Sensorless cardiac phase detection for synchronized control of ventricular assist devices using nonlinear kernel...

Area of Science:

• Biomedical Engineering
• Cardiovascular Devices
• Signal Processing

Background:

• Synchronizing ventricular assist devices (VADs) with heart rhythm is crucial for heart failure patients.
• Conventional synchronization methods often require invasive sensors, posing clinical challenges.
• There is a need for non-invasive methods to achieve VAD-heart rhythm synchronization.

Purpose of the Study:

• To extract heartbeat components from VAD signals without additional sensors.
• To synchronize the VAD's rotational speed with the patient's intrinsic heartbeat.
• To develop a sensor-free VAD synchronization system for clinical applicability.

Main Methods:

• Utilized nonlinear kernel regression to estimate instantaneous rotational speed from VAD raw signals (consumption current, rotational speed).
• Extracted heartbeat component by calculating the estimation error of the regression model.
• Determined model parameters using VAD signals unaffected by heartbeat for accurate estimation.

Main Results:

• Successfully extracted the heartbeat component from VAD signals using the proposed nonlinear kernel regression model.
• Validated the method's efficacy on a mock circulatory system, demonstrating accurate heartbeat extraction.
• Achieved successful heartbeat synchronization control of the VAD without requiring any additional sensors.

Conclusions:

• The developed method enables sensor-free extraction of heartbeat components for VAD synchronization.
• This approach offers a promising non-invasive solution for synchronizing VADs with heart rhythm in heart failure patients.
• The findings support the clinical feasibility of advanced VAD control strategies without additional instrumentation.