Hybrid BSS techniques for foetal ECG extraction using framework for stress-testing extraction algorithms

The non-invasive foetal ECG makes use of surface electrodes placed onto the maternal abdomen. The challenge is to extract the foetal signal from the ab-dominal mixture. Blind source separation is one way to do this using either ICA or PCA algorithms. COMBI and MULTICOMBI algorithms offer novel schemes for combining ICA and PCA. In this work, the performance of the algorithms COMBI, MULTICOMBI, EFICA, WASOBI and traditional JADE and ICA algorithms are compared, using a standard framework for benchmarking and regulatory testing of NI-FECG extraction algorithms. We use the F1-measure combining true positive, false negative and false positive detected peak for measure the accuracy of FQRS detection. Two experiments were carried out. One to determine a WASOBI algorithm input parameter, also necessary in COMBI, and other to compare performance using F1 measure. In first experiment it was established that the AR order required as input for COMBI is 10. In second experiment, the overall median FQRS detection accuracies (i.e. considering all no stationary events) are obtained. For the best performing methods in each group were 96.4% for COMBI, 95.8% for EFICA, 95.2% for JADEICA, 94.6% for PCA, 93,6% for ICAdef,, 92.9% for MULTICOMBI, 92,7 for ICAsym, and 89,7% for WASOBI.