Lower Limb Calf Muscle Segmentation from Diffusion-Weighted Magnetic Resonance Images Using Deep Learning

<p dir="ltr">Peripheral artery disease (PAD) affects blood flow to the limbs, and diffusion-weighted magnetic resonance imaging (DW-MRI) can quantify microvascular perfusion and diffusion in calf muscles, aiding diagnosis. However, manual segmentation is time consuming and subjective. We propose a conditional generative adversarial network (cGAN) with an enhanced U-Net architecture for automated segmentation of calf muscles from DW-MRI. Our method leverages data augmentation to address small dataset sizes, splitting images into left and right halves and applying flipping and progressive rotation. Evaluated on datasets of healthy and PAD patients, our approach achieves average Dice Similarity Coefficient (Dice) scores of 54.86% to 79.85% across muscle groups, significantly outperforming baseline models (original U-Net architecture with cGAN and no data augmentation). This work demonstrates the potential of deep learning for automating segmentation in PAD diagnosis, offering a scalable solution for clinical applications.</p>