Clara Sorensen is a Ph.D. candidate in Bioengineering in the UC Berkeley-UC San Francisco Joint Program. Her research develops domain-guided machine learning methods for high-dimensional multimodal neuroimaging and biomarker integration in Alzheimer's disease.
She holds dual B.A. degrees in Computer Science and Biological Sciences from Wellesley College. Her work intersects AI, neuroimaging, and biomarker discovery, largely inspired by prior experience at leading pharmaceutical companies including Novartis, Roche, and Genentech. Prior to joining the group, she specialized in building deep learning pipelines for image segmentation, classification, and analysis across neuroscience and immunotherapy applications.
As a graduate student, Clara’s work integrates AI with neuroimaging to develop predictive models of Alzheimer’s disease. She has led efforts to model tau pathology using graph neural networks on multimodal brain data, engineering MRI-based alternatives to PET by developing mixture-based regularized regression methods, and studied plasma biomarkers for early PET detection using partial least squares regression. Her research emphasizes scalable, non-invasive approaches to biomarker discovery and clinical translation.