A personalized approach for predicting the effect of aerobic exercise on blood pressure using a Fuzzy Inference System

Regular aerobic exercise is a recommended treatment for elevated blood pressure (BP). However, making permanent lifestyle changes is not easy. Having personally relevant information about the treatment, about its effects and importance, is a precondition for motivation. Thus, the first step towards a successful lifestyle change is appropriate education. This paper describes a Sugeno-type Fuzzy Inference System (FIS) that predicts the effect of regular aerobic exercise on blood pressure based on the exercise dose variables, exercise frequency and intensity, as well as demographics (age, gender, ethnicity), and the baseline BP of a person. Since BP response to exercise varies largely between individuals, the system takes an initial step towards personalized prediction. Hence, the system can be used to educate a person about the benefits of exercise on BP in a personally relevant way, providing more accurate information than traditional education materials. Furthermore, preliminary validation results of the performance of the FIS are promising. The predictions comply with the findings of medical research for populations, though the individual-level validation remains still to be done.