Abstract: Thyroid tumor is a common disease, of which the vast majorities are benign. However, there appears to be a recent increase in the occurrence of thyroid cancer. With the development of diagnostic techniques, acquisition of high-quality imaging and pathological data is no longer difficult. The difficulty of improving diagnosis now lies mainly in efficient and accurate interpretation of results. Given the increasing number of patients, treatment decision-making and follow-up management can be challenging. Machine learning has become a promising solution to the challenges in the field of medicine due to its high efficiency and accuracy. At present, its role in the diagnosis and treatment of thyroid tumors has been confirmed by several studies. A comprehensive understanding of machine learning and its role in the diagnosis and treatment of thyroid tumors can provide new ideas for improving diagnostic accuracy, guiding treatment decisions and improving patient management.

Key words: Thyroid tumor, Machine learning, Artificial neural network, Deep learning