The Thyroid Cancer Diagnose Model Based On Ultrasonic Characteristic Binary Logistic ? Composite Score Method

Objectives To screen out independent risk factors of regarding thyroid cancer.To establish ultrasonic diagnostic model for thyroid nodules by binary logistic regression coefficient ? composite score and to evaluate the value of the model and verify its working efficiency.TIRADS grading criteria for thyroid nodules were determined by this scoring method.Methods 1749 cases of patients with a total of 1988 thyroid nodules underwent ultrasound and confirmed by pathological result in affiliated hospital of North China university of technology,from January 2015 to August 2018,were included in the experimental set.The conventional sonographic features,reports and pathological result of these nodules were analyzed,ultrasonic characteristic include composition,echogenicity,shape,margin,extension beyond the thyroid border,echogenic foci,S/L ratio.To screen the ultrasound features with significant difference identified by the univariate analysis,then binary logistic regression analysis was performed to screen out independent risk factors regarding thyroid cancer.Then according to the regression coefficient ? value of the risk signs,the smallest ? deed 1 score,the ration of ? value of other partial regression coefficient to it was used to assign the else risk factors.A score of 10 times was used as the final weight of each factor.Finally to establish the conventional ultrasound characteristic diagnostic model of thyroid nodules based on the composite score,and draw the characteristic work curve of the test worker to evaluate the model.150 thyroid nodules underwent ultrasound and confirmed by pathological result in affiliated hospital of North China university of technology,from September 2018 to December 2018 were included in the validation set.The conventional sonographic features of these nodules were analyzed,ultrasonic characteristic include composition,echogenicity,shape,margin,extension beyond the thyroid border,echogenic foci,S/L ratio.Nodules were scored according to the weight score of each sign in the experiment study.then thyroid nodules were classified as benign or malignant with reference to cut-off points.The gold standard for the classification of benign and malignant thyroid nodules postoperative pathological results was adopted,and the ROC curve of the validation set was conducted to verify the ability of this model to distinguish benign and malignant thyroid nodules.Finally,the classification standard of thyroid nodules was studied by the experimental date,it also makes a comparative analysis with the internationally recognized TI-RADS classification standard.Results Suspected risk factors for thyroid cancer include: solid;hypoecho;very hypoecho;obscure boundary;irregular shape;S/L ratio?1;microcalcification and extension beyond the thyroid border.The weight scoring system based on the multivariate logistic regression ? value of ultrasonic signs is 10 for solid;13 for hypoecho;20 for very hypoecho;10.8 for obscure boundary;24.3 for irregular shape;15.4 for S/L ratio?1;27.7 for microcalcification.The ultrasound character diagnostic model of thyroid nodules on this basis has the higher diagnostic accuracy 95.3%[95% CI(0.942,0.964)],The best cut-off for the diagnosis of malignant lesions is 24.2 with a sensitivity of 88.6%,a specificity of 93.3%,a positive predictive value of 86.8% and a negative predictive value of 94.4%.In the validation set,the accuracy of this model in the differential diagnosis of benign and malignant thyroid nodules was 83.3%.The ultrasonic diagnostic criteria for thyroid nodules were divided into 0~6 grades according to the ultrasonic symptom weight score and malignant risk.Compared with other TI-RADS grading criteria,it was more perfect.Conclusions Suspected risk factors for thyroid cancer include: solid nodules;hypoecho;very hypoecho;obscure boundary;irregular shape;S/L ratio?1;microcalcification and extension beyond the thyroid border.The ultrasound characteristic diagnostic model of thyroid nodules based on ultrasound characteristic binary logistic regression coefficient ? composite score method has higher efficiency in differentiating benign and malignant thyroid nodules,and the TI-RADS grading standard with this weight score has better clinical value.Figure 2;Table 10;Reference 149.