Machine Learning for Mechanical Ventilation Control (Extended Abstract)

TL;DR

This paper introduces a data-driven control method for mechanical ventilation that outperforms traditional PID controllers and reinforcement learning algorithms, demonstrating improved accuracy and robustness in both simulation and real ventilator control.

Contribution

The paper presents a novel data-driven control approach trained on a simulator for invasive ventilator management, surpassing existing methods in performance and robustness.

Findings

Outperforms popular reinforcement learning algorithms.

Controls physical ventilator more accurately and robustly than PID.

Suggests potential for broader application to various ventilation types.

Abstract

Mechanical ventilation is one of the most widely used therapies in the ICU. However, despite broad application from anaesthesia to COVID-related life support, many injurious challenges remain. We frame these as a control problem: ventilators must let air in and out of the patient's lungs according to a prescribed trajectory of airway pressure. Industry-standard controllers, based on the PID method, are neither optimal nor robust. Our data-driven approach learns to control an invasive ventilator by training on a simulator itself trained on data collected from the ventilator. This method outperforms popular reinforcement learning algorithms and even controls the physical ventilator more accurately and robustly than PID. These results underscore how effective data-driven methodologies can be for invasive ventilation and suggest that more general forms of ventilation (e.g., non-invasive, adaptive) may also be amenable.