PURPOSE:The most widely used method in Europe to estimate breast dosimetry in mammography is based on Dance's publications. Recently, AAPM and EFOMP joint Task Group 282 (TG282) has published a new breast dosimetry model, which improves the accuracy in Average Glandular Dose (AGD) estimation. This multicentre study aims to evaluate variations in AGD between conventional and new dosimetry across different modalities and mammography systems. METHODS:Measurements were conducted on 74 mammography systems from six vendors, installed in hospitals across Italy. Each facility used its own equipment while adhering to the same protocol and spreadsheets. Tube output and Half-Value Layer (HVL) were measured in the two configurations for Digital Mammography (DM), Digital Breast Tomosynthesis (DBT), and Contrast-Enhanced Mammography (CEM). Exposures were performed using a breast-shaped PMMA phantom of different thicknesses. Data were centrally collected to compute AGD using appropriate correction factors for conventional dosimetry and the BreastAGD282 software for TG282 dosimetry, using EU legacy model. RESULTS:The sample allows for the identification of typical AGD values for both methodologies. The study highlights a reduction in median AGD with the new dosimetry model compared to Dance method across all modalities and vendors. Furthermore, the reduction magnitude depends on x-ray beam quality, modality and compressed breast thickness, ranging from 6.1-8.4 % at 21 mm to 27.3-38.3 % at 90 mm. CONCLUSIONS:The new TG282 dosimetry model showed a reduction in the estimated median AGD. Notably, the magnitude of this reduction depended on x-ray beam quality, modality and compressed breast thickness.

Neoadjuvant chemotherapy (NACT) improves surgical outcomes for breast cancer patients, with pathologic complete response (pCR) correlated with enhanced survival. The role of radiomics, particularly from peritumoral tissue, in predicting pCR remains under investigation. This retrospective study analyzed radiomic features from pretreatment dynamic contrast-enhanced breast MRI scans of 150 patients undergoing NACT. A proportional approach was used to define peritumoral zones, assessed both with a 10

PURPOSE:Digital Breast Tomosynthesis (DBT) is an advanced mammography technique for which there are currently no internationally agreed methods and reference values for image quality assessment. The aim of this multicentre study was to evaluate a simple method to assess the technical image quality of reconstructed and synthetic 2D (SM) images of different models of DBT systems using commercially available phantoms. METHODS:The signal difference to noise ratio (SDNR) was chosen as an index of technical image quality and was evaluated for three commercial phantoms, Tomophan, Tormam and CIRS model 015, on 55 DBT systems (six vendors, nine models). RESULTS:SDNR was found to depend on several factors: detail size, average glandular dose (AGD), reconstruction algorithm, software version and applied post-processing. In particular, an increase in SDNR was observed with increasing detail size, AGD, as well as with the use of contrast-enhanced post-processing and iterative reconstruction algorithms. Most systems showed higher SDNR values in SM images respect to DBT for the largest details and a decrease for smaller details. CONCLUSIONS:This study proposes a straightforward method to assess the technical image quality of reconstructed DBT and SM images using three different commercial phantoms. The algorithms used to generate SM and DBT images might perform differently when applied to phantoms with homogeneous backgrounds compared to clinical breast structures. However this method could be used to establish reference values for technical image quality during quality control (QC) which could be used to monitor the consistency of the system's performance over time.

PURPOSE:Contrast-enhanced mammography (CEM) is a technique that exploits the combination of two projections at different energies to produce an energy-subtracted (ES) image to highlight the uptake of iodinated contrast medium in breast lesions. The aim of this study is to analyze and compare the performance of different CEM systems. METHODS:Four mammography systems of different vendors (Fujifilm Amulet Innovality, Hologic 3Dimensions, IMS Giotto Class, GE Senographe Pristina) were compared employing a commercial dedicated phantom, equipped with a breast-equivalent target slab with four different iodine concentrations (IC) inserts. Acquisition parameters and average glandular dose (AGD) were collected at different phantom thicknesses (30-80 mm). Contrast-to-noise ratio (CNR) and linearity of IC response with phantom thickness were evaluated. RESULTS:AGDs (mGy) were in ranges 0.82-3.10, 0.90-6.25, 1.08-3.30, and 0.74-3.63 for Fujifilm, Hologic, GE, and IMS, respectively. High-energy AGD accounted for up to 41.5%, 25.6%, 40.5%, and 23.8% of the total dose for Fujifilm, Hologic, IMS, and GE, respectively. CNR increased with IC and generally decreased with increasing phantom thickness. In recombined images, Hologic, Fujifilm, and GE showed good linearity of IC response with phantom thickness and overlapping trends (maximum error = 1%) for all thicknesses. CONCLUSION:Similarities and differences in image quality and dose were found depending on the different technical and image processing characteristics of the different vendors. Linearity of IC at various thicknesses might be further exploited in clinical scenarios to differentiate between suspicious breast lesions.

Background: Computed tomography scans are widely used in everyday medical practice due to speed, image reliability, and detectability of a wide range of pathologies. Each scan exposes the patient to a radiation dose, and performing a fast estimation of the effective dose (E) is an important step for radiological safety. The aim of this work is to estimate E from patient and CT acquisition parameters in the absence of a dose-tracking software exploiting machine learning. Methods: In total, 69,037 CT acquisitions were collected with the dose-tracking software (DTS) available at our institution. E calculated by DTS was chosen as the target value for prediction. Different machine learning algorithms were selected, optimizing parameters to achieve the best performance for each algorithm. Effective dose was also estimated using DLP and k-factors, and with multiple linear regression. Mean absolute error (MAE, mean absolute percentage error (MAPE), and R2 were used to evaluate predictions in the test set and in an external dataset of 3800 acquisitions. Results: The random forest regressor (MAE: 0.416 mSv; MAPE: 7%; and R2: 0.98) showed best performances over the neural network and the support vector machine. However, all three machine learning algorithms outperformed effective dose estimation using k-factors (MAE: 2.06; MAPE: 26%) or multiple linear regression (MAE: 0.98; MAPE: 44.4%). The random forest regressor on the external dataset showed an MAE of 0.215 mSv and an MAPE of 7.1%. Conclusions: Our work demonstrated that machine learning models trained with data calculated by a dose-tracking software can provide good estimates of the effective dose just from patient and scanner parameters, without the need for a Monte Carlo approach.

BACKGROUND:Radiomics is changing clinical practice by providing quantitative information from images to improve diagnosis, prognosis, and treatment planning. This study aims to investigate a radiomics model developed from contrast-enhanced mammography (CEM) images to predict disease-free survival (DFS) and overall survival (OS) in breast cancer (BC) patients. METHODS:From January 2013 to December 2015, all consecutive BC patients who underwent CEM before biopsy at a referral center were enrolled. Clinical data included histological results, receptor profiles, and follow-up (DFS and OS). A region of interest (ROI) of the enhancing lesion was selected from recombined CEM images by experienced radiologists, and radiomic features were extracted. A Cox-LASSO model assigned coefficients to the features, generating patient radiomic scores (RSs), which were dichotomized for graphical representation. Model performance was assessed using the C index. RESULTS:The study included 126 BC patients with predominantly "mass"-type lesions (95%) and a median follow-up of 6.88 years (IQR 3.10-8.15). The median age of the patients at the time of examination was 49.2 years (IQR: [42.33-56.98]). Radiomic and clinical-radiomic models showed significant associations between RS, DFS, and OS, with patients with RS below the median showing a better prognosis (p < 0.001). Bootstrap testing confirmed a good model fit for OS prediction, with median C-index values of 0.82 for the clinical model and 0.84 for the clinical-radiomic model. CONCLUSIONS:Radiomic analysis of CEM images may predict DFS and OS in BC patients, offering additional prognostic value beyond clinical models alone.

To investigate whether methodological aspects may influence the performance of MRI-radiomic models to predict response to neoadjuvant treatment (NAT) in breast cancer (BC) patients. We conducted a systematic review until March 2023. A random-effects meta-analysis was performed to combine the area under the receiver operating characteristic curve (AUC) values. Publication bias was assessed using Egger’s test and heterogeneity was estimated by I2. A meta-regression was conducted to investigate the impact of various factors, including scanner, features’ number/transformation/type, pixel/voxel scaling, etc. Forty-two studies were included. The summary AUC was 0.77 (95

BACKGROUND:Breast cancer (BC) is the most common malignancy in women and the second cause of cancer death. In recent years, there has been a strong development in artificial intelligence (AI) applications in medical imaging for several tasks. Our aim was to evaluate the potential of transfer learning with convolutional neural networks (CNNs) in discriminating suspicious breast lesions on ultrasound images. METHODS:Transfer learning performances of five different CNNs (Inception V3, Xception, Densenet121, VGG 16, and ResNet50) were evaluated on a public and on an institutional dataset (526 and 392 images, respectively), customizing the top layers for the specific task. Institutional images were contoured by an expert radiologist and processed to feed the CNNs for training and testing. Postimaging biopsies were used as a reference standard for classification. The area under the receiver operating curve (AUROC) was used to assess diagnostic performance. RESULTS:Networks performed very well on the public dataset (AUROC 0.938-0.996). The direct generalization to the institutional dataset resulted in lower performances (max AUROC 0.676); however, when tested on BI-RADS 3 and BI-RADS 5 only, results were improved (max AUROC 0.792). Good results were achieved on the institutional dataset (AUROC 0.759-0.818) and, when selecting a threshold of 2% for classification, a sensitivity of 0.983 was obtained for three of five CNNs, with the potential to spare biopsy in 15.3%-18.6% of patients. CONCLUSION:In conclusion, transfer learning with CNNs may achieve high sensitivity and might be used as a support tool in managing suspicious breast lesions on ultrasound images. RELEVANCE STATEMENT:Transfer learning is a powerful technique to exploit the performances of well-trained CNNs for image classification. In a clinical scenario, it might be useful for the management of suspicious breast lesions on breast ultrasound, potentially sparing biopsy in a non-negligible number of patients. KEY POINTS:Properly trained CNNs with transfer learning are highly effective in differentiating benign and malignant lesions on breast ultrasound. Setting clinical thresholds increased sensitivity. CNNs might be useful as support tools in managing suspicious lesions on breast ultrasound.

The study aimed to evaluate the performance of radiomics features and one ultrasound CAD (computer-aided diagnosis) in the prediction of the malignancy of a breast lesion detected with ultrasound and to develop a nomogram incorporating radiomic score and available information on CAD performance, conventional Breast Imaging Reporting and Data System evaluation (BI-RADS), and clinical information. Data on 365 breast lesions referred for breast US with subsequent histologic analysis between January 2020 and March 2022 were retrospectively collected. Patients were randomly divided into a training group (n = 255) and a validation test group (n = 110). A radiomics score was generated from the US image. The CAD was performed in a subgroup of 209 cases. The radiomics score included seven radiomics features selected with the LASSO logistic regression model. The multivariable logistic model incorporating CAD performance, BI-RADS evaluation, clinical information, and radiomic score as covariates showed promising results in the prediction of the malignancy of breast lesions: Area under the receiver operating characteristic curve, [AUC]: 0.914; 95% Confidence Interval, [CI]: 0.876–0.951. A nomogram was developed based on these results for possible future applications in clinical practice.

Recent technological advances in the field of artificial intelligence hold promise in addressing medical challenges in breast cancer care, such as early diagnosis, cancer subtype determination and molecular profiling, prediction of lymph node metastases, and prognostication of treatment response and probability of recurrence. Radiomics is a quantitative approach to medical imaging, which aims to enhance the existing data available to clinicians by means of advanced mathematical analysis using artificial intelligence. Various published studies from different fields in imaging have highlighted the potential of radiomics to enhance clinical decision making. In this review, we describe the evolution of AI in breast imaging and its frontiers, focusing on handcrafted and deep learning radiomics. We present a typical workflow of a radiomics analysis and a practical "how-to" guide. Finally, we summarize the methodology and implementation of radiomics in breast cancer, based on the most recent scientific literature to help researchers and clinicians gain fundamental knowledge of this emerging technology. Alongside this, we discuss the current limitations of radiomics and challenges of integration into clinical practice with conceptual consistency, data curation, technical reproducibility, adequate accuracy, and clinical translation. The incorporation of radiomics with clinical, histopathological, and genomic information will enable physicians to move forward to a higher level of personalized management of patients with breast cancer.

This study aims to evaluate the Average Glandular Dose (AGD) and diagnostic performance of CEM versus Digital Mammography (DM) as well as versus DM plus one-view Digital Breast Tomosynthesis (DBT), which were performed in the same patients at short intervals of time. A preventive screening examination in high-risk asymptomatic patients between 2020 and 2022 was performed with two-view Digital Mammography (DM) projections (Cranio Caudal and Medio Lateral) plus one Digital Breast Tomosynthesis (DBT) projection (mediolateral oblique, MLO) in a single session examination. For all patients in whom we found a suspicious lesion by using DM + DBT, we performed (within two weeks) a CEM examination. AGD and compression force were compared between the diagnostic methods. All lesions identified by DM + DBT were biopsied; then, we assessed whether lesions found by DBT were also highlighted by DM alone and/or by CEM. We enrolled 49 patients with 49 lesions in the study. The median AGD was lower for DM alone than for CEM (3.41 mGy vs. 4.24 mGy, p = 0.015). The AGD for CEM was significantly lower than for the DM plus one single projection DBT protocol (4.24 mGy vs. 5.55 mGy, p < 0.001). We did not find a statistically significant difference in the median compression force between the CEM and DM + DBT. DM + DBT allows the identification of one more invasive neoplasm one in situ lesion and two high-risk lesions, compared to DM alone. The CEM, compared to DM + DBT, failed to identify only one of the high-risk lesions. According to these results, CEM could be used in the screening of asymptomatic high-risk patients.

PURPOSE:The aim of the present study, conducted by a working group of the Italian Association of Medical Physics (AIFM), was to define typical z-resolution values for different digital breast tomosynthesis (DBT) models to be used as a reference for quality control (QC). Currently, there are no typical values published in internationally agreed QC protocols.METHODS:To characterize the z-resolution of the DBT models, the full width at half maximum (FWHM) of the artifact spread function (ASF), a technical parameter that quantifies the signal intensity of a detail along reconstructed planes, was analyzed. Five different commercial phantoms, CIRS Model 011, CIRS Model 015, Modular DBT phantom, Pixmam 3-D, and Tomophan, were evaluated on reconstructed DBT images and 82 DBT systems (6 vendors, 9 models) in use at 39 centers in Italy were involved.RESULTS:The ASF was found to be dependent on the detail size, the DBT angular acquisition range, the reconstruction algorithm and applied image processing. In particular, a progressively greater signal spread was observed as the detail size increased and the acquisition angle decreased. However, a clear correlation between signal spread and angular range width was not observed due to the different signal reconstruction and image processing strategies implemented in the algorithms developed by the vendors studied.CONCLUSIONS:The analysis led to the identification of typical z-resolution values for different DBT model-phantom configurations that could be used as a reference during a QC program.

Abstract Background Breast cancer screening through mammography is crucial for early detection, yet the demand for mammography services surpasses the capacity of radiologists. Artificial intelligence (AI) can assist in evaluating microcalcifications on mammography. We developed and tested an AI model for localizing and characterizing microcalcifications. Methods Three expert radiologists annotated a dataset of mammograms using histology-based ground truth. The dataset was partitioned for training, validation, and testing. Three neural networks (AlexNet, ResNet18, and ResNet34) were trained and evaluated using specific metrics including receiver operating characteristics area under the curve (AUC), sensitivity, and specificity. The reported metrics were computed on the test set (10% of the whole dataset). Results The dataset included 1,000 patients aged 21–73 years and 1,986 mammograms (180 density A, 220 density B, 380 density C, and 220 density D), with 389 malignant and 611 benign groups of microcalcifications. AlexNet achieved the best performance with 0.98 sensitivity, 0.89 specificity of, and 0.98 AUC for microcalcifications detection and 0.85 sensitivity, 0.89 specificity, and 0.94 AUC of for microcalcifications classification. For microcalcifications detection, ResNet18 and ResNet34 achieved 0.96 and 0.97 sensitivity, 0.91 and 0.90 specificity and 0.98 and 0.98 AUC, retrospectively. For microcalcifications classification, ResNet18 and ResNet34 exhibited 0.75 and 0.84 sensitivity, 0.85 and 0.84 specificity, and 0.88 and 0.92 AUC, respectively. Conclusions The developed AI models accurately detect and characterize microcalcifications on mammography. Relevance statement AI-based systems have the potential to assist radiologists in interpreting microcalcifications on mammograms. The study highlights the importance of developing reliable deep learning models possibly applied to breast cancer screening. Key points • A novel AI tool was developed and tested to aid radiologists in the interpretation of mammography by accurately detecting and characterizing microcalcifications. • Three neural networks (AlexNet, ResNet18, and ResNet34) were trained, validated, and tested using an annotated dataset of 1,000 patients and 1,986 mammograms. • The AI tool demonstrated high accuracy in detecting/localizing and characterizing microcalcifications on mammography, highlighting the potential of AI-based systems to assist radiologists in the interpretation of mammograms. Graphical Abstract