Study Of SWE Combined With S-Detect In The Diagnosis Of Benign And Malignant Thyroid Nodules

Objective To investigate the diagnostic efficacy of S-Detect for thyroid nodules and the factors influencing it,as well as the diagnostic efficacy of the triple application of S-Detect software,radiologists applying ACR TI-RADS,and SWE.Methods（1）A total of 162 nodules that underwent routine ultrasound examination of the thyroid and S-Detect from November 2020 to December 2021 in our department with pathological results were selected,and the diagnostic efficacy of S-Detect analysis of nodules in maximum cross-sectional and longitudinal views was compared using pathological results as the gold standard.The results of the S-Detect analysis of thyroid nodules were also reassigned,and the diagnostic efficacy of the assigned values was analyzed.（2）The thyroid nodules were divided into ≤5 mm,5-10 mm,and >10 mm groups according to the maximum diameter of the nodules;diffuse and non-diffuse groups according to different diffuse lesion backgrounds;with and without calcifications groups according to calcifications in the nodules.The diagnostic efficacy of S-Detect for thyroid nodules in different groups was analyzed separately,using pathological findings as the gold standard.（3）The diagnostic efficacy of S-Detect software for thyroid nodules and radiologists applying ACR TI-RADS relative to each other,comparing the efficacy of combined diagnosis and each diagnosis alone.（4）Within the 162 nodules that underwent conventional ultrasound examination of the thyroid and S-Detect with pathological results obtained in our department from November 2020 to December 2021,a total of 98 nodules were performed SWE examination at the same time,and they were selected to compare the diagnostic efficacy of S-Detect software,radiologists applying ACR TI-RADS,and SWE on thyroid nodules and to observe the combined diagnostic efficacy.Results（1）When analyzing the maximum transverse section and the maximum longitudinal section of thyroid nodules,the sensitivity of S-Detect were 90.59% and91.76% respectively,and the specificity of S-Detect were 63.64% and 67.53%respectively,and there was no statistically significant difference in sensitivity and specificity between the two groups（P>0.05）.After S-Detect reassigned scores to the results of thyroid nodule analysis,the diagnostic efficacy was partially improved,with accuracy,specificity,PPV,and NPV of 80.86%,76.62%,80.00%,and 81.94%,respectively,and a slight decrease in sensitivity of 84.71%（2）When S-Detect diagnosed benign and malignant thyroid nodules,its diagnostic efficacy was correlated with nodule size,whether it was a diffuse lesion background,and whether it was calcified within the nodule.Among them,the diagnostic agreement was the worst when the maximum nodule diameter was ≤5 mm,and the highest when the maximum nodule diameter was 5-10 mm,and its sensitivity,specificity,and accuracy were higher than those of nodules with maximum diameter ≤5 mm and >10mm.S-Detect had a higher diagnostic agreement,higher specificity,PPV,and accuracy when comparing the diagnosis of thyroid nodules without diffuse background and those with diffuse background,and lower sensitivity.S-Detect has higher diagnostic consistency,higher specificity,accuracy,and lower sensitivity for thyroid nodules without calcification compared to those with calcification.（3）Compared with radiologists applying ACR TI-RADS to diagnose benign and malignant thyroid nodules,the sensitivity and AUC of S-Detect were higher than those of radiologists,the specificity was lower than those of radiologists,and the accuracy was comparable.The sensitivity,specificity,accuracy,PPV,and NPV of S-Detect software-assisted radiologist ACR TI-RADS in diagnosing thyroid nodules were96.47%,67.53%,82.72%,76.64%,and 94.55%,respectively.Compared with radiologist diagnosis alone,the specificity and PPV decreased,and the sensitivity,accuracy,and NPV increased to different degrees.（4）The sensitivity,specificity,accuracy and AUC of S-Detect software-assisted radiologist ACR TI-RADS combined with SWE in diagnosing thyroid nodules were90.91%,88.37%,89.80%,and 0.956,respectively,and compared with the sensitivity of radiologists applying ACR TI-RADS alone in diagnosing thyroid nodules,the differences were statistically significant（P < 0.001）;compared with the specificity of S-Detect alone for the diagnosis of thyroid nodules,the differences were statistically significant（P = 0.022）;and compared with the AUC of all three for the diagnosis of thyroid nodules alone,the differences were statistically significant（P < 0.0001;P =0.0006;P = 0.0001）.The combined Kappa coefficient of 0.793 was higher for all three than for all three diagnoses alone.Conclusions The choice of longitudinal and transverse views is not an independent factor in the diagnostic efficacy of S-Detect for thyroid nodules,which remains highly sensitive after diagnostic assignment and is more likely to be an adjunct to diagnosis and asymptomatic screening.Thyroid nodule size,diffuse background,and intra-nodular calcification have an impact on the diagnostic efficacy of thyroid nodules.S-Detect aiding radiologists can improve the sensitivity and accuracy of thyroid nodule diagnosis.The combination of S-Detect,radiologists applying ACR TI-RADS,and SWE should effectively improve the diagnostic efficacy of thyroid nodules,compensate for the low specificity of the S-Detect,avoid overdiagnosis,and compensate for the lower sensitivity of radiologists applying ACR TI-RADS to reduce the possibility of missed diagnosis.