Cross-Domain Semi-Supervised Organ Detection

Nian Li, Morteza Ghahremani, Bailiang Jian, Pascual Tejero Cervera, Benedikt Wiestler, Marcus Makowski, Christian Wachinger
Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, PMLR 315:107-120, 2026.

Abstract

Domain adaptation for 3D organ detection in CT imaging is challenging due to variations in scanner types, imaging protocols, and overall acquisition conditions. As supervised detection models require large, annotated datasets from diverse scanners and institutions, semi-supervised approaches have gained attention for their ability to leverage limited unlabeled target data. However, traditional semi-supervised methods typically fail to make effective use of the few labeled target samples and most often do not yield satisfactory results. To address this limitation, we introduce a novel cross-domain semi-supervised detection framework (CDSS-Det) built upon the Transformer-based Organ-DETR model. CDSS-Det is a cross-domain semi-supervised framework for 3D organ detection that addresses unreliable pseudo-labels and limited target supervision under domain shift. It introduces a curriculum-guided pseudo-labeling mechanism and domain-robust representation learning to enable effective knowledge transfer from a well-annotated source domain to a sparsely labeled target domain. Experiments on multi-domain CT datasets demonstrate that incorporating a small number of labeled target samples significantly boosts detection performance over conventional domain adaptation and semi-supervised methods. CDSS-Det consistently achieves higher mean Average Precision (mAP), with notable improvements in detecting small organs, and surpasses a fully supervised model trained solely on the labeled target domain by over 10%. These results underscore the potential of CDSS-Det in efficiently leveraging both labeled and unlabeled target data in cross-domain organ detection, advancing annotation-efficient deep learning models in medical imaging.

Cite this Paper

BibTeX
@InProceedings{pmlr-v315-li26a, title = {Cross-Domain Semi-Supervised Organ Detection}, author = {Li, Nian and Ghahremani, Morteza and Jian, Bailiang and Cervera, Pascual Tejero and Wiestler, Benedikt and Makowski, Marcus and Wachinger, Christian}, booktitle = {Proceedings of The 9th International Conference on Medical Imaging with Deep Learning}, pages = {107--120}, year = {2026}, editor = {Huo, Yuankai and Gao, Mingchen and Kuo, Chang-Fu and Jin, Yueming and Deng, Ruining}, volume = {315}, series = {Proceedings of Machine Learning Research}, month = {08--10 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v315/main/assets/li26a/li26a.pdf}, url = {https://proceedings.mlr.press/v315/li26a.html}, abstract = {Domain adaptation for 3D organ detection in CT imaging is challenging due to variations in scanner types, imaging protocols, and overall acquisition conditions. As supervised detection models require large, annotated datasets from diverse scanners and institutions, semi-supervised approaches have gained attention for their ability to leverage limited unlabeled target data. However, traditional semi-supervised methods typically fail to make effective use of the few labeled target samples and most often do not yield satisfactory results. To address this limitation, we introduce a novel cross-domain semi-supervised detection framework (CDSS-Det) built upon the Transformer-based Organ-DETR model. CDSS-Det is a cross-domain semi-supervised framework for 3D organ detection that addresses unreliable pseudo-labels and limited target supervision under domain shift. It introduces a curriculum-guided pseudo-labeling mechanism and domain-robust representation learning to enable effective knowledge transfer from a well-annotated source domain to a sparsely labeled target domain. Experiments on multi-domain CT datasets demonstrate that incorporating a small number of labeled target samples significantly boosts detection performance over conventional domain adaptation and semi-supervised methods. CDSS-Det consistently achieves higher mean Average Precision (mAP), with notable improvements in detecting small organs, and surpasses a fully supervised model trained solely on the labeled target domain by over 10%. These results underscore the potential of CDSS-Det in efficiently leveraging both labeled and unlabeled target data in cross-domain organ detection, advancing annotation-efficient deep learning models in medical imaging.} }
Endnote
%0 Conference Paper %T Cross-Domain Semi-Supervised Organ Detection %A Nian Li %A Morteza Ghahremani %A Bailiang Jian %A Pascual Tejero Cervera %A Benedikt Wiestler %A Marcus Makowski %A Christian Wachinger %B Proceedings of The 9th International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2026 %E Yuankai Huo %E Mingchen Gao %E Chang-Fu Kuo %E Yueming Jin %E Ruining Deng %F pmlr-v315-li26a %I PMLR %P 107--120 %U https://proceedings.mlr.press/v315/li26a.html %V 315 %X Domain adaptation for 3D organ detection in CT imaging is challenging due to variations in scanner types, imaging protocols, and overall acquisition conditions. As supervised detection models require large, annotated datasets from diverse scanners and institutions, semi-supervised approaches have gained attention for their ability to leverage limited unlabeled target data. However, traditional semi-supervised methods typically fail to make effective use of the few labeled target samples and most often do not yield satisfactory results. To address this limitation, we introduce a novel cross-domain semi-supervised detection framework (CDSS-Det) built upon the Transformer-based Organ-DETR model. CDSS-Det is a cross-domain semi-supervised framework for 3D organ detection that addresses unreliable pseudo-labels and limited target supervision under domain shift. It introduces a curriculum-guided pseudo-labeling mechanism and domain-robust representation learning to enable effective knowledge transfer from a well-annotated source domain to a sparsely labeled target domain. Experiments on multi-domain CT datasets demonstrate that incorporating a small number of labeled target samples significantly boosts detection performance over conventional domain adaptation and semi-supervised methods. CDSS-Det consistently achieves higher mean Average Precision (mAP), with notable improvements in detecting small organs, and surpasses a fully supervised model trained solely on the labeled target domain by over 10%. These results underscore the potential of CDSS-Det in efficiently leveraging both labeled and unlabeled target data in cross-domain organ detection, advancing annotation-efficient deep learning models in medical imaging.
APA
Li, N., Ghahremani, M., Jian, B., Cervera, P.T., Wiestler, B., Makowski, M. & Wachinger, C.. (2026). Cross-Domain Semi-Supervised Organ Detection. Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 315:107-120 Available from https://proceedings.mlr.press/v315/li26a.html.

Related Material