems globally. The diseases typically present pathologies in deficits of functions, vascular and metabolism, which caused chronic burdens for families and societies. Due to limited or no therapeutic available for many of the diseases, it is imperative to understand the pathogenesis, to improve methods of screening and early detection, and to develop interventions to delay or stop the progression of disease-related pathologies. Clinically, imaging technique is widely used as non-invasive approach to measure the parameters, detect the changes and monitor the progress of organs and lesions. However, in the current clinical practice, heavy workload is required to locate and delineate the organ and lesion regions in the medical images in order to find the changes, characterize the disease, and/or develop a precise planning and treatment model. In this talk, I will brief some of the research works by using machine learning and computational image analysis for medical applications.