Shared decision-making for supporting women's decisions about breast cancer screening.

BACKGROUND: In breast cancer screening programmes, women may have discussions with a healthcare provider to help them decide whether or not they wish to join the breast cancer screening programme. This process is called shared decision-making (SDM) and involves discussions and decisions based on the evidence and the person's values and preferences. SDM is becoming a recommended approach in clinical guidelines, extending beyond decision aids. However, the overall effect of SDM in women deciding to participate in breast cancer screening remains uncertain. OBJECTIVES: To assess the effect of SDM on women's satisfaction, confidence, and knowledge when deciding whether to participate in breast cancer screening. SEARCH METHODS: We searched the Cochrane Breast Cancer Group's Specialised Register, CENTRAL, MEDLINE, Embase, CINAHL, PsycINFO, ClinicalTrials.gov, and the World Health Organization International Clinical Trials Registry Platform on 8 August 2023. We also screened abstracts from two relevant conferences from 2020 to 2023. SELECTION CRITERIA: We included parallel randomised controlled trials (RCTs) and cluster-RCTs assessing interventions targeting various components of SDM. The focus was on supporting women aged 40 to 75 at average or above-average risk of breast cancer in their decision to participate in breast cancer screening. DATA COLLECTION AND ANALYSIS: Two review authors independently assessed studies for inclusion and conducted data extraction, risk of bias assessment, and GRADE assessment of the certainty of the evidence. Review outcomes included satisfaction with the decision-making process, confidence in the decision made, knowledge of all options, adherence to the chosen option, women's involvement in SDM, woman-clinician communication, and mental health. MAIN RESULTS: We identified 19 studies with 64,215 randomised women, mostly with an average to moderate risk of breast cancer. Two studies covered all aspects of SDM; six examined shortened forms of SDM involving communication on risks and personal values; and 11 focused on enhanced communication of risk without other SDM aspects. SDM involving all components compared to control The two eligible studies did not assess satisfaction with the SDM process or confidence in the decision. Based on a single study, SDM showed uncertain effects on participant knowledge regarding the age to start screening (risk ratio (RR) 1.18, 95% confidence interval (CI) 0.61 to 2.28; 133 women; very low certainty evidence) and frequency of testing (RR 0.84, 95% CI 0.68 to 1.04; 133 women; very low certainty evidence). Other review outcomes were not measured. Abbreviated forms of SDM with clarification of values and preferences compared to control Of the six included studies, none evaluated satisfaction with the SDM process. These interventions may reduce conflict in the decision made, based on two measures, Decisional Conflict Scale scores (mean difference (MD) -1.60, 95% CI -4.21 to 0.87; conflict scale from 0 to 100; 4 studies; 1714 women; very low certainty evidence) and the proportion of women with residual conflict compared to control at one to three months' follow-up (rate of women with a conflicted decision, RR 0.75, 95% CI 0.56 to 0.99; 1 study; 1001 women, very low certainty evidence). Knowledge of all options was assessed through knowledge scores and informed choice. The effect of SDM may enhance knowledge (MDs ranged from 0.47 to 1.44 higher scores on a scale from 0 to 10; 5 studies; 2114 women; low certainty evidence) and may lead to higher rates of informed choice (RR 1.24, 95% CI 0.95 to 1.63; 4 studies; 2449 women; low certainty evidence) compared to control at one to three months' follow-up. These interventions may result in little to no difference in anxiety (MD 0.54, 95% -0.96 to 2.14; scale from 20 to 80; 2 studies; 749 women; low certainty evidence) and the number of women with worries about cancer compared to control at four to six weeks' follow-up (RR 0.88, 95% CI 0.73 to 1.06; 1 study, 639 women; low certainty evidence). Other review outcomes were not measured. Enhanced communication about risks without other SDM aspects compared to control Of 11 studies, three did not report relevant outcomes for this review, and none assessed satisfaction with the SDM process. Confidence in the decision made was measured by decisional conflict and anticipated regret of participating in screening or not. These interventions, without addressing values and preferences, may result in lower confidence in the decision compared to regular communication strategies at two weeks' follow-up (MD 2.89, 95% CI -2.35 to 8.14; Decisional Conflict Scale from 0 to 100; 2 studies; 1191 women; low certainty evidence). They may result in higher anticipated regret if participating in screening (MD 0.28, 95% CI 0.15 to 0.41) and lower anticipated regret if not participating in screening (MD -0.28, 95% CI -0.42 to -0.14). These interventions increase knowledge (MD 1.14, 95% CI 0.61 to 1.62; scale from 0 to 10; 4 studies; 2510 women; high certainty evidence), while it is unclear if there is a higher rate of informed choice compared to regular communication strategies at two to four weeks' follow-up (RR 1.27, 95% CI 0.83 to 1.92; 2 studies; 1805 women; low certainty evidence). These interventions result in little to no difference in anxiety (MD 0.33, 95% CI -1.55 to 0.99; scale from 20 to 80) and depression (MD 0.02, 95% CI -0.41 to 0.45; scale from 0 to 21; 2 studies; 1193 women; high certainty evidence) and lower cancer worry compared to control (MD -0.17, 95% CI -0.26 to -0.08; scale from 1 to 4; 1 study; 838 women; high certainty evidence). Other review outcomes were not measured. AUTHORS' CONCLUSIONS: Studies using abbreviated forms of SDM and other forms of enhanced communications indicated improvements in knowledge and reduced decisional conflict. However, uncertainty remains about the effect of SDM on supporting women's decisions. Most studies did not evaluate outcomes considered important for this review topic, and those that did measured different concepts. High-quality randomised trials are needed to evaluate SDM in diverse cultural settings with a focus on outcomes such as women's satisfaction with choices aligned to their values.

Breast cancer diagnosis using support vector machine optimized by improved quantum inspired grey wolf optimization.

A prompt diagnosis of breast cancer in its earliest phases is necessary for effective treatment. While Computer-Aided Diagnosis systems play a crucial role in automated mammography image processing, interpretation, grading, and early detection of breast cancer, existing approaches face limitations in achieving optimal accuracy. This study addresses these limitations by hybridizing the improved quantum-inspired binary Grey Wolf Optimizer with the Support Vector Machines Radial Basis Function Kernel. This hybrid approach aims to enhance the accuracy of breast cancer classification by determining the optimal Support Vector Machine parameters. The motivation for this hybridization lies in the need for improved classification performance compared to existing optimizers such as Particle Swarm Optimization and Genetic Algorithm. Evaluate the efficacy of the proposed IQI-BGWO-SVM approach on the MIAS dataset, considering various metric parameters, including accuracy, sensitivity, and specificity. Furthermore, the application of IQI-BGWO-SVM for feature selection will be explored, and the results will be compared. Experimental findings demonstrate that the suggested IQI-BGWO-SVM technique outperforms state-of-the-art classification methods on the MIAS dataset, with a resulting mean accuracy, sensitivity, and specificity of 99.25%, 98.96%, and 100%, respectively, using a tenfold cross-validation datasets partition.

The incidence of breast cancer in Egyptian females in correlation to different mammographic ACR densities.

INTRODUCTION: The density of breast tissue, radiologically referred to as fibroglandular mammary tissue, was found to be a predisposing factor for breast cancer (BC). However, the stated degree of elevated BC risk varies widely in the literature.

Mammographic density mediates the protective effect of early-life body size on breast cancer risk.

The unexplained protective effect of childhood adiposity on breast cancer risk may be mediated via mammographic density (MD). Here, we investigate a complex relationship between adiposity in childhood and adulthood, puberty onset, MD phenotypes (dense area (DA), non-dense area (NDA), percent density (PD)), and their effects on breast cancer. We use Mendelian randomization (MR) and multivariable MR to estimate the total and direct effects of adiposity and age at menarche on MD phenotypes. Childhood adiposity has a decreasing effect on DA, while adulthood adiposity increases NDA. Later menarche increases DA/PD, but when accounting for childhood adiposity, this effect is attenuated. Next, we examine the effect of MD on breast cancer risk. DA/PD have a risk-increasing effect on breast cancer across all subtypes. The MD SNPs estimates are heterogeneous, and additional analyses suggest that different mechanisms may be linking MD and breast cancer. Finally, we evaluate the role of MD in the protective effect of childhood adiposity on breast cancer. Mediation MR analysis shows that 56% (95% CIs [32%-79%]) of this effect is mediated via DA. Our finding suggests that higher childhood adiposity decreases mammographic DA, subsequently reducing breast cancer risk. Understanding this mechanism is important for identifying potential intervention targets.

Strategies for Accessible Breast Screening for People With Intellectual Disability.

INTRODUCTION: People with intellectual disability are less likely to participate in breast screening than people without intellectual disability. They experience a range of barriers to accessing breast screening, however, there is no consensus on strategies to overcome these barriers. Our objective was to reach consensus on the strategies required for accessible breast screening for people with intellectual disability. METHODS: Fourteen experts participated in a modified on-line Delphi that used Levesque's model of health care access as the theoretical framework. At the end of each round descriptive and thematic analyses were completed. Data was then triangulated to determine if consensus was reached. RESULTS: After 3 rounds, 9 strategies were modified, 24 strategies were added and consensus was reached for 52 strategies across the 5 dimensions of access. Key areas of action related to (i) decision making and consent, (ii) accessible information, (iii) engagement of peer mentors, (iv) service navigators, and (v) equipping key stakeholders. CONCLUSIONS: The resulting strategies are the first to articulate how to make breast screening accessible and can be used to inform health policy and quality improvement practices.

Radiological features of screening-detected and interval breast cancers and subsequent survival in Eastern Finnish women.

Interval breast cancers are diagnosed between scheduled screenings and differ in many respects from screening-detected cancers. Studies comparing the survival of patients with interval and screening-detected cancers have reported differing results. The aim of this study was to investigate the radiological and histopathological features and growth rates of screening-detected and interval breast cancers and subsequent survival. This retrospective study included 942 female patients aged 50-69 years with breast cancers treated and followed-up at Kuopio University Hospital between January 2010 and December 2016. The screening-detected and interval cancers were classified as true, minimal-signs, missed, or occult. The radiological features were assessed on mammograms by one of two specialist breast radiologists with over 15 years of experience. A χ2 test was used to examine the association between radiological and pathological variables; an unpaired t test was used to compare the growth rates of missed and minimal-signs cancers; and the Kaplan-Meier estimator was used to examine survival after screening-detected and interval cancers. Sixty occult cancers were excluded, so a total of 882 women (mean age 60.4 ± 5.5 years) were included, in whom 581 had screening-detected cancers and 301 interval cancers. Disease-specific survival, overall survival and disease-free survival were all worse after interval cancer than after screening-detected cancer (p < 0.001), with a mean follow-up period of 8.2 years. There were no statistically significant differences in survival between the subgroups of screening-detected or interval cancers. Missed interval cancers had faster growth rates (0.47% ± 0.77%/day) than missed screening-detected cancers (0.21% ± 0.11%/day). Most cancers (77.2%) occurred in low-density breasts (< 25%). The most common lesion types were masses (73.9%) and calcifications (13.4%), whereas distortions (1.8%) and asymmetries (1.7%) were the least common. Survival was worse after interval cancers than after screening-detected cancers, attributed to their more-aggressive histopathological characteristics, more nodal and distant metastases, and faster growth rates.

Breast Cancer Screening: Can We Justify Deescalation?

Novel breast cancer screening methods that detect greater numbers of occult (nonpalpable) tumors have been rapidly incorporated into clinical practice, with the aim of reducing mortality. Yet, tumor detection has never been validated as a proper surrogate outcome measure for breast cancer mortality. Moreover, the detection of greater numbers of occult cancers increases the risk of overdiagnosis, which refers to detection of tumors that pose no threat to life and would never have been detected in the absence of screening. With recent advances in breast cancer therapy, many cancers that were previously curable only if detected as occult tumors with mammography screening are perhaps now curable even when detected as small palpable tumors, thereby giving us an opportunity to deescalate screening and mitigate the risk of overdiagnosis. Thus, a randomized trial comparing screening mammography versus screening clinical breast examination (CBE), with breast cancer mortality as the endpoint, is now warranted. In such a trial, hand-held ultrasound might aid in the interpretation of screening CBE findings. In conclusion, recent improvements in breast cancer therapy provide the justification to assess the deescalation of breast cancer screening. See related article by Farber et al., p. 671.

Functional hemodynamic imaging markers for the prediction of pathological outcomes in breast cancer patients treated with neoadjuvant chemotherapy.

Significance: Achieving pathologic complete response (pCR) after neoadjuvant chemotherapy (NACT) is a significant predictor of increased likelihood of survival in breast cancer patients. Early prediction of pCR is of high clinical value as it could allow personalized adjustment of treatment regimens in non-responding patients for improved outcomes. Aim: We aim to assess the association between hemoglobin-based functional imaging biomarkers derived from diffuse optical tomography (DOT) and the pathological outcome represented by pCR at different timepoints along the course of NACT. Approach: Twenty-two breast cancer patients undergoing NACT were enrolled in a multimodal DOT and X-ray digital breast tomosynthesis (DBT) imaging study in which their breasts were imaged at different compression levels. Logistic regressions were used to study the associations between DOT-derived imaging markers evaluated after the first and second cycles of chemotherapy, respectively, with pCR status determined after the conclusion of NACT at the time of surgery. Receiver operating characteristic curve analysis was also used to explore the predictive performance of selected DOT-derived markers. Results: Normalized tumor HbT under half compression was significantly lower in the pCR group compared to the non-pCR group after two chemotherapy cycles (p=0.042). In addition, the change in normalized tumor StO2 upon reducing compression from full to half mammographic force was identified as another potential indicator of pCR at an earlier time point, i.e., after the first chemo cycle (p=0.038). Exploratory predictive assessments showed that AUCs using DOT-derived functional imaging markers as predictors reach as high as 0.75 and 0.71, respectively, after the first and second chemo cycle, compared to AUCs of 0.50 and 0.53 using changes in tumor size measured on DBT and MRI. Conclusions: These findings suggest that breast DOT could be used to assist response assessment in women undergoing NACT, a critical but unmet clinical need, and potentially enable personalized adjustments of treatment regimens.

Breast arterial calcification is associated with sarcopenia in peri- and post-menopausal women.

PURPOSE: Breast arterial calcification (BAC) refers to medial calcium deposition in breast arteries and is detectable via mammography. Sarcopenia, which is characterised by low skeletal muscle mass and quality, is associated with several serious clinical conditions, increased morbidity, and mortality. Both BAC and sarcopenia share common pathologic pathways, including ageing, diabetes, and chronic kidney disease. Therefore, this study evaluated the relationship between BAC and sarcopenia as a potential indicator of sarcopenia. METHODS: This study involved women aged >40. BAC was evaluated using digital mammography and was defined as vascular calcification. Sarcopenia was assessed using abdominal computed tomography. The cross-sectional skeletal mass area was measured at the third lumbar vertebra level. The skeletal mass index was obtained by dividing the skeletal mass area by height in square meters(m2). Sarcopenia was defined as a skeletal mass index of ≤38.5 cm2/m2. A multivariable model was used to evaluate the relationship between BAC and sarcopenia. RESULTS: The study involved 240 participants. Of these, 36 (15 %) were patients with BAC and 204 (85 %) were without BAC. Sarcopenia was significantly higher among the patients with BAC than in those without BAC (72.2 % vs 17.2 %, P < 0.001). The multivariable model revealed that BAC and age were independently associated with sarcopenia (odds ratio[OR]: 7.719, 95 % confidence interval[CI]: 3.201-18.614, and P < 0.001 for BAC and OR: 1.039, 95 % CI: 1.007-1.073, P = 0.01 for age). CONCLUSION: BAC is independently associated with sarcopenia. BAC might be used as an indicator of sarcopenia on screening mammography.

BC-QNet: A quantum-infused ELM model for breast cancer diagnosis.

The timely and accurate diagnosis of breast cancer is pivotal for effective treatment, but current automated mammography classification methods have their constraints. In this study, we introduce an innovative hybrid model that marries the power of the Extreme Learning Machine (ELM) with FuNet transfer learning, harnessing the potential of the MIAS dataset. This novel approach leverages an Enhanced Quantum-Genetic Binary Grey Wolf Optimizer (Q-GBGWO) within the ELM framework, elevating its performance. Our contributions are twofold: firstly, we employ a feature fusion strategy to optimize feature extraction, significantly enhancing breast cancer classification accuracy. The proposed methodological motivation stems from optimizing feature extraction for improved breast cancer classification accuracy. The Q-GBGWO optimizes ELM parameters, demonstrating its efficacy within the ELM classifier. This innovation marks a considerable advancement beyond traditional methods. Through comparative evaluations against various optimization techniques, the exceptional performance of our Q-GBGWO-ELM model becomes evident. The classification accuracy of the model is exceptionally high, with rates of 96.54 % for Normal, 97.24 % for Benign, and 98.01 % for Malignant classes. Additionally, the model demonstrates a high sensitivity with rates of 96.02 % for Normal, 96.54 % for Benign, and 97.75 % for Malignant classes, and it exhibits impressive specificity with rates of 96.69 % for Normal, 97.38 % for Benign, and 98.16 % for Malignant classes. These metrics are reflected in its ability to classify three different types of breast cancer accurately. Our approach highlights the innovative integration of image data, deep feature extraction, and optimized ELM classification, marking a transformative step in advancing early breast cancer detection and enhancing patient outcomes.

Breast density quantification in dual-energy mammography using virtual anthropomorphic phantoms.

PURPOSE: Breast density is a significant risk factor for breast cancer and can impact the sensitivity of screening mammography. Area-based breast density measurements may not provide an accurate representation of the tissue distribution, therefore volumetric breast density (VBD) measurements are preferred. Dual-energy mammography enables volumetric measurements without additional assumptions about breast shape. In this work we evaluated the performance of a dual-energy decomposition technique for determining VBD by applying it to virtual anthropomorphic phantoms. METHODS: The dual-energy decomposition formalism was used to quantify VBD on simulated dual-energy images of anthropomorphic virtual phantoms with known tissue distributions. We simulated 150 phantoms with volumes ranging from 50 to 709 mL and VBD ranging from 15% to 60%. Using these results, we validated a correction for the presence of skin and assessed the method's intrinsic bias and variability. As a proof of concept, the method was applied to 14 sets of clinical dual-energy images, and the resulting breast densities were compared to magnetic resonance imaging (MRI) measurements. RESULTS: Virtual phantom VBD measurements exhibited a strong correlation (Pearson's r > 0.95 $r > 0.95$ ) with nominal values. The proposed skin correction eliminated the variability due to breast size and reduced the bias in VBD to a constant value of -2%. Disagreement between clinical VBD measurements using MRI and dual-energy mammography was under 10%, and the difference in the distributions was statistically non-significant. VBD measurements in both modalities had a moderate correlation (Spearman's ρ $\rho \ $ = 0.68). CONCLUSIONS: Our results in virtual phantoms indicate that the material decomposition method can produce accurate VBD measurements if the presence of a third material (skin) is considered. The results from our proof of concept showed agreement between MRI and dual-energy mammography VBD. Assessment of VBD using dual-energy images could provide complementary information in dual-energy mammography and tomosynthesis examinations.

Technical note: Characterization, validation, and spectral optimization of a dedicated breast CT system for contrast-enhanced imaging.

BACKGROUND: The development of a new imaging modality, such as 4D dynamic contrast-enhanced dedicated breast CT (4D DCE-bCT), requires optimization of the acquisition technique, particularly within the 2D contrast-enhanced imaging modality. Given the extensive parameter space, cascade-systems analysis is commonly used for such optimization. PURPOSE: To implement and validate a parallel-cascaded model for bCT, focusing on optimizing and characterizing system performance in the projection domain to enhance the quality of input data for image reconstruction. METHODS: A parallel-cascaded system model of a state-of-the-art bCT system was developed and model predictions of the presampled modulation transfer function (MTF) and the normalized noise power spectrum (NNPS) were compared with empirical data collected in the projection domain. Validation was performed using the default settings of 49 kV with 1.5 mm aluminum filter and at 65 kV and 0.257 mm copper filter. A 10 mm aluminum plate was added to replicate the breast attenuation. Air kerma at the isocenter was measured at different tube current levels. Discrepancies between the measured projection domain metrics and model-predicted values were quantified using percentage error and coefficient of variation (CoV) for MTF and NNPS, respectively. The optimal filtration was for a 5 mm iodine disk detection task at 49, 55, 60, and 65 kV. The detectability index was calculated for the default aluminum filtration and for copper thicknesses ranging from 0.05 to 0.4 mm. RESULTS: At 49 kV, MTF errors were +5.1% and -5.1% at 1 and 2 cycles/mm, respectively; NNPS CoV was 5.3% (min = 3.7%; max = 8.5%). At 65 kV, MTF errors were -0.8% and -3.2%; NNPS CoV was 13.1% (min = 11.4%; max = 16.9%). Air kerma output was linear, with 11.67 µGy/mA (R2 = 0.993) and 19.14 µGy/mA (R2 = 0.996) at 49 and 65 kV, respectively. For iodine detection, a 0.25 mm-thick copper filter at 65 kV was found optimal, outperforming the default technique by 90%. CONCLUSION: The model accurately predicts bCT system performance, specifically in the projection domain, under varied imaging conditions, potentially contributing to the enhancement of 2D contrast-enhanced imaging in 4D DCE-bCT.

Assessment of average glandular dose in mammography practice of a teaching hospital in Ghana.

Introduction: above the age of 40, women are advised to begin breast examinations and screenings for early detection of breast cancer. The average glandular dose (AGD) provides dosimetric information about the quantity of radiation received by the mammary glands during mammographic exposures. There is, therefore, the need to analyse the radiation dose received by patients presenting for mammography examinations. Methods: a retrospective cross-sectional design was carried out on the data of 663 participants, conveniently sampled between the months of July 2021 and June 2022. Paired T-test was used to compare imaging parameters for cranio-caudal (CC), medio-lateral (ML), automatic exposure control (AEC), manual exposure control (MEC), and left and right breast. Pearson´s correlation was used to test for relationship between imaging parameters and AGD. Results: the mean AGD per exposure was 1.9 ± 0.7 mGy for CC projections and 2.3 ± 1.2 mGy for ML projections. The mean AGD per examination for the study was 4.1 ± 1.4 mGy. A positive correlation was found between AGD per examination and exposure factors (tube loading and tube voltage), compressed breast thickness, and compression force. Patient age had no statistically significant relationship with the AGD per examination. Conclusion: average glandular dose (AGD) was consistent with other findings in literature studies. It was also observed that MEC yielded lower AGD per exposure values than AEC. There was no significant difference in the mean AGD per exposure for left and right breasts.

Performance of an Artificial Intelligence System for Breast Cancer Detection on Screening Mammograms from BreastScreen Norway.

Purpose To explore the stand-alone breast cancer detection performance, at different risk score thresholds, of a commercially available artificial intelligence (AI) system. Materials and Methods This retrospective study included information from 661 695 digital mammographic examinations performed among 242 629 female individuals screened as a part of BreastScreen Norway, 2004-2018. The study sample included 3807 screen-detected cancers and 1110 interval breast cancers. A continuous examination-level risk score by the AI system was used to measure performance as the area under the receiver operating characteristic curve (AUC) with 95% CIs and cancer detection at different AI risk score thresholds. Results The AUC of the AI system was 0.93 (95% CI: 0.92, 0.93) for screen-detected cancers and interval breast cancers combined and 0.97 (95% CI: 0.97, 0.97) for screen-detected cancers. In a setting where 10% of the examinations with the highest AI risk scores were defined as positive and 90% with the lowest scores as negative, 92.0% (3502 of 3807) of the screen-detected cancers and 44.6% (495 of 1110) of the interval breast cancers were identified with AI. In this scenario, 68.5% (10 987 of 16 040) of false-positive screening results (negative recall assessment) were considered negative by AI. When 50% was used as the cutoff, 99.3% (3781 of 3807) of the screen-detected cancers and 85.2% (946 of 1110) of the interval breast cancers were identified as positive by AI, whereas 17.0% (2725 of 16 040) of the false-positive results were considered negative. Conclusion The AI system showed high performance in detecting breast cancers within 2 years of screening mammography and a potential for use to triage low-risk mammograms to reduce radiologist workload. Keywords: Mammography, Breast, Screening, Convolutional Neural Network (CNN), Deep Learning Algorithms Supplemental material is available for this article. © RSNA, 2024 See also commentary by Bahl and Do in this issue.

Impact of AI for Digital Breast Tomosynthesis on Breast Cancer Detection and Interpretation Time.

Purpose To develop an artificial intelligence (AI) model for the diagnosis of breast cancer on digital breast tomosynthesis (DBT) images and to investigate whether it could improve diagnostic accuracy and reduce radiologist reading time. Materials and Methods A deep learning AI algorithm was developed and validated for DBT with retrospectively collected examinations (January 2010 to December 2021) from 14 institutions in the United States and South Korea. A multicenter reader study was performed to compare the performance of 15 radiologists (seven breast specialists, eight general radiologists) in interpreting DBT examinations in 258 women (mean age, 56 years ± 13.41 [SD]), including 65 cancer cases, with and without the use of AI. Area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and reading time were evaluated. Results The AUC for stand-alone AI performance was 0.93 (95% CI: 0.92, 0.94). With AI, radiologists' AUC improved from 0.90 (95% CI: 0.86, 0.93) to 0.92 (95% CI: 0.88, 0.96) (P = .003) in the reader study. AI showed higher specificity (89.64% [95% CI: 85.34%, 93.94%]) than radiologists (77.34% [95% CI: 75.82%, 78.87%]) (P < .001). When reading with AI, radiologists' sensitivity increased from 85.44% (95% CI: 83.22%, 87.65%) to 87.69% (95% CI: 85.63%, 89.75%) (P = .04), with no evidence of a difference in specificity. Reading time decreased from 54.41 seconds (95% CI: 52.56, 56.27) without AI to 48.52 seconds (95% CI: 46.79, 50.25) with AI (P < .001). Interreader agreement measured by Fleiss κ increased from 0.59 to 0.62. Conclusion The AI model showed better diagnostic accuracy than radiologists in breast cancer detection, as well as reduced reading times. The concurrent use of AI in DBT interpretation could improve both accuracy and efficiency. Keywords: Breast, Computer-Aided Diagnosis (CAD), Tomosynthesis, Artificial Intelligence, Digital Breast Tomosynthesis, Breast Cancer, Computer-Aided Detection, Screening Supplemental material is available for this article. © RSNA, 2024 See also the commentary by Bae in this issue.

Lesion attention guided neural network for contrast-enhanced mammography-based biomarker status prediction in breast cancer.

BACKGROUND AND OBJECTIVE: Accurate identification of molecular biomarker statuses is crucial in cancer diagnosis, treatment, and prognosis. Studies have demonstrated that medical images could be utilized for non-invasive prediction of biomarker statues. The biomarker status-associated features extracted from medical images are essential in developing medical image-based non-invasive prediction models. Contrast-enhanced mammography (CEM) is a promising imaging technique for breast cancer diagnosis. This study aims to develop a neural network-based method to extract biomarker-related image features from CEM images and evaluate the potential of CEM in non-invasive biomarker status prediction. METHODS: An end-to-end learning convolutional neural network with the whole breast images as inputs was proposed to extract CEM features for biomarker status prediction in breast cancer. The network focused on lesion regions and flexibly extracted image features from lesion and peri­tumor regions by employing supervised learning with a smooth L1-based consistency constraint. An image-level weakly supervised segmentation network based on Vision Transformer with cross attention to contrast images of breasts with lesions against the contralateral breast images was developed for automatic lesion segmentation. Finally, prediction models were developed following further selection of significant features and the implementation of random forest-based classification. Results were reported using the area under the curve (AUC), accuracy, sensitivity, and specificity. RESULTS: A dataset from 1203 breast cancer patients was utilized to develop and evaluate the proposed method. Compared to the method without lesion attention and with only lesion regions as inputs, the proposed method performed better at biomarker status prediction. Specifically, it achieved an AUC of 0.71 (95 % confidence interval [CI]: 0.65, 0.77) for Ki-67 and 0.73 (95 % CI: 0.65, 0.80) for human epidermal growth factor receptor 2 (HER2). CONCLUSIONS: A lesion attention-guided neural network was proposed in this work to extract CEM image features for biomarker status prediction in breast cancer. The promising results demonstrated the potential of CEM in non-invasively predicting the biomarker statuses in breast cancer.