To demonstrate a method of benchmarking the performance of two consecutive software releases of the same commercial artificial intelligence (AI) product to trained human readers using the Personal Performance in Mammographic Screening scheme (PERFORMS) external quality assurance scheme. In this retrospective study, ten PERFORMS test sets, each consisting of 60 challenging cases, were evaluated by human readers between 2012 and 2023 and were evaluated by Version 1 (V1) and Version 2 (V2) of the same AI model in 2022 and 2023 respectively. Both AI and humans considered each breast independently. Both AI and humans considered the highest suspicion of malignancy score per breast for non-malignant cases and per lesion for breasts with malignancy. Sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) were calculated for comparison, with the study powered to detect a medium-sized effect (odds ratio, 3.5 or 0.29) for sensitivity. The study included 1,254 human readers, with a total of 328 malignant lesions, 823 normal, and 55 benign breasts analysed. No significant difference was found between the AUCs for AI V1 (0.93) and V2 (0.94) (p = 0.13). In terms of sensitivity, no difference was observed between human readers and AI V1 (83.2 % vs 87.5 % respectively, p = 0.12), however V2 outperformed humans (88.7 %, p = 0.04). Specificity was higher for AI V1 (87.4 %) and V2 (88.2 %) compared to human readers (79.0 %, p < 0.01 respectively). The upgraded AI model showed no significant difference in diagnostic performance compared to its predecessor when evaluating mammograms from PERFORMS test sets.

Masked AutoEncoder (MAE) has demonstrated significant potential in medical image analysis by reducing the cost of manual annotations. However, MAE and its recent variants are not well-developed for ultrasound images in breast cancer diagnosis, as they struggle to generalize to the task of distinguishing ultrasound breast tumors of varying sizes. This limitation hinders the model's ability to adapt to the diverse morphological characteristics of breast tumors. In this paper, we propose a novel Breast UltraSound Multi-scale Masked AutoEncoder (BUS-M2AE) model to address the limitations of the general MAE. BUS-M2AE incorporates multi-scale masking methods at both the token level during the image patching stage and the feature level during the feature learning stage. These two multi-scale masking methods enable flexible strategies to match the explicit masked patches and the implicit features with varying tumor scales. By introducing these multi-scale masking methods in the image patching and feature learning phases, BUS-M2AE allows the pre-trained vision transformer to adaptively perceive and accurately distinguish breast tumors of different sizes, thereby improving the model's overall performance in handling diverse tumor morphologies. Comprehensive experiments demonstrate that BUS-M2AE outperforms recent MAE variants and commonly used supervised learning methods in breast cancer classification and tumor segmentation tasks.

To assess the impact of artificial intelligence (AI) on the diagnostic performance of radiologists with varying experience levels in mammography reading, considering single and simulated double reading approaches. In this retrospective study, 150 mammography examinations (30 with pathology-confirmed malignancies, 120 without malignancies [confirmed by 2-year follow-up]) were reviewed according to five approaches: A) human single reading by 26 radiologists of varying experience; B) AI single reading (Lunit INSIGHT MMG; C) human single reading with simultaneous AI support; D) simulated human-human double reading; E) simulated human-AI double reading, with AI as second independent reader flagging cases with a cancer probability ≥10 %. Sensitivity and specificity were calculated and compared using McNemar's test, univariate and multivariable logistic regression. Compared to single reading without AI support, single reading with simultaneous AI support improved mean sensitivity from 69.2 % (standard deviation [SD] 15.6) to 84.5 % (SD 8.1, p < 0.001), providing comparable mean specificity (91.8 % versus 90.8 %, p = 0.06). The sensitivity increase provided by the AI-supported single reading was largest in the group of radiologists with a sensitivity below the median in the non-supported single reading, from 56.7 % (SD 12.1) to 79.7 % (SD 10.2, p < 0.001). In the simulated human-AI double reading approach, sensitivity further increased to 91.8 % (SD 3.4), surpassing that of the human-human simulated double reading (87.4 %, SD 8.8, p = 0.016), with comparable mean specificity (from 84.0 % to 83.0 %, p = 0.17). AI support significantly enhanced sensitivity across all reading approaches, particularly benefiting worse performing radiologists. In the simulated double reading approaches, AI incorporation as independent second reader significantly increased sensitivity without compromising specificity.

Breast density, as derived from mammographic images and defined by the Breast Imaging Reporting & Data System (BI-RADS), is one of the strongest risk factors for breast cancer. Breast ultrasound is an alternative breast cancer screening modality, particularly useful in low-resource, rural contexts. To date, breast ultrasound has not been used to inform risk models that need breast density. The purpose of this study is to explore the use of artificial intelligence (AI) to predict BI-RADS breast density category from clinical breast ultrasound imaging. We compared deep learning methods for predicting breast density directly from breast ultrasound imaging, as well as machine learning models from breast ultrasound image gray-level histograms alone. The use of AI-derived breast ultrasound breast density as a breast cancer risk factor was compared to clinical BI-RADS breast density. Retrospective (2009-2022) breast ultrasound data were split by individual into 70/20/10% groups for training, validation, and held-out testing for reporting results. 405,120 clinical breast ultrasound images from 14,066 women (mean age 53 years, range 18-99 years) with clinical breast ultrasound exams were retrospectively selected for inclusion from three institutions: 10,393 training (302,574 images), 2593 validation (69,842), and 1074 testing (28,616). The AI model achieves AUROC 0.854 in breast density classification and statistically significantly outperforms all image statistic-based methods. In an existing clinical 5-year breast cancer risk model, breast ultrasound AI and clinical breast density predict 5-year breast cancer risk with 0.606 and 0.599 AUROC (DeLong's test p-value: 0.67), respectively. BI-RADS breast density can be estimated from breast ultrasound imaging with high accuracy. The AI model provided superior estimates to other machine learning approaches. Furthermore, we demonstrate that age-adjusted, AI-derived breast ultrasound breast density provides similar predictive power to mammographic breast density in our population. Estimated breast density from ultrasound may be useful in performing breast cancer risk assessment in areas where mammography may not be available. National Cancer Institute.

To assess the performance of an industry-developed deep learning (DL) algorithm to reconstruct low-resolution Cartesian T1-weighted dynamic contrast-enhanced (T1w) and T2-weighted turbo-spin-echo (T2w) sequences and compare them to standard sequences. Female patients with indications for breast MRI were included in this prospective study. The study protocol at 1.5 Tesla MRI included T1w and T2w. Both sequences were acquired in standard resolution (T1_S and T2_S) and in low-resolution with following DL reconstructions (T1_DL and T2_DL). For DL reconstruction, two convolutional networks were used: (1) Adaptive-CS-Net for denoising with compressed sensing, and (2) Precise-Image-Net for resolution upscaling of previously downscaled images. Overall image quality was assessed using 5-point-Likert scale (from 1=non-diagnostic to 5=excellent). Apparent signal-to-noise (aSNR) and contrast-to-noise (aCNR) ratios were calculated. Breast Imaging Reporting and Data System (BI-RADS) agreement between different sequence types was assessed. A total of 47 patients were included (mean age, 58±11 years). Acquisition time for T1_DL and T2_DL were reduced by 51% (44 vs. 90 s per dynamic phase) and 46% (102 vs. 192 s), respectively. T1_DL and T2_DL showed higher overall image quality (e.g., 4 [IQR, 4-4] for T1_S vs. 5 [IQR, 5-5] for T1_DL, P<0.001). Both, T1_DL and T2_DL revealed higher aSNR and aCNR than T1_S and T2_S (e.g., aSNR: 32.35±10.23 for T2_S vs. 27.88±6.86 for T2_DL, P=0.014). Cohen k agreement by BI-RADS assessment was excellent (0.962, P<0.001). DL for denoising and resolution upscaling reduces acquisition time and improves image quality for T1w and T2w breast MRI.

To explore the association between pretherapeutic contrast-enhanced cone beam breast CT (CE-CBBCT) features and pathological complete response (pCR), and to develop a predictive model that integrates clinicopathological and imaging features. In this prospective study, a cohort of 200 female patients who underwent CE-CBBCT prior to neoadjuvant therapy and surgery was divided into train (n=150) and test (n=50) sets in a 3:1 ratio. Optimal predictive features were identified using univariate logistic regression and recursive feature elimination with cross-validation (RFECV). Models were constructed using XGBoost and evaluated through the receiver operating characteristic (ROC) curve, calibration curves, and decision curve analysis. The performance of combined model was further evaluated across molecular subtypes. Feature significance within the combined model was determined using the SHapley Additive exPlanation (SHAP) algorithm. The model incorporating three clinicopathological and six CE-CBBCT imaging features demonstrated robust predictive performance for pCR, with area under curves (AUCs) of 0.924 in the train set and 0.870 in the test set. Molecular subtype, spiculation, and adjacent vascular sign (AVS) grade emerged as the most influential SHAP features. The highest AUCs were observed for HER2-positive subgroup (train: 0.935; test: 0.844), followed by luminal (train: 0.841; test: 0.717) and triple-negative breast cancer (TNBC; train: 0.760; test: 0.583). SHAP analysis indicated that spiculation was crucial for luminal breast cancer prediction, while AVS grade was critical for HER2-positive and TNBC cases. Integrating clinicopathological and CE-CBBCT imaging features enhanced pCR prediction accuracy, particularly in HER2-positive cases, underscoring its potential clinical applicability.

Detecting breast cancer early is of the utmost importance to effectively treat the millions of women afflicted by breast cancer worldwide every year. Although mammography is the primary imaging modality for screening breast cancer, there is an increasing interest in adding magnetic resonance imaging (MRI) to screening programmes, particularly for women at high risk. Recent guidelines by the European Society of Breast Imaging (EUSOBI) recommended breast MRI as a supplemental screening tool for women with dense breast tissue. However, acquiring and reading MRI scans requires significantly more time from expert radiologists. This highlights the need to develop new automated methods to detect cancer accurately using MRI and Artificial Intelligence (AI), which have the potential to support radiologists in breast MRI interpretation and classification and help detect cancer earlier. For this reason, the ODELIA consortium has made this multi-centre dataset publicly available to assist in developing AI tools for the detection of breast cancer on MRI.

Artificial intelligence (AI) in breast imaging has garnered significant attention given the numerous reports of improved efficiency, accuracy, and the potential to bridge the gap of expanded volume in the face of limited physician resources. While AI models are developed with specific data points, on specific equipment, and in specific populations, the real-world clinical environment is dynamic, and patient populations are diverse, which can impact generalizability and widespread adoption of AI in clinical practice. Implementation of AI models into clinical practice requires focused attention on the potential of AI bias impacting outcomes. The following review presents the concept, sources, and types of AI bias to be considered when implementing AI models and offers suggestions on strategies to mitigate AI bias in practice.

Mammographic breast cancer detection depends on high-quality positioning, which is traditionally assessed and monitored subjectively. This study used artificial intelligence (AI) to evaluate mammography positioning on digital screening mammograms to identify and quantify unmet mammography positioning quality (MPQ). Data were collected within an IRB-approved collaboration. In total, 126 367 digital mammography studies (553 339 images) were processed. Unmet MPQ criteria, including exaggeration, portion cutoff, posterior tissue missing, nipple not in profile, too high on image receptor, inadequate pectoralis length, sagging, and posterior nipple line (PNL) length difference, were evaluated using MPQ AI algorithms. The similarity of unmet MPQ occurrence and rank order was compared for each health system. Altogether, 163 759 and 219 785 unmet MPQ criteria were identified, respectively, at the health systems. The rank order and the probability distribution of the unmet MPQ criteria were not statistically significantly different between health systems (P = .844 and P = .92, respectively). The 3 most-common unmet MPQ criteria were: short PNL length on the craniocaudal (CC) view, inadequate pectoralis muscle, and excessive exaggeration on the CC view. The percentages of unmet positioning criteria out of the total potential unmet positioning criteria at health system 1 and health system 2 were 8.4% (163 759/1 949 922) and 7.3% (219 785/3 030 129), respectively. Artificial intelligence identified a similar distribution of unmet MPQ criteria in 2 health systems' daily work. Knowledge of current commonly unmet MPQ criteria can facilitate the improvement of mammography quality through tailored education strategies.