| Objective The purpose of this study was to differentiate benign and malignant thyroid nodules by using ACR Thyroid Imaging,Reporting and Data system(TI-RADS)and real-time Shear Wave Ultrasound Elastography(SWE),analyze the influence of factors such as nodule size on the diagnosis results,and the causes of false positive and false negative,to find out the role of SWE in assisting ACR TI-RADS,and to improve the clinical value of SWE in the diagnosis of thyroid nodules.Methods(1)From October 2016 to June 2017,412 patients were treated with conventional ultrasound and SWE,and 455 nodules were included in this study.Routine ultrasound and SWE ultrasound examination were carried out with SuperSonic Imaging Aixplorer ultrasound diagnostic apparatus.The TI-RADS score and the classification were evaluated and recorded.And we measured and recorded the Young's modulus of SWE-whole-mean,SWE-whole-min,SWE-2mm-mean,SWE-2mm-min,SWE-max,SWE-whole-ratio and SWE-max-ratio in each nodule.The ROC curve was plotted with the pathological results as the gold standard,and the best Cut-off value of each diagnostic index for diagnosis of malignancy was determined.(2)455 cases of nodules was divided into two groups according to whether the maximum diameter was greater than 1cm,and evaluate the diagnostic efficacy of SWE and TI-RADS respectively.(3)Of the 7 diagnostic indexes,the largest quantitative diagnostic index and semi quantitative diagnostic index of AUC were used to group the benign and malignant thyroid nodules with the cut-off value of ROC curve respectively.Every two groups were compared and the reasons for false positive and false negative were further analyzed.(4)The diagnostic value of SWE in TI-RADS 4 and TI-RADS 5 nodules was further evaluated according to the diagnostic threshold and influencing factors obtained in the previous study.The specific application of SWE in the diagnosis of thyroid nodules with ACR TI-RADS was discussed.Results(1)The total of 455 thyroid nodules included 298 malignant nodules and,157 benign nodules,the area under curve of TI-RADS classification and TI-RADS specific score in the diagnosis of thyroid malignant nodules were 0.767 and 0.822 respectively.According to the ROC curve,chose TI-RADS 5 as the diagnostic criteria of malignant nodules,sensitivity and specificity were 82.6%and 65.6%respectively;chose the TI-RADS score 9 points as the diagnostic criteria of malignant nodules,sensitivity and specificity were 71.5%and 77.7%respectively.The area under the ROC curve of SWE-whole-mean,SWE-whole-min,SWE-2mm-mean area,SWE-2mm-min,SWE-max,SWE-whole-ratio,and SWE-max-ratio were 0.771,0.667,0.772,0.762,0.764,0.779 and 0.768 respectively.And the Cut-off value of each diagnostic index for diagnosis of malignancy were 24.5kPa,16.5kPa,25.5kPa,19.5kPa,37.5kPa,1.25 and 1.95 respectively.Using the best cut-off value as the diagnostic criterion,the Sen and Spe of SWE-2mm-mean were 78.9%and 65.0%respectively,and the Sen and Spe of SWE-whole-ratio were 79.2%and 65.6%respectively.The AUC of the diagnosis of thyroid nodules combined with two methods of SWE and TI-RADS was 0.864,which was the largest.Except the AUC diagnosed by specific score of TI-RADS,there was significant difference between this AUC and the AUC of other diagnostic indexes(P<0.05).(2)As for the nodules with maximum diameter less than or equal to 1cm,the area under the ROC curve of TI-RADS grade and specific scores were 0.621 and 0.701;and as for the thyroid nodules with maximum diameter larger than 1cm,the area under the ROC curve of TI-RADS grade and specific scores were 0.884 and 0.921.For SWE measurements,the nodules of less than 1cm were not significantly changed in the diagnostic boundary value or the area under the ROC curve,and the diagnostic efficiency of the nodules with maximum diameter larger than 1cm was significantly increased.Except SWE-whole-min,the AUC ranged from 0.828 to 0.881,and the sensitivity and specificity of diagnosis were also increased,ranged from 78.5%to 94.9%and 72.1%to 83.7%respectively.(3)The quantitative evaluation index and semi quantitative evaluation index with the maximum AUC were SWE-2mm-mean and SWE-whole-ratio,and the diagnostic efficiency of these two indexes had no statistical difference(P=1.000),and the concordance of dignosis between these two indexes was general level(the Kappa value of diagnosis in malignant and benign nodules were 0.544 and 0.677 respectively,P<0.001).In the evaluation of malignant thyroid nodules,the nodules with the maximum diameter less than 1cm or extremely low echo were prone to false negative.When evaluated with SWE-2mm-mean,the malignant thyroid nodules without calcification or with the distance from the thyroid membrane larger than 2mm were prone to false negative(P<0.05).And when evaluated with SWE-whole-ratio,the malignant thyroid nodules on the middle or back of the gland,with the "taller-than-wide sign",or the deepest depth from the probe position>2 cm were more susceptible to false negatives.In the evaluation of benign thyroid nodules,hypoechoic or very hypoechoic,solid,aspect ratio>1,or CDFI showed that the nodules with no blood flow signal were prone to false positive.Benign thyroid nodules with microcalcification or coarse calcification were more likely to be false-positive in SWE-2mm-mean evaluation,and those with maximum diameter? 1cm were more likely to false-positive in SWE-whole-ratio evaluation.Multivariate analysis showed that the maximum diameter ? 1 cm and very low echo were the influencing factors of false-negative of SWE,and the distance from the capsule>2mm may be the factor that caused the false negative value of SWE-2mm-mean.The benign nodules with aspect ratio>1 may be false positive by SWE-2mm-mean.There was no independent factor that determined false positive of SWE-whole-ratio.(4)In the diagnosis of TI-RADS 4 type nodules,the Sen,Spe,accuracy,PPV,NPV of diagnosis with SWE-2mm-mean and SWE-whole-ratio were 76.9%,60.7%,68.5%,64.5%,73.9%,and 86.5%,67.9%,76.9%,71.4%,84.4%respectively.SWE-whole-ratio was used to evaluate the thyroid nodules of TI-RADS score 4,5 points,and the NPV was 93.3%.In the diagnosis of TI-RADS 5 type nodules,the Sen,Spe,accuracy,PPV,NPV of diagnosis with SWE-2mm-mean and SWE-whole-ratio were 79.3%,50.0%,74.0%,87.8%,34.6%and 77.6%,50.0%,72.7%,87.6%,32.9%respectively.The nodules of TI-RADS 5 with maximum diameter>1cm were divided into two groups:<9 and ? 9,for the nodules with<9 scores,the Sen,Spe,accuracy,PPV,NPV of diagnosis with SWE-2mm-mean and SWE-whole-ratio were 79.3%,50.0%,74.0%,87.8%,34.6%respectively,and for the nodules with ? 9 scores,the Spe and NPV of the two S WE indexes decreased significantly.Conclusions(1)ACR TI-RADS and SWE have high practical value in the differential diagnosis of thylroid nodules.The combination of ACR and SWE can improve the diagnostic efficacy of thyroid nodules.(2)Compared with the diagnosis in nodules with maximun diameter equal or less than 1cm,the diagnostic efficiency of ACR TI-RADS and SWE increased for nodules with maximum diameter larger than 1cm.(3)The diadynamic criteria SWE-2mm-mean(cut off value of 25.5kPa)and SWE-whole-ratio(cut off value of 1.25)can be used to diagnose thyroid nodules.The SWE diagnosis of thyroid malignant nodules with maximum diameter less than 1cm and extremely low echoes was more likely to be false-negative,(4)The negative assessment of TR4 nodules with ACR TI-RADS score of 4,5 by SWE could be considered as benign tendency,and may reduce the unnecessary puncture biopsy.(5)For the negative evaluation of TI-RADS 5 type nodules by SWE,we can consider the benign tendency and reduce the unnecessary puncture biopsy.But for the nodules ? 9 points,even if the SWE is negative,the biopsy should be recommended. |
PDF Full Text Request