A Boundary Element Approach for Estimating Cardiac Ejection Fraction Using Electrical Impedance Tomography

Abstract

Total heart volume change during cardiac cycle reflects the efficiency of energy used by heart. Cardiac ejection fraction (CEF) is one parameter which indicates the volume changes in heart due to pumping of blood. Determining CEF helps in monitoring the patient's heart condition and assist in diagnosis. In this study, electrical impedance tomography (EIT) is used as an imaging modality to detect the changes in ventricles due to heart beat. Situation is considered where a patient is told to hold his breath while the heart beats during the cardiac cycle. The resistivity of lungs, heart and background are assumed to be known priori therefore the conventional resistivity imaging is transformed to estimating the boundaries of ventricles. Boundary element method (BEM) with analytic boundary integration is formulated with complete electrode model to solve the forward problem of EIT. The boundary of ventricles is represented using truncated Fourier series with the coefficients describing the location and shape. The unknown Fourier coefficients are considered to be unknowns and are estimated using Levenberg-Marquardt method. Numerical experiments are performed with 2D human torso shape and the results show a promising performance of EIT in estimating the CEF.