Simulation, feature extraction and disorder detection (using fuzzy logic) of uterine contractions

With day to day changes in lifestyle, the probability of miscarriages, preterm delivery and various other structural and functional abnormalities in fetus have been observed to increase exponentially. With growing number of pregnancies, continual monitoring of the health status of the mother and the fetus has therefore become increasingly important. Quantitative assessment of the uterine activity by medical experts can be fundamental during both pregnancy and delivery. The uterine contractions can guide the physician to choose a particular therapy so as to relieve the mother. For the diagnosis of the uterine activity, a large number of samples were required. However, unavailability of the signals in the processable form led to an urgent need of simulating the uterine contraction like signals in the virtual environment. The signals were first simulated and values of the important diagnostic parameters of uterine activities were then extracted. A fuzzy expert model to infer the health condition of the patients was then developed and implemented. The developed system also reported the disorder present (if any). Our system enabled the patient to not to rush to her doctor everytime she underwent a contraction. The fuzzy system categorized the contractions as normal, suspicious or abnormal. Thus making it comfortable for the mother to contact her doctor only if urgently needed. Simulation of the signals was carried out successfully under the supervision of a medical expert. The developed fuzzy inference system, for detecting the normality/abnormality of the signals gave results with an efficiency of 0.98 and related the status of the contractions with the key features of the uterine contractions.