Effective Disjoint Representational Learning for Anatomical Segmentation

Priya Tomar, Aditya Parikh, Philipp Feodorovici, Jan Arensmeyer, Hanno Matthaei, Christian Bauckhage, Helen Schneider, Rafet Sifa
Proceedings of The 8th International Conference on Medical Imaging with Deep Learning, PMLR 301:1550-1567, 2026.

Abstract

In the wake of the limited availability of pertinent datasets, the application of computer vision methods for semantic segmentation of abdominal structures is mainly constrained to surgical instruments or organ-specific segmentations. Multi-organ segmentation has the potential to furnish supplementary assistance in multifarious domains in healthcare, for instance, robot-assisted laparoscopic surgery. However, in addition to the complexity involved in discriminating anatomical structures due to their visual attributes and operative conditions, the representation bias pertaining to organ size results in poor segmentation performance on organs with smaller pixel proportions. In this work, we focus on alleviating the influence of representation bias by involving different encoder-decoder frameworks for learning organ-specific features. In particular, we investigate the effect of organ-specific decoders on binary segmentation of anatomical structures in abdominal surgery. Additionally, we analyze the effect of organ-specific pretraining on the multi-label segmentation in two model training settings including knowledge sharing and disjoint learning, in relation to the contextual feature sharing between organ-specific decoders. Our results illustrate the significant gain in segmentation performance by incorporating organ-specific decoders, especially for less represented organs.

Cite this Paper

BibTeX
@InProceedings{pmlr-v301-tomar26a, title = {Effective Disjoint Representational Learning for Anatomical Segmentation}, author = {Tomar, Priya and Parikh, Aditya and Feodorovici, Philipp and Arensmeyer, Jan and Matthaei, Hanno and Bauckhage, Christian and Schneider, Helen and Sifa, Rafet}, booktitle = {Proceedings of The 8th International Conference on Medical Imaging with Deep Learning}, pages = {1550--1567}, year = {2026}, editor = {Tasdizen, Tolga and Elhabian, Shireen and Summers, Ronald and Chen, Chen and Koch, Lisa and Zhuang, Yan}, volume = {301}, series = {Proceedings of Machine Learning Research}, month = {09--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v301/main/assets/tomar26a/tomar26a.pdf}, url = {https://proceedings.mlr.press/v301/tomar26a.html}, abstract = {In the wake of the limited availability of pertinent datasets, the application of computer vision methods for semantic segmentation of abdominal structures is mainly constrained to surgical instruments or organ-specific segmentations. Multi-organ segmentation has the potential to furnish supplementary assistance in multifarious domains in healthcare, for instance, robot-assisted laparoscopic surgery. However, in addition to the complexity involved in discriminating anatomical structures due to their visual attributes and operative conditions, the representation bias pertaining to organ size results in poor segmentation performance on organs with smaller pixel proportions. In this work, we focus on alleviating the influence of representation bias by involving different encoder-decoder frameworks for learning organ-specific features. In particular, we investigate the effect of organ-specific decoders on binary segmentation of anatomical structures in abdominal surgery. Additionally, we analyze the effect of organ-specific pretraining on the multi-label segmentation in two model training settings including knowledge sharing and disjoint learning, in relation to the contextual feature sharing between organ-specific decoders. Our results illustrate the significant gain in segmentation performance by incorporating organ-specific decoders, especially for less represented organs.} }
Endnote
%0 Conference Paper %T Effective Disjoint Representational Learning for Anatomical Segmentation %A Priya Tomar %A Aditya Parikh %A Philipp Feodorovici %A Jan Arensmeyer %A Hanno Matthaei %A Christian Bauckhage %A Helen Schneider %A Rafet Sifa %B Proceedings of The 8th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2026 %E Tolga Tasdizen %E Shireen Elhabian %E Ronald Summers %E Chen Chen %E Lisa Koch %E Yan Zhuang %F pmlr-v301-tomar26a %I PMLR %P 1550--1567 %U https://proceedings.mlr.press/v301/tomar26a.html %V 301 %X In the wake of the limited availability of pertinent datasets, the application of computer vision methods for semantic segmentation of abdominal structures is mainly constrained to surgical instruments or organ-specific segmentations. Multi-organ segmentation has the potential to furnish supplementary assistance in multifarious domains in healthcare, for instance, robot-assisted laparoscopic surgery. However, in addition to the complexity involved in discriminating anatomical structures due to their visual attributes and operative conditions, the representation bias pertaining to organ size results in poor segmentation performance on organs with smaller pixel proportions. In this work, we focus on alleviating the influence of representation bias by involving different encoder-decoder frameworks for learning organ-specific features. In particular, we investigate the effect of organ-specific decoders on binary segmentation of anatomical structures in abdominal surgery. Additionally, we analyze the effect of organ-specific pretraining on the multi-label segmentation in two model training settings including knowledge sharing and disjoint learning, in relation to the contextual feature sharing between organ-specific decoders. Our results illustrate the significant gain in segmentation performance by incorporating organ-specific decoders, especially for less represented organs.
APA
Tomar, P., Parikh, A., Feodorovici, P., Arensmeyer, J., Matthaei, H., Bauckhage, C., Schneider, H. & Sifa, R.. (2026). Effective Disjoint Representational Learning for Anatomical Segmentation. Proceedings of The 8th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 301:1550-1567 Available from https://proceedings.mlr.press/v301/tomar26a.html.

Related Material