Providing effective treatment and making informed clinical decisions are essential goals of modern medicine and clinical care. We are interested in simulating disease dynamics for clinical decision-making, leveraging recent advances in large generative models. To this end, we introduce the Medical World Model (MeWM), the first world model in medicine that visually predicts future disease states based on clinical decisions. MeWM comprises (i) vision-language models to serve as policy models, and (ii) tumor generative models as dynamics models. The policy model generates action plans, such as clinical treatments, while the dynamics model simulates tumor progression or regression under given treatment conditions. Building on this, we propose the inverse dynamics model that applies survival analysis to the simulated post-treatment tumor, enabling the evaluation of treatment efficacy and the selection of the optimal clinical action plan. As a result, the proposed MeWM simulates disease dynamics by synthesizing post-treatment tumors, with state-of-the-art specificity in Turing tests evaluated by radiologists. Simultaneously, its inverse dynamics model outperforms medical-specialized GPTs in optimizing individualized treatment protocols across all metrics. Notably, MeWM improves clinical decision-making for interventional physicians, boosting F1-score in selecting the optimal TACE protocol by 13%, paving the way for future integration of medical world models as the second readers.

Multiparametric Magnetic Resonance Imaging (mpMRI) is the gold standard for diagnosing brain tumors, especially gliomas, which are difficult to segment due to their heterogeneity and varied sub-regions. While manual segmentation is time-consuming and error-prone, Deep Learning (DL) automates the process with greater accuracy and speed. We conducted ablation studies on surveyed articles to evaluate the impact of "add-on" modules-addressing challenges like spatial information loss, class imbalance, and overfitting-on glioma segmentation performance. Advanced modules-such as atrous (dilated) convolutions, inception, attention, transformer, and hybrid modules-significantly enhance segmentation accuracy, efficiency, multiscale feature extraction, and boundary delineation, while lightweight modules reduce computational complexity. Experiments on the Brain Tumor Segmentation (BraTS) dataset (comprising low- and high-grade gliomas) confirm their robustness, with top-performing models achieving high Dice score for tumor sub-regions. This survey underscores the need for optimal module selection and placement to balance speed, accuracy, and interpretability in glioma segmentation. Future work should focus on improving model interpretability, lowering computational costs, and boosting generalizability. Tools like NeuroQuant® and Raidionics demonstrate potential for clinical translation. Further refinement could enable regulatory approval, advancing precision in brain tumor diagnosis and treatment planning.

Multiple parametric imaging in positron emission tomography (PET) is challenging due to the noisy dynamic data and the complex mapping to kinetic parameters. Although methods like direct parametric reconstruction have been proposed to improve the image quality, limitations persist, particularly for nonlinear and small-value micro-parameters (e.g., k₂, k₃). This study presents a novel unsupervised deep learning approach to reconstruct and improve the quality of these micro-parameters. We proposed a direct parametric image reconstruction model, DIP-PM, integrating deep image prior (DIP) with a parameter magnification (PM) strategy. The model employs a U-Net generator to predict multiple parametric images using a CT image prior, with each output channel subsequently magnified by a factor to adjust the intensity. The model was optimized with a log-likelihood loss computed between the measured projection data and forward projected data. Two tracer datasets were simulated for evaluation: ⁸²Rb data using the 1-tissue compartment (1 TC) model and ¹⁸F-FDG data using the 2-tissue compartment (2 TC) model, with 10-fold magnification applied to the 1 TC k₂ and the 2 TC k₃, respectively. DIP-PM was compared to the indirect method, direct algorithm (OTEM) and the DIP method without parameter magnification (DIP-only). Performance was assessed on phantom data using peak signal-to-noise ratio (PSNR), normalized root mean square error (NRMSE) and structural similarity index (SSIM), as well as on real ¹⁸F-FDG scan from a male subject. For the 1 TC model, OTEM performed well in K₁ reconstruction, but both indirect and OTEM methods showed high noise and poor performance in k₂. The DIP-only method suppressed noise in k₂, but failed to reconstruct fine structures in the myocardium. DIP-PM outperformed other methods with well-preserved detailed structures, particularly in k₂, achieving the best metrics (PSNR: 19.00, NRMSE: 0.3002, SSIM: 0.9289). For the 2 TC model, traditional methods exhibited high noise and blurred structures in estimating all nonlinear parameters (K₁, k₂, k₃), while DIP-based methods significantly improved image quality. DIP-PM outperformed all methods in k₃ (PSNR: 21.89, NRMSE: 0.4054, SSIM: 0.8797), and consequently produced the most accurate 2 TC K_i images (PSNR: 22.74, NRMSE: 0.4897, SSIM: 0.8391). On real FDG data, DIP-PM also showed evident advantages in estimating K₁, k₂ and k₃ while preserving myocardial structures. The results underscore the efficacy of the DIP-based direct parametric imaging in generating and improving quality of PET parametric images. This study suggests that the proposed DIP-PM method with the parameter magnification strategy can enhance the fidelity of nonlinear micro-parameter images.

Diagnostic imaging interpretation involves distilling multimodal clinical information into text form, a task well-suited to augmentation by generative artificial intelligence (AI). However, to our knowledge, impacts of AI-based draft radiological reporting remain unstudied in clinical settings. To prospectively evaluate the association of radiologist use of a workflow-integrated generative model capable of providing draft radiological reports for plain radiographs across a tertiary health care system with documentation efficiency, the clinical accuracy and textual quality of final radiologist reports, and the model's potential for detecting unexpected, clinically significant pneumothorax. This prospective cohort study was conducted from November 15, 2023, to April 24, 2024, at a tertiary care academic health system. The association between use of the generative model and radiologist documentation efficiency was evaluated for radiographs documented with model assistance compared with a baseline set of radiographs without model use, matched by study type (chest or nonchest). Peer review was performed on model-assisted interpretations. Flagging of pneumothorax requiring intervention was performed on radiographs prospectively. The primary outcomes were association of use of the generative model with radiologist documentation efficiency, assessed by difference in documentation time with and without model use using a linear mixed-effects model; for peer review of model-assisted reports, the difference in Likert-scale ratings using a cumulative-link mixed model; and for flagging pneumothorax requiring intervention, sensitivity and specificity. A total of 23 960 radiographs (11 980 each with and without model use) were used to analyze documentation efficiency. Interpretations with model assistance (mean [SE], 159.8 [27.0] seconds) were faster than the baseline set of those without (mean [SE], 189.2 [36.2] seconds) (P = .02), representing a 15.5% documentation efficiency increase. Peer review of 800 studies showed no difference in clinical accuracy (χ2 = 0.68; P = .41) or textual quality (χ2 = 3.62; P = .06) between model-assisted interpretations and nonmodel interpretations. Moreover, the model flagged studies containing a clinically significant, unexpected pneumothorax with a sensitivity of 72.7% and specificity of 99.9% among 97 651 studies screened. In this prospective cohort study of clinical use of a generative model for draft radiological reporting, model use was associated with improved radiologist documentation efficiency while maintaining clinical quality and demonstrated potential to detect studies containing a pneumothorax requiring immediate intervention. This study suggests the potential for radiologist and generative AI collaboration to improve clinical care delivery.

Despite its potential to improve the assessment of cardiovascular diseases, 4D Flow CMR is hampered by long scan times. 4D Flow CMR is conventionally acquired with three motion encodings and one reference encoding, as the 3-dimensional velocity data are obtained by subtracting the phase of the reference from the phase of the motion encodings. In this study, we aim to use deep learning to predict the reference encoding from the three motion encodings for cardiovascular 4D Flow. A U-Net was trained with adversarial learning (U-Net_ADV) and with a velocity frequency-weighted loss function (U-Net_VEL) to predict the reference encoding from the three motion encodings obtained with a non-symmetric velocity-encoding scheme. Whole-heart 4D Flow datasets from 126 patients with different types of cardiomyopathies were retrospectively included. The models were trained on 113 patients with a 5-fold cross-validation, and tested on 13 patients. Flow volumes in the aorta and pulmonary artery, mean and maximum velocity, total and maximum turbulent kinetic energy at peak systole in the cardiac chambers and main vessels were assessed. 3-dimensional velocity data reconstructed with the reference encoding predicted by deep learning agreed well with the velocities obtained with the reference encoding acquired at the scanner for both models. U-Net_ADV performed more consistently throughout the cardiac cycle and across the test subjects, while U-Net_VEL performed better for systolic velocities. Comprehensively, the largest error for flow volumes, maximum and mean velocities was -6.031% for maximum velocities in the right ventricle for the U-Net_ADV, and -6.92% for mean velocities in the right ventricle for U-Net_VEL. For total turbulent kinetic energy, the highest errors were in the left ventricle (-77.17%) for the U-Net_ADV, and in the right ventricle (24.96%) for the U-Net_VEL, while for maximum turbulent kinetic energy were in the pulmonary artery for both models, with a value of -15.5% for U-Net_ADV and 15.38% for the U-Net_VEL. Deep learning-enabled referenceless 4D Flow CMR permits velocities and flow volumes quantification comparable to conventional 4D Flow. Omitting the reference encoding reduces the amount of acquired data by 25%, thus allowing shorter scan times or improved resolution, which is valuable for utilization in the clinical routine.

For breast cancer risk prediction to be clinically useful, it must be accurate and applicable to diverse groups of women across multiple settings. To examine whether a dynamic risk prediction model incorporating prior mammograms, previously validated in Black and White women, could predict future risk of breast cancer across a racially and ethnically diverse population in a population-based screening program. This prognostic study included women aged 40 to 74 years with 1 or more screening mammograms drawn from the British Columbia Breast Screening Program from January 1, 2013, to December 31, 2019, with follow-up via linkage to the British Columbia Cancer Registry through June 2023. This provincial, organized screening program offers screening mammography with full field digital mammography (FFDM) every 2 years. Data were analyzed from May to August 2024. FFDM-based, artificial intelligence-generated mammogram risk score (MRS), including up to 4 years of prior mammograms. The primary outcomes were 5-year risk of breast cancer (measured with the area under the receiver operating characteristic curve [AUROC]) and absolute risk of breast cancer calibrated to the US Surveillance, Epidemiology, and End Results incidence rates. Among 206 929 women (mean [SD] age, 56.1 [9.7] years; of 118 093 with data on race, there were 34 266 East Asian; 1946 Indigenous; 6116 South Asian; and 66 742 White women), there were 4168 pathology-confirmed incident breast cancers diagnosed through June 2023. Mean (SD) follow-up time was 5.3 (3.0) years. Using up to 4 years of prior mammogram images in addition to the most current mammogram, a 5-year AUROC of 0.78 (95% CI, 0.77-0.80) was obtained based on analysis of images alone. Performance was consistent across subgroups defined by race and ethnicity in East Asian (AUROC, 0.77; 95% CI, 0.75-0.79), Indigenous (AUROC, 0.77; 95% CI 0.71-0.83), and South Asian (AUROC, 0.75; 95% CI 0.71-0.79) women. Stratification by age gave a 5-year AUROC of 0.76 (95% CI, 0.74-0.78) for women aged 50 years or younger and 0.80 (95% CI, 0.78-0.82) for women older than 50 years. There were 18 839 participants (9.0%) with a 5-year risk greater than 3%, and the positive predictive value was 4.9% with an incidence of 11.8 per 1000 person-years. A dynamic MRS generated from both current and prior mammograms showed robust performance across diverse racial and ethnic populations in a province-wide screening program starting from age 40 years, reflecting improved accuracy for racially and ethnically diverse populations.

This study explores Artificial Intelligence (AI)s transformative role in diabetes care and monitoring, focusing on innovations that optimize patient outcomes. AI, particularly machine learning and deep learning, significantly enhances early detection of complications like diabetic retinopathy and improves screening efficacy. The methodology employs a bibliometric analysis using Scopus, VOSviewer, and Publish or Perish, analyzing 235 articles from 2023-2025. Results indicate a strong interdisciplinary focus, with Computer Science and Medicine being dominant subject areas (36.9% and 12.9% respectively). Bibliographic coupling reveals robust international collaborations led by the U.S. (1558.52 link strength), UK, and China, with key influential documents by Zhu (2023c) and Annuzzi (2023). This research highlights AIs impact on enhancing monitoring, personalized treatment, and proactive care, while acknowledging challenges in data privacy and ethical deployment. Future work should bridge technological advancements with real-world implementation to create equitable and efficient diabetes care systems.

BackgroundDelays in the diagnosis of transthyretin amyloid cardiomyopathy (ATTR-CM) contribute to the significant morbidity of the condition, especially in the era of disease-modifying therapies. Screening for ATTR-CM with AI and other algorithms may improve timely diagnosis, but these algorithms have not been directly compared. ObjectivesThe aim of this study was to compare the performance of four algorithms for ATTR-CM detection in a heart failure population and assess the risk for harms due to model bias. MethodsWe identified patients in an integrated health system from 2010-2022 with ATTR-CM and age- and sex-matched them to controls with heart failure to target 5% prevalence. We compared the performance of a claims-based random forest model (Huda et al. model), a regression-based score (Mayo ATTR-CM), and two deep learning echo models (EchoNet-LVH and EchoGo(R) Amyloidosis). We evaluated for bias using standard fairness metrics. ResultsThe analytical cohort included 176 confirmed cases of ATTR-CM and 3192 control patients with 79.2% self-identified as White and 9.0% as Black. The Huda et al. model performed poorly (AUC 0.49). Both deep learning echo models had a higher AUC when compared to the Mayo ATTR-CM Score (EchoNet-LVH 0.88; EchoGo Amyloidosis 0.92; Mayo ATTR-CM Score 0.79; DeLong P<0.001 for both). Bias auditing met fairness criteria for equal opportunity among patients who identified as Black. ConclusionsDeep learning, echo-based models to detect ATTR-CM demonstrated best overall discrimination when compared to two other models in external validation with low risk of harms due to racial bias.

The scarcity and imbalance of medical image datasets hinder the development of robust computer-aided diagnosis (CAD) systems for breast cancer. This study explores the application of advanced generative models, based on generative artificial intelligence (GenAI), for the synthesis of digital breast ultrasound images. Using a hybrid Conditional Variational Autoencoder-Wasserstein Generative Adversarial Network (CVAE-WGAN) architecture, we developed a system to generate high-quality synthetic images conditioned on the class (malignant vs. normal/benign). These synthetic images, generated from the low-resolution BreastMNIST dataset and filtered for quality, were systematically integrated with real training data at different mixing ratios (W). The performance of a CNN classifier trained on these mixed datasets was evaluated against a baseline model trained only on real data balanced with SMOTE. The optimal integration (mixing weight W=0.25) produced a significant performance increase on the real test set: +8.17% in macro-average F1-score and +4.58% in accuracy compared to using real data alone. Analysis confirmed the originality of the generated samples. This approach offers a promising solution for overcoming data limitations in image-based breast cancer diagnostics, potentially improving the capabilities of CAD systems.

BackgroundDifferent Natural Language Processing (NLP) techniques have demonstrated promising results for data extraction from radiological reports. Both traditional rule-based methods like regular expressions (Regex) and modern Large Language Models (LLMs) can extract structured information. However, comparison between these approaches for extraction of specific radiological data elements has not been widely conducted. MethodsWe compared accuracy and processing time between Regex and LLM-based approaches for extracting BI-RADS scores from 7,764 radiology reports (mammography, ultrasound, MRI, and biopsy). We developed a rule-based algorithm using Regex patterns and implemented an LLM-based extraction using the Rombos-LLM-V2.6-Qwen-14b model. A ground truth dataset of 199 manually classified reports was used for evaluation. ResultsThere was no statistically significant difference in the accuracy in extracting BI-RADS scores between Regex and an LLM-based method (accuracy of 89.20% for Regex versus 87.69% for the LLM-based method; p=0.56). Compared to the LLM-based method, Regex processing was more efficient, completing the task 28,120 times faster (0.06 seconds vs. 1687.20 seconds). Further analysis revealed LLMs favored common classifications (particularly BI-RADS value of 2) while Regex more frequently returned "unclear" values. We also could confirm in our sample an already known laterality bias for breast cancer (BI-RADS 6) and detected a slight laterality skew for suspected breast cancer (BI-RADS 5) as well. ConclusionFor structured, standardized data like BI-RADS, traditional NLP techniques seem to be superior, though future work should explore hybrid approaches combining Regex precision for standardized elements with LLM contextual understanding for more complex information extraction tasks.