Compliance modulation of a soft robotic atrioventricular model of heart failure with preserved ejection fraction

Figures

• 

  Fig. 1. Overview of the simulator and its biomimicry. A Factors influencing the development of HFpEF and its effect on myocardial stiffness, ventricular volume, and pressure during diastole. B The bio-inspiration of our LV simulator and its compliance modulation in diastole, allowing for diastolic augmentation mimicking HFpEF. Where LVP is left ventricular pressure, and AOP is aortic pressure.

• 

  Fig. 2. Simulation of impaired relaxation of the ventricle. A Visualization of healthy (blue) and impaired (red) relaxation of the ventricle. B–E Averaged LVP and AOP, muscle control signal, mitral and aortic flow, and PV loop over one cardiac cycle with increasingly impaired relaxation. F–H End-diastolic pressure (EDP) and minimum LVP, peak mitral flow rate, and end-diastolic volume with respect to muscle control signal relaxation gradient. B–H, excluding C and E, were plotted as mean with ±SD in a transparent shade. D includes a zoomed section revealing standard deviation in average flow rate. C is a feedforward input and therefore has no mean or SD. E contains mean data from B and D, where SD can be found. Number of cycles used to calculate mean is found in Supplementary Table 3.

• 

  Fig. 3. Results of closed-loop control over diastolic volume. A Visualization of pericardial restraint on the ventricle and the effective simulation strategy we used. B–E Averaged LVP and AOP, muscle control signal, mitral and aortic flow, and PV loop over one cardiac cycle with increasing restraint. F–J End-diastolic pressure (EDP) and minimum LVP, peak mitral flow rate, and end-diastolic volume with respect to control signal amplitude. B–H, excluding E, were plotted as mean with ±SD in a transparent shade. E contains mean data from B and D, where SD can be found. Number of cycles used to calculate mean is found in Supplementary Table 4.

• 

  Fig. 4. Results for closed-loop modulation of ventricular reaction force to filling. A Data is plotted on a scale of blue to red, representing high to low desired stiffness, respectively. B–E Averaged LVP and AOP, muscle control signal, mitral and aortic flow, and PV loop over one cardiac cycle with increasing restraint. F–H End-diastolic pressure (EDP) and minimum LVP, peak mitral flow rate, and end-diastolic volume with respect to muscle control signal stiffness gradient. B–H, excluding E, were plotted as mean with ±SD in a transparent shade. E contains mean data from B and D, where SD can be found. Number of cycles used to calculate mean is found in Supplementary Table 5.

• 

  Fig. 5. Results for closed-loop diastolic ventricular compliance modulation. A Data is plotted on a scale of blue to red, representing high to low diastolic compliance, respectively. B–E Averaged LVP and AOP, muscle control signal, mitral and aortic flow, and PV loop over one cardiac cycle with increasing restraint. F–H End-diastolic pressure (EDP) and minimum LVP, peak mitral flow rate, and end-diastolic volume with respect to muscle control signal stiffness gradient. B–H, excluding E, were plotted as mean with ±SD in a transparent shade. E contains mean data from B and D, where SD can be found. Number of cycles used to calculate mean is found in Supplementary Table 6.

• 

  Fig. 6. Atrioventricular model of the progression of HFpEF. A Expected progression of HFpEF from literature51,61, where IVR is isovolumetric relaxation, and the grey lines in the two right-most panels are copies of the healthy expectations in the left panel. B Illustration of expected atrioventricular coupling in HFpEF progression as they correspond to events in the expected data in (A). C System diagram of the atrioventricular model and echocardiographic imaging and physical device. D Results of simulated progression of HFpEF using the atrioventricular model, data were plotted as mean with ±SD in a transparent shade. Number of cycles used to calculate mean is found in Supplementary Table 7.

• 

  Fig. 7. LA particle image velocimetry (PIV) in healthy and HFpEF conditions. A 3D presentation of the LA membrane alongside the physical atrioventricular prototype. B AoP, LVP, and LAP under healthy and HFpEF conditions. C Instantaneous PIV velocity vectors under healthy and HFpEF conditions during diastolic E- and A-waves.

• 

  Fig. 8. Simulation of HeartMate 3 LVAD support for HFpEF. A Schematic of experimental setup. B Aortic (AOP), left ventricular (LVP), and left atrial pressure (LAP) under restrictive HFpEF conditions. C The same pressures under LVAD support with increasing pump speed. D Flow through the LVAD during support with increasing pump speed. E Mean left atrial pressure (MLAP), left atrial–left ventricular pressure gradient at onset of filling phase, mean arterial pressure (MAP), and pulse pressure (PP) with increasing LVAD pump speed.

• 

  Fig. 9. Demonstration of LA collapse under increased LVAD pump speed. A The atrioventricular model in its MCL setup for LVAD support of HFpEF. B 3D model of the anatomical LA membrane and the ultrasound imaging plane. C Ultrasound imaging of the radial cross-section of the LA along with LVP, LAP, and AoP with a ramped increase in LVAD pump speed from 5000 to 7000 rpm.