Enhancing stance robustness and jump height in bipedal muscle-actuated systems: a bioinspired morphological development approach

Recognizing humans' unmatched robustness, adaptability, and learning abilities across anthropomorphic movements compared to robots, we find inspiration in the simultaneous development of both morphology and cognition observed in humans. We utilize optimal control principles to train a muscle-actuated human model for both balance and squat jump tasks in simulation. Morphological development is introduced through abrupt transitions from a 4 year-old to a 12 year-old morphology, ultimately shifting to an adult morphology. We create two versions of the 4 year-old and 12 year-old models- one emulating human ontogenetic development and another uniformly scaling segment lengths and related parameters. Our results show that both morphological development strategies outperform the non-development path, showcasing enhanced robustness to perturbations in the balance task and increased jump height in the squat jump task. Our findings challenge existing research as they reveal that starting with initial robot designs that do not inherently facilitate learning and incorporating abrupt changes in their morphology can still lead to improved results, provided these morphological adaptations draw inspiration from biological principles.

Improvement in electrical synchrony during bi-ventricular vs. left ventricular pacing with dynamic atrioventricular delays may be predicted by conduction times

Funding Acknowledgements

Type of funding sources: Private company. Main funding source(s): Abbott

Introduction

Automatic adjustment of atrioventricular delay (AVD) with SyncAV has been shown to improve electrical synchrony with either biventricular (BiV) or left ventricular (LV) only pacing. Selecting the optimal pacing mode may be guided by measuring conduction delays between the right atrium (RA), right ventricle (RV), and LV.

Purpose

Evaluate correlations between inter-chamber conduction delays and the QRS duration (QRSd) reduction achieved by BiV or LV-only pacing with SyncAV.

Methods

CRT implant patients (LBBB, QRSd ≥ 150 ms) were prospectively enrolled. Blinded QRSd was measured from 12-lead ECG during BiV and LV-only pacing, with SyncAV enabled and optimized to minimize QRSd. Conduction delays were measured by the device from unipolar electrograms during pacing and sensing (e.g. RAp-RVs). Correlations between each delay and which mode had the narrower QRSd was assessed by binomial regression.

Results

In total, 68 patients were evaluated (66.1 yr, 67.1% male, 32.5% ischemic, 26.3% EF, 165.1 ms intrinsic QRSd). BiV+SyncAV and LV+SyncAV reduced QRSd by 23.8% and 21.3% (P<0.001) vs. intrinsic conduction. Of all delays, RVs-LVs and LVp-RVs were significantly correlated with patient probability of BiV+SyncAV yielding a narrower QRSd than LV+SyncAV. BiV+SyncAV was favored in 70.6% (48/68) of all patients, but was favored in 92.3% (24/26) of patients with LVp-RVs < median (143 ms) (P=0.030 vs all).

Conclusion

When combined with SyncAV, BiV yielded a narrower QRSd than LV-only in the majority of patients, but was favored in a greater proportion who exhibit LVp-RVs delays below the median. Such conduction delay cut-offs can be used to facilitate pacing mode selection.

Reducing clinical review burden for insertable cardiac monitors

Funding Acknowledgements

Type of funding sources: None.

Background

The insertable cardiac monitor (ICM) is an essential tool for the ambulatory diagnosis of arrhythmias. However, definitive diagnoses still rely on time-consuming, manual adjudication of electrograms (EGMs) transmitted to the patient care network. This EGM review burden may be minimized by automatically selecting a subset of EGMs for fast review without delaying the diagnosis.

Purpose

Develop EGM selection strategies to reduce the EGM review burden without delaying diagnoses.

Methods

A retrospective analysis of 1,000 randomly selected Abbott Confirm Rx devices with 90+ days of remote transmission history was performed, regardless of transmission frequency, and all EGMs were adjudicated as either true or false positive (TP, FP). Up to 3 EGMs per day per arrhythmia type were prioritized for review based on ventricular rate and episode duration, with rules specific to each arrhythmia type: atrial fibrillation (AF), tachycardia, bradycardia, and pause. The resulting reduction in EGM review burden and TP days (patient-days with at least 1 TP EGM), as well as any diagnostic delay from the first transmitted TP, were calculated relative to reviewing all transmitted EGMs.

Results

In this population and transmission period, at least one AF, tachycardia, bradycardia, and pause EGM was transmitted by 424, 343, 190, and 325 unique devices, respectively, with a total of 35,723, 12,239, 19,752, and 28,002 EGMs, and a total of 6,163, 1,572, 1,438, and 646 TP days. For these patients with ≥1 EGM, the median [IQR] EGM transmission rate was 2.6 [0.7, 11.6], 1.1 [0.4, 4.7], 2.1 [0.6, 10.7], and 3.4 [0.6, 29.9] EGMs/patient/month, respectively. The optimal EGM selection strategy reduced this EGM review burden by 43%, 67%, 76%, and 50%, while only missing 3.4%, 2.2%, 0.3%, and 0.2% of TP days, respectively. Ultimately, 97%, 99%, 99%, and 99% of devices with a TP AF, tachycardia, bradycardia, or pause EGM exhibited no diagnostic delay vs. reviewing all transmitted EGMs.

Conclusion

EGM prioritization rules for selecting up to 3 episodes/day significantly reduced EGM burden across all patients, not just "frequent fliers," with no delay-to-diagnosis in >97% of patients who exhibited a true arrhythmia. Implementing these rules on the patient care network may improve clinical workflow and ICM patient management. Abstract Figure.

Impact of pacing configuration and right ventricular lead location on dynamic atrioventricular delay optimization

Funding Acknowledgements

Type of funding sources: Private company. Main funding source(s): Abbott

Introduction

Automatic adjustment of atrioventricular delay (AVD) with SyncAV has been shown to improve electrical synchronization. However, the effect of pacing configuration and right ventricular (RV) lead location on SyncAV programming is unknown.

Purpose

  Evaluate the effect of pacing configuration and lead location on SyncAV optimization during biventricular (BiV) and LV-only pacing, with and without MultiPoint Pacing (MPP).

Methods

  Patients with LBBB and QRS duration (QRSd) ≥ 150 ms scheduled for CRT-P/D device implantation with quadripolar LV lead were enrolled in this prospective study. RV lead location was classified at implant by the operator via  fluoroscopy. QRSd was measured post-implant from 12-lead surface ECG by blinded experts during the following pacing modes: intrinsic conduction, BiV (BiV = RV + LV1), MPP (MPP = RV + LV1 + LV2), LV-only single-site (LVSS = LV1 only), and LV-only MPP (LVMPP = LV1 + LV2). For each mode, SyncAV was enabled (e.g. BiV + SyncAV) with the patient-tailored SyncAV offset that minimized QRSd. For BiV and LVSS, LV1 was the latest activating LV cathode; for MPP and LVMPP, LV1 + LV2 were the two LV cathodes with the widest possible separation (≥30mm). All modes used minimal RV-LV and LV1-LV2 delays.

Results

  Fifty-three patients (68% male, 36% ischemic, 26% ejection fraction, 169 ms intrinsic QRSd) completed device implant and QRSd assessment. RV leads were implanted in either the septum (48%) or apex (52%), according to implanting physician preference. Relative to intrinsic conduction, BiV + SyncAV and MPP + SyncAV reduced QRSd by 23% and 27%, respectively (p < 0.01). LVSS + SyncAV reduced QRSd by 22% (p < 0.01 vs BiV + SyncAV), and LVMPP + SyncAV reduced QRSd by 25% (p < 0.05 vs MPP + SyncAV). RV apex or septum lead location did not have a significant impact on QRS reduction for each pacing configuration. As a percent of PR interval, optimal SyncAV offsets were similar for BiV + SyncAV and MPP + SyncAV (16% vs 13%, p = 0.05), and for LVSS + SyncAV and LVMPP + SyncAV (18% vs 21%, p = 0.46), but were significantly higher for LV-only settings vs. corresponding BiV/MPP settings (p < 0.05 for both pairs). For BiV + SyncAV, apical vs septal RV leads required greater SyncAV offsets (22% vs 11%, p < 0.05). SyncAV offsets also tended to be higher in apical vs septal RV leads for MPP (21% vs 11%), LVSS (20% vs 15%), and LVMPP (25% vs 16%), but without statistical significance.

Conclusion

SyncAV improves acute electrical synchronization in CRT patients with LBBB, particularly with patient-specific SyncAV programming. Pacing configuration (RV + LV or LV only, with or without MPP) and RV lead location (apex or septum) could potentially influence optimal SyncAV programming. Abstract Figure.

862Syncav with multipoint pacing improves acute left ventricular hemodynamics

Funding Acknowledgements

Abbott

Introduction

SyncAV has been shown to improve electrical synchronization by automatically adjusting atrioventricular delay (AVD) according to the intrinsic atrioventricular conduction time. Additional incremental electrical synchronization may be gained by the addition of second left ventricular (LV) pulse with MultiPoint Pacing (MPP). While the electrical synchronization benefits of SyncAV have been previously explored, there has been no assessment of the acute hemodynamic impact of SyncAV with or without MPP.

Objective

 Evaluate the acute LV hemodynamic impact of SyncAV with and without MPP.

Methods

Heart failure patients with LBBB and QRS duration (QRSd) > 140 ms undergoing CRT-P/D implant with a quadripolar LV lead were enrolled in this prospective study. A guidewire or catheter with pressure transducer was placed in the LV chamber and the maximum pressure change (dP/dtmax) was recorded during the following pacing modes:  intrinsic conduction, conventional biventricular pacing with SyncAV (BiV + SyncAV), and MPP with SyncAV (MPP + SyncAV). Twelve-lead surface ECG was used to determine the patient-tailored SyncAV offset that minimized QRSd.

Results

Twenty-seven patients (67% male, 44% ischemic, 30 ± 7% ejection fraction) completed the acute recordings. Relative to the intrinsic QRSd of 163 ms, BiV + SyncAV reduced QRSd by 21.5% to 124 ms (p < 0.001 vs. intrinsic) and MPP + SyncAV reduced QRSd by 26.6% to 120 ms (p < 0.05 vs. BiV + SyncAV). Beyond electrical synchronization, SyncAV significantly improved acute hemodynamics. Relative to the intrinsic dP/dtmax of 842 mmHg/s, BiV + SyncAV elevated dP/dtmax by 6.3% to 900 mmHg/s (p < 0.001 vs. intrinsic) and MPP + SyncAV elevated dP/dtmax by 8.8% to 926 mmHg/s (p < 0.005 vs. BiV + SyncAV). Despite both QRSd and dP/dtmax improvement with SyncAV and MPP, correlation between electrical and hemodynamic measurements was poor (R2 = 0.0 for BiV + SyncAV, R2 = 0.1 for MPP + SyncAV).

Conclusion

SyncAV may significantly improve acute LV hemodynamics in addition to electrical synchrony in LBBB patients. Further incremental improvement was achieved by combining SyncAV with MPP.

Abstract Figure.

43Dynamic atrioventricular delay achieves superior electrical synchrony when pacing both ventricles rather than left ventricle alone

Funding Acknowledgements

Abbott

Introduction

Automatic adjustment of atrioventricular delay (AVD) with SyncAV has been shown to improve electrical synchronization when pacing one or two sites in the left ventricle together with the right ventricle. However, it is unknown if the same benefit can be gained by using SyncAV while pacing only the left ventricle without right ventricular pacing.

Purpose

  Evaluate the acute improvement in electrical synchrony provided by SyncAV with and without MultiPoint Pacing (MPP) during biventricular (BiV) and LV only pacing.

Methods

  Patients with LBBB and QRS duration (QRSd) ≥ 150 ms scheduled for CRT-P/D device implantation with quadripolar LV lead were enrolled in this prospective study. QRSd was measured post-implant from 12-lead surface electrograms by blinded experts during the following pacing configurations: intrinsic conduction, conventional BiV (BiV = RV + LV1), MPP (MPP = RV + LV1 + LV2), LV-only single-site (LVSS = LV1 only), and LV-only MPP (LVMPP = LV1 + LV2). For each pacing mode, SyncAV was enabled (e.g. BiV + SyncAV) with the patient-tailored SyncAV offset that minimized QRSd. As an additional reference, QRSd during BiV was also measured using the nominal static AVD (paced/sensed AVD = 140/110 ms). BiV and LVSS pacing used the latest activating LV cathode, whereas MPP and LVMPP used the two LV cathodes with the widest possible separation (>30mm). All configurations used the minimum programmable RV-LV and LV1-LV2 delays.

Results

  Thirty-five patients (78% male, 33% ischemic, 26% ejection fraction, 165 ms intrinsic QRSd) completed device implant and QRSd assessment. Relative to intrinsic conduction, BiV with nominal AVD reduced the QRSd by 17.5% (p < 0.001 vs intrinsic). Enabling SyncAV with a patient-optimized offset significantly improved QRSd reduction. BiV + SyncAV reduced QRSd by 25.2% (p < 0.001 vs. BiV). The greatest QRSd reduction of 28.9% was achieved by MPP + SyncAV (p < 0.01 vs. BiV + SyncAV). Single- and multi-site LV-only pacing reduced QRSd significantly less than corresponding biventricular modes. LVSS + SyncAV reduced QRSd by 22.5% (p < 0.05 vs. BiV + SyncAV), and LVMPP + SyncAV reduced QRSd by 24.3% (p < 0.05 vs. MPP + SyncAV). As a percent of PR interval, optimal SyncAV offsets were similar for BiV + SyncAV (median: 13%, mean: 17%) vs. MPP + SyncAV (median: 13%, mean 16%, p = 0.35 vs. BiV + SyncAV), and similar for LVSS + SyncAV (median: 20%, mean: 28%) and LVMPP + SyncAV (median: 23%, mean: 26%, p = 0.35 vs. LVSS + SyncAV), but were significantly higher for LV-only settings vs. corresponding BiV/MPP settings (p < 0.01 for both pairs). Conclusion: Greater improvement in electrical synchrony using SyncAV was observed when right ventricular pacing was included with left ventricular pacing. Additional benefit was gained by the addition of a second left ventricular pacing site with MPP in combination with SyncAV in both biventricular and LV only pacing modes.

Abstract Figure.

A man who inherited his SRY gene and Leri-Weill dyschondrosteosis from his mother and neurofibromatosis type 1 from his father

We report on a man with neurofibromatosis type 1 (NF1) and Leri-Weill dyschondrosteosis (LWD). His father had NF1. His mother had LWD plus additional findings of Turner syndrome (TS): high arched palate, bicuspid aortic valve, aortic stenosis, and premature ovarian failure. The proband's karyotype was 46,X,dic(X;Y)(p22.3;p11.32). Despite having almost the same genetic constitution as 47,XXY Klinefelter syndrome, he was normally virilized, although slight elevation of serum gonadotropins indicated gonadal dysfunction. His mother's karyotype was mosaic 45,X[17 cells]/46,X,dic(X;Y)(p22.3;p11.32)[3 cells].ish dic(X;Y)(DXZ1 +,DYZ1 + ). The dic(X;Y) chromosome was also positive for Y markers PABY, SRY, and DYZ5, but negative for SHOX. The dic(X;Y) chromosome was also positive for X markers DXZ1 and a sequence < 300 kb from PABX, suggesting that the deletion encompassed only pseudoautosomal sequences. Replication studies indicated that the normal X and the dic(X;Y) were randomly inactivated in the proband's lymphocytes. LWD in the proband and his mother was explained by SHOX haploinsufficiency. The mother's female phenotype was most likely due to 45,X mosaicism. This family segregating Mendelian and chromosomal disorders illustrates extreme sex chromosome variation compatible with normal male and female sexual differentiation. The case also highlights the importance of karyotyping for differentiating LWD and TS, especially in patients with findings such as premature ovarian failure or aortic abnormalities not associated with isolated SHOX haploinsufficiency.