Ultrasound classification of non-mass breast lesions following BI-RADS presents high positive predictive value

• Published in: PloS one 2022-11, Vol.17 (11), p.e0278299-e0278299
• Main Authors: Lin, Mingnan, Wu, Size
• Format: Article
• Language: English
• Subjects: Biology and Life Sciences; Biopsy; Breast; Breast - diagnostic imaging; Breast cancer; Breast carcinoma; Calcinosis - diagnostic imaging; Care and treatment; Classification; Clinical medicine; Diagnosis; Female; Humans; Informatics; Lesions; Lymphoma, Follicular; Malignancy; Mammography; Medicine and Health Sciences; Methods; Normal distribution; Patients; Physicians; Predictive Value of Tests; Research and Analysis Methods; Risk; Sensitivity; Ultrasonic imaging; Ultrasonography; Ultrasonography, Mammary; Ultrasound; Ultrasound imaging; Variables
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0278299

To investigate the positive predictive value of ultrasound classification of non-mass breast lesions (NMLs) following breast imaging reporting and data system (BI-RADS), and enhance understanding of NMLs. Fifty-nine women with 59 ultrasound-detected breast NMLs were finally enrolled. The ultrasound (US) features of breast NMLs were analyzed; the incidence of malignant NMLs was calculated; the malignancy risk stratification of US for breast NMLs was established using BI-RADS. The incidence of malignant NMLs was 4.59% of all breast carcinoma. Non-ductal hypoechoic area, microcalcifications and posterior shadowing are the main US features of malignant NMLs, and there were significant differences between malignant and benign NMLs for microcalcifications and posterior shadowing. Taking BI-RADS 4B as a cutoff value, the sensitivity, specificity, area under the receiver operating characteristic curve (AUC), positive and negative predictive values, and odds ratio of the BI-RADS category were 82.98%,41.67%,0.62,84.78%,38.46% and 3.48, respectively. Stratifying the malignancy risk of breast NMLs using the BI-RADS the sensitivity and positive and predictive value are promising, but the likelihood of malignancy of malignant NMLs is underestimated, and that of benign NMLs is overestimated. The solution may be that to separate NMLs from breast masses and use different malignancy risk stratification protocols.