GeoReg: Direct biplanar DSA-to-CTA registration with geodesic consistency for acute ischemic stroke

Rudolf L. M. van Herten, Robert Graf, Felix Bitzer, Jan S. Kirschke, Johannes C. Paetzold
Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, PMLR 315:153-171, 2026.

Abstract

The complementary nature of pre-procedural computed tomography angiography (CTA) and intraoperative digital subtraction angiography (DSA) has motivated significant interest in their registration to enhance therapeutic decision-making during stroke interventions. However, current methods depend on accurate vessel segmentation in both modalities, creating a deployment bottleneck due to the requirement for extensive annotated training data. Here, we present an alternative approach that establishes the feasibility of registration without this dependency. Instead of extracting vascular features using pre-trained models, we optimize a direct registration framework using maximum intensity projections of DSA sequences to align a silhouette of the subtracted X-ray image. We introduce a geodesic consistency formulation that jointly optimizes biplanar views, employing soft geometric constraints on SO(3) to maintain consistency while accommodating non-orthogonal scanner configurations. We demonstrate the effectiveness of this model on clinical stroke data and find that it outperforms existing methods, proving particularly effective in escaping local minima where single-view optimization fails. These results indicate that reliable DSA-to-CTA registration is achievable without vessel-specific training data, simplifying the path toward clinical integration.

Cite this Paper

BibTeX
@InProceedings{pmlr-v315-herten26a, title = {GeoReg: Direct biplanar DSA-to-CTA registration with geodesic consistency for acute ischemic stroke}, author = {van Herten, Rudolf L. M. and Graf, Robert and Bitzer, Felix and Kirschke, Jan S. and Paetzold, Johannes C.}, booktitle = {Proceedings of The 9th International Conference on Medical Imaging with Deep Learning}, pages = {153--171}, 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/herten26a/herten26a.pdf}, url = {https://proceedings.mlr.press/v315/herten26a.html}, abstract = {The complementary nature of pre-procedural computed tomography angiography (CTA) and intraoperative digital subtraction angiography (DSA) has motivated significant interest in their registration to enhance therapeutic decision-making during stroke interventions. However, current methods depend on accurate vessel segmentation in both modalities, creating a deployment bottleneck due to the requirement for extensive annotated training data. Here, we present an alternative approach that establishes the feasibility of registration without this dependency. Instead of extracting vascular features using pre-trained models, we optimize a direct registration framework using maximum intensity projections of DSA sequences to align a silhouette of the subtracted X-ray image. We introduce a geodesic consistency formulation that jointly optimizes biplanar views, employing soft geometric constraints on SO(3) to maintain consistency while accommodating non-orthogonal scanner configurations. We demonstrate the effectiveness of this model on clinical stroke data and find that it outperforms existing methods, proving particularly effective in escaping local minima where single-view optimization fails. These results indicate that reliable DSA-to-CTA registration is achievable without vessel-specific training data, simplifying the path toward clinical integration.} }
Endnote
%0 Conference Paper %T GeoReg: Direct biplanar DSA-to-CTA registration with geodesic consistency for acute ischemic stroke %A Rudolf L. M. van Herten %A Robert Graf %A Felix Bitzer %A Jan S. Kirschke %A Johannes C. Paetzold %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-herten26a %I PMLR %P 153--171 %U https://proceedings.mlr.press/v315/herten26a.html %V 315 %X The complementary nature of pre-procedural computed tomography angiography (CTA) and intraoperative digital subtraction angiography (DSA) has motivated significant interest in their registration to enhance therapeutic decision-making during stroke interventions. However, current methods depend on accurate vessel segmentation in both modalities, creating a deployment bottleneck due to the requirement for extensive annotated training data. Here, we present an alternative approach that establishes the feasibility of registration without this dependency. Instead of extracting vascular features using pre-trained models, we optimize a direct registration framework using maximum intensity projections of DSA sequences to align a silhouette of the subtracted X-ray image. We introduce a geodesic consistency formulation that jointly optimizes biplanar views, employing soft geometric constraints on SO(3) to maintain consistency while accommodating non-orthogonal scanner configurations. We demonstrate the effectiveness of this model on clinical stroke data and find that it outperforms existing methods, proving particularly effective in escaping local minima where single-view optimization fails. These results indicate that reliable DSA-to-CTA registration is achievable without vessel-specific training data, simplifying the path toward clinical integration.
APA
van Herten, R.L.M., Graf, R., Bitzer, F., Kirschke, J.S. & Paetzold, J.C.. (2026). GeoReg: Direct biplanar DSA-to-CTA registration with geodesic consistency for acute ischemic stroke. Proceedings of The 9th International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 315:153-171 Available from https://proceedings.mlr.press/v315/herten26a.html.

Related Material