Pulmonary disease can severely impair respiratory function and be life-threatening. Accurately recognizing pulmonary diseases in chest X-ray images is challenging due to overlapping body structures and the complex anatomy of the chest. We propose an adaptive multiscale feature fusion model for recognizing Chest X-ray images of pneumonia, tuberculosis, and COVID-19, which are common pulmonary diseases. We introduce an Adaptive Multiscale Fusion Network (AMFNet) for pulmonary disease classification in chest X-ray images. AMFNet consists of a lightweight Multiscale Fusion Network (MFNet) and ResNet50 as the secondary feature extraction network. MFNet employs Fusion Blocks with self-calibrated convolution (SCConv) and Attention Feature Fusion (AFF) to capture multiscale semantic features, and integrates a custom activation function, MFReLU, which is employed to reduce the model's memory access time. A fusion module adaptively combines features from both networks. Experimental results show that AMFNet achieves 97.48% accuracy and an F1 score of 0.9781 on public datasets, outperforming models like ResNet50, DenseNet121, ConvNeXt-Tiny, and Vision Transformer while using fewer parameters.

Generative image models have achieved remarkable progress in both natural and medical imaging. In the medical context, these techniques offer a potential solution to data scarcity-especially for low-prevalence anomalies that impair the performance of AI-driven diagnostic and segmentation tools. However, questions remain regarding the fidelity and clinical utility of synthetic images, since poor generation quality can undermine model generalizability and trust. In this study, we evaluate the effectiveness of state-of-the-art generative models-Generative Adversarial Networks (GANs) and Diffusion Models (DMs)-for synthesizing chest X-rays conditioned on four abnormalities: Atelectasis (AT), Lung Opacity (LO), Pleural Effusion (PE), and Enlarged Cardiac Silhouette (ECS). Using a benchmark composed of real images from the MIMIC-CXR dataset and synthetic images from both GANs and DMs, we conducted a reader study with three radiologists of varied experience. Participants were asked to distinguish real from synthetic images and assess the consistency between visual features and the target abnormality. Our results show that while DMs generate more visually realistic images overall, GANs can report better accuracy for specific conditions, such as absence of ECS. We further identify visual cues radiologists use to detect synthetic images, offering insights into the perceptual gaps in current models. These findings underscore the complementary strengths of GANs and DMs and point to the need for further refinement to ensure generative models can reliably augment training datasets for AI diagnostic systems.

Label scarcity remains a major challenge in deep learning-based medical image segmentation. Recent studies use strong-weak pseudo supervision to leverage unlabeled data. However, performance is often hindered by inconsistencies between pseudo labels and their corresponding unlabeled images. In this work, we propose \textbf{SynMatch}, a novel framework that sidesteps the need for improving pseudo labels by synthesizing images to match them instead. Specifically, SynMatch synthesizes images using texture and shape features extracted from the same segmentation model that generates the corresponding pseudo labels for unlabeled images. This design enables the generation of highly consistent synthesized-image-pseudo-label pairs without requiring any training parameters for image synthesis. We extensively evaluate SynMatch across diverse medical image segmentation tasks under semi-supervised learning (SSL), weakly-supervised learning (WSL), and barely-supervised learning (BSL) settings with increasingly limited annotations. The results demonstrate that SynMatch achieves superior performance, especially in the most challenging BSL setting. For example, it outperforms the recent strong-weak pseudo supervision-based method by 29.71\% and 10.05\% on the polyp segmentation task with 5\% and 10\% scribble annotations, respectively. The code will be released at https://github.com/Senyh/SynMatch.

Multi-sequence Magnetic Resonance Imaging (MRI) offers remarkable versatility, enabling the distinct visualization of different tissue types. Nevertheless, the inherent heterogeneity among MRI sequences poses significant challenges to the generalization capability of deep learning models. These challenges undermine model performance when faced with varying acquisition parameters, thereby severely restricting their clinical utility. In this study, we present PRISM, a foundation model PRe-trained with large-scale multI-Sequence MRI. We collected a total of 64 datasets from both public and private sources, encompassing a wide range of whole-body anatomical structures, with scans spanning diverse MRI sequences. Among them, 336,476 volumetric MRI scans from 34 datasets (8 public and 26 private) were curated to construct the largest multi-organ multi-sequence MRI pretraining corpus to date. We propose a novel pretraining paradigm that disentangles anatomically invariant features from sequence-specific variations in MRI, while preserving high-level semantic representations. We established a benchmark comprising 44 downstream tasks, including disease diagnosis, image segmentation, registration, progression prediction, and report generation. These tasks were evaluated on 32 public datasets and 5 private cohorts. PRISM consistently outperformed both non-pretrained models and existing foundation models, achieving first-rank results in 39 out of 44 downstream benchmarks with statistical significance improvements. These results underscore its ability to learn robust and generalizable representations across unseen data acquired under diverse MRI protocols. PRISM provides a scalable framework for multi-sequence MRI analysis, thereby enhancing the translational potential of AI in radiology. It delivers consistent performance across diverse imaging protocols, reinforcing its clinical applicability.

Stress is a natural response of the body to perceived threats, and it can have both positive and negative effects on brain hemodynamics. Stress-induced changes in pupil and eyelid size/shape have been used as a biomarker in several fMRI studies. However, there were limited knowledges regarding changes in behavior of pupil and eyelid dynamics, particularly on animal models. In the present study, the pupil and eyelid dynamics were carefully investigated and characterized in a newly developed awake rodent fMRI protocol. Leveraging deep learning techniques, the mouse pupil and eyelid diameters were extracted and analyzed during different training and imaging phases in the present project. Our findings demonstrate a consistent downwards trend in pupil and eyelid dynamics under a meticulously designed training protocol, suggesting that the behaviors of the pupil and eyelid can be served as reliable indicators of stress levels and motion artifacts in awake fMRI studies. The current recording platform not only enables the facilitation of awake animal MRI studies but also highlights its potential applications to numerous other research areas, owing to the non-invasive nature and straightforward implementation.

To visualize and assess hematoma growth trends by generating follow-up CT images within 24 h based on baseline CT images of spontaneous intracerebral hemorrhage (sICH) using Transformer-integrated Generative Adversarial Networks (GAN). Patients with sICH were retrospectively recruited from two medical centers. The imaging data included baseline non-contrast CT scans taken after onset and follow-up imaging within 24 h. In the test set, the peak signal-to-noise ratio (PSNR) and the structural similarity index measure (SSIM) were utilized to quantitatively assess the quality of the predicted images. Pearson's correlation analysis was performed to assess the agreement of semantic features and geometric properties of hematomas between true follow-up CT images and the predicted images. The consistency of hematoma expansion prediction between true and generated images was further examined. The PSNR of the predicted images was 26.73 ± 1.11, and the SSIM was 91.23 ± 1.10. The Pearson correlation coefficients (r) with 95 % confidence intervals (CI) for irregularity, satellite sign number, intraventricular or subarachnoid hemorrhage, midline shift, edema expansion, mean CT value, maximum cross-sectional area, and hematoma volume between the predicted and true follow-up images were as follows: 0.94 (0.91, 0.96), 0.87 (0.81, 0.91), 0.86 (0.80, 0.91), 0.89 (0.84, 0.92), 0.91 (0.87, 0.94), 0.78(0.68, 0.84), 0.94(0.91, 0.96), and 0.94 (0.91, 0.96), respectively. The correlation coefficient (r) for predicting hematoma expansion between predicted and true follow-up images was 0.86 (95 % CI: 0.79, 0.90; P < 0.001). The model constructed using a GAN integrated with Transformer modules can accurately visualize early hematoma changes in sICH.

Skull stripping is a common preprocessing step in Magnetic Resonance Imaging (MRI) pipelines and is often performed manually. Automating this process is challenging for preclinical data due to variations in brain geometry, resolution, and tissue contrast. Existing methods for MRI skull stripping often struggle with the low resolution and varying slice sizes found in preclinical functional MRI (fMRI) data. This study proposes a novel method that integrates a Dense UNet-based architecture with a feature extractor based on the Smart Swin Transformer (SST), called SST-DUNet. The Smart Shifted Window Multi-Head Self-Attention (SSW-MSA) module in SST replaces the mask-based module in the Swin Transformer (ST), enabling the learning of distinct channel-wise features while focusing on relevant dependencies within brain structures. This modification allows the model to better handle the complexities of fMRI skull stripping, such as low resolution and variable slice sizes. To address class imbalance in preclinical data, a combined loss function using Focal and Dice loss is applied. The model was trained on rat fMRI images and evaluated across three in-house datasets, achieving Dice similarity scores of 98.65%, 97.86%, and 98.04%. We compared our method with conventional and deep learning-based approaches, demonstrating its superiority over state-of-the-art methods. The fMRI results using SST-DUNet closely align with those from manual skull stripping for both seed-based and independent component analyses, indicating that SST-DUNet can effectively substitute manual brain extraction in rat fMRI analysis.

It is unclear whether coronary artery stenosis, plaque burden, and composition differ between major epicardial arteries supplying ischemic myocardial territories. We studied 837 symptomatic patients undergoing coronary computed tomography angiography (CTA) and ¹⁵O-water PET myocardial perfusion imaging for suspected obstructive coronary artery disease. Coronary CTA was analyzed using Artificial Intelligence-Guided Quantitative Computed Tomography (AI-QCT) to assess stenosis and atherosclerotic plaque characteristics. Myocardial ischemia was defined by regional PET perfusion in the left anterior descending (LAD), left circumflex (LCX), and right coronary artery (RCA) territories. Among arteries supplying ischemic territories, the LAD exhibited significantly higher stenosis and both absolute and normalized plaque volumes compared to LCX and RCA (p<0.001 for all). Multivariable logistic regression showed diameter stenosis (p=0.001-0.015), percent atheroma volume (PAV; p<0.001), and percent non-calcified plaque volume (p=0.001-0.017) were associated with ischemia across all three arteries. Percent calcified plaque volume was associated with ischemia only in the RCA (p=0.001). The degree of stenosis and atherosclerotic burden are significantly higher in LAD as compared to LCX and RCA, both in epicardial coronary arteries supplying non-ischemic or ischemic myocardial territories. In all the three main coronary arteries both luminal narrowing and plaque burden are independent predictors of ischemia, where the plaque burden is mainly driven by non-calcified plaque. However, many vessels supplying ischemic territories have relatively low stenosis degree and plaque burden, especially in the LCx and RCA, limiting the ability of diameter stenosis and PAV to predict myocardial ischemia.

Bronchial artery chemoembolization (BACE) is a new treatment method for lung cancer. This study aimed to investigate the ability of dual-energy computed tomography (DECT) to predict early recurrence (ER) after BACE among patients with non-small cell lung cancer (NSCLC) who failed first-line therapy. Clinical and imaging data from NSCLC patients undergoing BACE at Wenzhou Medical University Affiliated Fifth *** Hospital (10/2023-06/2024) were retrospectively analyzed. Logistic regression (LR) machine learning models were developed using 5 arterial-phase (AP) virtual monoenergetic images (VMIs; 40, 70, 100, 120, and 150 keV), while deep learning models utilized ResNet50/101/152 architectures with iodine maps. A combined model integrating optimal Rad-score, DL-score, and clinical features was established. Model performance was assessed via area under the receiver operating characteristic curve analysis (AUC), with SHapley Additive exPlanations (SHAP) framework applied for interpretability. A total of 196 patients were enrolled in this study (training cohort: n=158; testing cohort: n=38). The 100 keV machine learning model demonstrated superior performance (AUC=0.751) compared to other VMIs. The deep learning model based on the ResNet101 method (AUC=0.791) performed better than other approaches. The hybrid model combining Rad-score-100keV-A, Rad-score-100keV-V, DL-score-ResNet101-A, DL-score-ResNet101-V, and clinical features exhibited the best performance (AUC=0.798) among all models. DECT holds promise for predicting ER after BACE among NSCLC patients who have failed first-line therapy, offering valuable guidance for clinical treatment planning.

Large Language Models (LLMs) are increasingly applied to medical imaging tasks, including image interpretation and synthetic image generation. However, these models often produce hallucinations, which are confident but incorrect outputs that can mislead clinical decisions. This study examines hallucinations in two directions: image to text, where LLMs generate reports from X-ray, CT, or MRI scans, and text to image, where models create medical images from clinical prompts. We analyze errors such as factual inconsistencies and anatomical inaccuracies, evaluating outputs using expert informed criteria across imaging modalities. Our findings reveal common patterns of hallucination in both interpretive and generative tasks, with implications for clinical reliability. We also discuss factors contributing to these failures, including model architecture and training data. By systematically studying both image understanding and generation, this work provides insights into improving the safety and trustworthiness of LLM driven medical imaging systems.