Acoustic Characterization of Axial Flow Left Ventricular Assist Device Operation In Vitro and In Vivo

Abstract

The use of left ventricular assist devices (LVADs), implantable pumps used to supplement cardiac output, has become an increasingly common and effective treatment for advanced heart failure. Although modern continuous-flow LVADs improve quality of life and survival more than medical management of heart failure, device malfunction remains a common concern. Improved noninvasive methods for assessment of LVAD function are needed to detect device complications. An electronic stethoscope was used to record sounds from the HeartMate II axial flow pump in vitro and in vivo. The data were then uploaded to a computer and analyzed using two types of acoustic analysis software. Left ventricular assist device acoustics were quantified and were related to pump speed, acoustic environment, and inflow and outflow graft patency. Peak frequency values measured in vivo were found to correlate strongly with both predicted values and in vitro measurements (r > 0.999). Plots of the area under the acoustic spectrum curve, obtained by integrating over 50 Hz increments, showed strong correlations between in vivo and in vitro measurements (r > 0.966). Device thrombosis was found to be associated with reduced LVAD acoustic amplitude in two patients who underwent surgical device exchange. ASAIO Journal 2016; 62:46-55.

Acoustic and afterload evaluation of left ventricular assist devices

“…The difference in preload and afterload does not seem to affect the amplitude for peak harmonic frequency, which also indicates that sound analysis might be a valid method with low disturbance from other factors that might affect pump function. 274 The sound recordings from patients and mock loop are similar for both HMII and HM3, shown in paper II and III and by Yost. 263,274 However, all studies, including ours, are mostly based on a small number of patients or clinical situations which impact the validity of the method.…”

“…274 The sound recordings from patients and mock loop are similar for both HMII and HM3, shown in paper II and III and by Yost. 263,274 However, all studies, including ours, are mostly based on a small number of patients or clinical situations which impact the validity of the method. The first step to further assess the method could be to perform a larger scale simulation in mock loop circuit with daily recordings.…”

“…Different viscosities of the fluid within the mock loop circuit and different acoustic positions have previously been studied by Yost. 274 The power, flow, pulsatility and peak harmonic frequency were left unchanged as viscosity was increased. The results indicate that the use of water as a medium is sufficient to study the acoustics.…”

“…In presence of a thrombus, the spectral energy is reduced, resulting in lower amplitudes and lower area under the curve. 263 To evaluate if the location and mass of the thrombus have an impact on the acoustics, an artificial thrombus model with silicone simulating thrombosis in HVAD were developed. Both acoustics and vibration analysis were performed.…”

“…In Yost et al an electronic stethoscope, of the same brand as used in paper III, was used for sound recording. 263,274 Validity of the electronic stethoscope was never discussed. As shown in paper III, the electronic stethoscope is more susceptible for disturbances and less efficient in determining the fundamental frequency.…”