A Deep Learning Model for Tip Force Estimation on Steerable Catheters Via Learning-from-Simulation

EasyChair Preprint no. 10364

Abstract

Atrial Fibrillation (AFib) is the most common arrhythmia among the elderly population, where electrical activity becomes chaotic, leading to blood clots and strokes. During Radio Frequency Ablation (RFA), the arrhythmogenic sites within the cardiac tissue are burned off to reduce the undesired pulsation. Several studies showed excessive contact forces (more than 0.45 N) increase the incidence of tissue perforation, while inadequate force (less than 0.1 N) results in ineffective ablation. For the purpose of addressing the force estimation problem, finite element analysis can provide a useful tool to estimate the real-time tip contact force of the RFA catheters. In this work, a nonlinear planar finite element model of a steerable catheter was first developed with parametric material properties in ANSYS software. After that, a series of simulations based on each mechanical property was performed, and the deformed shape of the catheter was recorded. Next, validation was conducted by comparing the results of the simulation with experimental results between the range of 0-0.45 N to determine the material properties. The main contribution of this study was proposing a synthetic data generation, so as to train a light deep learning architecture for tip force estimation according to the finite element simulations. The proposed solution not only feeds the data-hungry methods based on deep learning with a sufficient amount of data, but also shows the feasibility of replacing the fast, accurate, and light-weight learning-based methods with slow finite element simulations.

Keyphrases: deep learning, finite element analysis, force estimation, Learning-From-Simulation, Steerable Catheter

@Booklet{EasyChair:10364,
  author = {Majid Roshanfar and Pedram Fekri and Javad Dargahi},
  title = {A Deep Learning Model for Tip Force Estimation on Steerable Catheters Via Learning-from-Simulation},
  howpublished = {EasyChair Preprint no. 10364},

  year = {EasyChair, 2023}}