Detection of physiological control inputs preload and afterload from intrinsic pump parameters in total artificial heart

• Published in: Artificial organs 2023-05, Vol.47 (5), p.817-827
• Main Authors: Kirn, Borut, Diedrich, Mario, Schmitz‐Rode, Thomas, Steinseifer, Ulrich, Jansen, Sebastian V.
• Format: Article
• Language: English
• Subjects: afterload; Artificial organs; Cardiovascular system; Coils; Ejection; Heart; Heart, Artificial; intrinsic parameters; Optimization; Outflow; Parameters; Pattern analysis; physiological control; Physiology; preload; Pulmonary Circulation; Reproducibility of Results; Systole; total artificial heart
• ISSN: 0160-564X; 1525-1594
• DOI: 10.1111/aor.14481

Background In the total artificial heart (TAH), the inputs to the physiological control unit, preload, and afterload, are detected from intrinsic pump parameters (e.g., motor current). Within this study, their detection techniques are developed, and their reliability in pre‐ and afterload prediction is mapped for a broad range of cardiovascular system states. Methods We used ReinHeart TAH which is a fully implantable TAH with a plunger coil drive that is alternately emptying the left and right chambers. From the coil currents we first derived a force generated by the piston with respect to its position and then analyzed its pattern to detect (1) preload—chamber filling, found as piston position at begin ejection and (2) afterload—mean outflow pressures, determined as linearly calibrated average piston force during ejection. TAH is then integrated into a mock loop circulation (MLC) which is set to 135 different steady operating points varying in chamber filling (0%–100%, five steps), mean outflow pressures (system circulation: 60–90–120 mm Hg, pulmonary circulation: 15–30‐45 mm Hg), and heart cycle duration (171–600 ms in seven non‐equidistant steps). The detected preload and afterload are compared to MLC set values, and the errors are mapped. Results Respectively for the left and right chambers, the preload was detectable in 134 and 118 operating points and the mean error was ±3% and ±2%. The afterload was detectable in 135 and 87 operating points and the mean error was 37% and 30% respectively for left and right circulation. The operational points that are further away from homeostatic equilibrium values generally yielded larger errors. The largest errors were observed for right circulation at long cycle duration, low afterload, and low filling. Conclusions The study yields reliable preload estimation in a broad range of physiological states, particularly for left circulation. Detection of afterload needs further improvements. The study revealed a need for piston movement optimization within the ReinHeart TAH during the early phase of systole. The intrinsic parameter (piston force) in total artificial heart (TAH) is studied to detect the preload and afterload as a ramp in the force profile curve and as a mean value during ejection, respectively. The TAH is integrated into a mock cardiovascular loop set to 135 different operational points varying in preload, afterload, and systole duration. The maps of detection errors reveal lower errors in detected prelo