OFELIA: Optical Flow-based Electrode LocalIzAtion

Xinyi Wang, Zikang Xu, Qingsong Yao, Yiyong Sun, S Kevin Zhou
Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning, PMLR 250:1652-1669, 2024.

Abstract

Catheter ablation is one of the most common cardiac ablation procedures for atrial fibrillation, which is mainly based on catheters with electrodes collecting electrophysiology signals.Catheter electrode localization facilitates intraoperative catheter positioning, surgical planning, and other applications such as 3D model reconstruction.In this paper, we propose a novel deep network for automatic electrode localization in an X-ray sequence, which integrates spatiotemporal features between adjacent frames, aided by optical flow maps.To improve the utility and robustness of the proposed method, we first design a saturation-based optical flow dataset construction pipeline, then finetune the optical flow estimation to obtain more realistic and contrasting optical flow maps for electrode localization.The extensive results on clinical-challenging test sequences reveal the effectiveness of our method, with a mean radial error (MRE) of 0.95 mm for radiofrequency catheters and an MRE of 0.71 mm for coronary sinus catheters, outperforming several state-of-the-art landmark detection methods.

Cite this Paper

BibTeX
@InProceedings{pmlr-v250-wang24b, title = {OFELIA: Optical Flow-based Electrode LocalIzAtion}, author = {Wang, Xinyi and Xu, Zikang and Yao, Qingsong and Sun, Yiyong and Zhou, S Kevin}, booktitle = {Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning}, pages = {1652--1669}, year = {2024}, editor = {Burgos, Ninon and Petitjean, Caroline and Vakalopoulou, Maria and Christodoulidis, Stergios and Coupe, Pierrick and Delingette, Hervé and Lartizien, Carole and Mateus, Diana}, volume = {250}, series = {Proceedings of Machine Learning Research}, month = {03--05 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v250/main/assets/wang24b/wang24b.pdf}, url = {https://proceedings.mlr.press/v250/wang24b.html}, abstract = {Catheter ablation is one of the most common cardiac ablation procedures for atrial fibrillation, which is mainly based on catheters with electrodes collecting electrophysiology signals.Catheter electrode localization facilitates intraoperative catheter positioning, surgical planning, and other applications such as 3D model reconstruction.In this paper, we propose a novel deep network for automatic electrode localization in an X-ray sequence, which integrates spatiotemporal features between adjacent frames, aided by optical flow maps.To improve the utility and robustness of the proposed method, we first design a saturation-based optical flow dataset construction pipeline, then finetune the optical flow estimation to obtain more realistic and contrasting optical flow maps for electrode localization.The extensive results on clinical-challenging test sequences reveal the effectiveness of our method, with a mean radial error (MRE) of 0.95 mm for radiofrequency catheters and an MRE of 0.71 mm for coronary sinus catheters, outperforming several state-of-the-art landmark detection methods.} }
Endnote
%0 Conference Paper %T OFELIA: Optical Flow-based Electrode LocalIzAtion %A Xinyi Wang %A Zikang Xu %A Qingsong Yao %A Yiyong Sun %A S Kevin Zhou %B Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2024 %E Ninon Burgos %E Caroline Petitjean %E Maria Vakalopoulou %E Stergios Christodoulidis %E Pierrick Coupe %E Hervé Delingette %E Carole Lartizien %E Diana Mateus %F pmlr-v250-wang24b %I PMLR %P 1652--1669 %U https://proceedings.mlr.press/v250/wang24b.html %V 250 %X Catheter ablation is one of the most common cardiac ablation procedures for atrial fibrillation, which is mainly based on catheters with electrodes collecting electrophysiology signals.Catheter electrode localization facilitates intraoperative catheter positioning, surgical planning, and other applications such as 3D model reconstruction.In this paper, we propose a novel deep network for automatic electrode localization in an X-ray sequence, which integrates spatiotemporal features between adjacent frames, aided by optical flow maps.To improve the utility and robustness of the proposed method, we first design a saturation-based optical flow dataset construction pipeline, then finetune the optical flow estimation to obtain more realistic and contrasting optical flow maps for electrode localization.The extensive results on clinical-challenging test sequences reveal the effectiveness of our method, with a mean radial error (MRE) of 0.95 mm for radiofrequency catheters and an MRE of 0.71 mm for coronary sinus catheters, outperforming several state-of-the-art landmark detection methods.
APA
Wang, X., Xu, Z., Yao, Q., Sun, Y. & Zhou, S.K.. (2024). OFELIA: Optical Flow-based Electrode LocalIzAtion. Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 250:1652-1669 Available from https://proceedings.mlr.press/v250/wang24b.html.

Related Material