What is the clinical evidence from this study?

Study design: Other. Population: Prenatal monitoring (fetal ECG extraction). Intervention: MCA-FM (minimal channel attention and flow matching). Primary outcome: Pearson correlation coefficient (PCC).

What question did this study set out to answer?

This research aims to improve non-invasive fetal ECG extraction methods for better prenatal monitoring.

June 16, 2026Open Access

MCA-FM: Robust Non-Invasive Fetal ECG Extraction via Minimal Channel Attention and Flow Matching

Key Result

The MCA-FM method achieved average Pearson correlation coefficients of 0.94 ± 0.050 on the ADDB dataset and 0.91 ± 0.122 on the BDDB dataset for non-invasive fetal ECG extraction.

Key Points

This research aims to improve non-invasive fetal ECG extraction methods for better prenatal monitoring.
Proposed a new method combining minimal channel attention and flow matching for FECG extraction.
Employs a bridge variance scheduling approach during the diffusion process.
Utilizes leave-one-out cross-validation on two public datasets (ADDB and BDDB).
Achieved average Pearson correlation coefficients of 0.94 ± 0.050 on ADDB and 0.91 ± 0.122 on BDDB.
Demonstrated improved waveform fidelity and stability in FECG extraction.
Balanced high accuracy with efficient feature extraction methods.

Structured PICO

Does the MCA-FM method accurately extract non-invasive fetal ECG from maternal abdominal ECG?

Population

Maternal abdominal ECG recordings from two public datasets (ADDB and BDDB)

Intervention

MCA-FM (minimal channel attention and flow matching) method

Outcome

Pearson correlation coefficients (PCCs) of extracted fetal ECGsurrogate

The MCA-FM method provides a robust and accurate approach for extracting fetal ECG from maternal abdominal recordings, potentially improving non-invasive fetal heart health monitoring.

Abstract

Non-invasive fetal electrocardiogram (FECG) extraction from maternal abdominal ECG (AECG) is crucial for prenatal monitoring but remains challenging due to strong interference from maternal ECG (MECG), baseline drift, and noise. We propose an FECG extraction method based on minimal channel attention (MCA) and flow matching (FM), learning a deterministic mapping from AECG to FECG via a probabilistic path. To balance the preservation of physiological signals and separation of interference, we employ bridge variance scheduling for the diffusion process. Target matching loss is introduced to regress the FECG directly, enhancing training stability and waveform fidelity. For feature selection, a minimal channel attention module with global average pooling and a single linear layer is embedded after feature extraction, capturing cross-channel dependencies with minimal parameters. Enhanced residual connections are incorporated to retain underlying features and optimize gradient flow in deep networks. Experiments on two public datasets (ADDB and BDDB) with a leave-one-out cross-validation strategy show that our method achieves average Pearson correlation coefficients (PCCs) of 0.94 ± 0.050 on ADDB and 0.91 ± 0.122 on BDDB, demonstrating robust performance across diverse real-world recording conditions. The method balances high accuracy with efficient feature extraction, offering a reliable solution for non-invasive fetal heart health monitoring.