Learning to control ventilators

We consider the problem of controlling an invasive mechanical ventilator for pressure-controlled ventilation: a controller must let air in and out of a sedated patient’s lungs according to a trajectory of airway pressures specified by a clinician. Hand-tuned PID controllers and similar variants have comprised the industry standard for decades, yet can behave poorly by over- or under-shooting their target or oscillating rapidly.

We consider a data-driven machine learning approach: First, we train a simulator based on data we collect from an artificial lung. Then, we train deep neural network controllers on these simulators.We show that our controllers are able to track target pressure waveforms significantly better than PID controllers. We further show that a learned controller generalizes across lungs with varying characteristics much more readily than PID controllers do.

Citation

@article{suo2021machine,
  title={Machine Learning for Mechanical Ventilation Control},
  author={Suo, Daniel and Ghai, Udaya and Minasyan, Edgar and Gradu, Paula and Chen, Xinyi and Agarwal, Naman and Zhang, Cyril and Singh, Karan and LaChance, Julienne and Zadjel, Tom and others},
  journal={arXiv preprint arXiv:2102.06779},
  year={2021}
}