Deep Vessel Segmentation Based on a New Combination of Vesselness Filters

• Published in: 2024 IEEE International Symposium on Biomedical Imaging (ISBI) pp. 1 - 5
• Main Authors: Garret, Guillaume, Vacavant, Antoine, Frindel, Carole
• Format: Conference Proceeding
• Language: English
• Published: IEEE 27-05-2024
• Subjects: Biological system modeling; Biomedical measurement; Brain; Computed tomography; Deep learning; Explainable AI; Image segmentation; Liver; Matched filters; Segmentation; Vesselness filters; Vessels
• ISSN: 1945-8452
• DOI: 10.1109/ISBI56570.2024.10635696

Vascular segmentation represents a crucial clinical task, yet its automation remains challenging. Because of the recent strides in deep learning, vesselness filters, which can significantly aid the learning process, have been overlooked. This study introduces an innovative filter fusion method crafted to amplify the effectiveness of vessel segmentation models. Our investigation seeks to establish the merits of a filter-based learning approach through a comparative analysis. Specifically, we contrast the performance of a U-Net model trained on original scans with an identical U-Net configuration trained on vesselness hyper-volumes using matching parameters. Based on two vascular datasets, our findings highlight improved segmentations, especially for small vessels, when the model's learning is exposed to vessel-enhanced inputs.