Abstract: Brain metastases (BrM’s) are a complication for many primary cancers and occur in over 20% of patients. The standard of care for multiple brain metastases has shifted in recent years from whole brain radiotherapy to single-fraction or hypofractionated stereotactic radiosurgery. Patients with BrM’s are living longer, often returning for additional courses of treatment. In this presentation, we will explore how we may facilitate more precise and efficient BrM treatment planning using automatic BrM detection and segmentation with deep learning techniques. Moreover, we will discuss an automatic method for longitudinal tracking of BrM’s. This begins to address an urgent need for clinical monitoring and decision-making with quantitative assessments, and could support longitudinal outcome studies with large patient cohorts in the future. 

About the speaker: Dr. Dylan Hsu is currently the chief therapy physics resident at the Memorial Sloan Kettering Cancer Center in New York City. Dr. Hsu obtained his Ph.D in high-energy physics at the Massachusetts Institute of Technology, studying dark matter production and precision Standard Model measurements in 13 TeV proton collisions with the CMS Experiment at the Large Hadron Collider. Subsequently, he joined the MSK Dept. of Medical Physics as a postdoctoral research associate in 2019, under the supervision of Drs. Michalis Aristophanous and Åse Ballangrud. His research work at MSK has focused on tumor segmentation, clinical automation, and long-term management for patients with metastatic brain cancer.