AUTOPILOT-BT: a system for knowledge and model based mechanical ventilation

Area of Science:

• Biomedical Engineering
• Critical Care Medicine
• Control Systems

Background:

• Mechanical ventilation is crucial for respiratory support but requires precise control.
• Optimizing ventilator settings is complex, influenced by patient-specific factors and disease states.
• Existing systems may lack adaptability and autonomous optimization capabilities.

Purpose of the Study:

• To develop and evaluate AUTOPILOT-BT, a closed-loop system for autonomous mechanical ventilation control.
• To achieve clinician-specified goals, optimize settings based on disease, and adapt to individual patient conditions.
• To improve precision and efficiency in reaching optimal ventilation states.

Main Methods:

• AUTOPILOT-BT integrates fuzzy logic control (expert knowledge) and a model-based system (individualized patient data).
• Focuses on arterial oxygen saturation (SpO2), end-tidal CO2 pressure (PetCO2), and positive end-expiratory pressure (PEEP) to maximize respiratory system compliance (Crs).
• Evaluated using a lung simulator with varying CO2 production rates, comparing AUTOPILOT-BT to clinician control.

Main Results:

• The model-based control system within AUTOPILOT-BT demonstrated superior precision compared to clinicians.
• AUTOPILOT-BT achieved faster settling times to the optimal ventilation state.
• The system effectively simulated different metabolic CO2 production rates.

Conclusions:

• The AUTOPILOT-BT closed-loop system shows significant potential for advanced mechanical ventilation management.
• Its model-based approach offers enhanced precision and efficiency over traditional clinician control.
• Further clinical validation is warranted to confirm benefits in real-world patient care.