Advanced signal processing techniques for fetal ECG analysis

In response to the PhysioNet/CinC Challenge 2013: Noninvasive Fetal ECG [1] we developed an algorithm for fetal QRS (fQRS) positions estimation based on a set of classic filters, which enhances the fetal ECG, combined with a robust QRS detection technique based on Christov's beat detection algorithm. These steps provides necessary information for the maternal ECG (mECG) cancellation, which is based on the technique provided by the Challenge organizers. Our work extends the provided algorithm with mECG reduction quality check and in case of insufficient reduction the mECG reduction algorithm is applied again until the criteria for sufficient reduction based on energy around the maternal QRS complex are satisfied. After noise reduction two techniques for fQRS were applied - one provided by the organizers and second based on entropy estimation. Results from both detectors are then corrected creating another set of fQRS positions estimates and from all sets of fQRS estimates there is selected one with the smallest standard deviation of fetal R-R distances. Our method results are 249.784 for Event 1/4 and 21.989 for Event 2/5 respectively. We did not participate in Event 3 - QT interval estimation.