Detection of fetal heart rate using ANFIS displayed on a smartphone

Abstract

Pregnant mothers can be monitored for accurate fetal ECG (FECG) from the abdominal ECG (aECG). This can help detect fetal asphyxia based on the decreasing fetal heart rate (FHR) before the fetus enters into the acidosis phase. In this paper, a technique is proposed based on the adaptive neuro fuzzy inference system (ANFIS) along with a fetal QRS detector. The proposed QRS detector is based on the Pan and Tompkins algorithm which is optimized to give the least number of incorrect maternal and fetal R peaks. The modified QRS detector adjusts the threshold for each record to obtain FHR. The proposed ANFIS system which was evaluated with the abdominal and direct fetal ECG (adfecgdb) physionet recordings, correctly estimated the FECG signal for most records. Using Matlab Mobile we could wirelessly connect to the laptop running Matlab files with our smartphone, accurately displaying simulations files and average FHR for each record.