FETAL ECG EXTRACTION AND CLASSIFICATION OF THE HEALTH CONDITION BASED ON BOOSTING CLASSIFIER

IMAGE PROCESSING

The non-invasive method of FECG monitoring is a promising one which can be used in all gestation weeks and also during delivery process, but there are some difficulties in this method. Since in non-invasive method the recording electrodes are placed on the abdomen region of the pregnant woman, they record both maternal electrocardiogram (MECG) and the fetal electrocardiogram (FECG), also in this case it may contain a relatively large amount of noise. The main sources of possible noise in the field of extraction of fetal ECG include the maternal electrocardiogram (MECG). In this work the proposed method includes dynamic ECG model to extract FECG from mixed ECG using adaptive filtering techniques. The extracted FECG signal is undergone for the analysis by decomposing with the help of empirical mode decomposition and the spectral features are extracted from these decomposed signals Finally adaptive boosting classifier is responsible for the FECG health analysis. The parameter analysis for different values of noise level, amplitude and heart rate ratios of fetal ECG's shows its effectiveness and is implemented using MATLAB.

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.