Classification of Multichannel Uterine EMG Signals by Using A Weighted Majority Voting Decision Fusion Rule

Recording the bioelectrical signals by using multiple sensors has been the subject of considerable research effort in the recent years. The multisensor recordings have opened the way to the application of more advanced signal processing techniques and the extraction of new parameters. The focus of this paper is to demonstrate the importance of multisensor recordings for classifying multichannel uterine EMG signals recorded by 16 electrodes. First, we showed that mapping the characteristics of the multichannel uterine EMG signals may allow to set some peculiar properties of these channels. Then, data recorded from each channel were individually classified. Based on the variability between the classification performances of each channel, a weighted majority voting (WMV) decision fusion rule was applied. The classification network yielded better classification accuracy than any individual channel could provide. We conclude that our multichannel-based approach can be very useful to gain insight into the modification of the uterine activity and can improve the classification accuracy of pregnancy and labor contractions.