Application Of Qualitative And Quantitative Optimized BI-RADS Classification In Differential Diagnosis Of Benign And Malignant Breast Nodules

Objective To investigate the specific standards of the qualitative and quantitative optimized breast imaging reporting and data system(BI-RADS)classification methods and the clinical application value in diagnosing breast nodules.Methods Ultrasonographic imaging data of 312 patients with 326 breast nodules were retrospectively analyzed.There were 8 ultrasonographic signs of breast nodules,including irregular shape,unparallel orientation,non-circumscribed margin(indistinct,angular,microlobulated,spiculated),complicated echogenicity,microcalcification,posterior acoustic shadow,structure distortions and duct ectasia,compared with the golden standard—biopsy or postoperative pathologic diagnosis and analyzed by the χ2 test.The ultrasonographic signs which had statistically significant difference were inserted into the multiple logistic regression analysis to screened out independent predictors.The valuable ultrasonographic signs formed the qualitative and quantitative optimized BI-RADS classification methods to classify 326 breast nodules and evaluate the diagnostic efficacy to find the cut-off point.Check consistency between qualitative and quantitative optimized BI-RADS classification methods.Results 1.In the χ2 test,irregular shape,unparallel orientation,non-circumscribed margin,complicated echogenicity,microcalcification,posterior acoustic shadow,structure distortions,duct ectasia,and indistinct,angular,microlobulated and spiculated margin had statistically significant difference(P<0.05).2.Multiple logistic regression analysis showed the independent predictors of malignant breast nodules were unparallel orientation,irregular margin,complicated echogenicity,microcalcification,posterior acoustic shadow and structure distortions,and their OR were 6.427、3.040、5.662、7.810、5.542、8.371.The AUC was 0.961.The above 6 ultrasonographic signs formed qualitative and quantitative optimized BI-RADS classification methods and 4b was the cut-off point which had the best efficacy to diagnose benign and malignant breast nodules.3.The consistency check between qualitative method and pathologic diagnosis showed Kappa was 0.804.The consistency check between quantitative method and pathologic diagnosis showed Kappa was 0.798.The consistency check between qualitative and quantitative methods showed Kappa was 0.897.Conclusion Qualitative and quantitative optimized BI-RADS classification methods have high clinical application value in diagnosing benign and malignant breast nodules.Qualitative optimized BI-RADS classification method is supposed to be the preferred method.