Artificial Intelligence in Adrenal Imaging

Adrenal lesions are common, occurring in approximately 5% of the population. Although the vast majority are benign, it can be challenging to differentiate clinically significant from clinically insignificant adrenal lesions given overlap in imaging features. Artificial intelligence (AI) may be able to aid radiologists in identifying adrenal masses and diagnosing their etiology. This review defines commonly used terminology in AI and summarizes AI based techniques for adrenal gland segmentation, adrenal lesion detection, and lesion characterization.