Abstract:Objective. Closed-loop mechanical ventilation has the potential to provide more effective ventilatory support to patients with less complexity than conventional ventilation. The purpose of this study was to investigate the effectiveness of an automatic technique for mechanical ventilation. Methods. Two closed-loop control systems for mechanical ventilation are combined in this study. In one of the control systems several physiological data are used to automatically adjust the frequency and tidal volume of breaths of a patient. This method, which is patented under US Patent number 4986268, uses the criterion of minimal respiratory work rate to provide the patient with a natural pattern of breathing. The inputs to the system include data representing CO2 and O2 levels of the patient as well as respiratory compliance and airway resistance. The I:E ratio is adjusted on the basis of the respiratory time constant to allow for effective emptying of the lungs in expiration and to avoid intrinsic positive end expiratory pressure (PEEP). This system is combined with another closed-loop control system for automatic adjustment of the inspired fraction of oxygen of the patient. This controller uses the feedback of arterial oxygen saturation of the patient and combines a rapid stepwise control procedure with a proportional-integral-derivative (PID) control algorithm to automatically adjust the oxygen concentration in the patient's inspired gas. The dual closed-loop control system has been examined by using mechanical lung studies, computer simulations and animal experiments. Results. In the mechanical lung studies, the ventilation controller adjusted the breathing frequency and tidal volume in a clinically appropriate manner in response to changes in respiratory mechanics. The results of computer simulations and animal studies under induced disturbances showed that blood gases were returned to the normal physiologic range in less than 25 s by the control system. In the animal experiments under steady-state conditions, the maximum standard deviations of arterial oxygen saturation and the end-tidal partial pressure of CO2 were ± 1.76% and ± 1.78 mmHg, respectively. Conclusion. The controller maintained the arterial blood gases within normal limits under steady-state conditions and the transient response of the system was robust under various disturbances. The results of the study have showed that the proposed dual closed-loop technique has effectively controlled mechanical ventilation under different test conditions.

Improving mechanical ventilation for patient care through repetitive control

Optimal Control of Mechanical Ventilators with Learned Respiratory Dynamics

Optimising mechanical ventilation through model-based methods and automation

A Dual Closed-Loop Control System for Mechanical Ventilation

Development and validation of a reinforcement learning algorithm to dynamically optimize mechanical ventilation in critical care

Volume Control of Low-Cost Ventilator with Automatic Set-Point Adaptation

An adaptive fractional order optimizer based optimal tilted controller design for artificial ventilator

Machine learning algorithm for ventilator mode selection, pressure and volume control

New Advances in Monitoring and Triggering of Mechanical Ventilation

[A computer-controlled closed circle system for ventilation during anesthesia and intensive care and its possibilities for patient monitoring]

Design and Performance Evaluation of a Low-Cost Non-Invasive Electromechanical Ventilator With Feedback Mechanism

Mechanical ventilation strategy for pulmonary rehabilitation based on patient-ventilator interaction

Adaptive Pressure Control Ventilation: Set It and Don't Forget It?

Methodology for Interpretable Reinforcement Learning for Optimizing Mechanical Ventilation

Telerobotic Operation of Intensive Care Unit Ventilators

A survey on patient synchronized electronic ventilator with optimized assistance

A Reinforcement Learning Approach to Weaning of Mechanical Ventilation in Intensive Care Units

Dynamic Characteristics Of A Mechanical Ventilation System With Spontaneous Breathing

Design and simulation of AI-based low-cost mechanical ventilator: An approach

Automation of Non-invasive Artificial Ventilation for Newborns Based on the Processing of Pressure Signals and Oxygen Concentration in the Gas-air mixture

Monitoring Asynchrony During Invasive Mechanical Ventilation