[Abstract] Objective To investigate the value of fully automated breast volume scanning (ABVS) for the differential diagnosis of breast masses. Methods The preoperative ABVS and conventional handheld probe ultrasound (HHUS) image characteristics of a total of 174 lesions in 129 patients with breast masses were retrospectively analyzed, and the masses were classified into five grades: benign, probably benign, indeterminate, probably malignant, and malignant, and the diagnostic effectiveness of both was evaluated by applying the subject operating characteristic (ROC) curve with histopathological findings. The accuracy of “convergence sign” and “hyperechoic border” in the diagnosis of benign and malignant breast masses was evaluated. The specificity and positive predictive value of the ABVS coronal “convergence sign” for the diagnosis of malignant breast masses was 100%, and the negative predictive value was 0. The specificity and positive predictive value of the ABVS coronal “convergence sign” for the diagnosis of malignant breast masses was 100%, and the negative predictive value of the ABVS coronal “convergence sign” was 0. The specificity and positive predictive value of “hyperechoic border” for the diagnosis of benign breast masses were 88.89% and 94.51%, respectively. Conclusion ABVS has high clinical application value, and its coronal breast mass “convergence sign” and “hyperechoic border” have high specificity for the differential diagnosis of benign and malignant breast masses. Ultrasonography (US) plays an important role in the detection and differential diagnosis of breast masses, and in recent years, a new ultrasonographic technique, the automated breast volume scanner (ABVS), has been used in clinical practice. In recent years, a new ultrasound technique, the automated breast volume scanner (ABVS), has begun to be used in clinical practice, but there are few reports on the clinical application value of ABVS in China and abroad. To this end, the author applied the receiver operating characteristic (ROC) curve to compare the value of ABVS with conventional handheld B-mode ultrasound (HHUS) for the diagnosis of benign and malignant breast masses. We also evaluated the “convergence sign” and “hyperechoic border” of the ABVS coronal breast masses, so as to investigate the clinical value of ABVS in the differential diagnosis of benign and malignant breast masses. The study excluded patients with breast deformity, thoracic deformity or history of breast surgery. 157 patients, all female, who underwent HHUS and ABVS of the breast from March 2011 to July 2012, had 188 lumps detected. Of these 188 breast lumps, 148 cases with a total of 174 breast lumps were enrolled in this study because pathological results were not obtained in 9 cases with 14 breast lumps, and the patients were aged 17 to 69 years with a mean of (41.3 ± 10.7) years. All enrolled cases had final histopathological results after HHUS and ABVS examination by ultrasound-guided puncture or surgery. II. Instruments and methods A Siemens ACUSON S2000 color Doppler ultrasound diagnostic instrument was applied, and a 14L5 linear array probe with a probe frequency of 5-14 MHz was used for HHUS examination; an ABVS system was used for ABVS examination with the ACUSON S2000, which used the S2000 as the main unit, a 14L5BV linear array probe and a touch screen fixed The system uses the S2000 as the host, the 14L5BV line array probe and the touch screen fixed at the end of the movable arm, and the probe frequency is also 5-14 MHz, and the post-processing and reconstruction of images are performed by the workstation. The HHUS examination is performed in the order of upper inner quadrant – lower inner quadrant – lower outer quadrant – upper outer quadrant – nipple area, with the probe placed lightly on the surface of the breast to ensure a close fit with the skin, and a continuous scan of the bilateral breast is performed to record the site, size, morphology, border, internal echogenicity, presence of calcification and other features of the mass. The site, size, morphology, boundary, internal echogenicity, presence of calcification and other features of the mass are recorded, and a single static image of the largest diameter of the mass and a dynamic image of the whole mass are stored on disk. Then, the patient is placed in the same position for ABVS scan. Depending on the size, shape and density of the patient’s breast, the device automatically adjusts the depth, total gain, frequency and acoustic beam focus band for image optimization. The usual scanning methods are anterior-posterior (AP), lateral (LAT) and medial (MED), with additional scanning in superior (SUP) and inferior (INF) orientation for larger breasts. A volume of 15.4 cm × 16.8 cm × 6 cm (38808px3) can be obtained with each scan. Each scan must include the nipple, which can be marked by a small yellow box to facilitate orientation and prevent confusion with other hypoechoic lesions in the breast. Data are acquired at an interval of 0.5 mm per scan section, and each ABVS scan takes 65 s. All collected data acquisition is automatically uploaded to the workstation, which performs multiplanar reconstruction (MPR) of the collected data, and the reconstructed images can be obtained from tri-orthogonal views (transverse, sagittal and coronal), the The reconstructed images can be displayed in three orthogonal views (transverse, sagittal and coronal), coronal multilevel images, coronal wide field of view or dynamic. The coronal images can be interpreted from the skin to the anatomical structures of the chest wall at a minimum interval of 0.5 mm. 3. Image analysis and evaluation The breast imaging reporting and data system (BI-RADS) developed by the American college of radiologists (ACR) is used as a reference. data system (BI-RADS) and clinical diagnostic experience, the HHUS images in the hard disk and ABVS data in the workstation were interpreted to evaluate the ultrasound characteristics of the mass, such as shape, growth direction, border, margin, internal echo, calcification, posterior echo and surrounding structures, and each mass was evaluated as benign, possibly benign, indeterminate, possibly malignant, and malignant in five grade. The marginal features of the breast masses on the ABVS coronal images were evaluated, and the convergence of hypoechoic and hypoechoic bands around the masses on several or more images was called the “convergence sign”, which was used as a feature to determine malignancy. The continuous or discontinuous hyperechoic border around the mass on several or more images is called “hyperechoic border” and is used as a feature to determine benign. The interpretation of HHUS and HBVS images of all masses was performed independently by two physicians with extensive experience in breast ultrasound; in case of disagreement, the final conclusion was reached by consensus between the two physicians, who were unaware of the physical examination results and the final histopathological findings of the patients. III. Statistical analysis Using MedCalc 11.2 software, the sensitivity degree and specificity of HHUS and ABVS grading for the diagnosis of breast masses were calculated separately, and the ROC curves were plotted with sensitivity as the vertical coordinate and 100-specificity as the horizontal coordinate to evaluate the diagnostic effect of both, and the area under the curves were compared (Z test at the α=0.05 level). RESULTS I. Histopathological results Each breast mass was subjected to ultrasound-guided puncture biopsy or surgery to obtain final histopathological results (Table 1). In all cases, 131 cases had 1 mass, 10 cases had 2 masses, 5 cases had 3 masses, and 2 cases had 4 masses. Out of the total 174 masses, 129 (74.14%) were benign and 45 (25.86%) were malignant. The maximum diameter of benign masses ranged from (0.7 to 3.9) cm, with an average of (1.76 ± 0.64) cm, and the maximum diameter of malignant masses ranged from 1.1 to 4.1 cm, with an average of (1.97 ± 0.73) cm. II. Diagnostic results of ROC curves The results of the graded diagnosis of breast masses by HHUS and ABVS are shown in Tables 2 and 3, and the results of the graded diagnosis of breast masses by HHUS and ABVS are shown in Figure 4. ROC curve results are shown in Figure 4, the ABVS curve is located at the upper right of the HHUS curve, and the area under the ROC curve for the graded diagnosis of breast masses by ABVS (0.927) is larger than that by HHUS (0.903) (Z=2.256, P=0.024). Typical features of ABVS coronal surface Among 174 breast masses, ABVS coronal surface “convergence sign” and “hyperechoic border” were observed. The specificity and positive predictive value of the “convergence sign” were 100%, and the accuracy and negative predictive value were 90.23% and 88.36%, respectively, with a false positive rate of 0 and a false negative rate of 37.78%, but the sensitivity was relatively low (62.22%). The specificity and negative predictive value of “hyperechoic border” for the diagnosis of benign breast masses were 88.89% and 94.51%, and the sensitivity, accuracy, false-positive rate, false-negative rate and positive predictive value were 66.67%, 72.41%, 11.11%, 33.33% and 31.01%, respectively. Discussion In recent years, the incidence of breast cancer in China has been increasing year by year with a trend of rejuvenation, and it is generally believed that early diagnosis and early treatment are the keys to a better prognosis of breast cancer. Currently, X-ray mammography is recognized by most clinicians and imaging physicians as the “gold standard” for breast cancer screening and diagnosis. HHUS is one of the most commonly used ultrasound methods to evaluate breast disease because of its convenience, high resolution and lack of ionizing radiation. However, HHUS has shortcomings such as high operator dependence and diagnostic real time. To reduce operator dependency, automated breast ultrasound has undergone more than 40 years of development. The concept was first proposed by Wells and his team in 1968, and a commercially available scanning device was produced in the 1970s, followed by several different automated breast volumetric scanning devices for clinical use. However, none of these devices were widely used in the clinic due to low sensitivity in detecting breast masses, inconvenience in use or patient discomfort. Recently, a new generation of automated breast ultrasound, ABVS, has begun to be used in clinical practice. The author previously reported a preliminary discussion of the value and feasibility of the clinical application of ABVS, and the results showed that ABVS has many advantages compared to conventional ultrasound. First, ABVS scanning minimizes operator dependence as the entire breast is captured and reconstructed without missing data. Second, ABVS reconstruction provides coronal image information that is not available with conventional ultrasound, known as the “surgical view” plane because it is the same plane as the surgeon’s visual orientation when the patient is lying on the operating table. Finally, because of the integrity of the data, ABVS images are useful for offline analysis, expert consultation, teleconsultation, and file management. The value of ABVS in the diagnosis of breast cancer has been studied by some scholars, but their perception of the clinical value of this method is not consistent. Therefore, the clinical application of ABVS is still in the experimental stage. In contrast, Chang et al. and Wang et al. held the opposite view, as they performed ABVS on 239 and 69 breast masses, respectively, and showed that the sensitivity, specificity, and accuracy were 100%, 95.0%, 97.1%, and 95.30%, 80.50%, and 85.80%, respectively, so they both concluded that ABVS has a high clinical application value. The author believes that the reasons for the differences in the results of different investigators may be related to the size of the sample size, the level of experience in ABVS sweeping and the level of interpretation of coronal images. In this study, the author performed a comparative diagnostic study of 174 breast masses with ABVS and HHUS grading, and the results showed that the area under the ROC curve for the diagnosis of benign and malignant breast with ABVS grading reached 0.927, which was higher than that of HHUS at 0.903, and the difference was statistically significant (Z=2.256, P=0.024), and this result indicates that ABVS has a high clinical application value. Many scholars have studied and discussed the characteristics of 3D coronal ultrasound images of breast masses, and the results have shown that the coronal “convergence sign” (or “sun sign”) of breast cancer has high specificity and can greatly improve the diagnostic accuracy of breast cancer. However, the acquisition and reconstruction of 3D ultrasound data are highly operator-dependent, and it is difficult to acquire all the breast tissue data because of the large size of the breast and the relatively small size of the 3D probe, thus limiting the wide application of 3D ultrasound in breast. In this study, the ABVS was used to automatically scan the entire breast volume and then reconstructed a good coronal image. The specificity and positive predictive value of the ABVS coronal “convergence sign” for the diagnosis of malignant breast masses was 100%, and the false-positive rate was 0. The specificity and negative predictive value of “high echogenicity border” for the diagnosis of benign breast masses were also as high as 88.89% and 94.51%, respectively. The author believes that the infiltration of malignant breast tumor cells into the surrounding tissues and the induction and pulling of the surrounding tissues are the pathological basis of the “convergence sign” of malignant breast masses; whereas benign masses generally grow in a distended manner, and the envelope of fibroadenoma or the duct wall of intraductal tumor becomes the reason for the formation of “hyperechoic borders”. The reason for the formation of “hyperechoic border” is the envelope of fibroadenoma or the ductal wall of intraductal tumor. At the same time, it is noteworthy that the sensitivity of the “convergence sign” for the diagnosis of malignant breast masses in this study was only 62.22%, while the false-negative rate was 37.78%, probably because some of the malignant masses were small or less malignant, and infiltrated and stretched the surrounding tissues to a lesser extent; whereas the “hyperechoic border” was used for the diagnosis of malignant breast masses. The sensitivity, false-negative rate and negative predictive value of “high echogenic border” in diagnosing benign breast masses were 66.67%, 33.33% and 31.01%, respectively, which may be due to the thin envelope of some fibroadenomas or the absence of envelope structure in other benign breast masses. However, overall, the “convergence sign” and “hyperechoic border” of ABVS coronal breast masses provide a new index for clinical diagnosis, which can improve the differential diagnosis of benign and malignant breast masses when combined with other signs. Compared with conventional ultrasound, ABVS also has some shortcomings. First, ABVS cannot be used for axillary sweeping and evaluation of axillary lymph nodes due to its large frame of fixed probe. Secondly, ABVS does not have the ability to detect the blood flow signal of the lesion and the elasticity of the tissue, while the blood flow information of the lesion and the elastic hardness of the tissue help to determine the nature of the lesion. In summary, ABVS has high clinical application value, and its unique coronal images can provide more useful information for the diagnosis of clinical breast masses, and the “convergence sign” and “hyperechoic border” of coronal breast masses are useful for the differential diagnosis of benign and malignant breast masses. The coronal breast mass “convergence sign” and “hyperechoic border” have high specificity for the differential diagnosis of benign and malignant breast masses.