CoroSAM: adaptation of the Segment Anything Model for interactive segmentation in Coronary angiograms
Integrating specialized Convolutional Adapters and channel-encoded point prompts enables accurate delineation of major coronary vessels with minimal user intervention. CoroSAM's architecture facilitates efficient inference on standard computing hardware without dedicated GPUs, providing a practical tool for clinical applications. This approach establishes an adaptation framework that effectively balances segmentation accuracy with computational efficiency, making it suitable for routine analysis workflows.
Authors
- Eloisa Arbustini (ORCID: https://orcid.org/0000-0003-2948-7994)
- Michela Ferrari (ORCID: https://orcid.org/0009-0005-6066-2051)
- Edoardo Spairani (ORCID: https://orcid.org/0009-0005-1113-0075)
- Francesco Prati (ORCID: https://orcid.org/0009-0001-9851-7862)
- Giovanni Magenes (ORCID: https://orcid.org/0000-0002-7558-1490)
- Mario Urtis (ORCID: https://orcid.org/0000-0003-3289-9228)
- Maurizia Grasso (ORCID: https://orcid.org/0000-0003-2611-4295)
- Francesca Sessa
- Antonio Tescari (ORCID: https://orcid.org/0009-0008-3297-8085)
Institutions
- University of Pavia (IT)
- Saint Camillus International University of Health and Medical Sciences (IT)
- Policlinico San Matteo Fondazione (IT)
Publication Details
- Journal
- Computer Methods and Programs in Biomedicine
- Published
- 2025-11-20
- DOI
- https://doi.org/10.1016/j.cmpb.2025.109172
- Primary Topic
- Medical Image Segmentation Techniques
- Type
- article
- Field-Weighted Citation Impact
- 0.00