Computer Science · Artificial Intelligence
Quantity research of sonographic images for breast imaging reporting and data reporting and data system grading in breast lesions
See on the mapAbstract
Objective To investigate the value of breast imaging reporting and data reporting and data system(BI-RADS)assessment category based on the scores of sonographic features in the breast lesions.Methods The ultrasonic images for 1005 breast lesions confirmed by pathology were retrospectively reviewed.The age and the features of sonographic images including lesion shape,margin,halo,taller-than-wider,posterior attenuation,microcalcification,flow,lymph node were analyzed,then the scores were obtained due to the features.And the lesions were classified according to the scores.In each group,the malignant predicted value was compared with theorized risk degree of malignancy by BI-RADS to obtain the relationship between score and BI-RADS.Results The proportion of malignant lesions in score ≤3,4,5,6,7,8,9,10,11,12,13 14 and ≥15 were 0,1.6%,4.3%,8.8%,29.5%,42.1%,55.3%,69.7%,79.2%,73.0%,77.8%,95.7% and 94.5%,respectively.The 2,3,4a,4b,4c and 5 by BI-RADS were corresponded to score≤3,4,5~7,8~9,10~13 and ≥14,respectively.Score 6 or BI-RADS 4b was the best cut-off value by ultrasonography for differentiating malignant and begin breast lesions.The sensitivity,specificity,accuracy,negative predictive value and positive predictive value were 90.2%,90.1%,90.2%,95.3% and 80.7%,respectively.Conclusion BI-RADS assessment category according to sonographic feature scores can partly standardize ultrasonic report and offer reliable guiding information for clinical diagnosis and treatment of breast disease.