In Vivo Acoustoelectric Imaging of Cardiac Currents in a Swine Model

Abstract

Acoustoelectric Cardiac Imaging (ACI) is a novel ultrasound-based electrical imaging modality capable of capturing spatiotemporal dynamics of electrical activation with high spatial and temporal resolution. This work describes the translation of ACI from a benchtop platform to an in vivo imaging modality; additionally, it demonstrates that ACI can capture beat-to-beat variability and 4D activation dynamics in a healthy farm swine model. After exploring the tradeoff in resolution and sensitivity of AE detection with different ultrasound (US) parameters in several commercial linear and custom matrix arrays, a custom 1.5D, 1.5 MHz US array crafted specifically for ACI was designed, fabricated, and tested. In addition to optimizing these US parameters and fabricating a new array, a new signal acquisition instrumentation setup was constructed to record ACI signals from a novel model of healthy cardiac activation wave. A proof-of-concept study using this instrumentation in a single swine demonstrated that after appropriate signal processing, ACI signals were time-synchronized with the cardiac activation wave and dependent on the level of US pressure pulsed into the tissue. Additionally, the proof-of-concept study provided validation of ACI with propagation velocity values measured from a recording electrode array placed directly on the epicardium. Finally, after proving the utility of ACI in this model, a study with nine swine was undertaken to analyze the variability in 1D beat-to-beat, 4D multi-beat averaged, and isochrone/isopotential mapping, which might hold relevance for physiological study. Beat-to-beat ACI M-Mode mapping on nine swine was realized with an SNR of 4.00 ± 0.36 dB. These maps demonstrated consistent activation magnitude, time (near the depolarization peak in the EGM), and depth on an intra-swine level, while the morphology of the ACI signal for each beat and between swine was unique. 4D activation cineloops exhibited a consistent FWHM at peak activation amongst four swine – 5.50 ± 0.30 mm along the apicobasal axis, 2.76 ± 0.10 mm along the mediolateral axis, and 6.16 ± 0.18 mm along the endo-epicardial axis. Finally, physiologic parameters extracted from ACI data highlighted a propagation velocity of the cardiac activation wave of 0.22 ± 0.03 m/s among nine swine. The work described in this dissertation demonstrates that ACI has great potential to be translated to map electrical activity in human patients with ventricular arrhythmias.