A novel approach to retinal blood vessel segmentation using self-supervised Masked Autoencoder (MAE) pretraining followed by supervised fine-tuning. This work demonstrates the viability of vision ...