Biosignal processing methods to guide cardiac patients to perform safe and beneficial exercise for rehabilitation

This paper describes biosignal interpretation methods in an application to support a CAD (Coronary Artery Disease) patient during rehabilitation. The methods use physiological signals from a wearable device (embedded in a shirt) to guide the patient through safe and beneficial exercise. The paper describes the different design considerations of the processing chain: signal quality verification; on-line analysis and feedback; recovery analysis and exercise summary analysis. The system components have been validated: differences in heart rate between the device and the reference were small, electrocardiograph (ECG) signal quality estimation algorithm had high sensitivity and specificity, and the on-line feedback algorithms perform well enough for use during exercise in practice. A study regarding the clinical usefulness of the overall system is carried in 2012 in the context of the EU project Heartcycle.