Skip to content
Home / Machine Learning in Medicine

Machine Learning in Medicine

Udo Hoffmann, MD (Harvard) recently delivered a distinguished lecture, titled “Machine Learning and Population Imaging – The Next Frontier of Cardiovascular Medicine.” In his talk, Prof. Hoffmann explained about many of the current challenges and implementation of machine learning in medical research, especially in cardiovascular and population studies. This is important as around 80% of all patients receive some sort of imaging, which greatly burdens health care professionals in clinical setting. He added that the current image segmentation and analytic strategies are inadequate to capture the complexity of diseases, even though recent advances have increased computer capacities and created opportunities to analyze multi-level patient data. In addition, Prof. Hoffmann discusses machine-learning-based Phenomaping as one of the most promising technologies to automate image assessment and identify new biomarkers, both of which are important in complex disease pathology. The lecture was then immediately followed by townhall discussion, where students were able to ask follow-up questions regarding the role of medical physicists in machine learning research and other career-related questions.