Coen de Vente successfully defended his PhD thesis, titled “Towards Robust Deep Learning for Medical Imaging”, on January 29, 2025, at the University of Amsterdam. His research focused on improving the robustness of deep learning models for medical imaging applications in radiology and ophthalmology.

The thesis addressed three key aspects of robustness: ensuring performance on data similar to the training set, detecting ungradable or out-of-distribution data, and maintaining performance under slight data misalignments in deployment settings. Coen developed and validated novel approaches for these challenges, applying them to tasks such as COVID-19 triage using chest CT scans and glaucoma screening with retinal imaging.