Thyroid Imaging Reporting And Data System In The Standardized Application Of Diagnosis And Treatment In Thyroid Nodular Disease

Objective: By retrospective analysis of the ultrasound image data ofpatients diagnosed with thyroid nodules through surgical operations andpathological results, in accordance with thyroid imaging reporting and datasystem (TI-RADS) classification standard, the study compared thyroid noduleclassification features of different pathological types to explore the value ofqualitative diagnosis and clinical widespread significance of TI-RADSclassification in thyroid nodules.Methods: The study involved in236patients with complete conventionalultrasound data and confirmed pathological types from October2011toOctober2012in the Second Hospital of Hebei Medical University. There were299patients with benign nodules and133with malignant nodules. Theultrasound data were analyzed retrospectively. Different pathological types ofthyroid nodules in ultrasound characteristics were compared, divided intodifferent grades depended on the TI-RADS classification standard:(1) grade1, benign lesions (mainly cystic with no malignant ultrasonic characteristics);(2)grade2, highly suggest of benign lesions (mainly solid, with/without afeature that suggested being malignant despite of microcalcification, abnormallymph nodes or anteroposterior to transverse diameter ratio(A/T)≥1);(3)grade3, indeterminate lesions:3A level, benign lesions tendency (with1-2ultrasonic characteristics that suggested being malignant but none ofmicrocalcifications, abnormal lymph nodes or anteroposterior to transversediameter ratio(A/T)≥1,or only one of microcalcification,abnormal lymphnodes and anteroposterior to transverse diameter ratio(A/T)≥1);3B level,malignant lesions tendency (with3-4ultrasonic characteristics that suggestedbeing malignant but none of microcalcifications, abnormal lymph nodes or anteroposterior to transverse diameter ratio(A/T)≥1, or one malignantfeatures with any of microcalcification, abnormal lymph nodes andanteroposterior to transverse diameter ratio(A/T)≥1);(4) grade4, malignantlesions at high risk (with5ultrasonic characteristics that suggested beingmalignant but none of microcalcifications, abnormal lymph nodes oranteroposterior to transverse diameter ratio(A/T)≥1,two malignant featureswith any of microcalcification, abnormal lymph nodes and anteroposterior totransverse diameter ratio(A/T)≥1)；(5) grade5, malignant lesions (withmore than6ultrasonic characteristics that suggested being malignant or4ultrasonic features including more than one of microcalcification,abnormallymph nodes and anteroposterior to transverse diameter ratio(A/T)≥1).Grade1-3A were dignosised as benign lesions, the3B-5grade were judged to bemalignant. All data were analyzed by SPSS19.0statistical software, countedby using the chi-square test. Statistical significance was set at P <0.05.Results: This study involved in236patients with a total of432nodules.There were299benign nodules (69.2%) which included241nodular goiter,50thyroid adenoma,8thyroid granulomatous inflammation;133malignantnodules (30.8%) including:124papillary thyroid carcinoma (including34thyroid microcarcinoma),9other types of thyroid cancer (4undifferentiatedcarcinoma,1poorly differentiated carcinoma,3medullary carcinoma,1follicular carcinoma). According to the statistical results of this study, weconcluded that obscure boundary（sensitivity67.7%，specificity87.6%,positive predictive value70.9%， negative predictive value85.9%,accuracy81.5%）; irregular shape（sensitivity75.9%，specificity89.0%,positive predictive value75.4%，negative predictive value89.3%，accuracy85.0%）; Solid（sensitivity91.7%，specificity20.4%，positive predictive value33.9%，negative predictive value84.7%，accuracy42.4%）; Uneven edge（sensitivity78.2%，specificity84.9%，positive predictive value69.8%,negative predictive value89.8%，accuracy82.9%）; low level echo（sensitivity78.2%，specificity61.5%，positive predictive value47.5%，negative predictivevalue86.4%， accuracy66.7%）; microcalcification （sensitivity53.4%, specificity96.7%，positive predictive value87.7%，negative predictive value82.3%， accuracy83.3%）; abnormal lymph nodes （sensitivity12.0%,specificity99.7%，positive predictive value94.1%，negative predictive value71.8%，accuracy72.7%）; irregular Vascular shape（sensitivity97.6%,specificity76.2%，positive predictive value66.3%，negative predictive value98.5%，accuracy83.1%）; anteroposterior to transverse diameter ratio(A/T)≥1（sensitivity17.3%，specificity99.7%，positive predictive value95.8%,negative predictive value73.0%， accuracy74.3%） lead to malignantindications. Halo, uniformity, blood flow ratio, blood flow locally abundantshowed no statistical differences between benign and malignant lesions.TI-RADS classification based on the ultrasound characteristics showedsensitivity83.5%，specificity91.0%，positive predictive value80.4%，negativepredictive value92.5%，accuracy88.7%. Despite of the uncertainty grade3,when only taking count on grade1,2and grade4,5the sensitivity was90.1%，specificity98.9%，positive predictive value97.3%，negative predictivevalue95.7%，accuracy96.2%.Conclusion:1. Thyroid Imaging Reporting and Data System (TI-RADS)showed impressively high sensitivity, specificity, and accuracy.It couldevaluate ultrasonic characteristics comprehensively, differentiate benign andmalignant nodules effectively. Thus, TI-RADS could provide objective,accessible standard to clinical examination of thyroid nodular disease, enhancestandardized ultrasound report to reduce the subjective error, to facilitatecommunication between physicians and guide on the clinical treatmentprotocols.2. Thyroid Imaging Reporting and Data System (TI-RADS) had highsensitivity, specificity and accuracy as a non-invasive examination afterremoving uncertainty lesions. Therefore, it could reduce some invasiveprocedures to save medical resources and identify lesions with other measureswhen there were uncertainty lesions.3. There were significant differences in obscure boundary,irregularshape, solid, uneven edge, low level echo,microcalcifications, abnormal lymph nodes, irregular Vascular shape, anteroposterior to transverse diameterratio(A/T)≥1between benign and malignant, Suggest of malignant lesions.However, halo, uniformity, blood flow ratio, blood flow locally abundantshowed no statistical differences between benign and malignant lesions.