| Objective1. To explore the application value of Breast Imaging Reporting and Data System contained Ultrasonography (BI-RADS-US) in the categorization of breast lesions.2. To estimate the clinical importance of the pattern on subdividing the breast lesions of Category 4 into Category 4a,4b and 4c according to the BI-RADS-US.3. To evaluate inter-and intra-observer variabilities in breast sonographic feature analysis and management, using the fourth edition BI-RADS-US.Materials and Methods1. The instrument was Accuson Sequoia 512、Alokaα10 And Hitachi EUB-8500 etc. with linear array probe, frequency from 8.0MHz to 14.0MHz.2. Breast ultrasound was performed in 6618 female patients with 7941 breast lesions between 2001 and 2004. Each lesion was described and classified into categories 0 to 5 according to the BI-RADS for breast ultrasound.All the lesions of Category 4 were subdivided into Category 4a,4b and 4c. Categorization and biopsy results were compared.3. We included 437 patients with 510 breast lesions who underwent breast ultrasound. Six observers who divided into two groups based on experience level retrospectively reviewed sonographic images of lesions twice within an 4-week interval. They were blinded to the clinical information and pathologic results of each case as well as the ratio of malignant to benign lesions included in the study. The observers described each lesion, using BI-RADS descriptors (such as lesion shape, orientation, margin,echo pattern,boundary and posterior acoustic features) and final assessment. Inter- and intra-observer variabilities were assessed with Cohen’s kappa statistic.Results1. In all 7941 breast lesions,59 lesions could not be evaluated, so they belonged to Category 0. Other 7882 lesions were classified into categories 1 to 5 according to the BI-RADS-US,5735 were confirmed benign and 2147 were confirmed malignant by histopathology. The sensitivity, specificity, accuracy, PPV and NPV of BI-RADS-US category were 88.50%,91.02%,90.36%,78.74% and 95.48%.2.2685 breast lesions belonging to categories 4 were observed,1495 were confirmed benign and 1190 were confirmed malignant by histopathology. The PPV of categories 4 was 44.32%. If all zhe lesions were subdivided into categories 4a,4b and 4c, zhe NPV of categories 4a was 90.72%, zhe PPV of categories 4b and 4c were 49.21% and 86.35%.3. For inter-observer agreements for sonographic descriptors, perfect agreement for lesion orientation, moderate agreement for final assessment,lesion shape, echo pattern,boundary and posterior acoustic features (κ=0.42,0.57,0.56,0.49 and 0.41, respectively),and fair agreement for lesion margin (κ=0.25) were obtained. For intra-observer agreement, substantial to perfect agreement was found for almost all lesion descriptors and final assessments (κ=0.52～0.91).Conclusions1.As a method of supervisory control quality, BI-RADS-US classification has high application value in the diagnosis of benigh and malignant breast lesions, otherwise, it can standardize the breast ultrasound report, it should be putted into use widely.2. The pattern on subdividing the breast lesions of Category 4 into Category 4a,4b and 4c according to the BI-RADS-US can provide clinical docters with more useful information, and it’s convenient to the clinical management.3. Because inter- and intra-observer agreement with the BI-RADS lexicon for US is good, the use of BI-RADS lexicon can provide accurate and consistent description and assessment for breast US, besides, it is less influenced by radiologists’ experience. |
PDF Full Text Request