AI in 3D Medical Images: Concepts, Milestones, and Opportunities
Speaker: Yangming Ou, PhD, Assistant Professor of Radiology; Affiliate Faculty, Computational Health Informatics Program; Faculty, Fetal-Neonatal Neuroimaging Data Science Center at Boston Children's Hospital
Dr. Yangming Ou briefly reviewed some major concepts and milestones of AI in medical images. The focus of Dr. Ou’s talk was on 3D medical images, for AI’s application in disease diagnosis, outcome prediction, early screening, neuroscience, and others. Dr. Ou then discussed some major challenges and potential opportunities, including further improving accuracy in detecting small diffuse lesions, and facilitating AI in small sample sizes.
Dr. Yangming Ou's research focuses on improving healthcare by medical image analysis, machine learning and imaging informatics. One pillar of Dr. Ou's lab is in developing MRI analysis and MRI machine learning algorithms. This includes algorithms for spatiotemporal MRI analysis, multi-modal MRI integration, quantification of normal brain development, detection of abnormalities, prediction of outcomes, treatment evaluation, and disease subtyping. Another pillar of Dr. Ou's lab is in applying advanced imaging informatics for clinical studies. The topics span mild, moderate and severe brain abnormalities. Mild conditions studied in the lab includes nutrition insufficiency and kids with in-utero exposure to opioid. Moderate conditions include hypoxic ischemic injury, rare diseases such as Sturge-Weber Syndrom, early screening of psychosis, dyslexia, etc. Severe diesease include pediatric brain tumor and how to improve neurocognitive outcomes in survivors. Dr. Ou holds a PhD degree in Medical Image Analysis and an MS degree in Applied Mathematics, both from the University of Pennsylvania. His BS degree was in Electrical Engineering (Biomedical Engineering Division) from Tsinghua University. His research interest lies in the intersection of big data, machine learning, data science, imaging informatics, and healthcare.
