Brigham and Women's Hospital/Dana-Farber
Boston, MA
Christopher Erkki Kehayias, PhD, is a postdoctoral fellow in medical physics at the Brigham and Women’s Hospital / Dana Farber Cancer Institute Department of Radiation Oncology. Christopher received his PhD in physics at the University of Pennsylvania in 2021. His dissertation involved volatile organic compound detection and disease diagnostics (including early-stage ovarian cancer) using DNA-functionalized carbon nanotube sensor arrays and AI algorithms. Christopher’s research focuses are: (1) automated medical image segmentation based on deep learning (DL), and (2) optimization of automated treatment planning systems.
Working in Dr. Christian Guthier’s research group, Christopher has developed a DL-enabled, fully-automated quality assurance (QA) tool that has been deployed at the BWH clinic with the aim of assisting clinicians with palliative spine radiation therapy (RT) treatments. The system leverages vertebral segmentations and associated anatomic labels from CT imaging generated by TotalSegmentator, a publicly available DL algorithm, and flags discrepancies between target levels indicated by a radiation oncologist and predicted levels based on computed dose coverage of the segmented vertebrae.
Christopher has also investigated workflows for improving site-specific treatment plans generated by automated treatment planning tools using prostate RT test cases to verify workflow performance. He has also developed DL segmentation models based on the nnU-Net model framework to enable automated segmentation of prostate and relevant organs at risk, and has incorporated these models with the auto-planning workflows to fully automate treatment planning for prostate RT. In addition, Christopher has developed tools for DL segmentation, dose estimation, and registration tasks for brachytherapy.
Disclosures: