“Cloud sign” in automated breast volume scanning coronal images for the differential diagnosis of benign and malignant breast lesions

Introduction

According to the 2020 Global Cancer Statistics Report by the International Agency for Research on Cancer (IARC), breast cancer ranks as the most prevalent cancer worldwide (1). Early-stage breast cancer generally has a favorable prognosis. Patients without axillary lymph node involvement have a 5-year overall survival rate exceeding 80%; however, in cases of distant metastasis, this rate can drop to 25% (2). Consequently, early diagnosis and treatment of breast cancer play a pivotal role in improving survival outcomes (3).

Currently, the primary imaging methods for breast cancer diagnosis include ultrasound, mammography, and magnetic resonance imaging (MRI) (4). Mammography is a non-invasive, efficient screening tool, particularly sensitive to microcalcifications, and can detect calcifications as small as 0.1 mm, making it indispensable for breast cancer screening. However, its sensitivity decreases in women with dense or small breasts, and it poorly evaluates axillary lymph nodes (5). MRI offers superior soft tissue resolution and sensitivity for microlesions, enabling detailed morphological and perfusion analysis, thus improving diagnostic accuracy. Despite its advantages, MRI’s high cost and long wait times limit its use to preoperative assessments rather than routine screening. Ultrasound, recognized for its convenience, cost-effectiveness, and absence of radiation, has become an indispensable tool for diagnosing breast diseases (4,6). Nonetheless, the traditional ultrasonography has several limitations, including operator dependence, difficulties in imaging specific structures, and reduced accuracy in obese patients (7).

Automated breast volume scanning (ABVS) is a dedicated 3D ultrasound imaging technology for breast examination, which uses a mechanical arm to automatically scan the entire breast, generating standardized and objective 3D volumetric images for diagnosis. The process is operator-independent, and stored images allow repeated review, facilitating expert consultations and long-term follow-up. Also, it provides a comprehensive and reproducible three-dimensional assessment of the breast, aiding in improved detection and characterization of breast lesions (8). The ABVS introduces a unique coronal plane sign, offering a fresh perspective for diagnosing breast lesions. This enables a comprehensive assessment of lesion boundaries and their relationship with surrounding tissues, as well as providing insights into the morphology and growth patterns of breast nodules (9,10).

Several studies have demonstrated that ABVS is associated with superior sensitivity, specificity, and accuracy in diagnosing breast tumors compared with conventional hand-held two-dimensional (2D) ultrasound (11,12). However, previous studies have mainly focused on the diagnostic value of the “convergence sign” (also called Retraction phenomenon) or “hyperechoic halo” for breast tumors (11-13). In addition to these signs, our team has identified a distinct pattern in certain breast lesions located between the ABVS coronal plane and the skin layer, characterized by patchy echogenic elevation with unclear boundaries, resembling a satellite cloud map. We initially termed this imaging feature the “cloud sign”. The aim of this study was to assess the diagnostic value of cloud sign in ABVS images for benign and malignant breast lesions, and to further compare it with convergence sign and hyperechoic halo. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2529/rc).

Methods

Study design and patients

This retrospective observational study included patients with breast lesions who underwent ABVS examinations at The Third Affiliated Hospital of Guangzhou Medical University between January 2020 and January 2022, comprising a total of 187 lesions.

The inclusion criteria were as follows: (I) individuals who underwent ABVS and had a Breast Imaging Reporting and Data System (BI-RADS) classification of 3 or higher; and (II) those who underwent postoperative or biopsy pathological examinations. The exclusion criteria were: (I) individuals who had received prior drug or surgical tumor treatment before ABVS examination; and (II) those with incomplete or poor-quality ultrasound imaging data.

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by Ethics Committee of The Third Affiliated Hospital of Guangzhou Medical University (No. LCYJ-2023-064) and informed consent was obtained from all individual participants.

Ultrasound scanning and image analysis

The Siemens Acuson S2000 ABVS color Doppler ultrasound diagnostic instrument, equipped with a high-resolution linear probe (14L5BV) operating in the frequency range of 5–14 MHz, was utilized. Patients were placed at the supine position on the examination bed with their hands placed above their heads. Breast scanning was performed, and images were acquired in a bottom-to-top direction, covering the nipple, within a scanning range of 15.4 cm × 16.8 cm × 6.0 cm. Standard anterior-posterior (AP) and lateral (LAT) images were routinely acquired, If the scans from the aforementioned two positions cannot clearly display the overall condition of the tumor, we have added scans in the medial (MED), superior (SUP), and inferior (INF) positions to improve visualization. After the probe is fixed, each scan can only capture one orientation, such as AP or LAT, and each scan takes 65 seconds. The total scanning time for a patient includes both preparation time and actual scanning time. Preparation involves adjusting the patient’s position and the probe’s placement, and this time is usually variable. For lesions in conventional locations, scanning in two orientations is generally sufficient, which takes approximately 130 seconds excluding preparation time. If additional scanning orientations are required, each additional orientation adds 65 seconds to the total scanning time. After image acquisition using the nipple as a reference point, images were transferred to the ABVS workplace image processing system. Following image reconstruction using multi-planar reconstruction (MPR) technique, ultrasound physicians accessed the reconstructed breast images for offline analysis. Subsequent to the ABVS scan, a consultant radiologist conducted conventional hand-held 2D ultrasound scanning to evaluate general tumor data, including location, size, morphology, edges, margins, internal echoes, vascularity, and axillary lymph nodes. The integrity and quality of the images were initially assessed by an ultrasound physician with three years of experience in ABVS scanning and diagnostics. If an image was deemed unsatisfactory, it was immediately rescanned to meet the required quality standards. Additional scans were performed only in special situations, such as when the lesion was too large to be fully captured in standard AP position and LAT position views, or when the lesion was located near the periphery of the breast, making standard views insufficient for complete visualization.

After image acquisition and data collection were completed, a unified image interpretation process was conducted. All images were reviewed by two ultrasound physicians, each with over five years of experience in ABVS image interpretation. Both radiologists received specialized training in identifying key imaging signs such as the cloud sign, convergence sign, and hyperechoic halo. Discrepancies between the two radiologists were resolved by a third radiologist with over 10 years of experience. The radiologists received specialized training to identify the cloud sign before the study. The training included the analysis of at least 30 cases with the cloud sign and 50 cases without the “cloud sign” to ensure accurate recognition. The analysis, conducted based on the fifth edition of the Bi-Rads classification published by the American College of Radiology (14), concentrated on signs, such as the cloud sign, convergence sign, and hyperechoic halo, documented in the anterior aspect of the tumor. In cases where opinions diverged between the two consultant radiologists, a third consultant radiologist with over 10 years of experience participated in a consensus discussion.

Data collection and definitions

The cloud sign was defined as a patchy echogenic elevation area observed on the coronal plane of ABVS between the tumor and the skin layer, characterized by unclear boundaries and resembling a mist or cloud-like appearance (Figure 1A-1C). The convergence sign was identified as a sonographic imaging feature that was radially distributed linear echogenic strands surrounding the mass and converging towards the center (Figure 1D). The hyperechoic halo was recognized as a bright echogenic band at the periphery of breast tumors on the coronal plane, bordering with normal tissue (Figure 1E). Tumor size was categorized into three groups based on the TNM staging for breast cancer (14): (I) ≤2 cm; (II) >2 to ≤5 cm; (III) >5 cm. Metastatic status was determined by lymph node aspiration or intraoperative sentinel lymph node biopsy results and was divided into two groups: the metastasis and non-metastasis group. All breast lesions underwent routine needle biopsy after completing ABVS and 2D ultrasound. The “cloud sign” area sampling specimens were immediately taken for sampling after mastectomy. The excised complete breast tissue was localized and sampled under ultrasound guidance and sent for pathological examination. The submitted specimens exhibiting the cloud sign were processed using routine hematoxylin and eosin (HE) staining for microscopic observation. The lesions were classified as benign or malignant according to the pathological results of needle biopsy or intraoperative frozen section. According to the 2013 International Breast Cancer Conference, breast cancer was categorized into four molecular subtypes based on estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER-2), and Ki-67 status (Table S1).

Figure 1 ABVS imaging features. (A) Satellite cloud map (Source: https://www.veer.com/). (B) A 72-year-old female patient was histopathologically confirmed to have invasive breast cancer. The ABVS coronal plane revealed a patchy echogenic elevation area in the upper outer quadrant of the right breast, with indistinct boundaries, extending to the skin (the region indicated by the red arrow). Resembling the cloud-like appearance observable in satellite cloud maps. (C) A 52-year-old female patient diagnosed with invasive non-special type breast cancer. In the ABVS coronal plane, a patchy echogenic area with unclear boundaries could be observed between the breast lesion in the upper outer quadrant of the right breast and the skin, resembling the cloud-like appearance in satellite cloud maps (the region indicated by the red arrow). (D) A 38-year-old female patient was histopathologically confirmed to have ductal carcinoma in situ. The ABVS coronal plane displays multiple radiating linear echogenic strands converging toward the center around the irregular low echogenic breast lesion (the region indicated by the red arrows). (E) A 77-year-old female patient was diagnosed with mixed invasive carcinoma. The ABVS coronal plane revealed an uneven high echogenic band at the border between the low echogenic breast lesion and the adjacent adipose tissue, along with the presence of spiculations and lobulation signs (the region indicated by the red arrows). ABVS, automated breast volume scanning.

Statistical analysis

Statistical analysis was conducted using SPSS 26.0 software (IBM Corp., Armonk, NY, USA). The continuous data conforming to the normal distribution were expressed as means ± standard deviation and compared by independent sample t-test. Multiple group comparisons were performed using one-way analysis of variance (ANOVA) and the least significant difference (LSD) test. The categorical data were expressed as n (%) and analyzed using the chi-square test or the Fisher’s exact test. Multivariate Logistic regression analysis was performed to identify risk factors for “cloud sign”. Taking sensitivity as the ordinate and 1-specificity as the abscissa to draw receiver operating characteristic (ROC). Sensitivity, specificity and the area under the curve (AUC) was calculated to judge the diagnostic efficiency of different signs for benign and malignant lesions. A two-sided P<0.05 were considered statistically significant.

Results

A total of 187 breast lesions from 185 patients were finally included in this study. In the malignant group (n=123; mean age, 55.4±11.2 years old; average tumor diameter, 26.7±13.7 mm). The pathological types of masses presenting with the cloud sign included invasive non-special type carcinoma in 72 cases, lobular carcinoma in 4 cases, invasive papillary carcinoma in 6 cases, ductal carcinoma in situ in 1 case, invasive micropapillary carcinoma in 3 cases, mixed type carcinoma in 3 cases, mucinous carcinoma in 1 case, and atypical lobular hyperplasia in 1 case. In the benign group (n=64; mean age, 39.5±10.1 years; average tumor diameter, 17.9±11.5 mm), the pathological types of masses presenting with the cloud sign included sclerosing adenosis in 4 cases, chronic mastitis in 3 cases, and benign phyllodes tumor in 1 case (Table 1).

The incidence of the cloud sign in the malignant group (91/123, 74.0%) of breast lesions was significantly higher than that in the benign group (74.0% vs. 12.5%, P<0.001). The incidence of the “cloud sign” differs significantly among different age groups (P<0.001) and different tumor size group (P=0.002). In malignant breast lesions, the “cloud sign” is more frequently presented in lesions with infiltrative nature (P=0.03), positive expression of ER (P<0.001), and positive expression of PR (P=0.011). However, there was no significant difference in the incidence of the cloud sign between the lymph node metastasis group and the non-metastasis group (P=0.283). The occurrence of the cloud sign also did not show a significant difference in relation to the presence or absence of Ki-67 (P=0.975) and HER-2 expression (P=0.433) (Table 2).

ROC curves for the three signs are presented in Figure 2. The of cloud sign, convergence sign and hyperechoic halo in the diagnosis of malignant lesions were 0.807, 0.668, and 0.841, respectively. The sensitivity, specificity, and accuracy of cloud sign were 74%, 87.5%, and 78.6%, respectively. The convergence sign was present in 45.5% of malignant tumors (56/123), with a sensitivity, specificity, and accuracy of 58.5%, 92.2%, and 70%, respectively. The hyperechoic halo was present in 83.7% (103/123) malignant breast lesions, with a sensitivity, specificity, and accuracy of 83.7%, 84.4%, and 84%, respectively (Tables 2,3).

Figure 2 ROC curves for the cloud sign, convergence sign, and hyperechoic halo. Blue, hyperechoic halo; red, convergence sign; green, cloud sign. AUC, area under the curve; ROC, receiver operating characteristic.

Totally, 10 specimens from different lesions were collected from regions exhibiting the cloud sign, all presenting a gross appearance similar to yellow adipose tissue. Microscopically, 9 cases exhibited cancerous tissue within the adjacent adipose tissue, of which, three cases featured fibrous connective tissue enveloping the cancerous tissue. Significant infiltration of inflammatory cells was found in the interstitial spaces of the adipose tissue. Some specimens exhibited detached cancer tissue and normal ductal glandular tissue (Figure 3).

Figure 3 The ABVS sonographic images from different patients along with corresponding microscopic images of the lesion areas exhibiting the Cloud Sign. All sections were stained with hematoxylin and eosin. (A-C) A 52-year-old female, depicting the ABVS coronal view, microscopic observations at 4× and 20× magnification of the cloud sign, respectively. (D-F) A 54-year-old female, depicting the ABVS coronal view, microscopic observations at 4× and 20× magnification of the cloud sign, respectively. (G-I) A 54-year-old female, depicting the ABVS coronal view, microscopic observations at 4× and 20× magnification of the cloud sign, respectively. ABVS, automated breast volume scanning.

As shown in Table S2, both the convergence sign and hyperechoic halo were significantly associated with increasing patient age (P<0.001). Additionally, the hyperechoic halo became more frequent with larger tumor diameter (P<0.001).

Out of 123 cases evaluated for ER, PR, HER-2, and Ki67 via immunohistochemistry, 28.5% were classified as luminal A, 51.2% as luminal B (comprising 22 luminal B1 and 41 luminal B2), 7.3% exhibited HER-2 overexpression, and 13% were of the Basal-like subtype (triple-negative). The occurrence of the cloud sign did not exhibit statistically significant differences among various molecular subtypes (P=0.204) (Table S3).

Discussion

The cloud sign in ABVS is a distinctive sign of breast lesions, showing better diagnostic value in distinguishing benign and malignant diseases than convergence sign and hyperechoic halo. Its pathological mechanism may be related to tissue inflammation, edema, and fibrous connective tissue hyperplasia in the lesion.

The study comprised 185 female patients with 187 breast lesions. The cloud sign was found in 74.0% of malignant cases and 12.5% of benign cases. The occurrence of the cloud sign was significantly higher in patients aged 40 years and above, in tumors larger than 2 cm, and in malignant tumors. It was significantly associated with invasiveness and positive ER and PR expressions. Invasive breast cancer and PR were identified as risk factors.

The convergence sign is characterized by radial, cord-like hyperechoic structures surrounding breast lesions in the coronal plane, converging towards the lesion center. It is demonstrated to originate from the infiltration and growth of cancerous tissue into surrounding normal tissue, leading to fibrous connective tissue proliferation and tension in adjacent normal tissue (12,13). Yang et al. reported a sensitivity of 61.1%, a specificity of 97.1%, and an accuracy of 78.6% for diagnosing breast cancer using the convergence sign (15). In the present study, the occurrence rate of the convergence sign in malignant breast lesions was 45.5% (56/123), and 50% (49/112) in malignant tumors, aligning with previous study (16). However, diagnostic sensitivity, specificity, and accuracy in the present study were 58.5%, 92.2%, and 70%, respectively. It is noteworthy that the sensitivity achieved in the current study differed from that of prior studies (15,17), potentially due to variations in lesion size. The majority of malignant breast lesions in the present study were larger than 2 cm, and previous research suggested a lower probability of observing the convergence sign with larger lesions (≥2 cm) (18). Larger cancer lesions are mainly more aggressive, rapidly growing, infiltrating surrounding stromal tissues, and spreading continuously, leading to a lower frequency of the convergence sign. Additionally, the size of the probe and adjustable angles may influence the outcomes, particularly for very large tumors (>3 cm), which could be affected during three-dimensional reconstruction due to limited inclusion of surrounding normal tissues (19).

Breast cancer, characterized by its highly invasive nature, displays a distinct growth pattern compared with benign breast lesions (20). Malignant tumors exhibit infiltrative growth, releasing growth factors that recruit inflammatory cells and trigger an inflammatory response, leading to stromal reactions, inflammation in the surrounding tissue, and the formation of new blood vessels and microlymphatics (21). Benign tumors, in contrast, primarily grow expansively, applying pressure on surrounding normal tissues. Non-infiltrating tumors, which have not breached the basement membrane, lack the invasive-reactive interface, making them less likely to manifest the cloud sign (22). The results of the present study revealed a significant difference in the occurrence of the cloud sign between benign and malignant tumors. Logistic regression analysis results strongly indicated that the presence of the cloud sign was highly indicative of infiltrating breast cancer, suggesting a correlation between its presence and the biological behavior of tumors. Tumor size, a crucial indicator for predicting breast cancer prognosis, exhibited a correlation with the cloud sign. Tumors larger than 2 cm had a higher occurrence rate of the cloud sign, reaching 100% in breast lesions exceeding 5 cm. This indicates that larger tumors are associated with increased invasiveness, more likely to induce stromal reactions in the surrounding tissue (23).

Notably, ER, PR, HER-2, and Ki-67 are widely utilized clinical biological markers for breast cancer treatment and prognosis assessment (24,25). Breast tumors expressing ER and PR positivity typically demonstrate superior tumor differentiation, lower invasiveness, slower disease progression, and a favorable response to endocrine therapy, such as tamoxifen, resulting in a relatively better prognosis (26,27). Previous research correlating breast cancer ultrasound features with ER and PR expression suggests that ER and PR-positive expression is associated with irregular tumor morphology, spiculated margins, angular or lobulated shapes, and posterior acoustic enhancement (28). Studies on the peripheral high-echo halo in 2D breast ultrasound demonstrated synchronous expression with ER and PR, which was also supported by Xu et al., who considered the high-echo halo as a positive prognostic indicator (29). In the present study, the occurrence rate of the cloud sign exhibited a statistically significant difference between ER and PR-negative and -positive groups. Logistic regression analysis results revealed a correlation between PR expression level and the presence of the cloud sign, aligning with the findings of previous study (29). Therefore, it was hypothesized that the cloud sign, similar to a high-echo halo, may indicate better tumor differentiation and a more favorable treatment response. However, whether it acts as a limiting or protective factor for breast cancer warrants further investigation.

Although both the cloud sign and the hyperechoic halo show peripheral high echogenicity around breast lesions, their pathological bases differ. The hyperechoic halo, seen on conventional 2D ultrasound, often reflects compressed adipose tissue or desmoplastic reaction and has been linked to hormone receptor-positive tumors and more accurate tumor size estimation (30). In contrast, the cloud sign on ABVS coronal images covers a wider area, reflecting not only desmoplasia but also edema and inflammatory infiltration, which may explain its stronger association with tumor invasiveness and size in this study. Our findings also show that these signs vary with patient age. With increasing patient age, the detection rate of peripheral echogenic signs such as cloud sign and convergence sign tends to increase. This may be explained by age-related changes in breast composition—namely, reduced glandular tissue and increased adiposity—leading to higher contrast on ultrasound (31). Jiang et al. (30) demonstrated that 3D ultrasound features, including convergence signs, improve diagnostic accuracy in breast nodules, further highlighting the potential of advanced imaging markers like the cloud sign. Still, due to the limited sample size of this study and the inclusion of only patients aged over 40 years, the generalizability and robustness of these findings require further validation.

Currently, there is no research on the pathological mechanism underlying the cloud sign in ABVS coronal images. Based on the pathological results, we speculate that breast cancer tissue secretes inflammatory factors that induce inflammation, edema, and fibrous connective tissue hyperplasia in the surrounding tissue, resulting in an increase in the number of acoustic interfaces in the originally homogeneous adipose tissue, and an increase in reflected energy, thus appearing as high echogenicity. Therefore, a large area of the cloud sign suggests the presence of extensive inflammation and edema in the surrounding tissue of the lesion, and to some extent, it can reflect the degree of invasion of tumor tissue into the surrounding normal tissue and stromal reaction hyperplasia. In addition, the distribution characteristics of the cloud sign may be related to the anatomical structure of the breast gland and lymphatic vessels.

Our study indeed identified a significant correlation between the presence of the cloud sign and age, with a notably higher occurrence of the cloud sign in patients aged 40 years and above. This finding suggests that the cloud sign may be more commonly associated with invasive breast cancers that tend to occur in this age group, consistent with previous research on tumor biology in older women (32).

Similarly, we have now included a comparative evaluation of the correlation between the cloud sign and tumor invasiveness. As noted in our study, the cloud sign was significantly more frequent in invasive tumors compared to benign or non-invasive tumors, aligning with its role in reflecting stromal reactions and aggressive tumor behavior (33). Regarding the overlap of cloud sign with the convergence sign and hyperechoic halo, both the convergence sign and hyperechoic halo appeared significantly more frequently in lesions exhibiting the cloud sign in our analysis. This indicates a possible correlation among these features. However, the co-occurrence does not exclude the potential diagnostic value of the cloud sign as a distinct imaging characteristic. Further studies using multivariate models and larger datasets are warranted to explore the independence and combined diagnostic value of these features.

The present study has some limitations. Firstly, its retrospective design might introduce inherent biases, including potential selection bias and a reliance on historical data, impacting the generalizability and ability to establish causation. Secondly, a relatively small sample size of 187 breast lesions could compromise statistical power, particularly for rare outcomes or subgroups. Exclusion criteria, such as excluding patients with prior tumor treatment, might introduce selection bias. Additionally, this study was conducted using data from a single ABVS system, which may limit the generalizability of the findings. Therefore, the results should be interpreted with caution. Thirdly, the study lacks longitudinal data, hindering the presentations of insights into the evolution of observed ultrasound features over time. Fourthly, a key limitation is the limited number of readers involved in image interpretation, which may restrict the assessment of inter-reader consistency. Finally, the insufficient exploration of the implications of molecular subtypes on ultrasound features adds a layer of complexity to the interpretation of results. Future multi-center studies involving different ABVS systems are needed to validate the diagnostic value of the cloud sign.

In summary, the cloud sign demonstrates diagnostic value comparable to the hyperechoic halo and convergence signs. Its underlying mechanism may be associated with lesion-induced tissue inflammation, edema, and fibrous connective tissue proliferation. Based on our clinical observations, we believe the “cloud sign” holds potential clinical value in specific contexts. The presence of the “cloud sign” often correlates with lesions situated in the posterior breast tissue, where conventional imaging may be less effective. In cases lacking a well-defined mass, subtle parenchymal changes such as localized edema—manifesting as the “cloud sign”—may be the only detectable sonographic abnormality. These subtle findings are frequently missed by conventional 2D ultrasound due to its limited imaging planes. In contrast, ABVS enables full-volume cross-sectional imaging and facilitates the identification of hyperechoic abnormalities through systematic comparison with surrounding tissues. Given the limited sample size in this preliminary investigation, we were unable to conduct a robust statistical analysis of the diagnostic performance of the “cloud sign.” We fully recognize the need for further validation and plan to explore its diagnostic utility in detecting non-mass-forming breast cancers in future studies with larger, well-characterized cohorts.

Conclusions

The presence of the “cloud sign” on ABVS coronal plane imaging demonstrates diagnostic value in differentiating benign from malignant breast masses, while also showing clinical significance for assessing invasiveness and ER, PR expression status in malignant lesions. These findings provide additional theoretical basis for clinical diagnosis of breast tumors. The formation mechanism of the “cloud sign” may be attributed to tumor-induced inflammatory responses, resulting in inflammatory cell infiltration and proliferative fibrotic connective tissue changes.

Acknowledgments

We would like to thank our colleagues from both participating departments for supporting our research.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2529/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2529/dss

Funding: The study was supported by Medical Science and Technology Research Fund of Guangdong Province (Guangdong Health Science and Education Letter [2024]) (No. 2-A2024599); Guangzhou Science and Technology Program Project (2024A03J0937); Guangdong Province Undergraduate Teaching Quality and Reform Project Construction Project (Guangdong Education High Letter [2021]) (No. 29-454); Guangzhou Health Science and Technology General Guidance Project (Suiwei Science and Education [2022]) (No. 3-20231A011092); Undergraduate Teaching Quality and Teaching Reform Project of Guangzhou Medical University (Guangzhou Medical Development [2023]) (No. 166-174).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2024-2529/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by Ethics Committee of the Third Affiliated Hospital of Guangzhou Medical University (No. LCYJ-2023-064) and informed consent was obtained from all individual participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
2. Valastyan S, Weinberg RA. Tumor metastasis: molecular insights and evolving paradigms. Cell 2011;147:275-92. [Crossref] [PubMed]
3. Cardoso F, Kyriakides S, Ohno S, Penault-Llorca F, Poortmans P, Rubio IT, Zackrisson S, Senkus EESMO Guidelines Committee. Electronic address: clinicalguidelines@esmo.org. Early breast cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2019;30:1194-220. Erratum in: Ann Oncol 2019;30:1674 Erratum in: Ann Oncol 2021;32:284. [Crossref] [PubMed]
4. Jafari SH, Saadatpour Z, Salmaninejad A, Momeni F, Mokhtari M, Nahand JS, Rahmati M, Mirzaei H, Kianmehr M. Breast cancer diagnosis: Imaging techniques and biochemical markers. J Cell Physiol 2018;233:5200-13. [Crossref] [PubMed]
5. Baltzer PA, Dietzel M, Burmeister HP, Zoubi R, Gajda M, Camara O, Kaiser WA. Application of MR mammography beyond local staging: is there a potential to accurately assess axillary lymph nodes? evaluation of an extended protocol in an initial prospective study. AJR Am J Roentgenol 2011;196:W641-7. [Crossref] [PubMed]
6. Yan J, Liu Z, Du S, Li J, Ma L, Li L. Diagnosis and Treatment of Breast Cancer in the Precision Medicine Era. Methods Mol Biol 2020;2204:53-61. [Crossref] [PubMed]
7. Independent UK Panel on Breast Cancer Screening. The benefits and harms of breast cancer screening: an independent review. Lancet 2012;380:1778-86. [Crossref] [PubMed]
8. Spick C, Bickel H, Polanec SH, Baltzer PA. Breast lesions classified as probably benign (BI-RADS 3) on magnetic resonance imaging: a systematic review and meta-analysis. Eur Radiol 2018;28:1919-28. [Crossref] [PubMed]
9. Huang CS, Yang YW, Chen RT, Lo CM, Lo C, Cheng CF, Lee CS, Chang RF. Whole-Breast Ultrasound for Breast Screening and Archiving. Ultrasound Med Biol 2017;43:926-33. [Crossref] [PubMed]
10. Spear GG, Mendelson EB. Automated breast ultrasound: Supplemental screening for average-risk women with dense breasts. Clin Imaging 2021;76:15-25. [Crossref] [PubMed]
11. Schmachtenberg C, Fischer T, Hamm B, Bick U. Diagnostic Performance of Automated Breast Volume Scanning (ABVS) Compared to Handheld Ultrasonography With Breast MRI as the Gold Standard. Acad Radiol 2017;24:954-61. [Crossref] [PubMed]
12. Chen L, Chen Y, Diao XH, Fang L, Pang Y, Cheng AQ, Li WP, Wang Y. Comparative study of automated breast 3-D ultrasound and handheld B-mode ultrasound for differentiation of benign and malignant breast masses. Ultrasound Med Biol 2013;39:1735-42. [Crossref] [PubMed]
13. Spak DA, Plaxco JS, Santiago L, Dryden MJ, Dogan BE. BI-RADS® fifth edition: A summary of changes. Diagn Interv Imaging 2017;98:179-90.
14. Xu G, Han T, Yao MH, Xie J, Xu HX, Wu R. Three-dimensional ultrasonography for the prediction of breast cancer prognosis. J BUON 2014;19:643-9.
15. Yang L, Zhao YZ, Du JJ, Chen L, He Q. Diagnostic value of ABVS convergence sign combined with conventional ultrasound irregular margin for breast cancer. Chinese Journal of Ultrasound in Medicine 2017;33:587-90.
16. Tan YJ, Bao LY, Huang AQ, Zhu LQ, Han GJ, Liu J. The Relationship between the Retraction Phenomenon on Coronal Plane and Clinical Pathological Characteristics in Breast Carcinoma Patients. Chinese Journal of Ultrasound in Medicine 2015;31:587-9.
17. D'Angelo A, Rinaldi P, Belli P, D'Amico R, Carlino G, Grippo C, Giuliani M, Orlandi A, Infante A, Manfredi R. Usefulness of automated breast volume scanner (ABVS) for monitoring tumor response to neoadjuvant treatment in breast cancer patients: preliminary results. Eur Rev Med Pharmacol Sci 2019;23:225-31. [Crossref] [PubMed]
18. Xu G, Wu R, Ma F, Guo LH, Xing CY, Yao MH. Relationship of the characteristic performance of Three-dimensional ultrasound- convergence sign with prognostic index of breast cancer. Chin J Clinicians. 2013;7:5116-8.
19. Wang XY, Zhang SH. Assessment of the malignant risk of breast nodules using risk scoring method based on ultrasound images. Chinese Journal of Cancer Prevention and Treatment 2019;26:6.
20. Yokotani T, Ikeda N, Hirao T, Tanaka Y, Morita K, Fujii T, Ohbayashi C, Nakamura T, Kobayashi T, Sho M. Predictive value of tumor-infiltrating lymphocytes for pathological response to neoadjuvant chemotherapy in breast cancer patients with axillary lymph node metastasis. Surg Today 2021;51:595-604. [Crossref] [PubMed]
21. Schoppmann SF, Bayer G, Aumayr K, Taucher S, Geleff S, Rudas M, Kubista E, Hausmaninger H, Samonigg H, Gnant M, Jakesz R, Horvat RAustrian Breast and Colorectal Cancer Study Group. Prognostic value of lymphangiogenesis and lymphovascular invasion in invasive breast cancer. Ann Surg 2004;240:306-12. [Crossref] [PubMed]
22. Yoshida A, Hayashi N, Akiyama F, Yamauchi H, Uruno T, Kikuchi M, Yagata H, Tsugawa K, Suzuki K, Nakamura S, Tsunoda H. Ductal carcinoma in situ that involves sclerosing adenosis: high frequency of bilateral breast cancer occurrence. Clin Breast Cancer 2012;12:398-403. [Crossref] [PubMed]
23. Zhuo JW, He YM, Zhang ML, Ye X, Xue E, Lin L. Relationship between conventional ultrasonography combined with shear wave elastography features and lymph node metastasis in breast cancer. Chinese Journal of Ultrasonography 2018;27:709-13.
24. Hadgu E, Seifu D, Tigneh W, Bokretsion Y, Bekele A, Abebe M, Sollie T, Merajver SD, Karlsson C, Karlsson MG. Breast cancer in Ethiopia: evidence for geographic difference in the distribution of molecular subtypes in Africa. BMC Womens Health 2018;18:40. [Crossref] [PubMed]
25. Loibl S, Poortmans P, Morrow M, Denkert C, Curigliano G. Breast cancer. Lancet 2021;397:1750-69. Erratum in: Lancet 2021;397:1710. [Crossref] [PubMed]
26. Mirmalek SA, Hajilou M, Salimi Tabatabaee SA, Parsa Y, Yadollah-Damavandi S, Parsa T. Prevalence of HER-2 and Hormone Receptors and P53 Mutations in the Pathologic Specimens of Breast Cancer Patients. Int J Breast Cancer 2014;2014:564308. [Crossref] [PubMed]
27. Aho M, Irshad A, Ackerman SJ, Lewis M, Leddy R, Pope TL, Campbell AS, Cluver A, Wolf BJ, Cunningham JE. Correlation of sonographic features of invasive ductal mammary carcinoma with age, tumor grade, and hormone-receptor status. J Clin Ultrasound 2013;41:10-7. [Crossref] [PubMed]
28. Irshad A, Leddy R, Pisano E, Baker N, Lewis M, Ackerman S, Campbell A. Assessing the role of ultrasound in predicting the biological behavior of breast cancer. AJR Am J Roentgenol 2013;200:284-90. [Crossref] [PubMed]
29. Xu J, Li F, Chang F. Correlation of the ultrasound imaging of breast cancer and the expression of molecular biological indexes. Pak J Pharm Sci 2017;30:1425-30.
30. Jiang X, Chen C, Yao J, Wang L, Yang C, Li W, Ou D, Jin Z, Liu Y, Peng C, Wang Y, Xu D. A nomogram for diagnosis of BI-RADS 4 breast nodules based on three-dimensional volume ultrasound. BMC Med Imaging 2025;25:48. [Crossref] [PubMed]
31. Ohmaru A, Maeda K, Ono H, Kamimura S, Iwasaki K, Mori K, Kai M. Age-related change in mammographic breast density of women without history of breast cancer over a 10-year retrospective study. PeerJ 2023;11:e14836. [Crossref] [PubMed]
32. Zhang Q, Ma B, Kang M. A retrospective comparative study of clinicopathological features between young and elderly women with breast cancer. Int J Clin Exp Med 2015;8:5869-75.
33. Kim SH, Seo BK, Lee J, Kim SJ, Cho KR, Lee KY, Je BK, Kim HY, Kim YS, Lee JH. Correlation of ultrasound findings with histology, tumor grade, and biological markers in breast cancer. Acta Oncol 2008;47:1531-8. [Crossref] [PubMed]