In recent years

the incidence of thyroid cancer has leaped to the top of endocrine malignancies. Thyroid fine-needle aspiration biopsy (FNAB) has become the only gold standard for preoperative diagnosis

but it is an invasive test and requires high medical conditions

so it is still not widely available nationwide. Ultrasound is the first choice for thyroid examination. The limitations of two-dimensional ultrasound examinations

which are subjective and more dependent on physician experience

mean that inexperienced physicians are more likely to misdiagnose cancer and increase the probability of FNAB

resulting in a waste of medical resources. Artificial intelligence based on deep learning of big data can give fast and objective diagnosis. Elastography can objectively reflect the softness of thyroid nodules

and contrast-enhanced ultrasound can provide changes in the subtle blood supply to the nodules

both of which can be used as a complementary tool to regular ultrasound. Artificial intelligence based on big data deep learning can capture more subtle picture information

and the resulting trained deep learning model can quickly analyze the captured information and give an objective diagnosis. The high incidence of thyroid nodules has caused a wave of research on the application of new ultrasound techniques in thyroid nodules. This paper summarized the above new ultrasound technologies and introduced latest research progress.