Correlation of ACR TI-RADS ultrasound classification with cytopathology: a retrospective evaluation of diagnostic accuracy

Abstract

Aims: To evaluate the diagnostic performance of the American College of Radiology Thyroid Imaging Reporting and Data System (ACR TI-RADS) in predicting malignancy in thyroid nodules and to assess the correlation between TI-RADS categories, cytopathological outcomes, and guideline-recommended biopsy thresholds. Methods: This retrospective single-center study included 567 thyroid nodules evaluated with ultrasound and Doppler imaging between January 2023 and November 2025. TI-RADS scores were retrospectively assigned according to ACR criteria based on sonographic features. Cytopathology served as the reference standard, with nodules classified as benign (Bethesda II) or malignant/suspicious for malignancy (Bethesda V–VI). Diagnostic performance metrics of TI-RADS ?4 were calculated, including sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy, and area under the receiver operating characteristic curve (AUC). Multivariate logistic regression was performed to identify independent predictors of malignancy. Results: Of the 567 nodules, 33 (5.8%) were malignant. Malignancy rates increased across TI-RADS categories, reaching 31.6% in TR5 nodules. Using TI-RADS ?4 as the threshold, sensitivity was 78.8%, specificity 61.0%, PPV 11.1%, and NPV 97.9%, with an overall accuracy of 62.1% and an AUC of 0.75 (p<0.001). Multivariate analysis demonstrated that both TI-RADS score (OR=4.09 per category increase, p<0.001) and nodule size (OR=1.04 per mm, p=0.013) were independent predictors of malignancy. Nodules meeting guideline-based biopsy size criteria showed significantly higher malignancy odds. Conclusion: ACR TI-RADS provides reliable malignancy risk stratification with a high NPV in routine clinical practice. The system effectively reduces unnecessary biopsies while maintaining diagnostic safety, supporting its continued use as a standardized ultrasound-based risk stratification tool for thyroid nodules.