Monitoring uterine activity is a vital diagnostic activity to observe one’s adaptations to the physiological variations during pregnancy. Electromyography has been shown to be a potential non-invasive method to analyze changes in uterine contractions during Term and Preterm deliveries. The fluctuations associated with these contractions are complex in nature and non-linear signal processing methods are to be employed to extract useful diagnostic information. In this work, uterine EMG signals recorded during Term pregnancies are analyzed using adaptive fractal features and machine learning methods. The characteristics of the signal and the significant results would be presented in detail in this seminar.