Machine Learning Predictive Model for Thyroid Disease Detection

Abstract

The thyroid, a butterfly-shaped gland located in the neck, plays a crucial role in regulating metabolism, energy, and hormonal balance, meeting the peripheral tissues’ needs. Thyroid diseases pose a significant global health problem, affecting millions with varying degrees of severity. Conditions such as hyperthyroidism and hypothyroidism can manifest symptoms due to fluctuations in thyroid hormone levels, impacting individuals’ well-being. Additionally, thyroid disorders may involve the enlargement of the gland, known as goiter, or the formation of thyroid nodules, which can have functional or neoplastic implications. This article explores the development and implementation of a predictive model using advanced Machine Learning techniques to forecast thyroid diseases. By analyzing clinical and relevant biomedical data, we employ sophisticated Machine Learning algorithms to identify hidden patterns and correlations within the data. The goal is to enhance precision and anticipation in diagnosing thyroid diseases, offering healthcare professionals a powerful tool for improved patient care, and optimized clinical outcomes. The results highlight that the Random Forest model achieved remarkable performance with 100% on the accuracy, precision, recall F1 Score and AUC metrics. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.