Research Progress of Deep Learning in Thyroid Nodule Imaging Examination

Deep learning can extract high-throughput information that is difficult for human eyes to detect in images and can learn autonomously. In recent years, it has been widely used in the field of medical image recognition. Research based on deep learning can help doctors improve their ability to detect small lesions in images, reducing the impact of misdiagnosis or missed diagnoses on patients. Accurate diagnosis of thyroid nodules is a difficult problem in clinical treatment. Although the malignant probability of thyroid nodules is low, early neglect of diagnosis and treatment can lead to the deterioration of thyroid nodules and even thyroid cancer, which affects human life safety. This paper first introduces the application and research progress of deep learning in medical imaging, summarizes the current medical image processing algorithms, and then discusses and summarizes the development process and current limitations of deep learning in different imaging diagnosis of thyroid nodules, so as to provide a reference for researchers to further study this field.