Background echotexture classification in breast ultrasound: inter-observer agreement study

Accuracy of breast ultrasound image analysis software in feature analysis: a comparative study with sonographers

Breast ultrasound is recommended for early breast cancer detection in China, but the rapid increase in imaging data burdens sonographers. This study evaluated the agreement between artificial intelligence (AI) software and sonographers in analyzing breast nodule features. Breast ultrasound images from two hospitals in Shanghai were analyzed by both the software and the sonographers for features including echotexture, echo pattern, orientation, shape, margin, calcification, and posterior echo attenuation. Agreement between software and sonographers was compared using the proportion of agreement and Kappa, with analysis time also evaluated. A total of 493 images were analyzed. The proportion of agreement between software and sonographers in assessing features was 80.5% for echotexture, 84.4% for echo pattern, 93.7% for orientation, 85.8% for shape, 88.6% for margin, 80.5% for calcification, and 90.5% for posterior echo attenuation, highlighting software's high accuracy. Cohen's kappa for other features indicated moderate to substantial agreement (0.411-0.674), with calcification showing fair agreement (0.335). The software significantly reduced analysis time compared to sonographers (P < 0.001). The software showed high accuracy and time efficiency. AI software presents a viable solution for reducing sonographers' workload and enhance healthcare in underserved areas by automating feature analysis in breast ultrasound images.

Nonmass Lesions on Breast US: An International Perspective on Clinical Use and Outcomes

Nonmass lesions (NMLs) on breast US are defined as discrete areas of altered echotexture compared to surrounding breast tissue and lack the 3-dimensionality of a mass. They are not a component of American College of Radiology BI-RADS, but they are a finding type included in the Japan Association of Breast and Thyroid Sonology lexicon. Use of the NML finding is routine in many Asian practices, including the Samsung Medical Center and Seoul National University Hospital, and their features and outcomes have been investigated in multiple studies. Nonmass lesions are most often observed when US is used to evaluate mammographic asymmetries, suspicious calcifications, and nonmass enhancement on MRI and contrast-enhanced mammography. Nonmass lesions can be described by their echogenicity, distribution, presence or absence of associated calcifications, abnormal duct changes, architectural distortion, posterior shadowing, small cysts, and hypervascularity. Malignant lesions, especially ductal carcinoma in situ, can manifest as NMLs on US. There is considerable overlap between the US features of benign and malignant NMLs, and they also must be distinguished from normal variants. The literature indicates that NMLs with linear or segmental distribution, associated calcifications, abnormal duct changes, posterior shadowing, and hypervascularity are suggestive of malignancy, whereas NMLs with only interspersed small cysts are usually benign fibrocystic changes. In this article, we introduce the concepts of NMLs, illustrate US features suggestive of benign and malignant etiologies, and discuss our institutional approach for evaluating NMLs and an algorithm that we use to guide interpretation in clinical practice.

Comparison of the diagnostic efficacy of mathematical models in distinguishing ultrasound imaging of breast nodules

This study compared the diagnostic efficiency of benign and malignant breast nodules using ultrasonographic characteristics coupled with several machine-learning models, including logistic regression (Logistics), partial least squares discriminant analysis (PLS-DA), linear support vector machine (Linear SVM), linear discriminant analysis (LDA), K-nearest neighbor (KNN), artificial neural network (ANN) and random forest (RF). The clinical information and ultrasonographic characteristics of 926 female patients undergoing breast nodule surgery were collected and their relationships were analyzed using Pearson's correlation. The stepwise regression method was used for variable selection and the Monte Carlo cross-validation method was used to randomly divide these nodule cases into training and prediction sets. Our results showed that six independent variables could be used for building models, including age, background echotexture, shape, calcification, resistance index, and axillary lymph node. In the prediction set, Linear SVM had the highest diagnosis rate of benign nodules (0.881), and Logistics, ANN and LDA had the highest diagnosis rate of malignant nodules (0.910~0.912). The area under the ROC curve (AUC) of Linear SVM was the highest (0.890), followed by ANN (0.883), LDA (0.880), Logistics (0.878), RF (0.874), PLS-DA (0.866), and KNN (0.855), all of which were better than that of individual variances. On the whole, the diagnostic efficacy of Linear SVM was better than other methods.

Analysis of background echotexture on automated breast ultrasound using BI-RADS and modified classification: Association with clinical features and mammographic density

PURPOSE

To analyze BE on ABUS using BI-RADS and a modified classification in association with mammographic density and clinical features.

METHODS

Menopausal status, parity, and family history of breast cancer were collected for 496 women who underwent ABUS and mammography. Three radiologists independently reviewed all ABUS BE and mammographic density. Statistical analyses including kappa statistics (κ) for interobserver agreement, Fisher's exact test, and univariate and multivariate multinomial logistic regression were performed.

RESULTS

BE distribution between the two classifications and between each classification and mammographic density were associated (P < 0.001). BI-RADS homogeneous-fibroglandular (76.8%) and modified heterogeneous BE (71.3%, 75.7%, and 87.5% of mild, moderate, and marked heterogeneous background echotexture, respectively) tended to be dense. BE was correlated between BI-RADS homogeneous-fat and modified homogeneous background (95.1%) and between BI-RADS homogeneous-fibroglandular or heterogeneous (90.6%) and modified heterogeneous (86.9%) (P < 0.001). In multinomial logistic regression, age < 50 years was independently associated with heterogeneous BE (OR, 8.89, P = 0.003, in BI-RADS; OR, 3.74; P = 0.020 in modified classification).

CONCLUSION

BI-RADS homogeneous-fat and modified homogeneous BE on ABUS was likely to be mammographically fatty. However, BI-RADS homogeneous-fibroglandular or heterogeneous BE might be classified as any modified BE. Younger age was independently associated with heterogeneous BE.

Comparison of the background echotexture between automated breast ultrasound and handheld breast ultrasound

This study aimed to compare the background echotexture (BE) between automated breast ultrasound (ABUS) and handheld breast ultrasound (HHUS) and evaluate the correlation of BE with mammographic (MG) density and background parenchymal enhancement (BPE) on magnetic resonance imaging (MRI). A total of 212 women with newly diagnosed breast cancer who had undergone preoperative ABUS, HHUS, MG, and MRI were included. Two breast radiologists blinded to the menopausal status analyzed the BE of the contralateral breasts of the patients with breast cancer in consensus. The MG density and BPE of breast MRI on the radiologic reports were compared with the BE in the ultrasound. We used the cumulative link mixed model to compare the BE and Spearman rank correlation to evaluate the association between BE with MG density and BPE. BE was more heterogeneous in ABUS than in HHUS (P < .001) and in the premenopausal group than in the postmenopausal group (P < .001). The heterogeneity of BE in the premenopausal group was higher with ABUS than with HHUS (P = .013). BE and MG density showed a moderate correlation in the postmenopausal group, but a weak correlation in the premenopausal group. BE and BPE showed moderate correlations only in the premenopausal group. ABUS showed a more heterogeneous BE, especially in the premenopausal group. Therefore, more attention is required to interpret ABUS screening in premenopausal women.

Evaluation of Computer-Aided Detection (CAD) in Screening Automated Breast Ultrasound Based on Characteristics of CAD Marks and False-Positive Marks

The present study evaluated the effectiveness of computer-aided detection (CAD) system in screening automated breast ultrasound (ABUS) and analyzed the characteristics of CAD marks and the causes of false-positive marks. A total of 846 women who underwent ABUS for screening from January 2017 to December 2017 were included. Commercial CAD was used in all ABUS examinations, and its diagnostic performance and efficacy in shortening the reading time (RT) were evaluated. In addition, we analyzed the characteristics of CAD marks and the causes of false-positive marks. A total of 1032 CAD marks were displayed based on the patient and 534 CAD marks on the lesion. Five cases of breast cancer were diagnosed. The sensitivity, specificity, PPV, and NPV of CAD were 60.0%, 59.0%, 0.9%, and 99.6% for 846 patients. In the case of a negative study, it was less time-consuming and easier to make a decision. Among 530 false-positive marks, 459 were identified clearly for pseudo-lesions; the most common cause was marginal shadowing, followed by Cooper’s ligament shadowing, peri-areolar shadowing, rib, and skin lesions. Even though CAD does not improve the performance of ABUS and a large number of false-positive marks were detected, the addition of CAD reduces RT, especially in the case of negative screening ultrasound.

The lesion detection efficacy of deep learning on automatic breast ultrasound and factors affecting its efficacy: a pilot study

OBJECTIVES

The aim of this study was to investigate the detection efficacy of deep learning (DL) for automatic breast ultrasound (ABUS) and factors affecting its efficacy.

METHODS

Females who underwent ABUS and handheld ultrasound from May 2016 to June 2017 (N = 397) were enrolled and divided into training (n = 163 patients with breast cancer and 33 with benign lesions), test (n = 57) and control (n = 144) groups. A convolutional neural network was optimized to detect lesions in ABUS. The sensitivity and false positives (FPs) were evaluated and compared for different breast tissue compositions, lesion sizes, morphologies and echo patterns.

RESULTS

In the training set, with 688 lesion regions (LRs), the network achieved sensitivities of 93.8%, 97.2% and 100%, based on volume, lesion and patient, respectively, with 1.9 FPs per volume. In the test group with 247 LRs, the sensitivities were 92.7%, 94.5% and 96.5%, respectively, with 2.4 FPs per volume. The control group, with 900 volumes, showed 0.24 FPs per volume. The sensitivity was 98% for lesions > 1 cm³, but 87% for those ≤1 cm³ (p < 0.05). Similar sensitivities and FPs were observed for different breast tissue compositions (homogeneous, 97.5%, 2.1; heterogeneous, 93.6%, 2.1), lesion morphologies (mass, 96.3%, 2.1; non-mass, 95.8%, 2.0) and echo patterns (homogeneous, 96.1%, 2.1; heterogeneous 96.8%, 2.1).

CONCLUSIONS

DL had high detection sensitivity with a low FP but was affected by lesion size.

ADVANCES IN KNOWLEDGE

DL is technically feasible for the automatic detection of lesions in ABUS.

Assessment of MRI-Detected Breast Lesions: A Benign Correlate on Second-Look Ultrasound Can Safely Exclude Malignancy

Introduction

Due to the increasing use of dynamic breast MRI and the limited availability of MR-guided interventions, MRI-detected lesions usually undergo a second-look ultrasound (SLUS). We investigated the safety of a negative SLUS and a benign SLUS correlate in excluding malignant and high-risk lesions (B3) and evaluated criteria for the rate of detection on SLUS.

Methods

In the retrospective analysis, all breast MRIs performed between 2011 and 2013 were screened for newly detected lesions. We analyzed the SLUS detection rate dependent on breast density, mass character, lesion size, and histology. We calculated the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of a negative and benign SLUS for malignant lesions (B5) and lesions requiring surgical excision (including high-risk and B5 lesions).

Results

We successfully correlated 110 of 397 lesions. The detection rate was significantly higher for mass than for non-mass lesions and correlated with lesion size for mass lesions only. Lesions without/with a benign SLUS correlate were more frequently benign (including B3) or required no further procedure (B2). The sensitivity of SLUS in the detection of B3 and B5 lesions was 58%, and 73% in the detection of B5 lesions. The NPV of a negative or benign SLUS for B3 and B5 lesions was 89%, and 96% for B5 lesions.

Discussion

SLUS is a safe diagnostic tool for the management of MRI-detected lesions and can spare patients from undergoing invasive procedures.

Evidence and assessment of parenchymal patterns of ultrasonography for breast cancer detection among Chinese women: a cross-sectional study

Background

Screening of breast cancer in asymptomatic women is important to evaluate for early diagnosis. In China ultrasound is a more frequently used method than mammography for the detection of breast cancer. The objectives of the study were to provide evidence and assessment of parenchymal patterns of ultrasonography for breast cancer detection among Chinese women.

Methods

Breast ultrasound examinations including the parenchymatous pattern of cytopathological confirmed breast cancer (n = 541) and age-matched cytopathological not confirmed breast cancer (n = 849) women were retrospectively reviewed by seven sonographer physicians. According to compositions of ducts, the thickness of the breast, diameter of ducts, fat lobules, and fibro glandular tissues, the breast parenchymatous pattern was categorized into heterogeneous (high percentage of fatty tissues), ductal (the inner diameters of ducts > 50% of the thick mass of the breast), mixed (the inner diameters of ducts was 50% of the thick mass of the breast), and fibrous categories (a dense classification of the breast).

Results

Heterogeneous (p < 0.0001, OR = 3.972) and fibrous categories (p < 0.0001, OR = 2.702) were higher among women who have cytopathological confirmed breast cancer than those who have not cytopathological confirmed breast cancer. The heterogeneous category was high-risk ultrasonographic examination category followed by the fibrous category. Agreements between sonographer physicians for categories of ultrasonic examinations were fair to good (Cohen’s k = 0.591).

Conclusions

Breast cancer risk in Chinese asymptomatic women differ according to the ultrasonographic breast parenchymal pattern.

Level of Evidence: III.

Technical efficacy stage: 2.

Glandular Tissue Component and Breast Cancer Risk in Mammographically Dense Breasts at Screening Breast US

Background Breast density at mammography is an established risk factor for breast cancer, but it cannot be used to distinguish between glandular and fibrous tissue. Purpose To evaluate the association between the glandular tissue component (GTC) at screening breast US and the risk of future breast cancer in women with dense breasts and the association between the GTC and lobular involution. Materials and Methods Screening breast US examinations performed in women with no prior history of breast cancer and with dense breasts with negative findings from mammography from January 2012 to December 2015 were retrospectively identified. The GTC was reported as being minimal, mild, moderate, or marked at the time of the US examination. In women who had benign breast biopsy results, the degree of lobular involution in normal background tissue was categorized as not present, mild, moderate, or complete. The GTC-related breast cancer risk in women with a cancer diagnosis or follow-up after 6 months was estimated by using Cox proportional hazards regression. Cumulative logistic regression was used to evaluate the association between the GTC and lobular involution. Results Among 8483 women (mean age, 49 years ± 8 [standard deviation]), 137 developed breast cancer over a median follow-up time of 5.3 years. Compared with a minimal or mild GTC, a moderate or marked GTC was associated with an increased cancer risk (hazard ratio, 1.5; 95% CI: 1.05, 2.1; P = .03) after adjusting for age and breast density. The GTC had an inverse association with lobular involution; women with no, mild, or moderate involution had greater odds (odds ratios of 4.9 [95% CI: 1.5, 16.6], 2.6 [95% CI: 0.95, 7.2], and 1.8 [95% CI: 0.7, 4.6], respectively) of a moderate or marked GTC than those with complete involution (P = .004). Conclusion The glandular tissue component was independently associated with the future breast cancer risk in women with dense breasts and reflects the lobular involution. It should be considered for risk stratification during screening breast US. © RSNA, 2021 Online supplemental material is available for this article.

Automated breast US as the primary screening test for breast cancer among East Asian women aged 40-49 years: a multicenter prospective study

OBJECTIVES

To prospectively evaluate the diagnostic performance of screening ABUS as the primary screening test for breast cancer among Korean women aged 40-49 years.

METHODS

This prospective, multicenter study included asymptomatic Korean women aged 40-49 years from three academic centers between February 2017 and October 2019. Each participant underwent ABUS without mammography, and the ABUS images were interpreted at each hospital with double-reading by two breast radiologists. Biopsy and at least 1 year of follow-up was considered the reference standard. Diagnostic performance of ABUS screening and subgroup analyses according to patient and tumor characteristics were evaluated.

RESULTS

Reference standard data were available for 959 women. The recall rate was 9.8% (95% confidence interval [CI]: 7.9%, 11.7%; 94 of 959 women) and the cancer detection yield was 5.2 per 1000 women (95% CI: -0.6, 11.1; 5 of 959 women). There was only one interval cancer. The sensitivity was 83.3% (95% CI: 53.5%, 100%; 5 of 6 cancers) and the specificity was 90.7% (95% CI: 88.8%, 92.5%; 864 of 95. women). The positive predictive values of biopsies performed (PPV₃) was 20.0% (95% CI: 4.3%, 35.7%; 5 of 25 women). Women with heterogeneous background echotexture had a higher recall rate (p = .009) and lower specificity (p = .036). Women with body mass index values < 25 kg/m² had a higher mean recall rate (p = .046).

CONCLUSION

In East Asia, screening automated breast US may be an alternative to screening mammography for detecting breast cancers in women aged 40-49 years.

KEY POINTS

• Automated breast US screening for breast cancer in asymptomatic women aged 40-49 is effective with 5.2 per 1000 cancer detection yield. • Women with heterogeneous background echotexture had a higher recall rate and lower specificity. • Women with body mass index < 25 kg/m² had a higher recall rate.

Relationship Between Breast Ultrasound Background Echotexture and Magnetic Resonance Imaging Background Parenchymal Enhancement and the Effect of Hormonal Status Thereon

We studied the relationship between breast ultrasound background echotexture (BET) and magnetic resonance imaging (MRI) background parenchymal enhancement (BPE) and whether this relationship varied with hormonal status and amount of fibroglandular tissue (FGT) on MRI. Two hundred eighty-three Korean women (52.1 years; range = 27-79 years) with newly diagnosed primary breast cancer who underwent preoperative breast ultrasound and MRI were retrospectively studied. Background echotexture, BPE, and FGT were classified into 4 categories, and age, menopausal status, menstrual cycle regularity, and menstrual cycle stage at MRI were recorded. Background echotexture and BPE relationship was assessed overall, and in menopausal, FGT, menstrual cycle regularity, and menstrual cycle stage subgroups. Background echotexture and BPE correlated in women overall, and menopausal, FGT, and menstrual cycle subgroups and those in the first half of the cycle (all P < 0.001). Background echotexture reflects BPE, regardless of menopausal status, menstrual cycle regularity, and FGT and may be a biomarker of breast cancer risk.