The goal of this intervention study is to learn about the diagnostic performance of artificial intelligence to assist the contrast-enhanced ultrasound diagnosis of thyroid nodules. The main questions it aims to answer are:

1. Can we develop an artificial intelligence model to assist the contrast-enhanced ultrasound diagnosis of thyroid nodules?
2. Can artificial intelligent reduce the unnecessary biopsy of thyroid nodules?

Participants will be asked to undergo contrast-enhanced ultrasound examination and ultrasound-guided fine-needle aspiration of thyroid nodules. Researchers will calculate the diagnostic performance of the contrast-enhanced ultrasound diagnosis of thyroid nodules with and without artificial intelligent assistance.

Inclusion Criteria:

• Patients with thyroid nodules with a solid component ≥5 mm confirmed by conventional ultrasound;
• Patients who underwent conventional ultrasound, contrast-enhanced ultrasound, and fine-needle aspiration biopsy;
• Patients with a final benign or malignant pathological results.

Exclusion Criteria:

• Patients with cytopathology of Bethesda I, III, or IV and without final benign or malignant pathology；
• Patients with a history of thyroid ablation or surgery;
• Patients with low-quality ultrasound images.

Data generated or analyzed during the study are available from the principal investigator by request.

The first week after the publication of the research papers.

Data generated or analyzed during the study are available from the principal investigator by request.

Cytopathologists, histopathologists and statistical analysts were blinded to the intervention.