Mengjin Dong earned a Ph.D. in Bioengineering at Penn, advised by Paul Yushkevich, with the dissertation “Detecting and Localizing Progressive Changes in Longitudinal MRI of the Hippocampal Region in Alzheimer’s Disease” (2023). Mengjin developed deep learning methods that measure disease progression directly from pairs of longitudinal MRI scans: DeepAtrophy (NeuroImage, 2021), which learns to tell real progressive atrophy from noise, and Regional Deep Atrophy (Imaging Neuroscience, 2024), which locates the regions where change is associated with Alzheimer’s disease. Earlier work at PICSL applied joint label fusion to multiple sclerosis lesion segmentation (2017).
Penn Image Computing and Science Laboratory