To develop a deep learning model using orbital computed tomography (CT) imaging to accurately distinguish thyroid eye disease (TED) and orbital myositis, two conditions with overlapping clinical presentations. Retrospective, single-center cohort study spanning 12 years including normal controls, TED, and orbital myositis patients with orbital imaging and examination by an oculoplastic surgeon. A deep learning model employing a Visual Geometry Group-16 network was trained on various binary combinations of TED, orbital myositis, and controls using single slices of coronal orbital CT images. A total of 1628 images from 192 patients (110 TED, 51 orbital myositis, 31 controls) were included. The primary model comparing orbital myositis and TED had accuracy of 98.4% and area under the receiver operating characteristic curve (AUC) of 0.999. In detecting orbital myositis, it had a sensitivity, specificity, and F1 score of 0.964, 0.994, and 0.984, respectively. Deep learning models can differentiate TED and orbital myositis based on a single, coronal orbital CT image with high accuracy. Their ability to distinguish these conditions based not only on extraocular muscle enlargement but also other salient features suggests potential applications in diagnostics and treatment beyond these conditions.

Accurate preoperative discrimination of salivary gland pleomorphic adenoma (SPA) stromal subtypes is essential for therapeutic plannings. We aimed to establish and test machine learning (ML) models for classification of stromal subtypes in SPA based on ultrasound histogram analysis. A total of 256 SPA patients were enrolled in the study and categorized into two groups: stroma-low and stroma-high. The dataset was split into a training cohort with 177 patients and a validation cohort with 79 patients. The least absolute shrinkage and selection operator (LASSO) regression identified optimal features, which were then utilized to build predictive models using logistic regression (LR) and eight ML algorithms. The effectiveness of the models was evaluated using a range of performance metrics, with a particular focus on the area under the receiver operating characteristic curve (AUC). After LASSO regression, six key features (lesion size, shape, cystic areas, vascularity, mean, and skewness) were selected to develop predictive models. The AUCs ranged from 0.575 to 0.827 for the nine models. The support vector machine (SVM) algorithm achieved the highest performance with an AUC of 0.827, accompanied by an accuracy of 0.798, precision of 0.792, recall of 0.862, and an F1 score of 0.826. The LR algorithm also exhibited robust performance, achieving an AUC of 0.818, slightly trailing behind the SVM algorithm. Decision curve analysis indicated that the SVM-based model provided superior clinical utility compared to other models. The ML model based on ultrasound histogram analysis offers a precise and non-invasive approach for preoperative categorization of stromal subtypes in SPA.

This paper describes PARADIM, a digital infrastructure designed to support research at the interface of data science and medical imaging, with a focus on Research Data Management best practices. The platform is built from open-source components and rooted in the FAIR principles through strict compliance with the DICOM standard. It addresses key needs in data curation, governance, privacy, and scalable resource management. Supporting every stage of the data science discovery cycle, the platform offers robust functionalities for user identity and access management, data de-identification, storage, annotation, as well as model training and evaluation. Rich metadata are generated all along the research lifecycle to ensure the traceability and reproducibility of results. PARADIM hosts several medical image collections and allows the automation of large-scale, computationally intensive pipelines (e.g., automatic segmentation, dose calculations, AI model evaluation). The platform fills a gap at the interface of data science and medical imaging, where digital infrastructures are key in the development, evaluation, and deployment of innovative solutions in the real world.

Magnetic resonance imaging (MRI) is essential in clinical and research contexts, providing exceptional soft-tissue contrast. However, prolonged acquisition times often lead to patient discomfort and motion artifacts. Diffusion-based deep learning super-resolution (SR) techniques reconstruct high-resolution (HR) images from low-resolution (LR) pairs, but they involve extensive sampling steps, limiting real-time application. To overcome these issues, this study introduces a residual error-shifting mechanism markedly reducing sampling steps while maintaining vital anatomical details, thereby accelerating MRI reconstruction. We developed Res-SRDiff, a novel diffusion-based SR framework incorporating residual error shifting into the forward diffusion process. This integration aligns the degraded HR and LR distributions, enabling efficient HR image reconstruction. We evaluated Res-SRDiff using ultra-high-field brain T1 MP2RAGE maps and T2-weighted prostate images, benchmarking it against Bicubic, Pix2pix, CycleGAN, SPSR, I2SB, and TM-DDPM methods. Quantitative assessments employed peak signal-to-noise ratio (PSNR), structural similarity index (SSIM), gradient magnitude similarity deviation (GMSD), and learned perceptual image patch similarity (LPIPS). Additionally, we qualitatively and quantitatively assessed the proposed framework's individual components through an ablation study and conducted a Likert-based image quality evaluation. Res-SRDiff significantly surpassed most comparison methods regarding PSNR, SSIM, and GMSD for both datasets, with statistically significant improvements (p-values≪0.05). The model achieved high-fidelity image reconstruction using only four sampling steps, drastically reducing computation time to under one second per slice. In contrast, traditional methods like TM-DDPM and I2SB required approximately 20 and 38 seconds per slice, respectively. Qualitative analysis showed Res-SRDiff effectively preserved fine anatomical details and lesion morphologies. The Likert study indicated that our method received the highest scores, 4.14±0.77(brain) and 4.80±0.40(prostate). Res-SRDiff demonstrates efficiency and accuracy, markedly improving computational speed and image quality. Incorporating residual error shifting into diffusion-based SR facilitates rapid, robust HR image reconstruction, enhancing clinical MRI workflow and advancing medical imaging research. Code available at https://github.com/mosaf/Res-SRDiff.

Evaluating the minimum distance (dTICA) between the internal carotid artery (ICA) and tonsillar tumors (TT) on imaging is essential for preoperative planning; we propose a tool to automatically extract dTICA. CT scans of 96 patients with TT were selected from the cancer imaging archive. nnU-Net, a deep learning framework, was implemented to automatically segment both the TT and ICA from these scans. Dice similarity coefficient (DSC) and average hausdorff distance (AHD) were used to evaluate the performance of the nnU-Net. Thereafter, an automated tool was built to calculate the magnitude of dTICA from these segmentations. The average DSC and AHD were 0.67, 2.44 mm, and 0.83, 0.49 mm for the TT and ICA, respectively. The mean dTICA was 6.66 mm and statistically varied by tumor T stage (p = 0.00456). The proposed pipeline can accurately and automatically capture dTICA, potentially assisting clinicians in preoperative evaluation.

Compound figures, which are multi-panel composites containing diverse subfigures, are ubiquitous in biomedical literature, yet large-scale subfigure extraction remains largely unaddressed. Prior work on subfigure extraction has been limited in both dataset size and generalizability, leaving a critical open question: How does high-fidelity image-text alignment via large-scale subfigure extraction impact representation learning in vision-language models? We address this gap by introducing a scalable subfigure extraction pipeline based on transformer-based object detection, trained on a synthetic corpus of 500,000 compound figures, and achieving state-of-the-art performance on both ImageCLEF 2016 and synthetic benchmarks. Using this pipeline, we release OPEN-PMC-18M, a large-scale high quality biomedical vision-language dataset comprising 18 million clinically relevant subfigure-caption pairs spanning radiology, microscopy, and visible light photography. We train and evaluate vision-language models on our curated datasets and show improved performance across retrieval, zero-shot classification, and robustness benchmarks, outperforming existing baselines. We release our dataset, models, and code to support reproducible benchmarks and further study into biomedical vision-language modeling and representation learning.

Compound figures, which are multi-panel composites containing diverse subfigures, are ubiquitous in biomedical literature, yet large-scale subfigure extraction remains largely unaddressed. Prior work on subfigure extraction has been limited in both dataset size and generalizability, leaving a critical open question: How does high-fidelity image-text alignment via large-scale subfigure extraction impact representation learning in vision-language models? We address this gap by introducing a scalable subfigure extraction pipeline based on transformer-based object detection, trained on a synthetic corpus of 500,000 compound figures, and achieving state-of-the-art performance on both ImageCLEF 2016 and synthetic benchmarks. Using this pipeline, we release OPEN-PMC-18M, a large-scale high quality biomedical vision-language dataset comprising 18 million clinically relevant subfigure-caption pairs spanning radiology, microscopy, and visible light photography. We train and evaluate vision-language models on our curated datasets and show improved performance across retrieval, zero-shot classification, and robustness benchmarks, outperforming existing baselines. We release our dataset, models, and code to support reproducible benchmarks and further study into biomedical vision-language modeling and representation learning.

Deep learning-based disease grading technologies facilitate timely medical intervention due to their high efficiency and accuracy. Recent advancements have enhanced grading performance by incorporating the ordinal relationships of disease labels. However, existing methods often assume same probability distributions for disease labels across instances within the same category, overlooking variations in label distributions. Additionally, the hyperparameters of these distributions are typically determined empirically, which may not accurately reflect the true distribution. To address these limitations, we propose a disease grading network utilizing a sample-aware asymmetric Gaussian label distribution, termed DGN-AGLD. This approach includes a variance predictor designed to learn and predict parameters that control the asymmetry of the Gaussian distribution, enabling distinct label distributions within the same category. This module can be seamlessly integrated into standard deep learning networks. Experimental results on four disease datasets validate the effectiveness and superiority of the proposed method, particularly on the IDRiD dataset, where it achieves a diabetic retinopathy accuracy of 77.67%. Furthermore, our method extends to joint disease grading tasks, yielding superior results and demonstrating significant generalization capabilities. Visual analysis indicates that our method more accurately captures the trend of disease progression by leveraging the asymmetry in label distribution. Our code is publicly available on https://github.com/ahtwq/AGNet.

Standard deep learning (DL) models often suffer significant performance degradation on out-of-distribution (OOD) data, where test data differs from training data, a common challenge in medical imaging due to real-world variations. We propose a unified self-censorship framework as an alternative to the standard DL models for glaucoma classification using deep evidential uncertainty quantification. Our approach detects OOD samples at both the dataset and image levels. Dataset-level self-censorship enables users to accept or reject predictions for an entire new dataset based on model uncertainty, whereas image-level self-censorship refrains from making predictions on individual OOD images rather than risking incorrect classifications. We validated our approach across diverse datasets. Our dataset-level self-censorship method outperforms the standard DL model in OOD detection, achieving an average 11.93% higher area under the curve (AUC) across 14 OOD datasets. Similarly, our image-level self-censorship model improves glaucoma classification accuracy by an average of 17.22% across 4 external glaucoma datasets against baselines while censoring 28.25% more data. Our approach addresses the challenge of generalization in standard DL models for glaucoma classification across diverse datasets by selectively withholding predictions when the model is uncertain. This method reduces misclassification errors compared to state-of-the-art baselines, particularly for OOD cases. This study introduces a tunable framework that explores the trade-off between prediction accuracy and data retention in glaucoma prediction. By managing uncertainty in model outputs, the approach lays a foundation for future decision support tools aimed at improving the reliability of automated glaucoma diagnosis.

The purpose of this study was to investigate the impact of optic nerve tortuosity (ONT), and the interaction of globe proptosis and size on retinal ganglion cell (RGC) thickness, using retinal nerve fiber layer (RNFL) thickness, across general, glaucoma, and myopic populations. This study analyzed 17,940 eyes from the UKBiobank cohort (ID 76442), including 72 glaucomatous and 2475 myopic eyes. Artificial intelligence models were developed to derive RNFL thickness corrected for ocular magnification from 3D optical coherence tomography scans and orbit features from 3D magnetic resonance images, including ONT, globe proptosis, axial length, and a novel feature: the interzygomatic line-to-posterior pole (ILPP) distance - a composite marker of globe proptosis and size. Generalized estimating equation (GEE) models evaluated associations between orbital and retinal features. RNFL thickness was positively correlated with ONT and ILPP distance (r = 0.065, P < 0.001 and r = 0.206, P < 0.001, respectively) in the general population. The same was true for glaucoma (r = 0.040, P = 0.74 and r = 0.224, P = 0.059), and for myopia (r = 0.069, P < 0.001 and r = 0.100, P < 0.001). GEE models revealed that straighter optic nerves and shorter ILPP distance were predictive of thinner RNFL in all populations. Straighter optic nerves and decreased ILPP distance could cause RNFL thinning, possibly due to greater traction forces. ILPP distance emerged as a potential biomarker of axonal health. These findings underscore the importance of orbit structures in RGC axonal health and warrant further research into orbit biomechanics.