Receive a weekly summary and discussion of the top papers of the week by leading researchers in the field.

In Frontiers in bioengineering and biotechnology

Background : Cardiac conduction disturbance requiring new permanent pacemaker implantation (PPI) is an important complication of TAVR that has been associated with increased mortality. It is extremely challenging to optimize the valve size alone to prevent a complete atrioventricular block (AVB).

Methods : In this study, we randomly took 48 patients who underwent TAVR and had been followed for at least 2 years to assess the risk of AVB. CT images of 48 patients with TAVR were analyzed using three-dimensional (3D) anatomical models of the aortic valve apparatus. The stresses were formulated according to loading force and tissue properties. Support vector regression (SVR) was used to model the relationship between AVB risk and biomechanical stresses. To avoid AVB, overlapping regions on the prosthetic valve where AV bundle passes will be removed as cylindrical sector with the angle θ. Thus, the optimization of the valve shape will be predicted with the joint optimization of the θ and valve size R.

Results : The average AVB risk prediction accuracy was 83.33% in the range from 0.8-0.85 with 95% CI for all cases; specifically, 85.71% for Group A (no AVB), and 80.0% for Group B (undergoing AVB after the TAVR).

Conclusions : This model can estimate the optimal valve size and shape to avoid the risk of AVB after TAVR. This optimization may eliminate the excessive stresses to keep the normal function of both AV bundle and valve leaflets, leading to a favorable clinical outcome. The combination of biomechanical properties and machine learning method substantially improved prediction of surgical results.

Zhang Guangming, Liu Rong, Pu Min, Zhou Xiaobo


atrioventricular block, calcification, finite element method, stress, transcatheter aortic valve replacement