Enhancing Impaired Waist-to-Height Ratio Classification Using Neural Networks

Obesity is a condition characterized by the excessive accumulation of adipose tissue. However, directly measuring adiposity can be challenging, especially in epidemiological and clinical settings. Therefore, simple anthropometric measurements are commonly used to assess fat quantity and distribution. The Body Mass Index (BMI) is a widely used measure for estimating total fat quantity. Additionally, indicators such as waist circumference and waist-to-height ratio (WHtR) provide valuable insights into the distribution of visceral, central, or abdominal fat. These measurements play a crucial role in understanding and evaluating health risks associated with obesity. This study utilized a dataset consisting of 1978 participants, anthropometric measurements, including height, weight, body circumferences, and body folds, were collected from each participant. The research aimed to classify individuals with impaired WHtR using artificial neural networks (ANNs) based on anthropometric parameters. Multiple tests were conducted using Monte Carlo cross-validation with different training and testing ratios. The architecture of the ANN was modified by varying the number of hidden layers. The results showed an accuracy exceeding 82.4%. The sensitivity values consistently surpassed 79.9%, indicating the model’s effective detection of positive cases. The model also demonstrated excellent specificity, with a score exceeding 85%. Positive and negative predictive values showed slight improvements as the training data expanded. The F1 score, which considers both precision and sensitivity, was above 0.794, indicating a favorable balance in classifying individuals with impaired WHtR. The model’s performance remained consistent across different training-test splits, suggesting stability and reliability in its predictions.