Solution of ECG Inverse Problem Using Artificial Neural Network

In this paper, a solution of the ECG inverse problem is computed using an Artificial Neural Network. The ECG inverse problem is formulated by using a simple example which is in the form of a Laplace equation in a unit square and two boundary conditions, which are like the boundary conditions of the inverse problem of ECG. We proposed an Artificial Neural Network model using Tikhonov Regularization techniques (ANNTR) for solving the ECG inverse problem. In the first step Artificial Neural Network will be used to approximate the forward problem solution and in the next step Tikhonov Regularization techniques by vary the value of the regularization parameter ϵ starting from 0.001. The root mean squared error of the inverse problem with the measured data is computed to obtain an optimal value of ϵ that accurately determines the solution of the inverse problem.