Diagnostic accuracy of the ultrasonographic features for subcentimeter thyroid nodules suggested by the revised American Thyroid Association guidelines
Related Experiment Video
Updated: May 11, 2026
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
Published on: April 21, 2023
Hyun Gi Kim1, Hee-Jung Moon, Jin Young Kwak
1Department of Radiology and Research Institute of Radiological Science, Severance Hospital, Yonsei University , College of Medicine, Seoul, Korea.
Background:
Recommendations for subcentimeter thyroid nodules that need fine-needle aspiration biopsy are renewed in the revised American Thyroid Association (ATA) guidelines published in 2009. We applied these recommendations to analyze the diagnostic performance of the ATA guidelines and compared it to that of other modified guidelines.
Methods:
We evaluated 1054 nodules with sizes of 6-10 mm in 991 patients. A total of 713 nodules were included in the study population by excluding nodules with insufficient results for deciding whether they had a benign or malignant cytology. Frequencies of ultrasonographic features in benign and malignant nodules were compared, and odds ratios of suspicious ultrasonographic features were obtained with univariate and multivariate analysis. Seven modified guidelines were made based on the revised ATA guidelines and from multivariate analysis results. Diagnostic performances of the guidelines were compared by sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and the area under the receiver operating characteristic curve (Az) value.
Results:
In addition to hypoechogenicity, infiltrative margin, microcalcification, and taller-than-wide shape that were suggested by the ATA guidelines, solid composition and macrocalcification were significantly associated with malignancy on multivariate analysis (p=0.001 and 0.003, respectively). Increased vascularity, however, was not significantly associated with malignant nodules (odds ratio 0.729, p=0.212). Among the eight guidelines, the ATA guidelines showed the lowest diagnostic performance (Az=0.616). Excluding increased vascularity and including solid composition with or without macrocalcification to the suspicious ultrasonographic features of the ATA guidelines improved sensitivity (96.6% vs. 97.0%), specificity (26.6% vs. 42.9%), PPV (48.3% vs. 54.7%), and NPV (91.7% vs. 95.2%), thereby resulting in the highest Az value (Az=0.699, p<0.001).
Conclusions:
This study suggests that excluding increased vascularity and adding solid composition to the suspicious ultrasonographic features of the ATA guidelines would significantly improve the diagnostic performance in subcentimeter nodules for the identification of malignant lesions.
