Mojtaba Safari

Yuheng Li, Yuan Gao, Haoyu Dong et al.

Computed tomography (CT) is a central to three-dimensional medical imaging, yet CT-based artificial intelligence remains fragmented across task-specific models for segmentation, classification, registration, and report analysis. Here we present FlexiCT, a family of CT foundation models trained by agglomerative continual pretraining on 266,227 CT volumes from 56 publicly available datasets, forming a large-scale public resource for CT representation learning. FlexiCT uses agglomerative pretraining across three stages: two-dimensional axial pretraining, three-dimensional anatomical pretraining and report-guided semantic alignment. This training strategy supports slice-level, volume-level and vision-language analysis. Across five downstream task families (segmentation, classification, registration, vision-language understanding and clinical retrieval), FlexiCT matches or exceeds prior task-specific approaches on multiple benchmarks. Its embeddings further organize CT scans along gradients associated with various tumor stages, suggesting that CT foundation models can capture imaging features relevant to disease phenotype characterization. Code is available at https://github.com/ricklisz/FlexiCT

Zach Eidex, Mojtaba Safari, Jacob Wynne et al.

Purpose: Apparent diffusion coefficient (ADC) maps derived from diffusion weighted (DWI) MRI provides functional measurements about the water molecules in tissues. However, DWI is time consuming and very susceptible to image artifacts, leading to inaccurate ADC measurements. This study aims to develop a deep learning framework to synthesize ADC maps from multi-parametric MR images. Methods: We proposed the multiparametric residual vision transformer model (MPR-ViT) that leverages the long-range context of ViT layers along with the precision of convolutional operators. Residual blocks throughout the network significantly increasing the representational power of the model. The MPR-ViT model was applied to T1w and T2- fluid attenuated inversion recovery images of 501 glioma cases from a publicly available dataset including preprocessed ADC maps. Selected patients were divided into training (N=400), validation (N=50) and test (N=51) sets, respectively. Using the preprocessed ADC maps as ground truth, model performance was evaluated and compared against the Vision Convolutional Transformer (VCT) and residual vision transformer (ResViT) models. Results: The results are as follows using T1w + T2-FLAIR MRI as inputs: MPR-ViT - PSNR: 31.0 +/- 2.1, MSE: 0.009 +/- 0.0005, SSIM: 0.950 +/- 0.015. In addition, ablation studies showed the relative impact on performance of each input sequence. Both qualitative and quantitative results indicate that the proposed MR- ViT model performs favorably against the ground truth data. Conclusion: We show that high-quality ADC maps can be synthesized from structural MRI using a MPR- VCT model. Our predicted images show better conformality to the ground truth volume than ResViT and VCT predictions. These high-quality synthetic ADC maps would be particularly useful for disease diagnosis and intervention, especially when ADC maps have artifacts or are unavailable.

Zach Eidex, Mojtaba Safari, Richard L. J. Qiu et al.

Objective: Gadolinium-based contrast agents (GBCAs) are commonly used in MRI scans of patients with gliomas to enhance brain tumor characterization using T1-weighted (T1W) MRI. However, there is growing concern about GBCA toxicity. This study develops a deep-learning framework to generate T1-postcontrast (T1C) from pre-contrast multiparametric MRI. Approach: We propose the tumor-aware vision transformer (TA-ViT) model that predicts high-quality T1C images. The predicted tumor region is significantly improved (P < .001) by conditioning the transformer layers from predicted segmentation maps through adaptive layer norm zero mechanism. The predicted segmentation maps were generated with the multi-parametric residual (MPR) ViT model and transformed into a latent space to produce compressed, feature-rich representations. The TA-ViT model predicted T1C MRI images of 501 glioma cases. Selected patients were split into training (N=400), validation (N=50), and test (N=51) sets. Main Results: Both qualitative and quantitative results demonstrate that the TA-ViT model performs superior against the benchmark MRP-ViT model. Our method produces synthetic T1C MRI with high soft tissue contrast and more accurately reconstructs both the tumor and whole brain volumes. The synthesized T1C images achieved remarkable improvements in both tumor and healthy tissue regions compared to the MRP-ViT model. For healthy tissue and tumor regions, the results were as follows: NMSE: 8.53 +/- 4.61E-4; PSNR: 31.2 +/- 2.2; NCC: 0.908 +/- .041 and NMSE: 1.22 +/- 1.27E-4, PSNR: 41.3 +/- 4.7, and NCC: 0.879 +/- 0.042, respectively. Significance: The proposed method generates synthetic T1C images that closely resemble real T1C images. Future development and application of this approach may enable contrast-agent-free MRI for brain tumor patients, eliminating the risk of GBCA toxicity and simplifying the MRI scan protocol.

Mojtaba Safari, Shansong Wang, Vanessa L Wildman et al.

Background: High-resolution MRI is critical for diagnosis, but long acquisition times limit clinical use. Super-resolution (SR) can enhance resolution post-scan, yet existing deep learning methods face fidelity-efficiency trade-offs. Purpose: To develop a computationally efficient and accurate deep learning framework for MRI SR that preserves anatomical detail for clinical integration. Materials and Methods: We propose a novel SR framework combining multi-head selective state-space models (MHSSM) with a lightweight channel MLP. The model uses 2D patch extraction with hybrid scanning to capture long-range dependencies. Each MambaFormer block integrates MHSSM, depthwise convolutions, and gated channel mixing. Evaluation used 7T brain T1 MP2RAGE maps (n=142) and 1.5T prostate T2w MRI (n=334). Comparisons included Bicubic interpolation, GANs (CycleGAN, Pix2pix, SPSR), transformers (SwinIR), Mamba (MambaIR), and diffusion models (I2SB, Res-SRDiff). Results: Our model achieved superior performance with exceptional efficiency. For 7T brain data: SSIM=0.951+-0.021, PSNR=26.90+-1.41 dB, LPIPS=0.076+-0.022, GMSD=0.083+-0.017, significantly outperforming all baselines (p<0.001). For prostate data: SSIM=0.770+-0.049, PSNR=27.15+-2.19 dB, LPIPS=0.190+-0.095, GMSD=0.087+-0.013. The framework used only 0.9M parameters and 57 GFLOPs, reducing parameters by 99.8% and computation by 97.5% versus Res-SRDiff, while outperforming SwinIR and MambaIR in accuracy and efficiency. Conclusion: The proposed framework provides an efficient, accurate MRI SR solution, delivering enhanced anatomical detail across datasets. Its low computational demand and state-of-the-art performance show strong potential for clinical translation.

Mojtaba Safari, Zach Eidex, Chih-Wei Chang et al.

Magnetic resonance imaging (MRI) is a non-invasive imaging modality and provides comprehensive anatomical and functional insights into the human body. However, its long acquisition times can lead to patient discomfort, motion artifacts, and limiting real-time applications. To address these challenges, strategies such as parallel imaging have been applied, which utilize multiple receiver coils to speed up the data acquisition process. Additionally, compressed sensing (CS) is a method that facilitates image reconstruction from sparse data, significantly reducing image acquisition time by minimizing the amount of data collection needed. Recently, deep learning (DL) has emerged as a powerful tool for improving MRI reconstruction. It has been integrated with parallel imaging and CS principles to achieve faster and more accurate MRI reconstructions. This review comprehensively examines DL-based techniques for MRI reconstruction. We categorize and discuss various DL-based methods, including end-to-end approaches, unrolled optimization, and federated learning, highlighting their potential benefits. Our systematic review highlights significant contributions and underscores the potential of DL in MRI reconstruction. Additionally, we summarize key results and trends in DL-based MRI reconstruction, including quantitative metrics, the dataset, acceleration factors, and the progress of and research interest in DL techniques over time. Finally, we discuss potential future directions and the importance of DL-based MRI reconstruction in advancing medical imaging. To facilitate further research in this area, we provide a GitHub repository that includes up-to-date DL-based MRI reconstruction publications and public datasets-https://github.com/mosaf/Awesome-DL-based-CS-MRI.

Mojtaba Safari, Shansong Wang, Zach Eidex et al.

Objective:This study introduces a residual error-shifting mechanism that drastically reduces sampling steps while preserving critical anatomical details, thus accelerating MRI reconstruction. Approach:We propose a novel diffusion-based SR framework called Res-SRDiff, which integrates residual error shifting into the forward diffusion process. This enables efficient HR image reconstruction by aligning the degraded HR and LR distributions.We evaluated Res-SRDiff on ultra-high-field brain T1 MP2RAGE maps and T2-weighted prostate images, comparing it with Bicubic, Pix2pix, CycleGAN, and a conventional denoising diffusion probabilistic model with vision transformer backbone (TM-DDPM), using quantitative metrics such as peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), gradient magnitude similarity deviation (GMSD), and learned perceptual image patch similarity (LPIPS). Main results: Res-SRDiff significantly outperformed all comparative methods in terms of PSNR, SSIM, and GMSD across both datasets, with statistically significant improvements (p-values<<0.05). The model achieved high-fidelity image restoration with only four sampling steps, drastically reducing computational time to under one second per slice, which is substantially faster than conventional TM-DDPM with around 20 seconds per slice. Qualitative analyses further demonstrated that Res-SRDiff effectively preserved fine anatomical details and lesion morphology in both brain and pelvic MRI images. Significance: Our findings show that Res-SRDiff is an efficient and accurate MRI SR method, markedly improving computational efficiency and image quality. Integrating residual error shifting into the diffusion process allows for rapid and robust HR image reconstruction, enhancing clinical MRI workflows and advancing medical imaging research. The source at:https://github.com/mosaf/Res-SRDiff

Yifei Zhang, Jiashuo Zhang, Mojtaba Safari et al.

Low-dose chest computed tomography (LDCT) inherently captures both pulmonary and cardiac structures, offering a unique opportunity for joint assessment of lung and cardiovascular health. However, most existing approaches treat these domains as independent tasks, overlooking their physiological interplay and shared imaging biomarkers. We propose an Explainable Cross-Disease Reasoning Framework that enables interpretable cardiopulmonary risk assessment from a single LDCT scan. The framework introduces an agentic reasoning process that emulates clinical diagnostic thinking-first perceiving pulmonary findings, then reasoning through established medical knowledge, and finally deriving a cardiovascular judgment with explanatory rationale. It integrates three synergistic components: a pulmonary perception module that summarizes lung abnormalities, a knowledge-guided reasoning module that infers their cardiovascular implications, and a cardiac representation module that encodes structural biomarkers. Their outputs are fused to produce a holistic cardiovascular risk prediction that is both accurate and physiologically grounded. Experiments on the NLST cohort demonstrate that the proposed framework achieves state-of-the-art performance for CVD screening and mortality prediction, outperforming single-disease and purely image-based baselines. Beyond quantitative gains, the framework provides human-verifiable reasoning that aligns with cardiological understanding, revealing coherent links between pulmonary abnormalities and cardiac stress mechanisms. Overall, this work establishes a unified and explainable paradigm for cardiovascular analysis from LDCT, bridging the gap between image-based prediction and mechanism-based medical interpretation.

Mojtaba Safari, Shansong Wang, Zach Eidex et al.

Background: MRI is crucial for brain imaging but is highly susceptible to motion artifacts due to long acquisition times. This study introduces PI-MoCoNet, a physics-informed motion correction network that integrates spatial and k-space information to remove motion artifacts without explicit motion parameter estimation, enhancing image fidelity and diagnostic reliability. Materials and Methods: PI-MoCoNet consists of a motion detection network (U-net with spatial averaging) to identify corrupted k-space lines and a motion correction network (U-net with Swin Transformer blocks) to reconstruct motion-free images. The correction is guided by three loss functions: reconstruction (L1), perceptual (LPIPS), and data consistency (Ldc). Motion artifacts were simulated via rigid phase encoding perturbations and evaluated on IXI and MR-ART datasets against Pix2Pix, CycleGAN, and U-net using PSNR, SSIM, and NMSE. Results: PI-MoCoNet significantly improved image quality. On IXI, for minor artifacts, PSNR increased from 34.15 dB to 45.95 dB, SSIM from 0.87 to 1.00, and NMSE reduced from 0.55% to 0.04%. For moderate artifacts, PSNR improved from 30.23 dB to 42.16 dB, SSIM from 0.80 to 0.99, and NMSE from 1.32% to 0.09%. For heavy artifacts, PSNR rose from 27.99 dB to 36.01 dB, SSIM from 0.75 to 0.97, and NMSE decreased from 2.21% to 0.36%. On MR-ART, PI-MoCoNet achieved PSNR gains of ~10 dB and SSIM improvements of up to 0.20, with NMSE reductions of ~6%. Ablation studies confirmed the importance of data consistency and perceptual losses, yielding a 1 dB PSNR gain and 0.17% NMSE reduction. Conclusions: PI-MoCoNet effectively mitigates motion artifacts in brain MRI, outperforming existing methods. Its ability to integrate spatial and k-space information makes it a promising tool for clinical use in motion-prone settings. Code: https://github.com/mosaf/PI-MoCoNet.git.

Shansong Wang, Mingzhe Hu, Qiang Li et al.

Recent advances in large language models (LLMs) have enabled general-purpose systems to perform increasingly complex domain-specific reasoning without extensive fine-tuning. In the medical domain, decision-making often requires integrating heterogeneous information sources, including patient narratives, structured data, and medical images. This study positions GPT-5 as a generalist multimodal reasoner for medical decision support and systematically evaluates its zero-shot chain-of-thought reasoning performance on both text-based question answering and visual question answering tasks under a unified protocol. We benchmark GPT-5, GPT-5-mini, GPT-5-nano, and GPT-4o-2024-11-20 against standardized splits of MedQA, MedXpertQA (text and multimodal), MMLU medical subsets, USMLE self-assessment exams, and VQA-RAD. Results show that GPT-5 consistently outperforms all baselines, achieving state-of-the-art accuracy across all QA benchmarks and delivering substantial gains in multimodal reasoning. On MedXpertQA MM, GPT-5 improves reasoning and understanding scores by +29.26% and +26.18% over GPT-4o, respectively, and surpasses pre-licensed human experts by +24.23% in reasoning and +29.40% in understanding. In contrast, GPT-4o remains below human expert performance in most dimensions. A representative case study demonstrates GPT-5's ability to integrate visual and textual cues into a coherent diagnostic reasoning chain, recommending appropriate high-stakes interventions. Our results show that, on these controlled multimodal reasoning benchmarks, GPT-5 moves from human-comparable to above human-expert performance. This improvement may substantially inform the design of future clinical decision-support systems.

Yizhou Wu, Shansong Wang, Yuheng Li et al.

Brain MRI underpins a wide range of neuroscientific and clinical applications, yet most learning-based methods remain task-specific and require substantial labeled data. Here we show that a single self-supervised representation can generalize across heterogeneous brain MRI endpoints. We trained BrainDINO, a self-distilled foundation model, on approximately 6.6 million unlabeled axial slices from 20 datasets encompassing broad variation in population, disease, and acquisition setting. Using a frozen encoder with lightweight task heads, BrainDINO supported transfer across tumor segmentation, neurodegenerative and neurodevelopmental conditions classification, brain age estimation, post-stroke temporal prediction, molecular status prediction, MRI sequence classification, and survival modeling. Across tasks and supervision regimes, BrainDINO consistently equaled or exceeded natural-image and MRI-specific self-supervised baselines, with particularly strong advantages under label scarcity. Representation analyses further showed anatomically organized and pathology-sensitive feature structure in the absence of task-specific supervision. Our findings indicate that large-scale slice-wise self-supervised learning can yield a unified brain MRI representation that supports diverse neuroimaging tasks without volumetric pretraining or full-network fine-tuning, establishing a scalable foundation for robust and data-efficient brain imaging analysis.

Shansong Wang, Mojtaba Safari, Qiang Li et al.

Vision foundation models (VFMs) are pre-trained on extensive image datasets to learn general representations for diverse types of data. These models can subsequently be fine-tuned for specific downstream tasks, significantly boosting performance across a broad range of applications. However, existing vision foundation models that claim to be applicable to various clinical tasks are mostly pre-trained on 3D computed tomography (CT), which benefits from the availability of extensive 3D CT databases. Significant differences between CT and magnetic resonance imaging (MRI) in imaging principles, signal characteristics, and data distribution may hinder their practical performance and versatility in MRI-specific applications. Here, we propose Triad, a vision foundation model for 3D MRI. Triad adopts a widely used autoencoder architecture to learn robust representations from 131,170 3D MRI volumes and uses organ-independent imaging descriptions to constrain the semantic distribution of the visual modality. The above pre-training dataset is called Triad-131K, which is currently the largest 3D MRI pre-training dataset. We evaluate Triad across three tasks, namely, organ/tumor segmentation, organ/cancer classification, and medical image registration, in two data modalities (within-domain and out-of-domain) settings using 25 downstream datasets. By initializing models with Triad's pre-trained weights, nnUNet-Triad improves segmentation performance by 2.51% compared to nnUNet-Scratch across 17 datasets. Swin-B-Triad achieves a 3.97% improvement over Swin-B-Scratch in classification tasks across five datasets. SwinUNETR-Triad improves by 4.00% compared to SwinUNETR-Scratch in registration tasks across two datasets. Our study demonstrates that pre-training can improve performance when the data modalities and organs of upstream and downstream tasks are consistent.

Shansong Wang, Zhecheng Jin, Mingzhe Hu et al.

CLIP models pretrained on natural images with billion-scale image-text pairs have demonstrated impressive capabilities in zero-shot classification, cross-modal retrieval, and open-ended visual answering. However, transferring this success to biomedicine is hindered by the scarcity of large-scale biomedical image-text corpora, the heterogeneity of image modalities, and fragmented data standards across institutions. These limitations hinder the development of a unified and generalizable biomedical foundation model trained from scratch. To overcome this, we introduce MMKD-CLIP, a generalist biomedical foundation model developed via Multiple Medical CLIP Knowledge Distillation. Rather than relying on billion-scale raw data, MMKD-CLIP distills knowledge from nine state-of-the-art domain-specific or generalist biomedical CLIP models, each pretrained on millions of biomedical image-text pairs. Our two-stage training pipeline first performs CLIP-style pretraining on over 2.9 million biomedical image-text pairs from 26 image modalities, followed by feature-level distillation using over 19.2 million feature pairs extracted from teacher models. We evaluate MMKD-CLIP on 58 diverse biomedical datasets, encompassing over 10.8 million biomedical images across nine image modalities. The evaluation spans six core task types: zero-shot classification, linear probing, cross-modal retrieval, visual question answering, survival prediction, and cancer diagnosis. MMKD-CLIP consistently outperforms all teacher models while demonstrating remarkable robustness and generalization across image domains and task settings. These results underscore that multi-teacher knowledge distillation is a scalable and effective paradigm for building high-performing biomedical foundation models under the practical constraints of real-world data availability.

Shansong Wang, Mojtaba Safari, Mingzhe Hu et al.

Prior medical image registration approaches, particularly learning-based methods, often require large amounts of training data, which constrains clinical adoption. To overcome this limitation, we propose a training-free pipeline that relies on a frozen DINOv3 encoder and test-time optimization of the deformation field in feature space. Across two representative benchmarks, the method is accurate and yields regular deformations. On Abdomen MR-CT, it attained the best mean Dice score (DSC) of 0.790 together with the lowest 95th percentile Hausdorff Distance (HD95) of 4.9+-5.0 and the lowest standard deviation of Log-Jacobian (SDLogJ) of 0.08+-0.02. On ACDC cardiac MRI, it improves mean DSC to 0.769 and reduces SDLogJ to 0.11 and HD95 to 4.8, a marked gain over the initial alignment. The results indicate that operating in a compact foundation feature space at test time offers a practical and general solution for clinical registration without additional training.

Mingzhe Hu, Zach Eidex, Shansong Wang et al.

Radiology, radiation oncology, and medical physics require decision-making that integrates medical images, textual reports, and quantitative data under high-stakes conditions. With the introduction of GPT-5, it is critical to assess whether recent advances in large multimodal models translate into measurable gains in these safety-critical domains. We present a targeted zero-shot evaluation of GPT-5 and its smaller variants (GPT-5-mini, GPT-5-nano) against GPT-4o across three representative tasks. We present a targeted zero-shot evaluation of GPT-5 and its smaller variants (GPT-5-mini, GPT-5-nano) against GPT-4o across three representative tasks: (1) VQA-RAD, a benchmark for visual question answering in radiology; (2) SLAKE, a semantically annotated, multilingual VQA dataset testing cross-modal grounding; and (3) a curated Medical Physics Board Examination-style dataset of 150 multiple-choice questions spanning treatment planning, dosimetry, imaging, and quality assurance. Across all datasets, GPT-5 achieved the highest accuracy, with substantial gains over GPT-4o up to +20.00% in challenging anatomical regions such as the chest-mediastinal, +13.60% in lung-focused questions, and +11.44% in brain-tissue interpretation. On the board-style physics questions, GPT-5 attained 90.7% accuracy (136/150), exceeding the estimated human passing threshold, while GPT-4o trailed at 78.0%. These results demonstrate that GPT-5 delivers consistent and often pronounced performance improvements over GPT-4o in both image-grounded reasoning and domain-specific numerical problem-solving, highlighting its potential to augment expert workflows in medical imaging and therapeutic physics.

Mojtaba Safari, Shansong Wang, Mingzhe Hu et al.

Accurate differentiation of brain tumor types on magnetic resonance imaging (MRI) is critical for guiding treatment planning in neuro-oncology. Recent advances in large language models (LLMs) have enabled visual question answering (VQA) approaches that integrate image interpretation with natural language reasoning. In this study, we evaluated GPT-4o, GPT-5-nano, GPT-5-mini, and GPT-5 on a curated brain tumor VQA benchmark derived from 3 Brain Tumor Segmentation (BraTS) datasets - glioblastoma (GLI), meningioma (MEN), and brain metastases (MET). Each case included multi-sequence MRI triplanar mosaics and structured clinical features transformed into standardized VQA items. Models were assessed in a zero-shot chain-of-thought setting for accuracy on both visual and reasoning tasks. Results showed that GPT-5-mini achieved the highest macro-average accuracy (44.19%), followed by GPT-5 (43.71%), GPT-4o (41.49%), and GPT-5-nano (35.85%). Performance varied by tumor subtype, with no single model dominating across all cohorts. These findings suggest that GPT-5 family models can achieve moderate accuracy in structured neuro-oncological VQA tasks, but not at a level acceptable for clinical use.

Zach Eidex, Mojtaba Safari, Tonghe Wang et al.

Purpose: Ultra-high-field 7T MRI offers improved resolution and contrast over standard clinical field strengths (1.5T, 3T). However, 7T scanners are costly, scarce, and introduce additional challenges such as susceptibility artifacts. We propose an efficient transformer-based model (7T-Restormer) to synthesize 7T-quality T1-maps from routine 1.5T or 3T T1-weighted (T1W) images. Methods: Our model was validated on 35 1.5T and 108 3T T1w MRI paired with corresponding 7T T1 maps of patients with confirmed MS. A total of 141 patient cases (32,128 slices) were randomly divided into 105 (25; 80) training cases (19,204 slices), 19 (5; 14) validation cases (3,476 slices), and 17 (5; 14) test cases (3,145 slices) where (X; Y) denotes the patients with 1.5T and 3T T1W scans, respectively. The synthetic 7T T1 maps were compared against the ResViT and ResShift models. Results: The 7T-Restormer model achieved a PSNR of 26.0 +/- 4.6 dB, SSIM of 0.861 +/- 0.072, and NMSE of 0.019 +/- 0.011 for 1.5T inputs, and 25.9 +/- 4.9 dB, and 0.866 +/- 0.077 for 3T inputs, respectively. Using 10.5 M parameters, our model reduced NMSE by 64 % relative to 56.7M parameter ResShift (0.019 vs 0.052, p = <.001 and by 41 % relative to 70.4M parameter ResViT (0.019 vs 0.032, p = <.001) at 1.5T, with similar advantages at 3T (0.021 vs 0.060 and 0.033; p < .001). Training with a mixed 1.5 T + 3 T corpus was superior to single-field strategies. Restricting the model to 1.5T increased the 1.5T NMSE from 0.019 to 0.021 (p = 1.1E-3) while training solely on 3T resulted in lower performance on input 1.5T T1W MRI. Conclusion: We propose a novel method for predicting quantitative 7T MP2RAGE maps from 1.5T and 3T T1W scans with higher quality than existing state-of-the-art methods. Our approach makes the benefits of 7T MRI more accessible to standard clinical workflows.

Alexandru Florea, Shansong Wang, Mingzhe Hu et al.

The transition from task-specific artificial intelligence toward general-purpose foundation models raises fundamental questions about their capacity to support the integrated reasoning required in clinical medicine, where diagnosis demands synthesis of ambiguous patient narratives, laboratory data, and multimodal imaging. This landscape commentary provides the first controlled, cross-sectional evaluation of the GPT-5 family (GPT-5, GPT-5 Mini, GPT-5 Nano) against its predecessor GPT-4o across a diverse spectrum of clinically grounded tasks, including medical education examinations, text-based reasoning benchmarks, and visual question-answering in neuroradiology, digital pathology, and mammography using a standardized zero-shot chain-of-thought protocol. GPT-5 demonstrated substantial gains in expert-level textual reasoning, with absolute improvements exceeding 25 percentage-points on MedXpertQA. When tasked with multimodal synthesis, GPT-5 effectively leveraged this enhanced reasoning capacity to ground uncertain clinical narratives in concrete imaging evidence, achieving state-of-the-art or competitive performance across most VQA benchmarks and outperforming GPT-4o by margins of 10-40% in mammography tasks requiring fine-grained lesion characterization. However, performance remained moderate in neuroradiology (44% macro-average accuracy) and lagged behind domain-specific models in mammography, where specialized systems exceed 80% accuracy compared to GPT-5's 52-64%. These findings indicate that while GPT-5 represents a meaningful advance toward integrated multimodal clinical reasoning, mirroring the clinician's cognitive process of biasing uncertain information with objective findings, generalist models are not yet substitutes for purpose-built systems in highly specialized, perception-critical tasks.

Praveenbalaji Rajendran, Mojtaba Safari, Wenfeng He et al.

Recent advancements in artificial intelligence (AI), particularly foundation models (FMs), have revolutionized medical image analysis, demonstrating strong zero- and few-shot performance across diverse medical imaging tasks, from segmentation to report generation. Unlike traditional task-specific AI models, FMs leverage large corpora of labeled and unlabeled multimodal datasets to learn generalized representations that can be adapted to various downstream clinical applications with minimal fine-tuning. However, despite the rapid proliferation of FM research in medical imaging, the field remains fragmented, lacking a unified synthesis that systematically maps the evolution of architectures, training paradigms, and clinical applications across modalities. To address this gap, this review article provides a comprehensive and structured analysis of FMs in medical image analysis. We systematically categorize studies into vision-only and vision-language FMs based on their architectural foundations, training strategies, and downstream clinical tasks. Additionally, a quantitative meta-analysis of the studies was conducted to characterize temporal trends in dataset utilization and application domains. We also critically discuss persistent challenges, including domain adaptation, efficient fine-tuning, computational constraints, and interpretability along with emerging solutions such as federated learning, knowledge distillation, and advanced prompting. Finally, we identify key future research directions aimed at enhancing the robustness, explainability, and clinical integration of FMs, thereby accelerating their translation into real-world medical practice.

Zach Eidex, Mojtaba Safari, Jie Ding et al.

Objective: Gadolinium-based contrast agents (GBCAs) are commonly employed with T1w MRI to enhance lesion visualization but are restricted in patients at risk of nephrogenic systemic fibrosis and variations in GBCA administration can introduce imaging inconsistencies. This study develops an efficient 3D deep-learning framework to generate T1-contrast enhanced images (T1C) from pre-contrast multiparametric MRI. Approach: We propose the 3D latent rectified flow (T1C-RFlow) model for generating high-quality T1C images. First, T1w and T2-FLAIR images are input into a pretrained autoencoder to acquire an efficient latent space representation. A rectified flow diffusion model is then trained in this latent space representation. The T1C-RFlow model was trained on a curated dataset comprised of the BraTS 2024 glioma (GLI; 1480 patients), meningioma (MEN; 1141 patients), and metastases (MET; 1475 patients) datasets. Selected patients were split into train (N=2860), validation (N=612), and test (N=614) sets. Results: Both qualitative and quantitative results demonstrate that the T1C-RFlow model outperforms benchmark 3D models (pix2pix, DDPM, Diffusion Transformers (DiT-3D)) trained in the same latent space. T1C-RFlow achieved the following metrics - GLI: NMSE 0.044 +/- 0.047, SSIM 0.935 +/- 0.025; MEN: NMSE 0.046 +/- 0.029, SSIM 0.937 +/- 0.021; MET: NMSE 0.098 +/- 0.088, SSIM 0.905 +/- 0.082. T1C-RFlow had the best tumor reconstruction performance and significantly faster denoising times (6.9 s/volume, 200 steps) than conventional DDPM models in both latent space (37.7s, 1000 steps) and patch-based in image space (4.3 hr/volume). Significance: Our proposed method generates synthetic T1C images that closely resemble ground truth T1C in much less time than previous diffusion models. Further development may permit a practical method for contrast-agent-free MRI for brain tumors.

Mojtaba Safari, Zach Eidex, Richard L. J. Qiu et al.

Background: To systematically review and perform a meta-analysis of artificial intelligence (AI)-driven methods for detecting and correcting magnetic resonance imaging (MRI) motion artifacts, assessing current developments, effectiveness, challenges, and future research directions. Methods: A comprehensive systematic review and meta-analysis were conducted, focusing on deep learning (DL) approaches, particularly generative models, for the detection and correction of MRI motion artifacts. Quantitative data were extracted regarding utilized datasets, DL architectures, and performance metrics. Results: DL, particularly generative models, show promise for reducing motion artifacts and improving image quality; however, limited generalizability, reliance on paired training data, and risk of visual distortions remain key challenges that motivate standardized datasets and reporting. Conclusions: AI-driven methods, particularly DL generative models, show significant potential for improving MRI image quality by effectively addressing motion artifacts. However, critical challenges must be addressed, including the need for comprehensive public datasets, standardized reporting protocols for artifact levels, and more advanced, adaptable DL techniques to reduce reliance on extensive paired datasets. Addressing these aspects could substantially enhance MRI diagnostic accuracy, reduce healthcare costs, and improve patient care outcomes.

Qiang Li, Shansong Wang, Mingzhe Hu et al.

Mammogram visual question answering (VQA) integrates image interpretation with clinical reasoning and has potential to support breast cancer screening. We systematically evaluated the GPT-5 family and GPT-4o model on four public mammography datasets (EMBED, InBreast, CMMD, CBIS-DDSM) for BI-RADS assessment, abnormality detection, and malignancy classification tasks. GPT-5 consistently was the best performing model but lagged behind both human experts and domain-specific fine-tuned models. On EMBED, GPT-5 achieved the highest scores among GPT variants in density (56.8%), distortion (52.5%), mass (64.5%), calcification (63.5%), and malignancy (52.8%) classification. On InBreast, it attained 36.9% BI-RADS accuracy, 45.9% abnormality detection, and 35.0% malignancy classification. On CMMD, GPT-5 reached 32.3% abnormality detection and 55.0% malignancy accuracy. On CBIS-DDSM, it achieved 69.3% BI-RADS accuracy, 66.0% abnormality detection, and 58.2% malignancy accuracy. Compared with human expert estimations, GPT-5 exhibited lower sensitivity (63.5%) and specificity (52.3%). While GPT-5 exhibits promising capabilities for screening tasks, its performance remains insufficient for high-stakes clinical imaging applications without targeted domain adaptation and optimization. However, the tremendous improvements in performance from GPT-4o to GPT-5 show a promising trend in the potential for general large language models (LLMs) to assist with mammography VQA tasks.

Mojtaba Safari, Shansong Wang, Qiang Li et al.

Objective. Motion artifacts in brain MRI, mainly from rigid head motion, degrade image quality and hinder downstream applications. Conventional methods to mitigate these artifacts, including repeated acquisitions or motion tracking, impose workflow burdens. This study introduces Res-MoCoDiff, an efficient denoising diffusion probabilistic model specifically designed for MRI motion artifact correction.Approach.Res-MoCoDiff exploits a novel residual error shifting mechanism during the forward diffusion process to incorporate information from motion-corrupted images. This mechanism allows the model to simulate the evolution of noise with a probability distribution closely matching that of the corrupted data, enabling a reverse diffusion process that requires only four steps. The model employs a U-net backbone, with attention layers replaced by Swin Transformer blocks, to enhance robustness across resolutions. Furthermore, the training process integrates a combined l1+l2 loss function, which promotes image sharpness and reduces pixel-level errors. Res-MoCoDiff was evaluated on both an in-silico dataset generated using a realistic motion simulation framework and an in-vivo MR-ART dataset. Comparative analyses were conducted against established methods, including CycleGAN, Pix2pix, and a diffusion model with a vision transformer backbone, using quantitative metrics such as PSNR, SSIM, and NMSE.Main results. The proposed method demonstrated superior performance in removing motion artifacts across minor, moderate, and heavy distortion levels. Res-MoCoDiff consistently achieved the highest SSIM and the lowest NMSE values, with a PSNR of up to 41.91+-2.94 dB for minor distortions. Notably, the average sampling time was reduced to 0.37 seconds per batch of two image slices, compared with 101.74 seconds for conventional approaches.

Mojtaba Safari, Zach Eidex, Shaoyan Pan et al.

Purpose: To propose a self-supervised deep learning-based compressed sensing MRI (DL-based CS-MRI) method named "Adaptive Self-Supervised Consistency Guided Diffusion Model (ASSCGD)" to accelerate data acquisition without requiring fully sampled datasets. Materials and Methods: We used the fastMRI multi-coil brain axial T2-weighted (T2-w) dataset from 1,376 cases and single-coil brain quantitative magnetization prepared 2 rapid acquisition gradient echoes (MP2RAGE) T1 maps from 318 cases to train and test our model. Robustness against domain shift was evaluated using two out-of-distribution (OOD) datasets: multi-coil brain axial postcontrast T1 -weighted (T1c) dataset from 50 cases and axial T1-weighted (T1-w) dataset from 50 patients. Data were retrospectively subsampled at acceleration rates R in {2x, 4x, 8x}. ASSCGD partitions a random sampling pattern into two disjoint sets, ensuring data consistency during training. We compared our method with ReconFormer Transformer and SS-MRI, assessing performance using normalized mean squared error (NMSE), peak signal-to-noise ratio (PSNR), and structural similarity index (SSIM). Statistical tests included one-way analysis of variance (ANOVA) and multi-comparison Tukey's Honesty Significant Difference (HSD) tests. Results: ASSCGD preserved fine structures and brain abnormalities visually better than comparative methods at R = 8x for both multi-coil and single-coil datasets. It achieved the lowest NMSE at R in {4x, 8x}, and the highest PSNR and SSIM values at all acceleration rates for the multi-coil dataset. Similar trends were observed for the single-coil dataset, though SSIM values were comparable to ReconFormer at R in {2x, 8x}. These results were further confirmed by the voxel-wise correlation scatter plots. OOD results showed significant (p << 10^-5 ) improvements in undersampled image quality after reconstruction.