Related Experiment Video
Updated: Nov 19, 2025
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
Published on: April 21, 2023
Comparison of Different Ultrasound Classification Systems of Thyroid Nodules for Identifying Malignant Potential: A
Hua Chen1,2, Jun Ye2, Jianming Song2
1Department of Medical imaging and nuclear medicine, Medical College of Nanchang University, Nanchang, Nanchang, Jiangxi, China.
Objective:
In our organization, it has been necessary in our organization to calculate the risk categories according to the American Thyroid Association (ATA), the American Association of Clinical Endocrinologists/American College of Endocrinology/Associazione Medici Endocrinologi (AACE/ACE/AME), and the American College of Radiology Thyroid Imaging Reporting and Data System (ACR TIRADS) classification systems for each patient, from the year 2019; these are also required to be registered in the database. This creates a barrier to medical collaboration in everyday radiological practice because using multiple rating systems can be confusing for both readers and patients. For the change in routine practice, this study aimed to compare diagnostic parameters of the ATA, AACE/ACE/AME, and ACR TIRADS classification systems for the detection of suspicious thyroid nodule(s) considering the results of fine-needle aspiration cytopathology as the reference standard.
Methods:
Data on ultrasound characteristics (2,000 nodules) and fine-needle aspiration cytopathology (39 nodules) were included in the analysis. The decision making of fine-needle aspiration biopsies was evaluated from the ultrasound characteristics as per the ATA, AACE/ACE/AME, and ACR TIRADS classification systems.
Results:
The ATA, AACE/ACE/AME, and ACR TIRADS recommended 26, 32, and 37 nodules for fine-needle aspiration biopsies, respectively. Considering the results of fine-needle aspiration cytopathology as the reference standard, the ATA, AACE/ACE/AME, and ACR TIRADS classification systems had 0.993, 0.996, and 0.998 sensitivity, respectively. The accuracies were 0.641, 0.795, and 0.923, respectively.
Conclusion:
The ACR TIRADS classification system is less invasive and can identify suspicious nodules more accurately than that of ATA and AACE/ACE/AME.
