Elimination of Isoline Drift when Analysis of the Electrocardiosignal of the Vehicle Driver

The paper considers a new method of processing an electrocardiosignal (ECS), which increases the reliability of its analysis. The purpose of our work is to eliminate the drift of the ECS isoline during automatic analysis of the ECS. This interference is one of the difficult-to-remove ones and the solution of this problem significantly increases the accuracy of the subsequent analysis of the ECS in a number of tasks, especially in the task of detecting cardiac arrhythmias. To solve this problem, we propose to extract the isoline drift signal from the recorded ECS using spline approximation based on the regularization method. To do this we use the signal samples at the TP-segments of the ECS as reference points for the spline. This method uses a number of natural approaches to approximating such functions that generate splines whose values are given with errors. Then we subtract the allocated signal of the isoline drift from the registered ECS. The implementation of the proposed method for eliminating the isoline drift using spline approximation based on the regularization method on real ECS showed that the residual drift does not exceed 0.97% of the initial one with an increase in the isoline drift frequency up to half the heart rate. This allows us to conclude that the accuracy of the ECS parameter estimation obtained using the proposed method is increased in comparison with the “classical” method of spline approximation. The originality of the proposed method lies in the fact that the samples of the TP-segments of the ECS were used to construct a spline approximation of the isoline drift.