Student Conference Proceedings
Vol. 1 No. 1 (2025): Stud Conf Proc
https://doi.org/10.18416/SCP.2025.1923

Biomedical Engineering, ID 1923

Analyzing Patient-Ventilator Interaction with Neural Networks

Main Article Content

Shrawan Kumar Mehta (Study program Biomedical Engineering, Luebeck University of Applied Sciences, Lübeck, Germany), Franziska Bilda (1)Institute for Electrical Engineering in Medicine, Universität zu Lübeck, Lübeck, Germany; 2) Fraunhofer IMTE, Lübeck, Germany), Jan Graßhoff (1) Institute for Electrical Engineering in Medicine, Universität zu Lübeck, Lübeck, Germany; 2) Fraunhofer IMTE, Lübeck, Germany)

Copyright (c) 2025 Shrawan Mehta; Franziska Bilda, Jan Graßhoff

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

Abstract

This study investigates the integration of computational modeling and machine learning to analyze patientventilator interaction (PVI) in mechanical ventilation. One-compartment, multi-compartment and nonlinear model were developed to generate synthetic data that account for lung mechanics under various conditions. Utilizing synthetic data addresses the limitations of real clinical data availability. These datasets were used to train a residual neural network (ResNet) model for time-series classification which allows to predict PVI events effectively. The ResNet model that consists of convolutional layers and skip connections which captured complex relationships between physiological parameters to achieve a test accuracy of 89.27% and a test loss of 0.24. The framework offers a promising direction for personalized ventilation strategies that aims to enhance patient care in critical respiratory conditions. This integration makes a significant advancement in the management and optimization of mechanical ventilation approach using artificial intelligence (AI).

Article Details