科研之友
微信
新浪微博
Facebook
分享链接
摘要
Rationale and Objectives: To evaluate the reliability of cone-beam breast computed tomography (CBBCT) in visual assessment of breast density referring to the fifth edition of the Breast Imaging Reporting and Data System compared to digital mammography. Materials and Methods: Breast density assessments of 130 female patients were performed by five radiologists referring to the fifth edition of Breast Imaging Reporting and Data System atlas both on two-view mammograms and CBBCT images. Assessments were repeated by three radiologists with different experience more than 1 month after the initial evaluation. The inter- and intrareader agreements were compared by using the Cohen's weighted Kappa statistic and intraclass correlation coefficient. Weighted Kappa statistic was also used to analyze the agreement between CBBCT images and mammograms. The influence of radiologist experience for breast density assessment was analyzed using a chi-square test. Results: For CBBCT images, the inter-reader agreement was 0.781, whereas the agreement on mammograms was 0.744, both demonstrating moderate agreement. The level of intrareader reliability was higher on the CBBCT images than mammograms for breast density evaluation, 0.856 versus 0.786. Based on the majority report, the agreement between these two modalities was on substantial agreement degree. There was a statistically significant difference among radiologists with different levels of experience, and higher density categories were reported more often by experienced reader. Conclusion: CBBCT showed equal aptitude and better agreement for the breast density evaluation compared to mammography. CBBCT could be an effective modality for breast density assessment and breast cancer risk evaluation in routine diagnosis and breast cancer screening.
- 出版日期2019-6
- 单位天津医科大学
