OBJECTIVETo assess the accuracy of detection by automated breast volume scanner (ABVS) in diagnosis of high-risk and small breast lesions.

METHODSOne hundred and twelve patients with solid high-risk and small breast lesions were identified by ABVS. The patients were divided into benign lesion group and cancer group after pathological examination. The clinicopathological findings and ultrasonographic features of the lesions were compared.

RESULTSAmong the 112 lesions there were 49 benign and 63 malignant lesions. The mean size on ABVS and pathology were (1.59 ± 0.52) cm and (1.52 ± 0.58) cm. There was no significant difference in tumor sizes determined by ABVS and pathology (P = 0.194). The mean age of patients with benign lesions was (38.5 ± 7.4) years and that of malignant lesions was (52.4 ± 13.6) years, showing a significant difference between the two groups (P < 0.001) . The mass shape, orientation, margin, lesion boundary, echo pattern, calcification, BI-RADS category and retraction phenomenon were significantly different of the malignant and benign masses (P < 0.05). But there was no significant difference in the location of lesions and posterior acoustic features (P > 0.05) . Retraction phenomenon was significantly associated with pathological type and histologic grade of the breast cancer (P < 0.01). The specificity, sensitivity and accuracy of retraction phenomenon were 100% (46/46), 73.0% (46/63), and 84.8% (95/112), respectively.

CONCLUSIONSABVS provides advantages of better size prediction of high-risk and small breast lesions. Furthermore, the retraction phenomenon in coronal plane shows high specificity and sensitivity in detecting breast cancer.