Accelerating physics-informed neural fields for fast CT perfusion analysis in acute ischemic stroke

Lucas de Vries, Rudolf Leonardus Mirjam Van Herten, Jan W. Hoving, Ivana Isgum, Bart Emmer, Charles B. Majoie, Henk Marquering, Stratis Gavves
Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning, PMLR 250:1606-1626, 2024.

Abstract

Spatio-temporal perfusion physics-informed neural networks were introduced as a new method (SPPINN) for CT perfusion (CTP) analysis in acute ischemic stroke. SPPINN leverages physics-informed learning and neural fields to perform a robust analysis of noisy CTP data. However, SPPINN faces limitations that hinder its application in practice, namely its implementation as a slice-based (2D) method, lengthy computation times, and the lack of infarct core segmentation. To address these challenges, we introduce a new approach to accelerate physics-informed neural fields for fast, volume-based (3D), CTP analysis including infarct core segmentation: ReSPPINN. To accommodate 3D data while simultaneously reducing computation times, we integrate efficient coordinate encodings. Furthermore, to ensure even faster model convergence, we use a meta-learning strategy. In addition, we also segment the infarct core. We employ acute MRI reference standard infarct core segmentations to evaluate ReSPPINN and we compare the performance with two commercial software packages. We show that meta-learning allows for full-volume perfusion map generation in 1.2 minutes without comprising quality, compared to over 40 minutes required by SPPINN. Moreover, ReSPPINNś infarct core segmentation outperforms commercial software.

Cite this Paper

BibTeX
@InProceedings{pmlr-v250-vries24a, title = {Accelerating physics-informed neural fields for fast CT perfusion analysis in acute ischemic stroke}, author = {de Vries, Lucas and Herten, Rudolf Leonardus Mirjam Van and Hoving, Jan W. and Isgum, Ivana and Emmer, Bart and Majoie, Charles B. and Marquering, Henk and Gavves, Stratis}, booktitle = {Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning}, pages = {1606--1626}, 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/vries24a/vries24a.pdf}, url = {https://proceedings.mlr.press/v250/vries24a.html}, abstract = {Spatio-temporal perfusion physics-informed neural networks were introduced as a new method (SPPINN) for CT perfusion (CTP) analysis in acute ischemic stroke. SPPINN leverages physics-informed learning and neural fields to perform a robust analysis of noisy CTP data. However, SPPINN faces limitations that hinder its application in practice, namely its implementation as a slice-based (2D) method, lengthy computation times, and the lack of infarct core segmentation. To address these challenges, we introduce a new approach to accelerate physics-informed neural fields for fast, volume-based (3D), CTP analysis including infarct core segmentation: ReSPPINN. To accommodate 3D data while simultaneously reducing computation times, we integrate efficient coordinate encodings. Furthermore, to ensure even faster model convergence, we use a meta-learning strategy. In addition, we also segment the infarct core. We employ acute MRI reference standard infarct core segmentations to evaluate ReSPPINN and we compare the performance with two commercial software packages. We show that meta-learning allows for full-volume perfusion map generation in 1.2 minutes without comprising quality, compared to over 40 minutes required by SPPINN. Moreover, ReSPPINNś infarct core segmentation outperforms commercial software.} }
Endnote
%0 Conference Paper %T Accelerating physics-informed neural fields for fast CT perfusion analysis in acute ischemic stroke %A Lucas de Vries %A Rudolf Leonardus Mirjam Van Herten %A Jan W. Hoving %A Ivana Isgum %A Bart Emmer %A Charles B. Majoie %A Henk Marquering %A Stratis Gavves %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-vries24a %I PMLR %P 1606--1626 %U https://proceedings.mlr.press/v250/vries24a.html %V 250 %X Spatio-temporal perfusion physics-informed neural networks were introduced as a new method (SPPINN) for CT perfusion (CTP) analysis in acute ischemic stroke. SPPINN leverages physics-informed learning and neural fields to perform a robust analysis of noisy CTP data. However, SPPINN faces limitations that hinder its application in practice, namely its implementation as a slice-based (2D) method, lengthy computation times, and the lack of infarct core segmentation. To address these challenges, we introduce a new approach to accelerate physics-informed neural fields for fast, volume-based (3D), CTP analysis including infarct core segmentation: ReSPPINN. To accommodate 3D data while simultaneously reducing computation times, we integrate efficient coordinate encodings. Furthermore, to ensure even faster model convergence, we use a meta-learning strategy. In addition, we also segment the infarct core. We employ acute MRI reference standard infarct core segmentations to evaluate ReSPPINN and we compare the performance with two commercial software packages. We show that meta-learning allows for full-volume perfusion map generation in 1.2 minutes without comprising quality, compared to over 40 minutes required by SPPINN. Moreover, ReSPPINNś infarct core segmentation outperforms commercial software.
APA
de Vries, L., Herten, R.L.M.V., Hoving, J.W., Isgum, I., Emmer, B., Majoie, C.B., Marquering, H. & Gavves, S.. (2024). Accelerating physics-informed neural fields for fast CT perfusion analysis in acute ischemic stroke. Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 250:1606-1626 Available from https://proceedings.mlr.press/v250/vries24a.html.

Related Material