Enhanced sugeno fuzzy inference system with fuzzy AHP and coefficient of variation to diagnose cardiovascular disease during pregnancy

Abstract

Accurate prediction of cardiovascular disease (CVD) risks during pregnancy is vital for preventing and treating life-threatening condition. Existing research has explored fuzzy inference systems and machine-learning techniques for predicting cardiovascular disease risks. However, these models have not explicitly addressed pregnancy-related cardiovascular disease. This study proposes a novel hybrid system that combines the fuzzy analytic hierarchy process (fuzzy-AHP), coefficient of variation (CV) and Sugeno fuzzy inference system with the optimized rule to diagnose cardiovascular disease during pregnancy. The study considers 12 cardiovascular disease risk factors and involves an experienced cardiac clinician and gynaecologist in determining their contributions. Fuzzy-AHP is employed for assigning weights to risk factors and distributing them among sub-risk factors based on their relative contributions. The coefficient of variation is utilized for calculating the weights of the strings, while the nearest neighboring method is employed to cluster the potential strings. The obtained rules are incorporated into the Sugeno fuzzy inference system to compute the output of CVD during pregnancy. The system’s performance is evaluated using an online clinical dataset of 1015 maternal health risks, performance measures, receiving operating system, and statistical analysis. The hybrid system detects cardiovascular disease with 98.71% accuracy, 98.73% sensitivity, and 98.91% precision. This suggest that the proposed technique could be an accurate and valuable tool for predicting cardiovascular disease risks during pregnancy in clinical settings.