Fetal ECG enhancement: Adaptive power line interference cancellation based on Hilbert Huang Transform

• An adaptive PLI canceller based on Hilbert Huang Transform, IPRAC, is proposed.
• It is compared with high performance adaptive, and EMD based algorithms.
• The evaluation considers three scenarios: variable/fixed PLI frequency (and harmonics).
• IPRAC shows the best results in all scenarios.
• It preserves abdominal fetal ECG morphology being appropriate for fetal monitoring.

In this paper the fetal electrocardiogram (fECG) enhancement in abdominal recordings acquired by electrodes placed on the maternal belly is considered, a new algorithm being proposed to remove the power line inference (PLI). Typically, PLI affects the evaluation of physiological signals, e.g., (fECG), for diagnostic purposes, therefore different hardware and software approaches to reduce/remove it from biomedical measurements have been proposed, some considering even the case of very low energy physiological signals. The remaining PLI still impairs the analysis of the signal of interest in some specific cases like fetal monitoring, where the morphology of fECG is essential. This paper proposes an adaptive filter based on Hilbert Huang Transform (HHT) to remove both the PLI fundamental frequency and its harmonics from the abdominal recordings, allowing the further fECG processing. The proposed algorithm, the Internal Powerline Reference Adaptive Canceler (IPRAC), shows very good performance in PLI cancellation without affecting the fECG morphology. The IPRAC performance is evaluated on both real and simulated signals including also the worst case scenario when the PLI signal does not have an absolutely constant fundamental frequency. It outperforms two recently investigated algorithms as proved by the evaluation of four different quantitative performance indexes analyzed in this study.