Machine Learning Based Weight Prediction System for Bariatric Patients

Obesity has become a global issue, both in Sri Lanka and in elsewhere. It has been shown beyond any doubt that bariatric surgery (BS) is the finest effective treatment for morbid obesity patients with a Body Mass Index (BMI) greater than 40.0. After undergoing the surgery, it is required to monitor patients’ weight for eighteen months until they reach a healthy weight that falls within the normal BMI range (18.5— 24.9). In this study, we have analyzed records of bariatric patients registered at Colombo South Teaching Hospital, Kalubowila under four surgery types. Most of patients have lost track of their weight loss before reaching the eighteenth month, according to records, due to their incapacity to track their weight loss during the post-surgery period and the lack of continuing examination after the Bariatric Surgery. Therefore, some of the patients have to go through the same operation more than once, which will create a threat to their lives. The main aim of this study is to remotely track pre and post-surgery bariatric patients and allow them to keep track of their weight loss until they achieve their expected weight using a web based weight prediction system using Machine Learning. To predict the final weight, they might get after the surgery three machine Learning Algorithms are examined to get the highest accuracy. Feed Forward, Liner Regression and Random Forest Algorithms are taken into consideration to give the two output predictions. Pre-surgery and post-surgery data are taken as inputs. The final month’s weight is presented as the output in this system prior to the surgery and post to the surgery. In comparison with the algorithms, the random forest algorithm gives 97 and 95 percent accuracy after examining the data inputs under three algorithms in each prediction.