A Comparative Study On Three Different TI-RADS Classification Standards

Objective:To compare the diagnostic value of thyroid imaging report and data system（Kwak-TIRADS）developed by Korean scholars in 2011,thyroid imaging and data system（ACR-TIRADS）developed by American Institute of Radiology in 2017 and Chinese version of thyroid imaging and data system（C-TIRADS）developed by Ultrasonic Medicine Branch of Chinese Medical Association in 2020.Materials and methods:From January 2018 to June 2019,611 patients（625 nodules）with thyroid gland confirmed by operation and pathology in Anhui cancer hospital were analyzed retrospectively.Three different TI-RADS（Kwak-TIRADS,China-TIRADS and ACR TI-RADS）were used to classify thyroid nodules.The number and percentage of each classification were calculated based on the surgical pathological results as the gold standard.The receiver operating characteristic（ROC）curve was constructed and the area under the ROC curve was obtained The sensitivity,specificity,accuracy,positive predictive value（PPV）and negative predictive value（NPV）of the three diagnostic methods were calculated.Classification of C-TIRADS and ACR-TIRADS is used to evaluate nodules requiring FNAB,and the FNAB rate and unnecessary biopsy rate of malignant nodules are calculated by the following formula:Malignant nodule FNAB rate=number of malignant nodules in FNAB nodules/number of summary nodes;Unnecessary FNAB rate=number of benign nodules in FNAB nodules/number of summary nodes.Results:Among 625 thyroid nodules,the malignant rates of Kwak-TIRADS classified into 2-5 categories（categories 2,3,4a,4b,4c and 5）were 0%,2.9%,16.5%,51.8%,90.6%and 95.8%respectively.The malignant rates of ACR-TIRADS classified as 1-5 were 0%,3.3%,7.3%,45.3%and 91.5%,respectively.The malignant rates of C-TIRADS classified into 2-5 categories（categories 2,3,4a,4b,4c and 5）were 0%,5.1%,34.1%,82.8%,90.3%and 100%respectively.The optimal cutoff value of Kwak-TIRADS classification is 4b,The sensitivity,specificity,positive predictive value（PPV）,negative predictive value（NPV）,accuracy and area under the curve（AUC）of diagnosing benign and malignant thyroid nodules were 96.1%,66.7%,84.5%,90.0%,85.9%,0.897,0.897(95%CI:The best cut-off value of ACR-TIRADS classification is 5,and the sensitivity,specificity,PPV,NPV,accuracy and AUC of diagnosing benign and malignant thyroid nodules are 81.9%,85.6%,91.5%,71.4%,83.2%,0.876（95%CI:0.848-0.901）。The best cut-off value of C-TIRADS classification is 4b,and the sensitivity,specificity,PPV,NPV,accuracy and AUC of diagnosing benign and malignant thyroid nodules are 88.5%,79.6%,89.2%,78.5%,85.4%and 0.881（95%CI:0.853-0.905）,respectively.According to the classification of ACR-TIRADS and C-TIRADS,the biopsy rates of malignant nodules were 36.4%and 39.5%,respectively,with no significant difference（P=0.268）.The unnecessary biopsy rates were 19.2%and 15.2%,respectively,and the difference was statistically significant（P<0.05）.Conclusion:Kwak-TIRADS,C-TIRADS and ACR-TIRADS have good diagnostic value for benign and malignant thyroid nodules.The area under curve（AUC）of Kwak-TIRADS classification is the largest and has the highest sensitivity,while ACR-TIRADS classification has the highest specificity,and the unnecessary biopsy rate of C-TIRADS classification is lower than that of ACR-TIRADS classification.