Comparative Assessment of CNN and Transformer U-Nets in Multiple Sclerosis Lesion Segmentation

Abstract

Multiple sclerosis (MS) lesion segmentation in magnetic resonance imaging (MRI) is essential for patient diagnosis and treatment tracking. In this study, we conduct a systematic comparative assessment evaluating CNN-based U-Net architectures against modern Transformer-based U-Net variants across standard benchmark datasets. The empirical findings highlight architectural trade-offs in local feature precision versus global context representation in neuroimaging.

Yunus Serhat Bıçakçı

Assistant Professor

Assistant Professor specializing in GeoAI, Multimodal Vision-Language Models, and Spatial Data Science.