Medical Imaging
Faculty in the department have been contributing to medical imaging research for several decades, and it is a rapidly expanding focus area in the Department of Computational Medicine. Current work includes clinical collaborations with Ophthalmology and Neurology, as well as basic science collaborations with Brain Mapping. Research in imaging science spans several subject areas: from the physics of image formation to 3D reconstruction and processing, to decision making based on machine learning and statistical inference. Computational techniques in medical imaging have the potential to lead to improved diagnostic accuracy and medical decision making, and a better understanding of both disease and normal biology.
One goal of medical image analysis is to quantify complex patterns in imaging data in order to make statements that are useful for clinicians and scientists. As an example, research in the Department involves studying patterns of brain tissue loss observed in MRI scans in older adults. Quantifying subtle differences between early Alzheimer’s disease and normal aging is leading to disease state biomarkers to be used in clinical trials. Another example involves studying the location and degree of brain lesions due to stroke and predicting which patients are likely to require medical intervention in the future.
This field draws on many different mathematical modeling, estimation, and inference techniques, with a unique emphasis on geometry. Computational challenges include designing algorithms that will scale effectively to high resolutions and large sample sizes. As imaging systems and clinical practice evolve, computational approaches are becoming essential to learn from this wealth of new data.