Description of Invention
Well-differentiated thyroid cancer (WDTC) has a generally favorable prognosis, yet recurrence remains a clinical challenge. Traditional risk stratification methods may lead to overtreatment or missed recurrence risks. Machine learning (ML) offers a promising approach to improve predictive accuracy. This study explores support vector machines (SVM), random forest (RF), k-nearest neighbors (KNN), and stacked models to predict recurrence using clinicopathologic features. This study aims to integrate ML into thyroid cancer recurrence prediction, improving clinical decision-making by reducing unnecessary interventions and enabling more precise patient management. Findings could support the development of AI-driven decision-support tools in oncology.