Abstract 19451: Is a CardioAlarm Using Ambulatory Subcutaneous Electrocardiography from Closely Spaced Electrodes Feasible?
Methods: Eight dogs were implanted for ambulatory testing. The SQ sensor was a 4cm square device containing 4 electrodes at its corners, 2-channel ECG amplifier, microprocessor for detecting VF or asystole, and 433 MHz transmitter for sending an alarm and SQ ECG to a receiver nearby. ECGs were recorded from multiple sites on the left chest. Q and SQ ECGs were simultaneously compared for QRS amplitude and morphology. ECGs were transmitted from the SQ implant during normal ambulatory activity and during induced VF. These were processed by an algorithm with an 8–17Hz band-pass filter, an adaptive-threshold R wave detector, and typical ICD VF detection algorithm.
Results: Recordings from surface and immediately underlying SQ electrodes were equivalent (average r= 0.94; range 0.81–0.98). At acute implant, Q and SQ QRS amplitudes averaged 1.1 and 1.0 milliVolts (mV) respectively. In long term implant (6 months), the mean QRS amplitude was stable at 1.3 mV in sinus and 0.8mV in VF. ECGs were successfully transmitted from 3 to 6 meters during ambulatory activity with little artifact. Orthogonal leads were useful to adjust for different ambulatory positions. R-wave detection had a sensitivity of 99.9% and a positive predictive accuracy of 97.2%. The alarm sensitivity and specificity were both 100%.
Conclusions: Cutaneous and SQ ECGs from the same location are equivalent in amplitude and morphology. When the site of implantation is guided by surface recordings during sinus rhythm, a SQ sensor with electrodes at its corners can obtain and transmit ECGs accurate to detect, store and distinguish VF and bradyarrhythmia from sinus rhythm. Orthogonal ECG leads were helpful to adjust for and maintain adequate QRS amplitude during ambulatory position changes.
- © 2010 by American Heart Association, Inc.