Abstract: Abstract: Objective To explore the diagnostic value of artificial intelligence computer aided system (AI-CAD) in benign and malignant thyroid nodules. Methods To analyzed 473 patients with 517 thyroid nodule in Xi 'an Jiaotong University the second affiliated hospital from June 2019 to December 2019， all nodules were confirmed by pathology, based on AI-CAD, the judgment standard is divided into three groups, the AI score of benign group was 0~0.40, the AI score of suspicious group score was
0.41~0.60 and the AI score of malignant group was 0.61~1.00. The distribution of scores in each group was observed by boxplot, and the diagnostic accuracy was compared among the three groups with pathological results as the gold standard. Results The diagnostic accuracy rates of benign and malignant thyroid nodules were respectively 88.94%, 69.75% and 84.53%. There was significant difference between groups (P<0.001). Conclusion Artificial intelligence assisted system has high value for the differential diagnosis of benign and malignant thyroid nodules, especially for the determination of benign nodules, which is beneficial to help beginners to improve their confidence in diagnosis.

Key words: Key words: Artificial intelligence computer aided system, Thyroid nodule, Ultrasound