FEM-based volume estimation using electrode catheter measurements.

Abstract

Cardiovascular diseases are the leading cause of death worldwide, and accurate assessment of left ventricular volume (LVV) is crucial for diagnosis, therapy guidance, and long-term monitoring in these patients. Existing clinical techniques either require substantial resources, depend strongly on operator expertise, or lack the precision required for continuous monitoring. In this work we investigate the feasibility of a finite-element- method-based (FEM) algorithm for estimating volume and catheter offset from catheter-based admittance measurements to provide the basis for a new LVV estimation method. The present paper describes the underlying ventricle FEM model, a sensitivity-guided selection of informative 4-electrode configurations, and a constrained Gauss-Newton estimation algorithm. The algorithm was evaluated using numerical simulations and measurements obtained with a dynamic 3D-printed ventricle phantom with controllable volume changes. In both simulation and experiment, the reduced measurement set achieved volume reconstructions comparable to the full measurement set. In the phantom experiments, volume estimation errors of approximately 3-4% were observed over a wide range of volumes. These results demonstrate the feasibility of FEM-based parameter estimation and show that a small number of optimized electrode configurations can provide sufficient information for robust volume estimation. While the present study uses simplified geometries and a controlled phantom setup, the proposed approach provides a methodological basis for future investigations using more realistic anatomical models and biological experiments.