RPS 202 - Beyond the basics: the new era of breast ultrasound

Author Block: T. Uematsu, U. G. T. C. (. R. C. Japan Association of Breast and Thyroid Sonology; Shizuoka/JP
Purpose: The sonographic glandular tissue component (GTC) has emerged as a potential imaging biomarker for stratifying breast cancer risk in women with dense breasts. Anticipating its inclusion in the upcoming BI-RADS 6th edition, this study aimed to assess the reliability and clinical utility of GTC classification and sonographic breast composition through a multi-institutional observer performance study.
Methods or Background: Fifteen readers (9 physicians, 6 sonographers) from 12 institutions independently assessed 100 paired ultrasound images twice, 1 month apart. GTC was qualitatively classified into four categories by fibroglandular tissue proportion (minimal, mild, moderate, marked); minimal/mild were grouped as low GTC and moderate/marked as high GTC. Breast composition was categorized as dense (high GTC), nondense (fat lobules visible), or intermediate, standardized by an FGT-to-FAT (subcutaneous and retromammary fat) ratio cutoff of 1. Inter- and intra-observer agreement was evaluated using κ statistics.
Results or Findings: Inter-observer agreement for GTC classifications was substantial for both the four-category (κ = 0.66 and 0.73 for the first and second readings, respectively) and dichotomized (κ = 0.61 and 0.64) schemes. Agreement for sonographic breast composition was similarly substantial (κ = 0.65 and 0.67). Intra-observer agreement was also substantial for GTC (four-category κ = 0.78; two-category κ = 0.75) and for breast composition (κ = 0.77).
Conclusion: Both inter- and intra-observer agreement were substantial, indicating that qualitative assessment of GTC and sonographic breast composition is reproducible and reliable. These findings support the integration of GTC into routine clinical practice and research as a practical tool for risk stratification in women with dense breasts.
Limitations: First, its retrospective design may have introduced selection bias. Second, evaluations were based on static images, which do not fully capture the dynamic information available in real-time ultrasound.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This multi-institutional study was approved by the Institutional Review Board of our hospital (Approval No. T2024-11-2024-10-3).

Author Block: E. Jain, R. BANSAL, Z. Khan, B. Aggarwal; New Delhi/IN
Purpose: Non-mass lesions (NMLs) are frequently encountered in breast imaging but are not formally defined in the 5th edition of the BI-RADS ultrasound lexicon. This study aimed to characterise their ultrasound features, evaluate corresponding findings on mammography and MRI when available, and identify predictors of malignancy through clinical and pathological correlation.
Methods or Background: This prospective single-centre study included 120 women with non-mass lesions (NMLs) detected on ultrasound between December 2024 and August 2025. Ultrasound served as the index test, with features assessed according to the Japan Association of Breast and Thyroid Sonology (JABTS) criteria, classifying NMLs as ductal abnormalities, hypoechoic areas, architectural distortion, microcysts, or echogenic foci without a hypoechoic area. Histopathology of ultrasound-guided biopsies was the reference standard, with malignant lesions further characterised by immunohistochemistry (ER, PR, HER2, Ki-67). Benign and malignant NMLs were compared for demographics, clinical presentation, and imaging features. Logistic regression was used to identify independent predictors, and receiver operating characteristic (ROC) analysis compared the diagnostic accuracy of BI-RADS with the predictive model.
Results or Findings: Of 120 NMLs, 62.5% were malignant and 37.5% benign. Independent predictors of malignancy were posterior shadowing (p < 0.001), echogenic foci (p = 0.008), and architectural distortion (p = 0.006), while hypoechoic areas were highly sensitive but non-specific. Malignancy rates increased progressively across BI-RADS categories, with BI-RADS ≥4B yielding a sensitivity of 75%, specificity of 80%, and accuracy of 77%. The regression-based predictive model demonstrated superior diagnostic performance (AUC 0.87) compared with BI-RADS alone (AUC 0.78).
Conclusion: Posterior shadowing, echogenic foci, and architectural distortion are key predictors of malignancy in NMLs. As these lesions remain undefined in the BI-RADS lexicon, this study underscores the need for standardised descriptors and integrated imaging to improve diagnostic accuracy.
Limitations: Not applicable.
Funding for this study: No funding was provided for this study.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: The ethics committee UID number is 27731.

Author Block: Z. h. Temiz, M. A. Nazli; Istanbul/TR
Purpose: Non-mass breast lesions present diagnostic difficulties due to their subtle imaging features and varied pathology. Although 14-gauge (14G) core needle biopsy (CNB) is standard, 12-gauge (12G) devices may offer better tissue yield. This study aims to compare the diagnostic performance of 12G and 14G CNB in non-mass breast lesions, focusing on tissue yield, sample adequacy, radiology–pathology concordance, upgrade rates of high-risk lesions (B3).
Methods or Background: This retrospective study analyzed 106 non-mass breast lesions (12G: 38; 14G: 68) detected by mammography, ultrasound, or MRI. Biopsies were ultrasound and histopathology was categorized as benign, high-risk, or malignant. Evaluated outcomes included radiology–pathology concordance, calcification, upgrade after surgery/re-biopsy, and management. Statistical analysis used Fisher’s exact test with ORs and 95% CI; p<0.05 was significant.
Results or Findings: Radiology–pathology discordance was lower in the 12G group (10.5%) compared to the 14G group (20.6%) (OR=0.45, p=0.23). Calcification sampling was significantly higher with 12G (39.5%) versus 14G (17.6%) (p=0.02). Upgrade rates were similar: 12G 13.2% vs 14G 10.3% (OR=1.33, p=0.75). Among calcification-positive lesions, discordance was lower in the 12G group (13.3%) than in the 14G group (33.3%) (OR=0.31, p=0.36), and upgrade rates were also lower (20.0% vs 41.7%; OR=0.35, p=0.40).
Conclusion: The 12G core needle biopsy demonstrated a higher success rate in sampling calcifications and showed a trend toward lower radiology–pathology discordance compared to 14G device, although the difference was not statistically significant. While upgrade rates remained similar between the two groups, these preliminary findings indicate that the 12G needle may provide more reliable sampling, particularly in calcification-rich non-mass breast lesions.
Limitations: This study is limited by its retrospective single-center design, small sample size, and incomplete follow-up, which may affect the accuracy of upgrade rates. Additionally, needle selection bias could have influenced group distribution.
Funding for this study: This study received no funding.
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Ethics approval was obtained from the IRB of Başakşehir Çam and Sakura City Hospital.

Author Block: A. K. K. Soydan, M. Demir; Istanbul/TR
Purpose: To evaluate the cumulative diagnostic adequacy of sequential 14-gauge tru-cut biopsy samples in breast lesions visible on ultrasonography and determine the optimal number of core samples required for diagnostic accuracy.
Methods or Background: Prospective study conducted June-December 2024 including 86 patients with 86 breast lesions (BI-RADS 3, 4, or 5). Three sequential core biopsy samples were obtained from each lesion under US guidance using 14-G automated biopsy needle. Each sample was sent separately for pathological evaluation. Cumulative diagnostic adequacy was assessed for one, two, and three samples.
Results or Findings: Cumulative diagnostic adequacy: 84.9% (73/86) with first core, 94.2% (81/86) with two cores, 98.8% (85/86) with three cores. All malignant lesions (n=15) were diagnosed with first two samples (100%). Diagnostic adequacy increase was significant from first to second sample (p<0.05) but not from second to third (p>0.05). First core diagnostic rate was higher for solid lesions (87.3%) versus cystic lesions (50.0%) (p=0.049).
Conclusion: Three core samples in US-guided tru-cut biopsy provide very high diagnostic adequacy (98.8%). Two high-quality core samples may be sufficient for diagnosing malignant lesions, suggesting biopsy protocols could be optimized to be less invasive without compromising oncologic safety.
Limitations: Single-center study limiting generalizability. Small sample size of malignant lesions (n=15). Single pathologist evaluation precluding inter-observer reliability assessment. Benign lesion reference standard included follow-up rather than surgical excision in all cases.
Funding for this study: No
Not specified
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Study approved by Clinical Research Ethics Committee of Ümraniye Training and Research Hospital. Conducted according to Declaration of Helsinki principles. Written informed consent obtained from all participants.

Author Block: U. Chouksey, P. B. Thakkar, A. Katdare, P. Haria, N. Shetty, S. Kulkarni; Mumbai/IN
Purpose: Evaluation of qualitative and quantitative characteristics of Shear wave elastography (SWE) among malignant breast lesions according to their receptor status.
Methods or Background: This study was a subset analysis of a larger prospective observational study including 468 lesions in 404 patients. Among these, hormonal and molecular receptor statuses were available for 388 malignant cases. These lesions were broadly categorized into three groups, i.e. first group being receptor positive, that is either ER or PR positive, the second group being HER – 2 enriched and the
third group being triple negative. The mean values along with measures of distribution for Emax and Emean was calculated for each category. Qualitative characteristics were also evaluated, like presence of perilesional stiff rim.

These lesions were accordingly differently classified based on the qualitative internal stiffness and perilesional stiffness into four categories: perilesional stiff rim present and qualitatively soft, perilesional stiff rim present and qualitatively hard, perilesional stiff rim absent and qualitatively soft, and perilesional stiff rim absent and qualitatively hard.
Results or Findings: Majority of the lesions with triple negative receptor status showed a perilesional stiff ring and were internally soft. While 90 percent of the lesions in the HER-2 subgroup demonstrated a perilesional stiff ring (including both hard and soft lesions). No such trend was observed in the receptor-positive subgroup.
Conclusion: 1. A significant difference was found among the values of Emean among the three classifications of molecular and hormonal subtypes but not among Emax values.
2. There was significant association of the perilesional ring present, qualitative soft subset with both TNBC and HER-2 enriched subset. There was no significant association of any of the four categories with the Receptor positive subgroup.
Limitations: Evaluation of qualitative parameters suffered from observer bias.
Funding for this study: None
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This was a subset analysis of a prospective observational study, approved by the institutional ethics committee.

Author Block: M. Renda, C. Catalano, G. Bonito, C. Solito, F. Raponi, V. Dolcetti, G. Maglia, G. Lo Conte, V. Cantisani; Rome/IT
Purpose: The purpose of our study was to compare the diagnostic performance of the new ultrasound techniques in the differentiation between benign and malignant breast lesions, in accordance to the BI-RADS classification, evaluating the agreement between the software “Live Breast Assist”, human investigator, and the diagnostic accuracy of Elastography performed with Strain Ratio (SR), 2D-Shear Wave (STE) Point Shear Wave (STQ), and advanced Doppler technique MV-Flow. Inter-operator variability will also be analyzed.
Methods or Background: 92 patients have been enrolled, with a total of 110 focal breast lesions evaluated by an experienced breast imaging radiologist who assigned BI-RADS categories.
All patients were analyzed using Strain Elastography (SE), Shear Wave Elastography (STE),using B-mode ultrasound, color Doppler, MV-FLOW, semi-automatic “Live Breast Assist” software analysis.
All patients underwent biopsy, with histopathology serving as the gold standard.
Results or Findings: A total of 110 breast lesions were analyzed (14% benign, 7.6% B3, 78.4% malignant), with fibroadenomas and fibrocystic mastopathy being the most common benign findings, and infiltrating ductal carcinoma the predominant malignancy. The Live Breast Assist software demonstrated excellent agreement with the operator in assessing size, margins, orientation, and echostructure. Concordance with histological BI-RADS classification was high, while operator–histology agreement was good. Elastography identified a Strain Ratio cut-off of 1.89, with strain elastography outperforming 2D SWE and pSWE.
The MV FLOW software, which provides advantages over conventional color Doppler, has demonstrated that higher measured values in breast lesions are correlated with the histopathological grade of lesion severity
Conclusion: New advanced ultrasound techniques show promising results for differentiating breast lesions, potentially reducing the need for biopsies, and serving as a support tool in discriminating a suspicious lesion.
Limitations: The limitations of this study include the small sample size, which, however, can be readily increased.
Funding for this study: No
Has your study been approved by an ethics committee? Not applicable
Ethics committee - additional information:

Author Block: Z. L. Wang; Beijing/CN
Purpose: To develop a deep learning radiomics nomogram (DLRN) that integrated B-mode ultrasound (BMUS) and contrast-enhanced ultrasound (CEUS) images for preoperative lymphovascular invasion (LVI) prediction in invasive breast cancer (IBC).
Methods or Background: 981 patients with IBC from three hospitals were retrospectively enrolled. Of 834 patients recruited from Hospital I, 688 were designated as the training cohort and 146 as the internal test cohort, whereas 147 patients from Hospitals II and III were assigned to constitute the external test cohort. Deep learning and handcrafted radiomics features of BMUS and CEUS images were extracted from breast cancer to construct a deep learning radiomics (DLR) signature. The DLRN was developed by integrating the DLR signature and independent clinicopathological parameters. The performance of the DLRN was evaluated with respect to discrimination, calibration, and clinical benefit.
Results or Findings: The DLRN exhibited good performance in predicting LVI, with areas under the receiver operating characteristic curves (AUCs) of 0.885 (95% confidence interval [CI], 0.858-0.912), 0.914 (95% CI, 0.868-0.960) and 0.914 (95% CI, 0.867-0.960) in the training, internal test, and external test cohorts, respectively. The DLRN exhibited good stability and clinical practicability, as demonstrated by the calibration curve and decision curve analysis. In addition, the DLRN outperformed the traditional clinical model and the DLR signature for LVI prediction in the internal and external test cohorts (all P < 0.05).
Conclusion: The DLRN exhibited good performance in predicting LVI, representing a non-invasive approach to preoperatively determining LVI status in IBC.
Limitations: This study only included patients with unifocal IBC due to the difficulty in identifying the lesion causing ALN metastasis in multifocal and multicentric lesions. The DLRN was developed based on static breast CEUS images rather than CEUS videos due to the retrospective study.
Funding for this study: This study has received funding by National Key Research and Development Program of China (2023YFC2414203), and National Natural Science Foundation of China (No. 82371972).
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: This study was authorized by the institutional review board of the Hospital and adhered to the Declaration of Helsinki (S2024-801-01).

Author Block: J. T. Lee¹, B. K. Seo¹, J. Y. Lee¹, M. S. Bae¹, Y. Moon¹, K. R. Cho², O. Woo², S. E. Song², S-Y. Kim²; ¹Ansan/KR, ²Seoul/KR
Purpose: Breast cancer is the most common female malignancy worldwide, and ultrasound (US) is widely used for diagnosis and screening. While its clinical utility is well established, the environmental impact has rarely been assessed. This study compared the energy consumption of automated breast ultrasound (ABUS) and handheld ultrasound (HHUS) systems and explored strategies for optimized ultrasound management.
Methods or Background: In this prospective study, one ABUS and two HHUS systems (HHUS-1 and HHUS-2) from different vendors were evaluated using a power meter (1-Hz sampling). A total of 240 patients without prior breast surgery underwent B-mode US (80 ABUS, 80 HHUS-1, 80 HHUS-2). For HHUS, four breast radiologists (16–26 years of experience) each performed 20 examinations per system. We measured average power (kW) during active, idle, and off modes, scan time (min), energy use (kWh), and carbon emissions (kgCO₂e). Group comparisons were performed using the Mann-Whitney U test.
Results or Findings: Compared with HHUS-1, ABUS required 28%, 35%, and 100% less average power in active, idle, and off modes, respectively, and 24% shorter scan time, resulting in a 46% reduction in energy consumption and carbon emissions (all p < 0.05). Compared with HHUS-2, ABUS showed similar active-mode power but 26% and 100% reductions in idle and off modes and 18% shorter scan time. These changes yielded a 7% overall reduction in energy and emissions; the decreases in idle, off, and scan time were statistically significant (p < 0.05). Reducing idle-mode duration further lowered energy consumption by 12–16%.
Conclusion: ABUS demonstrated lower power demand, energy consumption, and carbon emissions compared with HHUS. Optimizing ultrasound workflows, particularly minimizing idle time, may improve environmental sustainability while maintaining clinical utility.
Limitations: Not applicable
Funding for this study: National Research Foundation of Korea funded by the Korea government (RS-2024-00347290)
Has your study been approved by an ethics committee? Yes
Ethics committee - additional information: Not applicable because no patient data are involved