Project Summary
Atrial Fibrillation (AF) is the most common sustained arrhythmia in adults, carries a lifetime risk of 25-33%,
and cost the US healthcare system > 6 Billion dollars a year. Morbidities from AF include congestive heart failure,
dementia, symptoms due to decreased cardiac output, and most importantly an elevated thrombotic stroke risk
due to blood stasis. Recent advances in wearable technology have increased the capacity and accuracy of AF
diagnosis through monitoring heart rate variability in an ambulatory setting. However, there remains no
available ambulatory technologies that permit continuous assessment of the hemodynamic effects of
AF nor measurements of cardiac repolarization that determine the proarrhythmic effects of
antiarrhythmic drugs. Some treatment aspects of AF have remained controversial due to limited efficacy of
rhythm control interventions including cardiac ablation and pharmacologic therapy, both of which have significant
risks. Pharmacologic interventions using Class III antiarrhythmics are especially risky due to the proarrhythmic
effects on ventricular repolarization which can lead to life-threatening ventricular tachyarrhythmias such as
torsades de pointes. As a result of these issues, careful assessment of ventricular repolarization, as represented
by the QT interval on the electrocardiogram (ECG), is required during drug initiation. This requires lengthy and
costly inpatient monitoring during drug loading and difficulties in maintaining the narrow therapeutic window after
discharge. Having a reliable method to quantify measures of hemodynamic instability that leads to some AF-
related symptoms, as well as continuously monitoring the QT interval and other clinically important parameters
in an ambulatory setting would allow clinicians to tailor their care to the individual. This proposal aims to refine
a flexible, non-obtrusive wearable system to continuously monitor these key clinical parameters, help in
predicting heart rhythm complications and regulate medication intervention for both rhythm control and
rate control of AF. A sensor suite consisting of a small, wireless and flexible chest-worn patch that measures
both ECG and accelerometer based seismocardiography (SCG) and a wirelessly time synchronized flexible limb
unit capable of monitoring photoplethysmographic (PPG) signals will be used. Aim 1 will quantitatively define the
optimal location for QT interval monitoring on the surface of the chest. Aim 2 will combine the ECG, SCG, and
PPG data streams and extract both unique and time-relational parameters from the signal streams known to be
related to cardiac function and hemodynamics, using machine learning to obtain predictions of hemodynamic
instability from these complex relationships. Aim 3 will address the use and compliance of this wearable system
in a clinical setting against the standard of care during Class III antiarrhythmic drug loading and pharmacologic
cardioversion. The results from this study will provide the first ambulatory measures of AF symptoms and
responses to treatment, informing future studies and clinical care teams of the dominant bioelectrical and
biomechanical parameters that are the most important in the care of the growing, aging population with AF.
