TI-RADS may help guide biopsy decisions for thyroid nodules in children



The retrospective analysis included images of thyroid nodules of 139 patients with an average age of 17.5 years, each of whom had definitive pathological results available. Three radiologists characterized the nodules as malignant or benign based on their impressions, then again using TI-RADS. After a deep learning algorithm was also applied to the imaging, the researchers compared the sensitivity and specificity of all interpretations.

The radiologists had an average sensitivity of 58.3% and a specificity of 79.9%. For ACR TI-RADS, these figures were 85.1% and 50.6%, and for the algorithm, the researchers recorded an average sensitivity of 87.5% and a specificity of 36.1%.

Although the ACR TI-RADS and deep learning algorithm had lower specificity than the global impressions, they recorded higher sensitivity. Additionally, compared to rad impressions, ACR TI-RADS had better inter-observer agreement, leading the authors to suggest that use of the system might be applicable in a pediatric setting.

“Given the increased priority given to sensitivity when evaluating thyroid nodules in children compared to adults, the results support the continued exploration in children of the ACR TI-RADS and the algorithm of ‘deep learning’,” the authors concluded, adding that the algorithm would require more. training and validation using pediatric data before clinical implementation.

To learn more, Click here.

Source link


Comments are closed.