Robust Coronary Artery Delineation in XCA Images using Enhanced Pre-processing and ResNet-U-Net segmentation

We used X-ray Coronary Angiograms (XCA) dataset, publicly available. In this research work we applied XCA dataset that includes, 134 X-ray coronary angiogram images, provided with precisely annotated ground-truth images from an experienced cardiologist. Each angiographic image in the dataset is a 300 × 300 grayscale image formatted in Portable Gray Map (PGM). The data were obtained from the Cardiology Department of the Mexican Social Security Institute.