The application of artificial intelligence to thyroid nodule assessment

Abstract

Artificial intelligence (AI) is of considerable interest in the healthcare community including its diagnostic applications for thyroid nodules in assisting both radiology and FNA assessment. Fine-needle aspiration (FNA) helps distinguishing benign from malignant thyroid nodules and is a crucial step in the initial diagnosis of cancer. The classification of some lesions can be challenging, and the use of AI in some cases may become essential in order not to give an indeterminate result to the lesion. In this review, we summarize the available evidence regarding the application of AI in thyroid imaging and cytopathology. There are now considerable applications in digital waiting to be approved that will save time and cut costs. The published literature to date has shown promising results. However, future work is required to better define how this technology can be exploited in routine cytopathology practice.