Machine Learning-based Obesity Classification using 3D Body Scanner

Abstract Knowing one's obesity group is very important for healthcare. Obesity can cause various diseases; however, BMI, which is the currently popular standard for judging obesity, does not accurately classify the obese group. This is because BMI just reflects height and weight, ignoring the characteristics of body type. Therefore, we present the idea that reflecting the three-dimensional (3D) measurements of the human body can better classify the obese group than BMI. To prove this, we recruited a total of 160 subjects and collected 3D body scans, Dual-energy X-ray absorptiometry (DXA), and Bioelectrical Impedance Analysis (BIA) data pairwise. Through this, 3D body scan data could be expanded clinically. We proposed a machine learning-based obesity classification framework using 3D body scan data, validated it through Accuracy, Recall, Precision, and F1 score, and compared it with BMI and BIA. BMI had the following values: Accuracy: 0.529, Recall: 0.472, Precision: 0.458, and F1 score: 0.462 while BIA had the following values: Accuracy: 0.752, Recall: 0.742, Precision: 0.751, and F1 score: 0.739. Our proposed model had the following values: Accuracy: 0.800, Recall: 0.767, Precision: 0.842, and F1 score: 0.792 and our accuracy was 80%, which is higher than either BMI at 52.9% or BIA at 75.2%. Our model can be used for obesity management through 3D body scans.