Development of a fuzzy-based predictive model for the risk of postpartum hypertension in nursing mothers

Abstract

Fuzzy Logic has grown in prominence in recent years as its capacity to address a variety of problems in the field of medicine has been demonstrated to be exceptional. This study developed a model to predict risk of postpartum hypertension among nursing mothers. In this study, the variables associated with the risk of postpartum hypertension were identified, formulated a fuzzy-logic model and simulated the predictive model. The variables that were common symptoms associated with the risk of postpartum hypertension were elicited by review of related works on the body of knowledge of postpartum hypertension following which the variables were validated by mental health experts at a Nigerian hospital using the interview method.  The variables identified were classified into their respective linguistic labels based on the crisp values assigned to them within the interval of acceptable values while the risk of postpartum hypertension was classified as low, moderate and high risk.  Fuzzy-based related functions were used to formulate the fuzzy model for the input variables and the risk of postpartum hypertension.  It was also observed that the number of triangular membership function formulated for each variable is a function of the number of classified labels for each variable.  The results also showed that a total number of 288 rules were defined by the IF-THEN statements created from the variables identified by the experts. The predictive model developed and simulated can be utilized to aid expert obstetricians in the early detection and treatment of postpartum hypertension in the future.