Learning Prostate Anatomy at Test Time for Cancer Detection in Micro-Ultrasound

Obed Korshie Dzikunu, Mohammad Mahdi Abootorabi, Mohamed Harmanani, Paul F R Wilson, Emma Willis, Ferdinand Luger, Adam Kinnaird, Brian Wodlinger, Parvin Mousavi, Purang Abolmaesumi
Proceedings of the 11th Machine Learning for Healthcare Conference, PMLR 340:495-520, 2026.

Abstract

Domain shift across clinical centers using different imaging hardware or acquisition protocols remains a fundamental barrier to deploying deep learning models for prostate cancer (PCa) detection. Existing test-time adaptation (TTA) methods address distribution shift through entropy minimization or augmentation-based self-supervision, correcting for statistical differences in image appearance but ignoring the anatomical structure of the target domain. We propose ANT, a segmentation-guided TTA framework that adapts a pretrained cancer detection encoder to the target domain by solving an auxiliary prostate segmentation task at test time, supervised by pseudo-masks from a frozen pretrained segmentation network. By aligning encoder representations to prostate anatomy in the target domain, ANT corrects domain-specific feature drift while preserving cancer-discriminative structure. The model was trained on 693 patients imaged with an earlier-generation micro-ultrasound scanner in a multi-center clinical trial, and evaluated on 118 patients acquired with a newer-generation system across two centers in another clinical trial. Under a leave-one-center-out protocol with identical evaluation conditions across all methods, ANT improves mean AUC by 2.9% and 3.6% at the biopsy-core and patient levels, respectively, over no adaptation, outperforming TTA baselines. Code is available at: https://github.com/ObedDzik/ant.git.

Cite this Paper

BibTeX
@InProceedings{pmlr-v340-dzikunu26a, title = {Learning Prostate Anatomy at Test Time for Cancer Detection in Micro-Ultrasound}, author = {Dzikunu, Obed Korshie and Abootorabi, Mohammad Mahdi and Harmanani, Mohamed and Wilson, Paul F R and Willis, Emma and Luger, Ferdinand and Kinnaird, Adam and Wodlinger, Brian and Mousavi, Parvin and Abolmaesumi, Purang}, booktitle = {Proceedings of the 11th Machine Learning for Healthcare Conference}, pages = {495--520}, year = {2026}, editor = {Krishnan, Rahul G. and van Amsterdam, Wouter A. C. and Chopra, Sumit and Overgaard, Shauna and Hughes, Michael and Ötleş, Erkin and Shen, Yiqiu and Shanmugam, Divya and Nayan, Madhur and Engelhard, Matthew and Fackler, Jim and Oberst, Michael}, volume = {340}, series = {Proceedings of Machine Learning Research}, month = {12--14 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v340/main/assets/dzikunu26a/dzikunu26a.pdf}, url = {https://proceedings.mlr.press/v340/dzikunu26a.html}, abstract = {Domain shift across clinical centers using different imaging hardware or acquisition protocols remains a fundamental barrier to deploying deep learning models for prostate cancer (PCa) detection. Existing test-time adaptation (TTA) methods address distribution shift through entropy minimization or augmentation-based self-supervision, correcting for statistical differences in image appearance but ignoring the anatomical structure of the target domain. We propose ANT, a segmentation-guided TTA framework that adapts a pretrained cancer detection encoder to the target domain by solving an auxiliary prostate segmentation task at test time, supervised by pseudo-masks from a frozen pretrained segmentation network. By aligning encoder representations to prostate anatomy in the target domain, ANT corrects domain-specific feature drift while preserving cancer-discriminative structure. The model was trained on 693 patients imaged with an earlier-generation micro-ultrasound scanner in a multi-center clinical trial, and evaluated on 118 patients acquired with a newer-generation system across two centers in another clinical trial. Under a leave-one-center-out protocol with identical evaluation conditions across all methods, ANT improves mean AUC by 2.9% and 3.6% at the biopsy-core and patient levels, respectively, over no adaptation, outperforming TTA baselines. Code is available at: https://github.com/ObedDzik/ant.git.} }
Endnote
%0 Conference Paper %T Learning Prostate Anatomy at Test Time for Cancer Detection in Micro-Ultrasound %A Obed Korshie Dzikunu %A Mohammad Mahdi Abootorabi %A Mohamed Harmanani %A Paul F R Wilson %A Emma Willis %A Ferdinand Luger %A Adam Kinnaird %A Brian Wodlinger %A Parvin Mousavi %A Purang Abolmaesumi %B Proceedings of the 11th Machine Learning for Healthcare Conference %C Proceedings of Machine Learning Research %D 2026 %E Rahul G. Krishnan %E Wouter A. C. van Amsterdam %E Sumit Chopra %E Shauna Overgaard %E Michael Hughes %E Erkin Ötleş %E Yiqiu Shen %E Divya Shanmugam %E Madhur Nayan %E Matthew Engelhard %E Jim Fackler %E Michael Oberst %F pmlr-v340-dzikunu26a %I PMLR %P 495--520 %U https://proceedings.mlr.press/v340/dzikunu26a.html %V 340 %X Domain shift across clinical centers using different imaging hardware or acquisition protocols remains a fundamental barrier to deploying deep learning models for prostate cancer (PCa) detection. Existing test-time adaptation (TTA) methods address distribution shift through entropy minimization or augmentation-based self-supervision, correcting for statistical differences in image appearance but ignoring the anatomical structure of the target domain. We propose ANT, a segmentation-guided TTA framework that adapts a pretrained cancer detection encoder to the target domain by solving an auxiliary prostate segmentation task at test time, supervised by pseudo-masks from a frozen pretrained segmentation network. By aligning encoder representations to prostate anatomy in the target domain, ANT corrects domain-specific feature drift while preserving cancer-discriminative structure. The model was trained on 693 patients imaged with an earlier-generation micro-ultrasound scanner in a multi-center clinical trial, and evaluated on 118 patients acquired with a newer-generation system across two centers in another clinical trial. Under a leave-one-center-out protocol with identical evaluation conditions across all methods, ANT improves mean AUC by 2.9% and 3.6% at the biopsy-core and patient levels, respectively, over no adaptation, outperforming TTA baselines. Code is available at: https://github.com/ObedDzik/ant.git.
APA
Dzikunu, O.K., Abootorabi, M.M., Harmanani, M., Wilson, P.F.R., Willis, E., Luger, F., Kinnaird, A., Wodlinger, B., Mousavi, P. & Abolmaesumi, P.. (2026). Learning Prostate Anatomy at Test Time for Cancer Detection in Micro-Ultrasound. Proceedings of the 11th Machine Learning for Healthcare Conference, in Proceedings of Machine Learning Research 340:495-520 Available from https://proceedings.mlr.press/v340/dzikunu26a.html.

Related Material