Method on Fetal Electrocardiogram Extraction Utilizing LightGBM Combined with EKF

Abstract

A novel method on fetal electrocardiogram (FECG) extraction utilizing LightGBM combined with Extended Kalman Filter (EKF) was proposed. Firstly, the nonlinear transform between maternal electrocardiogram (MECG) and MECG component in the abdominal signal was estimated using LightGBM. Then the optimal estimation of MECG component was obtained by MECG undergoing the estimated nonlinear transform. And noisy FECG was extracted by suppressing the estimated MECG component. At last, the FECG denoising was performed by use of EKF. The clinical data are adopted to validate the proposed FECG extraction method. The experimental results show that the proposed FECG extraction method is better than other conventional methods both on visual results and objective assessment.