Comparison of Intelligent Non-invasive Fetal Health Monitoring Schemes

Fetal echocardiography (FECG) analysis was rarely used earlier in clinical practice using ECG surface electrodes. With recent research studies have been extensively monitoring the wellbeing of the fetus using non-invasive FECG (NIFECG). The reliable assessment of changes in fetus condition is of high importance today. Monitoring changes in the fetal heart rate (FHR) can be used as an excellent diagnostic aid to access fetal health status. NIFECG can be extracted by intelligent detection techniques. In this article, three known FECG extraction algorithms namely, Independent Component Analysis (ICA), Adaptive neuro-fuzzy inference system (ANFIS) and hybrid model of ICA-ANFIS have been evaluated. The results are compared with the direct scalp ECG of the Physionet FECG database. ICA technique with 2 iteration stages matched the scalp FECG FHR. However using ANFIS, the FHR matched almost 100% with the invasive fetal ECG. It shows that the hybrid ICA-ANFIS is a promising non-invasive solution for producing FECG, which is close to invasive scalp FECG peaks.