Evaluating waist-to-hip ratio in youth using frequency-modulated continuous wave radar and machine learning.

Waist-to-hip ratio (WHR) is an essential predictor of cardiometabolic diseases, but traditional tape-based WHR measurements in children and adolescents can cause discomfort due to direct contact and are prone to measurer variation. This study aimed to develop a non-invasive, precise, and convenient alternative for WHR measurement and central obesity assessment using frequency modulated continuous wave (FMCW) radar, and to evaluate its accuracy by comparing it with traditional measurement methods. We included 100 participants aged 7-18 and radar data were analyzed using point cloud generation processed through convolutional neural networks for estimating WHR. The radar-based WHR measurements were compared to conventional clinician measurements. Participants were classified into low (WHR < 0.86), moderate (≥ 0.86, < 0.91) and high WHR (≥ 0.91) groups, and the classifications were compared. Strong agreement was observed between the two methods, with an intraclass correlation coefficient of 0.83 (p = 0.023995). The radar system achieved 82% accuracy in classifying participants into the correct abdominal obesity risk groups. Our findings demonstrate that FMCW radar can be a reliable tool for routine monitoring of central obesity. This technology addresses concerns about privacy and discomfort, making it suitable for widespread application in both clinical and non-clinical settings.