An ECG forward solution by the boundary element method.

概要

論文の詳細を見る
It has already been found that the inverse problem in estimating epicardial potentials from measured body surface potentials is seriously disturbed by noise in the measured data and a small discrepancy of the transfer matrix from the subject concerned. The number of parameters which can be decided independently of one another is between ten and fifteen.To improve this situation, we should come back to the forward problem again and try to find a cardiac model which can simulate the observed evolution of the body surface potential. This approach requires finding of a technique for calculating the body surface potential from the assumed cardiac potential so quickly as to allow a trial-and-error approach. As a preliminary step to this approach, we have tried to solve a forward problem by means of a boundary element method which permits us much shorter computing time than the finite element method in solving the same problem at the same accuracy. In the present calculations, it has been found that inhomogeneity in the electric conductivity must be taken into consideration and that to assure computation errors of less than 0.1%, the body surface must at least be divided into 150 points which result in about 500 triangular elements. In this error estimation the lung surfaces were divided into 86 points which was determined from the capacity of our computer.
一般社団法人日本生体医工学会の論文