Abstract
Research background and purpose: Currently, most percutaneous ventricular assist devices (pVADs) use a constant flow infusion method. Long-term use of a constant flow infusion device may increase the risk of adverse outcomes. Investigating pulsation methods for pVADs have significant importance for the development of cardiac digital twins. Methods: A hydraulic experimental platform was established to capture the physical characteristics of pVADs across multidimensional operating conditions. Based on the experimental data, we propose a novel NPQ model defining the mathematical relationship among motor speed, pressure, and flow rate, alongside an LSTM-Transformer model to accurately predict pulsation characteristic time points. Results: Experimental pressure-flow curves successfully determined the NPQ model parameters, which achieved a maximum pressure prediction error of only 2.15 mmHg. Furthermore, the LSTM-Transformer model demonstrated substantially lower prediction errors than conventional methods and maintained robust predictive performance across varying noise levels. Conclusions: This study elucidates that pVAD motor speed exhibits a linear relationship with pressure and a quadratic relationship with flow rate. Both proposed models facilitate accurate predictions of pVAD pulsatile characteristics. Validated by extensive hydrodynamic experiments, they provide a solid foundation for realizing pVAD pulsatile function and developing cardiac digital twins.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/DT4H_003.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: Not Submitted
Link to Open Review
Open Review Page: https://openreview.net/forum?id=XDhbMn0j3e
BibTex
@InProceedings{ECha_An_MICCAISAT2026,
author = { E, Chaoran AND Chen, Chenghan AND Qi, Kaijie AND Cheng, Tiefeng AND Wang, Haiy AND Xiong, Weiqing AND Wang, Liuying AND Zhao, Gezheng AND Liu, Ziyu AND Hao, Qikai AND Yue, Qiu},
title = { { An AI-Driven Pulsation Method of pVADs for Cardiac Digital Twins } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17275},
month = {pending},
page = {pending}
}