Data-driven simulator for mechanical circulatory support with domain adversarial neural process

Sophia Sun, Wenyuan Chen, Zihao Zhou, Sonia Fereidooni, Elise Jortberg, Rose Yu
Proceedings of the 6th Annual Learning for Dynamics & Control Conference, PMLR 242:1513-1525, 2024.

Abstract

We propose a data-driven simulator for Mechanical Circulatory Support (MCS) devices, implemented as a probabilistic deep sequence model. Existing mechanical simulators for MCS rely on oversimplifying assumptions and are insensitive to patient-specific behavior, limiting their applicability to real-world treatment scenarios. To address these shortcomings, our model Domain Adversarial Neural Process (DANP) employs a neural process architecture, allowing it to capture the probabilistic relationship between MCS pump levels and aortic pressure measurements with uncertainty. We use domain adversarial training to combine real-world and simulation data, resulting in a more realistic and diverse representation of potential outcomes. Empirical results with an improvement of 19% in non-stationary trend prediction establish DANP as an effective tool for clinicians to understand and make informed decisions regarding MCS patient treatment.

Cite this Paper

BibTeX
@InProceedings{pmlr-v242-sun24a, title = {Data-driven simulator for mechanical circulatory support with domain adversarial neural process}, author = {Sun, Sophia and Chen, Wenyuan and Zhou, Zihao and Fereidooni, Sonia and Jortberg, Elise and Yu, Rose}, booktitle = {Proceedings of the 6th Annual Learning for Dynamics & Control Conference}, pages = {1513--1525}, year = {2024}, editor = {Abate, Alessandro and Cannon, Mark and Margellos, Kostas and Papachristodoulou, Antonis}, volume = {242}, series = {Proceedings of Machine Learning Research}, month = {15--17 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v242/sun24a/sun24a.pdf}, url = {https://proceedings.mlr.press/v242/sun24a.html}, abstract = {We propose a data-driven simulator for Mechanical Circulatory Support (MCS) devices, implemented as a probabilistic deep sequence model. Existing mechanical simulators for MCS rely on oversimplifying assumptions and are insensitive to patient-specific behavior, limiting their applicability to real-world treatment scenarios. To address these shortcomings, our model Domain Adversarial Neural Process (DANP) employs a neural process architecture, allowing it to capture the probabilistic relationship between MCS pump levels and aortic pressure measurements with uncertainty. We use domain adversarial training to combine real-world and simulation data, resulting in a more realistic and diverse representation of potential outcomes. Empirical results with an improvement of 19% in non-stationary trend prediction establish DANP as an effective tool for clinicians to understand and make informed decisions regarding MCS patient treatment.} }
Endnote
%0 Conference Paper %T Data-driven simulator for mechanical circulatory support with domain adversarial neural process %A Sophia Sun %A Wenyuan Chen %A Zihao Zhou %A Sonia Fereidooni %A Elise Jortberg %A Rose Yu %B Proceedings of the 6th Annual Learning for Dynamics & Control Conference %C Proceedings of Machine Learning Research %D 2024 %E Alessandro Abate %E Mark Cannon %E Kostas Margellos %E Antonis Papachristodoulou %F pmlr-v242-sun24a %I PMLR %P 1513--1525 %U https://proceedings.mlr.press/v242/sun24a.html %V 242 %X We propose a data-driven simulator for Mechanical Circulatory Support (MCS) devices, implemented as a probabilistic deep sequence model. Existing mechanical simulators for MCS rely on oversimplifying assumptions and are insensitive to patient-specific behavior, limiting their applicability to real-world treatment scenarios. To address these shortcomings, our model Domain Adversarial Neural Process (DANP) employs a neural process architecture, allowing it to capture the probabilistic relationship between MCS pump levels and aortic pressure measurements with uncertainty. We use domain adversarial training to combine real-world and simulation data, resulting in a more realistic and diverse representation of potential outcomes. Empirical results with an improvement of 19% in non-stationary trend prediction establish DANP as an effective tool for clinicians to understand and make informed decisions regarding MCS patient treatment.
APA
Sun, S., Chen, W., Zhou, Z., Fereidooni, S., Jortberg, E. & Yu, R.. (2024). Data-driven simulator for mechanical circulatory support with domain adversarial neural process. Proceedings of the 6th Annual Learning for Dynamics & Control Conference, in Proceedings of Machine Learning Research 242:1513-1525 Available from https://proceedings.mlr.press/v242/sun24a.html.

Related Material