The Study Of Computer Aided Diagnosis Combined With Shear-wave Elastography For Diagnosis Of Malignant And Benign Thyroid Nodules

ObjectiveThe computer-aided diagnosis software and shear wave elastography are used to diagnose benign and malignant thyroid nodules.This study will explore: 1.The diagnostic efficiency of computer-aided diagnosis software for benign and malignant thyroid nodules;2.The use of computer-aided diagnosis software The diagnostic efficiency of various guidelines for benign and malignant thyroid nodules;3.The diagnostic efficiency of shear wave elastography for benign and malignant thyroid nodules;4.Three different combinations of BTA guidelines,ACR guidelines and shear wave elasticity are used Analyze the diagnostic performance of the imaging results.Methods1.This study included 62 patients with benign and malignant thyroid nodules who were admitted to the First Affiliated Hospital of Guangzhou University of Traditional Chinese Medicine(hereinafter referred to as "our hospital")from December 2018 to July 2021.A total of 67 nodules were the research objects,and retrospective research was carried out.Inclusion criteria:(1)solitary thyroid nodule;(2)nodule diameter ≤ 3cm;(3)surgical treatment and pathological results;(4)solid nodules;exclusion criteria:(1)multiple nodules in unilateral lobe of thyroid;(2)nodules with large calcification,Circular calcification or calcification acoustic shadow affects the observation of rear echo(the largest diameter exceeds 5mm);(3)The position of the nodule is more than 4cm away from the skin.All ultrasound images of the included nodule should contain one longitudinal and transverse two-dimensional image,and four longitudinal and transverse shear wave elastic images.2.Then use Am Cad Bio Med Corporation thyroid ultrasound image processing software2.2(c)to analyze the collected cross-sections and longitudinal sections,and analyze the degree of hypoechoic nodules,strong echo points,unevenness of internal echo,and degree of edge blur,Aspect ratio,anechoic area and other six aspects to identify and judge,and get the thyroid graphic report classification system(Kwak2011)released by Kwak et al.in Korea in 2011,and the British TI-thyroid graphic report classification system in 2014(British TIRADS),the thyroid graphic report classification system(Seo2015)released by Seo et al.in2015,the thyroid graphic report classification system of the American Thyroid Association in 2015(ATA TI-RADS),the 2016 American Association of Clinical Endocrinologists and American Endocrine Society Thyroid Graphic Report Classification System(AACE2016),2016 Consensus Statement and Recommendations of the Korean Society of Thyroid Radiology(KSR2016)2017 American Society of Radiological Thyroid Graphic Report Classification System(ACR 2017),2017 European Thyroid Association released The thyroid graphic report classification system(ETi RADS2017)and other 8 guidelines are concluded,and the best guide results for Area Under Curve(AUC)are selected through statistical software.3.In addition,for the results of elastography,the average shear wave velocity Mean(m/s),the minimum shear wave velocity Min(m/s),and the maximum shear wave velocity Max of the transverse and longitudinal sections(m/s),the average value of Young’s modulus Mean(k Pa),the minimum value of Young’s modulus Min(k Pa),and the maximum value of Young’s modulus Max(k Pa)are entered into statistical analysis,and the best is analyzed by statistical software Calculate the cut-off value of each parameter of elastography.4.Use series method,parallel method and partial combination method to combine the results of computer-aided diagnosis software and the results of shear wave elastography to analyze their diagnostic efficacy.Results1.Among the 67 cases of nodules included this time,49 cases were malignant nodules and 18 cases were benign nodules.2.Calculate the results of the binary classification data of each guideline and the pathological benign and malignant comparison through SPSS25.0,analyze and statistics the respective receiver operating characteristic curve(Receiver Operating Characteristic Curve,ROC),and calculate the AUC,Kwak2011: 0.573;ACR2017: 0.636;ATA2015: 0.618;KSR2016: 0.636;Seo2015: 0.618;AACE2016: 0.626;BTA2014: 0.666;ETi RADS2017:0.629.The results of BTA2014 from each guide were higher,higher than the other guides,P<0.05,there was a statistical difference;the results of the other guides were not significantly different,P >0.05,there was no statistical difference.3.By calculating the average value of shear wave velocity Mean(m/s),the minimum value of shear wave velocity Min(m/s),the maximum value of shear wave velocity Max(m/s),Yang The average value of the modulus Mean(k Pa),the minimum value of the Young’s modulus Min(k Pa),and the maximum value of the Young’s modulus Max(k Pa).The result is: Mean(k Pa): 0.788;Min(k Pa): 0.634,Max(k Pa): 0.730;Mean(m/s): 0.753;Min(m/s):0.650;Max(m/s): 0.718.It is found that the AUC of Mean(k Pa)is the highest.4.Through SPSS25.0 calculation statistics each guide.Analyze the results of the binary classification data and the pathological benign and malignant comparison,analyze and statistics the respective receiver operating characteristic curve(Receiver Operating Characteristic Curve,ROC),and calculate the AUC,Kwak2011: 0.573;ACR2017: 0.636;ATA2015: 0.618;KSR2016: 0.636;Seo2015: 0.618;AACE2016: 0.626;BTA2014: 0.666;ETi RADS2017: 0.629.The results of BTA2014 from each guide were higher,higher than the other guides,P <0.05,there was a statistical difference;the results of the other guides were not significantly different,P >0.05,there was no statistical difference.In addition,by calculating the average value of shear wave velocity Mean(m/s),minimum value of shear wave velocity Min(m/s),maximum value of shear wave velocity Max(m/s),and Young’s The AUC of 6 parameters including the average value of the modulus Mean(k Pa),the minimum value of Young’s modulus Min(k Pa),and the maximum value of Young’s modulus Max(k Pa).The result is: Mean(k Pa): 0.788;Min(k Pa)): 0.634,Max(k Pa): 0.730;Mean(m/s): 0.753;Min(m/s): 0.650;Max(m/s): 0.718.It is found that the AUC of Mean(k Pa)is the highest.Finally,the two diagnostic methods were combined through three methods: series combination,parallel combination and partial combination,and finally obtained the AUC value of series method: 0.710;parallel method AUC value: 0.643;partial combination method AUC value: 0.651.The AUC value of the series combination is the largest,and the AUC value of each single method is increased.The AUC values of the parallel combination and partial combination method are approximately the same.This shows that the two can have complementary effects.The sensitivity of the partial combination method and the parallel method is the highest,and the diagnostic efficacy of the combination of the two is roughly the same.Conclusion1.The diagnostic efficiency of the thyroid graphic report classification system used by the computer-aided diagnosis system is roughly the same,and the BTA guidelines have the highest diagnostic efficiency.2.Mean(k Pa)of Young’s modulus in shear wave elastography has high diagnostic efficiency.3.Shear wave elastography can be used for the differential diagnosis of benign and malignant thyroid nodules.4.The combination of two diagnostic methods in tandem will maximize the diagnostic efficiency.5.Shear wave elastography has a supplementary effect on the thyroid graphic report classification system.6.The diagnostic efficiency of computer-aided diagnosis combined with shear wave elastography in the diagnosis of benign and malignant thyroid nodules is higher than that of a single method.Computer-aided diagnosis combined with shear wave elastography for the diagnosis of benign and malignant thyroid nodules is worthy of promotion in clinical work.