A Predictive Model for TSH Levels Based on Thyroid Disorders Symptoms

Machine learning models plays a major role in enhancing the diagnosis in thyroid disorders. However, the available models have several limitations including relying on laboratory measures that are expensive and inconvenient. Consequently, the aim of this study is to investigate the symptoms’ role in predicting TSH levels. Four machine learning models, including Multiple Linear Regression, Support Vector Machine, Decision Tree, and Artificial Neural Network, were employed to predict the TSH level. Data from the Jordanian population was utilized, and 34 input variables, including 32 symptoms and 2 extracted symptom scoring variables (calculated using statistical methods) were utilized. Results suggested that the symptoms factor is an essential input variable that should be incorporated in TSH prediction models. The best-performing model was Artificial Neural Network with R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> value of 0.86 and RMSE value 0.02.