Background Reducing the amount of contrast agent needed for contrast-enhanced breast MRI is desirable. Purpose To investigate if generative adversarial networks (GANs) can recover contrast-enhanced breast MRI scans from unenhanced images and virtual low-contrast-enhanced images. Materials and Methods In this retrospective study of breast MRI performed from January 2010 to December 2019, simulated low-contrast images were produced by adding virtual noise to the existing contrast-enhanced images. GANs were then trained to recover the contrast-enhanced images from the simulated low-contrast images (approach A) or from the unenhanced T1- and T2-weighted images (approach B). Two experienced radiologists were tasked with distinguishing between real and synthesized contrast-enhanced images using both approaches. Image appearance and conspicuity of enhancing lesions on the real versus synthesized contrast-enhanced images were independently compared and rated on a five-point Likert scale. P values were calculated by using bootstrapping. Results A total of 9751 breast MRI examinations from 5086 patients (mean age, 56 years ± 10 [SD]) were included. Readers who were blinded to the nature of the images could not distinguish real from synthetic contrast-enhanced images (average accuracy of differentiation: approach A, 52 of 100; approach B, 61 of 100). The test set included images with and without enhancing lesions (29 enhancing masses and 21 nonmass enhancement; 50 total). When readers who were not blinded compared the appearance of the real versus synthetic contrast-enhanced images side by side, approach A image ratings were significantly higher than those of approach B (mean rating, 4.6 ± 0.1 vs 3.0 ± 0.2; P < .001), with the noninferiority margin met by synthetic images from approach A (P < .001) but not B (P > .99). Conclusion Generative adversarial networks may be useful to enable breast MRI with reduced contrast agent dose. © RSNA, 2023 Supplemental material is available for this article. See also the editorial by Bahl in this issue.

To investigate the integration of differential privacy (DP) and analyze its impact on model performance as compared with models trained without DP. Leveraging more than 590 000 chest radiographs from five institutions, including VinDr-CXR from Vietnam, ChestX-ray14 and CheXpert from the United States, UKA-CXR from Germany, and PadChest from Spain, the authors evaluated the efficacy of DP-enhanced domain transfer (DP-DT) in classifying cardiomegaly, pleural effusion, pneumonia, atelectasis, and healthy individuals. Diagnostic performance and sex-specific and age-specific demographic fairness of DP-DT and of non–DP-DT models were compared using the area under the receiver operating characteristic curve (AUC) as the main metric, as well as accuracy, sensitivity, and specificity as secondary metrics, and evaluated for statistical significance using paired Student t tests. Even with high privacy levels (ε ≈ 1), DP-DT showed no evidence of differences compared with non–DP-DT in terms of a decrease in AUC of cross-institutional performance as compared with single-institutional performance (VinDr-CXR: 0.07 vs 0.07, P = .96; ChestX-ray14: 0.07 vs 0.06, P = .12; CheXpert: 0.07 vs 0.07, P = .18; UKA-CXR: 0.18 vs 0.18, P = .90; and PadChest: 0.07 vs 0.07, P = .35). Furthermore, AUC differences between DP-DT and non–DP-DT models were less than 1% for all sex subgroups (P > .33 for female and P > .22 for male, for all domains) and nearly all age subgroups (P > .16 for younger participants, P > .33 for adults, and P > .27 for older adults, for nearly all domains). Cross-institutional performance of artificial intelligence models was not affected by DP. Keywords: Convolutional Neural Network (CNN), Transfer Learning, Supervised Learning, Diagnosis, Forensics, Computer Applications–General, Image Postprocessing, Informatics, Neural Networks, Thorax, Computer-Aided Diagnosis, Deep Learning, Domain Transfer, Differential Privacy, Privacy-Preserving AI, Chest Radiograph Supplemental material is available for this article. © RSNA, 2023 See also the commentary by Suri and Summers in this issue.

Background: To compare denoising diffusion probabilistic models (DDPM) and generative adversarial networks (GAN) for recovering contrast-enhanced breast magnetic resonance imaging (MRI) subtraction images from virtual low-dose subtraction images. Methods: Retrospective, ethically approved study. DDPM- and GAN-reconstructed single-slice subtraction images of 50 breasts with enhancing lesions were compared to original ones at three dose levels (25%, 10%, 5%) using quantitative measures and radiologic evaluations. Two radiologists stated their preference based on the reconstruction quality and scored the lesion conspicuity as compared to the original, blinded to the model. Fifty lesion-free maximum intensity projections were evaluated for the presence of false-positives. Results were compared between models and dose levels, using generalized linear mixed models. Results: At 5% dose, both radiologists preferred the GAN-generated images, whereas at 25% dose, both radiologists preferred the DDPM-generated images. Median lesion conspicuity scores did not differ between GAN and DDPM at 25% dose (5 versus 5, p = 1.000) and 10% dose (4 versus 4, p = 1.000). At 5% dose, both readers assigned higher conspicuity to the GAN than to the DDPM (3 versus 2, p = 0.007). In the lesion-free examinations, DDPM and GAN showed no differences in the false-positive rate at 5% (15% versus 22%), 10% (10% versus 6%), and 25% (6% versus 4%) (p = 1.000). Conclusions: Both GAN and DDPM yielded promising results in low-dose image reconstruction. However, neither of them showed superior results over the other model for all dose levels and evaluation metrics. Further development is needed to counteract false-positives. Relevance statement: For MRI-based breast cancer screening, reducing the contrast agent dose is desirable. Diffusion probabilistic models and generative adversarial networks were capable of retrospectively enhancing the signal of low-dose images. Hence, they may supplement imaging with reduced doses in the future. Key points: • Deep learning may help recover signal in low-dose contrast-enhanced breast MRI. • Two models (DDPM and GAN) were trained at different dose levels. • Radiologists preferred DDPM at 25%, and GAN images at 5% dose. • Lesion conspicuity between DDPM and GAN was similar, except at 5% dose. • GAN and DDPM yield promising results in low-dose image reconstruction. Graphical

Purpose: Magnetic resonance fingerprinting (MRF) with spiral readout enables rapid quantification of tissue relaxation times. However, it is prone to blurring because of off‐resonance effects. Hence, fat blurring into adjacent regions might prevent identification of small tumors by their quantitative T₁ and T₂ values. This study aims to correct for the blurring artifacts, thereby enabling fast quantitative mapping in the female breast.Methods: The impact of fat blurring on spiral MRF results was first assessed by simulations. Then, MRF was combined with 3‐point Dixon water–fat separation and spiral blurring correction based on conjugate phase reconstruction. The approach was assessed in phantom experiments and compared to Cartesian reference measurements, namely inversion recovery (IR), multi‐echo spin echo (MESE), and Cartesian MRF, by normalized root‐mean‐square error (NRMSE) and SD calculations. Feasibility is further demonstrated in vivo for quantitative breast measurements of 6 healthy female volunteers, age range 24–31 y.Results: In the phantom experiment, the blurring correction reduced the NRMSE per phantom vial on average from 16% to 8% for T₁ and from 18% to 11% for T₂ when comparing spiral MRF to IR/MESE sequences. When comparing to Cartesian MRF, the NRMSE reduced from 15% to 8% for T₁ and from 12% to 7% for T₂. Furthermore, SDs decreased. In vivo, the blurring correction removed fat bias on T₁/T₂ from a rim of ~7–8 mm width adjacent to fatty structures.Conclusion: The blurring correction for spiral MRF yields improved quantitative maps in the presence of water and fat.

Electroporation (EP) is widely used in medicine, such as cancer treatment, in form of electrochemotherapy or irreversible electroporation (IRE). For EP device testing, living cells or tissue inside a living organism (including animals) are needed. Plant-based models seem to be a promising alternative to substitute animal models in research. The aim of this study is to find a suitable plant-based model for visual evaluation of IRE, and to compare the geometry of electroporated areas with in-vivo animal data. For this purpose, a variety of fruit and vegetables were selected and visually evaluated after 0/1/2/4/6/8/12/16/24 h post-EP. Apple and potato were found to be suitable models as they enabled a visual evaluation of the electroporated area. For these models, the size of the electroporated area was determined after 0/1/2/4/6/8/12/16/24 h. For apples, a well-defined electroporated area was visual within two hours, while in potatoes it reached a plateau after eight hours only. The electroporated area of apple, which showed the fastest visual results was then compared to a retrospectively evaluated swine liver IRE dataset which had been obtained for similar conditions. The electroporated area of the apple and swine liver both showed a spherical geometry of comparable size. For all experiments, the standard protocol for human liver IRE was followed. To conclude, potato and apple were found to be suitable plant-based models for the visual evaluation of electroporated area after irreversible EP, with apple being the best choice for fast visual results. Given the comparable range, the size of the electroporated area of the apple may be promising as a quantitative predictor in animal tissue. Even if plant-based models cannot completely replace animal experiments, they can be used in the early stages of EP device development and testing, decreasing animal experiments to the necessary minimum.