Unsupervised Single-source Domain Generalization for Robust Quantification of Lymphatic Perfusion

Fischer LK, Müller J, Schröder C, Hanser A, Cuomo M, Day T, Ouyang C, Dewald O, Rompel O, Dittrich S, Küstner T, Kainz B (2025)

Publication Language: English

Publication Type: Conference contribution

Publication year: 2025

Journal

Informatik aktuell Gesellschaft für Informatik (GI)

Publisher: Springer

Series: Informatik aktuell

City/Town: Cham

Pages Range: 178-184

Conference Proceedings Title: Bildverarbeitung für die Medizin 2025. Book SubtitleProceedings, German Conference on Medical Image Computing, Regensburg March 09-11, 2025

Event location: Regensburg DE

ISBN: 9783658474218

DOI: 10.1007/978-3-658-47422-5_40

Abstract

Infants and young children with Fontan circulation, resulting from lifesaving heart surgery, experience altered hemodynamics that can lead to severe, life-threatening complications, often driven by congestion in the lymphatic system. These complications typically become more apparent as the patients grow older. Early detection and management of these complications require precise assessment of lymphatic perfusion through segmentation of lymphatic fluids. However, manual evaluation is time-intensive and subject to variability, underscoring the need for automated, reliable segmentation in clinical practice. Yet, domain shifts arising from variations in MRI acquisition protocols (e.g., BLADE versus SPACE sequences) and scanner differences present significant challenges for current segmentation models. In this work, we propose the use of causalityinspired single-source domain generalization (CISDG) to develop a robust and accurate segmentation network for lymphatic perfusion patterns across diverse imaging domains. Using a dataset of T2-weighted MR images from 71 patients, we demonstrate that our CISDG model outperforms conventional segmentation networks, including nnU-Net, in both source and target domains. Our results indicate that the proposed method not only generalizes effectively across domain shifts but also holds promise for enhancing diagnostic efficiency and reliability in clinical settings.

Authors with CRIS profile

Johanna Müller Professur für Image Data Exploration and Analysis Christian Schröder Department of Paediatric Cardiology Michela Cuomo Einrichtungen, die zum Universitätsklinikum Erlangen gehören Oliver Dewald Lehrstuhl für Herzchirurgie Oliver Rompel Lehrstuhl für Diagnostische Radiologie Sven Dittrich Professur für Kinderkardiologie Bernhard Kainz Professur für Image Data Exploration and Analysis

Involved external institutions

Universitätsklinikum Tübingen DE Germany (DE) King’s College London GB United Kingdom (GB) University of Oxford GB United Kingdom (GB)

How to cite

APA:

Fischer, L.K., Müller, J., Schröder, C., Hanser, A., Cuomo, M., Day, T.,... Kainz, B. (2025). Unsupervised Single-source Domain Generalization for Robust Quantification of Lymphatic Perfusion. In Christoph Palm, Katharina Breininger, Thomas Deserno, Heinz Handels, Andreas Maier, Klaus H. Maier-Hein, Thomas M. Tolxdorff (Eds.), Bildverarbeitung für die Medizin 2025. Book SubtitleProceedings, German Conference on Medical Image Computing, Regensburg March 09-11, 2025 (pp. 178-184). Regensburg, DE: Cham: Springer.

MLA:

Fischer, Lisa K., et al. "Unsupervised Single-source Domain Generalization for Robust Quantification of Lymphatic Perfusion." Proceedings of the German Conference on Medical Image Computing, 2025, Regensburg Ed. Christoph Palm, Katharina Breininger, Thomas Deserno, Heinz Handels, Andreas Maier, Klaus H. Maier-Hein, Thomas M. Tolxdorff, Cham: Springer, 2025. 178-184.

BibTeX: Download