Hybrid Vessel Wall Segmentation for Assisted Annotation in CT Angiography

Bracci J, Katzmann A, Rist L, Vorberg L, Sühling M, Maier A (2026)

Publication Type: Conference contribution

Publication year: 2026

Journal

Informatik aktuell Gesellschaft für Informatik (GI)

Publisher: Springer Science and Business Media Deutschland GmbH

Pages Range: 285-291

Conference Proceedings Title: Informatik aktuell

Event location: Lübeck DE

ISBN: 9783658510992

DOI: 10.1007/978-3-658-51100-5_57

Abstract

Cardiovascular diseases are the leading cause of death worldwide, responsible for over 19.8 million deaths annually. Accurate delineation of the vessel lumen and wall is essential for quantifying vascular health, diagnosing disease, and guiding treatment. However, manual annotation, the clinical gold standard, remains time consuming, inconsistent, and prone to human bias, limiting reproducibility and comparability across studies. To address this, we present a fully automatic pipeline designed to assist and standardize vessel annotation. A multi-purpose segmentation network first provides an estimate of the outer vessel boundary, which is then refined through a gradient based algorithm that delineates the inner lumen contour. On a clinically annotated test set of n = 46 slices, the method achieved a Mean Absolute Error of 0.197 ± 0.109mm for the inner lumen and 0.284 ± 0.237mm for the outer wall. Only 12.16% of the inner and 19.97% of the outer contour lines required manual adjustment. This demonstrates the potential of the proposed approach to substantially reduce annotation effort and improve standardization for quantitative vascular analysis.

Authors with CRIS profile

Jacopo Bracci Lehrstuhl für Informatik 5 (Mustererkennung) Linda Vorberg Lehrstuhl für Informatik 5 (Mustererkennung) Andreas Maier Lehrstuhl für Informatik 5 (Mustererkennung)

Involved external institutions

Siemens Healthineers (Germany) DE Germany (DE)

How to cite

APA:

Bracci, J., Katzmann, A., Rist, L., Vorberg, L., Sühling, M., & Maier, A. (2026). Hybrid Vessel Wall Segmentation for Assisted Annotation in CT Angiography. In Heinz Handels, Katharina Breininger, Thomas Deserno, Andreas Maier, Klaus Maier-Hein, Christoph Palm, Thomas Tolxdorff (Eds.), Informatik aktuell (pp. 285-291). Lübeck, DE: Springer Science and Business Media Deutschland GmbH.

MLA:

Bracci, Jacopo, et al. "Hybrid Vessel Wall Segmentation for Assisted Annotation in CT Angiography." Proceedings of the Bildverarbeitung für die Medizin Workshop, BVM 2026, Lübeck Ed. Heinz Handels, Katharina Breininger, Thomas Deserno, Andreas Maier, Klaus Maier-Hein, Christoph Palm, Thomas Tolxdorff, Springer Science and Business Media Deutschland GmbH, 2026. 285-291.

BibTeX: Download