Affiliation:
1. Department of Biomedical Engineering, SNS College of Technology (Autonomous), Coimbatore
2. Department of Electronics and Communication Engineering, Mahendra Engineering College (Autonomous), Namakkal
Abstract
Fetal Electrocardiogram (FECG) analysis helps in diagnosis of fetal heart. Extracting FECG from composite abdominal signal that contains noises like maternal ECG (MECG), electrical interference etc is a topic of great research interest, and several approaches have been reported. The proposed method is Heuristic RNN-based Kalman Filter for Fetal Electrocardiogram Extraction (HRKFFEE) which is based on redundant noise and signal patterns in the residual signal of FECG and MECG. Two functional blocks are used in the proposed method. The first functional block is based on Heuristic RNN equipped with legacy Long Short-Term Memory (LSTM) for assembling a knowledgebase and the second functional block is RNN-based Kalman filter. Upon testing, the proposed method delivers better average values of accuracy, F Score, Precision and Specificity as 93.118%, 93.106%, 92.9495 % and 92.98% respectively.
Subject
Artificial Intelligence,General Engineering,Statistics and Probability
Reference50 articles.
1. A review of fetal ecg signal processing issues and promising directions;Sameni;The Open Pacing, Electrophsiology and Therapy Journal,2010
2. The future of fetal monitoring;Wolfberg;Reviews in Obstetrics & Gynecology,2012
3. Fetal magnetocardiogram waveform characteristics;Strand;Physiol Meas,2019
4. The Future of Fetal Monitoring;Wolfberg;Rev Obstet Gynecol,2012
5. Noninvasive Fetal ECG analysis;Clifford;Physiol Meas,2014