This study centers around the competition between Convolutional Neural Networks (CNNs) with large convolutional kernels and Vision Transformers in the domain of computer vision, delving deeply into the issues pertaining to parameters and computational complexity that stem from the utilization of large convolutional kernels. Even though the size of the convolutional kernels has been extended up to 51×51, the enhancement of performance has hit a plateau, and moreover, striped convolution incurs a performance degradation. Enlightened by the hierarchical visual processing mechanism inherent in humans, this research innovatively incorporates a shared parameter mechanism for large convolutional kernels. It synergizes the expansion of the receptive field enabled by large convolutional kernels with the extraction of fine-grained features facilitated by small convolutional kernels. To address the surging number of parameters, a meticulously designed parameter sharing mechanism is employed, featuring fine-grained processing in the central region of the convolutional kernel and wide-ranging parameter sharing in the periphery. This not only curtails the parameter count and mitigates the model complexity but also sustains the model's capacity to capture extensive spatial relationships. Additionally, in light of the problems of spatial feature information loss and augmented memory access during the 1×1 convolutional channel compression phase, this study further puts forward a dynamic channel sampling approach, which markedly elevates the accuracy of tumor subregion segmentation. To authenticate the efficacy of the proposed methodology, a comprehensive evaluation has been conducted on three brain tumor segmentation datasets, namely BraTs2020, BraTs2024, and Medical Segmentation Decathlon Brain 2018. The experimental results evince that the proposed model surpasses the current mainstream ConvNet and Transformer architectures across all performance metrics, proffering novel research perspectives and technical stratagems for the realm of medical image segmentation.

Surface-based cortical analysis is valuable for a variety of neuroimaging tasks, such as spatial normalization, parcellation, and gray matter (GM) thickness estimation. However, most tools for estimating cortical surfaces work exclusively on scans with at least 1 mm isotropic resolution and are tuned to a specific magnetic resonance (MR) contrast, often T1-weighted (T1w). This precludes application using most clinical MR scans, which are very heterogeneous in terms of contrast and resolution. Here, we use synthetic domain-randomized data to train the first neural network for explicit estimation of cortical surfaces from scans of any contrast and resolution, without retraining. Our method deforms a template mesh to the white matter (WM) surface, which guarantees topological correctness. This mesh is further deformed to estimate the GM surface. We compare our method to recon-all-clinical (RAC), an implicit surface reconstruction method which is currently the only other tool capable of processing heterogeneous clinical MR scans, on ADNI and a large clinical dataset (n=1,332). We show a approximately 50 % reduction in cortical thickness error (from 0.50 to 0.24 mm) with respect to RAC and better recovery of the aging-related cortical thinning patterns detected by FreeSurfer on high-resolution T1w scans. Our method enables fast and accurate surface reconstruction of clinical scans, allowing studies (1) with sample sizes far beyond what is feasible in a research setting, and (2) of clinical populations that are difficult to enroll in research studies. The code is publicly available at https://github.com/simnibs/brainnet.

In many real-world applications, deployed models encounter inputs that differ from the data seen during training. Out-of-distribution detection identifies whether an input stems from an unseen distribution, while open-world recognition flags such inputs to ensure the system remains robust as ever-emerging, previously $unknown$ categories appear and must be addressed without retraining. Foundation and vision-language models are pre-trained on large and diverse datasets with the expectation of broad generalization across domains, including medical imaging. However, benchmarking these models on test sets with only a few common outlier types silently collapses the evaluation back to a closed-set problem, masking failures on rare or truly novel conditions encountered in clinical use. We therefore present $NOVA$, a challenging, real-life $evaluation-only$ benchmark of $\sim$900 brain MRI scans that span 281 rare pathologies and heterogeneous acquisition protocols. Each case includes rich clinical narratives and double-blinded expert bounding-box annotations. Together, these enable joint assessment of anomaly localisation, visual captioning, and diagnostic reasoning. Because NOVA is never used for training, it serves as an $extreme$ stress-test of out-of-distribution generalisation: models must bridge a distribution gap both in sample appearance and in semantic space. Baseline results with leading vision-language models (GPT-4o, Gemini 2.0 Flash, and Qwen2.5-VL-72B) reveal substantial performance drops across all tasks, establishing NOVA as a rigorous testbed for advancing models that can detect, localize, and reason about truly unknown anomalies.

Brain-to-image decoding has been recently propelled by the progress in generative AI models and the availability of large ultra-high field functional Magnetic Resonance Imaging (fMRI). However, current approaches depend on complicated multi-stage pipelines and preprocessing steps that typically collapse the temporal dimension of brain recordings, thereby limiting time-resolved brain decoders. Here, we introduce Dynadiff (Dynamic Neural Activity Diffusion for Image Reconstruction), a new single-stage diffusion model designed for reconstructing images from dynamically evolving fMRI recordings. Our approach offers three main contributions. First, Dynadiff simplifies training as compared to existing approaches. Second, our model outperforms state-of-the-art models on time-resolved fMRI signals, especially on high-level semantic image reconstruction metrics, while remaining competitive on preprocessed fMRI data that collapse time. Third, this approach allows a precise characterization of the evolution of image representations in brain activity. Overall, this work lays the foundation for time-resolved brain-to-image decoding.

Neuroimaging screening and surveillance is one of the first frontline diagnostic tools leveraged in the acute assessment (first 24 h postinjury) of patients suspected to have traumatic brain injury (TBI). While imaging, in particular computed tomography, is used almost universally in emergency departments worldwide to evaluate possible features of TBI, there is no currently agreed-upon reporting system, standard terminology, or framework to contextualize brain imaging findings with other available medical, psychosocial, and environmental data. In 2023, the NIH-National Institute of Neurological Disorders and Stroke convened six working groups of international experts in TBI to develop a new framework for nomenclature and classification. The goal of this effort was to propose a more granular system of injury classification that incorporates recent progress in imaging biomarkers, blood-based biomarkers, and injury and recovery modifiers to replace the commonly used Glasgow Coma Scale-based diagnosis groups of mild, moderate, and severe TBI, which have shown relatively poor diagnostic, prognostic, and therapeutic utility. Motivated by prior efforts to standardize the nomenclature for pathoanatomic imaging findings of TBI for research and clinical trials, along with more recent studies supporting the refinement of the originally proposed definitions, the Imaging Working Group sought to update and expand this application specifically for consideration of use in clinical practice. Here we report the recommendations of this working group to enable the translation of structured imaging common data elements to the standard of care. These leverage recent advances in imaging technology, electronic medical record (EMR) systems, and artificial intelligence (AI), along with input from key stakeholders, including patients with lived experience, caretakers, providers across medical disciplines, radiology industry partners, and policymakers. It was recommended that (1) there would be updates to the definitions of key imaging features used for this system of classification and that these should be further refined as new evidence of the underlying pathology driving the signal change is identified; (2) there would be an efficient, integrated tool embedded in the EMR imaging reporting system developed in collaboration with industry partners; (3) this would include AI-generated evidence-based feature clusters with diagnostic, prognostic, and therapeutic implications; and (4) a "patient translator" would be developed in parallel to assist patients and families in understanding these imaging features. In addition, important disclaimers would be provided regarding known limitations of current technology until such time as they are overcome, such as resolution and sequence parameter considerations. The end goal is a multifaceted TBI characterization model incorporating clinical, imaging, blood biomarker, and psychosocial and environmental modifiers to better serve patients not only acutely but also through the postinjury continuum in the days, months, and years that follow TBI.

Deep learning (DL) reconstruction has been successful in realizing otherwise impracticable acceleration factors and improving image quality in conventional MRI field strengths; however, there has been limited application to ultra-high field MRI.The objective of this study was to evaluate the performance of a prototype DL-based image reconstruction technique in 7T MRI of the brain utilizing MP2RAGE and SPACE acquisitions, in comparison to reconstructions in conventional compressed sensing (CS) and controlled aliasing in parallel imaging (CAIPIRINHA) techniques. This retrospective study involved 60 patients who underwent 7T brain MRI between June 2024 and October 2024, comprised of 30 patients with MP2RAGE data and 30 patients with SPACE FLAIR data. Each set of raw data was reconstructed with DL-based reconstruction and conventional reconstruction. Image quality was independently assessed by two neuroradiologists using a 5-point Likert scale, which included overall image quality, artifacts, sharpness, structural conspicuity, and noise level. Inter-observer agreement was determined using top-box analysis. Contrast-to-noise ratio (CNR) and noise levels were quantitatively evaluated and compared using the Wilcoxon signed-rank test. DL-based reconstruction resulted in a significant increase in overall image quality and a reduction in subjective noise level for both MP2RAGE and SPACE FLAIR data (all P<0.001), with no significant differences in image artifacts (all P>0.05). When compared to standard reconstruction, the implementation of DL-based reconstruction yielded an increase in CNR of 49.5% [95% CI 33.0-59.0%] for MP2RAGE data and 90.6% [95% CI 73.2-117.7%] for SPACE FLAIR data, along with a decrease in noise of 33.5% [95% CI 23.0-38.0%] for MP2RAGE data and 47.5% [95% CI 41.9-52.6%] for SPACE FLAIR data. DL-based reconstruction of 7T MRI significantly enhanced image quality compared to conventional reconstruction without introducing image artifacts. The achievable high acceleration factors have the potential to substantially improve image quality and resolution in 7T MRI. CAIPIRINHA = Controlled Aliasing In Parallel Imaging Results IN Higher Acceleration; CNR = contrast-to-noise ratio; CS = compressed sensing; DL = deep learning; MNI = Montreal Neurological Institute; MP2RAGE = Magnetization-Prepared 2 Rapid Acquisition Gradient Echoes; SPACE = Sampling Perfection with Application-Optimized Contrasts using Different Flip Angle Evolutions.

Glioblastoma (GBM) is a highly aggressive brain tumor with poor prognosis. This study aimed to construct and validate a radiomics-based machine learning model for predicting overall survival (OS) in IDH-wildtype GBM after maximal safe surgical resection using magnetic resonance imaging. A total of 582 patients were retrospectively enrolled, comprising 301 in the training cohort, 128 in the internal validation cohort, and 153 in the external validation cohort. Volumes of interest (VOIs) from contrast-enhanced T1-weighted imaging (CE-T1WI) were segmented into three regions: contrast-enhancing tumor, necrotic non-enhancing core, and peritumoral edema using an ResNet-based segmentation network. A total of 4,227 radiomic features were extracted and filtered using LASSO-Cox regression to identify signatures. The prognostic model was constructed using the Mime prediction framework, categorizing patients into high- and low-risk groups based on the median OS. Model performance was assessed using the concordance index (CI) and Kaplan-Meier survival analysis. Independent prognostic factors were identified through multivariable Cox regression analysis, and a nomogram was developed for individualized risk assessment. The Step Cox [backward] + RSF model achieved CIs of 0.89, 0.81, and 0.76 in the training, internal and external validation cohorts. Log-rank tests demonstrated significant survival differences between high- and low-risk groups across all cohorts (P < 0.05). Multivariate Cox analysis identified age (HR: 1.022; 95% CI: 0.979, 1.009, P < 0.05), KPS score (HR: 0.970, 95% CI: 0.960, 0.978, P < 0.05), rad-scores of the necrotic non-enhancing core (HR: 8.164; 95% CI: 2.439, 27.331, P < 0.05), and peritumoral edema (HR: 3.748; 95% CI: 1.212, 11.594, P < 0.05) as independent predictors of OS. A nomogram integrating these predictors provided individualized risk assessment. This deep learning segmentation-based radiomics model demonstrated robust performance in predicting OS in GBM after maximal safe surgical resection. By incorporating radiomic signatures and advanced machine learning algorithms, it offers a non-invasive tool for personalized prognostic assessment and supports clinical decision-making.

Cortical regions such as parietal area H (PH) and the fundus of the superior temporal sulcus (FST) are involved in higher visual function and may play a role in dementia with Lewy bodies (DLB), which is frequently associated with hallucinations. The authors evaluated functional connectivity between these two regions for distinguishing participants with DLB from those with Alzheimer's disease (AD) or mild cognitive impairment (MCI) and from cognitively normal (CN) individuals to identify a functional connectivity MRI signature for DLB. Eighteen DLB participants completed cognitive testing and functional MRI scans and were matched to AD or MCI and CN individuals whose data were obtained from the Alzheimer's Disease Neuroimaging Initiative database (https://adni.loni.usc.edu). Images were analyzed with data from Human Connectome Project (HCP) comparison individuals by using a machine learning-based subject-specific HCP atlas based on diffusion tractography. Bihemispheric functional connectivity of the PH to left FST regions was reduced in the DLB group compared with the AD and CN groups (mean±SD connectivity score=0.307±0.009 vs. 0.456±0.006 and 0.433±0.006, respectively). No significant differences were detected among the groups in connectivity within basal ganglia structures, and no significant correlations were observed between neuropsychological testing results and functional connectivity between the PH and FST regions. Performances on clock-drawing and number-cancelation tests were significantly and negatively correlated with connectivity between the right caudate nucleus and right substantia nigra for DLB participants but not for AD or CN participants. The functional connectivity between PH and FST regions is uniquely affected by DLB and may help distinguish this condition from AD.

To summarize recent advancements in artificial intelligence-driven lesion segmentation and novel neuroimaging modalities that enhance the identification and characterization of multiple sclerosis (MS) lesions, emphasizing their implications for clinical use and research. Artificial intelligence, particularly deep learning approaches, are revolutionizing MS lesion assessment and segmentation, improving accuracy, reproducibility, and efficiency. Artificial intelligence-based tools now enable automated detection not only of T2-hyperintense white matter lesions, but also of specific lesion subtypes, including gadolinium-enhancing, central vein sign-positive, paramagnetic rim, cortical, and spinal cord lesions, which hold diagnostic and prognostic value. Novel neuroimaging techniques such as quantitative susceptibility mapping (QSM), χ-separation imaging, and soma and neurite density imaging (SANDI), together with PET, are providing deeper insights into lesion pathology, better disentangling their heterogeneities and clinical relevance. Artificial intelligence-powered lesion segmentation tools hold great potential for improving fast, accurate and reproducible lesional assessment in the clinical scenario, thus improving MS diagnosis, monitoring, and treatment response assessment. Emerging neuroimaging modalities may contribute to advance the understanding MS pathophysiology, provide more specific markers of disease progression, and novel potential therapeutic targets.

Understanding the impact of epilepsy on pediatric brain tumors is crucial to diagnostic precision and optimal treatment selection. This study investigated MRI radiomics features, tumor location, voxel-based morphometry (VBM) for gray matter density, and tumor volumetry to differentiate between children with low grade glioma (LGG)-associated epilepsies and those without, and further identified key radiomics features for predicting of epilepsy risk in children with supratentorial LGG to construct an epilepsy prediction model. A total of 206 radiomics features of tumors and voxel-based morphometric analysis of tumor location features were extracted from T2-FLAIR images in a primary cohort of 48 children with LGG with epilepsy (N = 23) or without epilepsy (N = 25), prior to surgery. Feature selection was performed using the minimum redundancy maximum relevance algorithm, and leave-one-out cross-validation was applied to assess the predictive performance of radiomics and tumor location signatures in differentiating epilepsy-associated LGG from non-epilepsy cases. Voxel-based morphometric analysis showed significant positive t-scores within bilateral temporal cortex and negative t-scores in basal ganglia between epilepsy and non-epilepsy groups. Eight radiomics features were identified as significant predictors of epilepsy in LGG, encompassing characteristics of 2 locations, 2 shapes, 1 image gray scale intensity, and 3 textures. The most important predictor was temporal lobe involvement, followed by high dependence high grey level emphasis, elongation, area density, information correlation 1, midbrain and intensity range. The Linear Support Vector Machine (SVM) model yielded the best prediction performance, when implemented with a combination of radiomics features and tumor location features, as evidenced by the following metrics: precision (0.955), recall (0.913), specificity (0.960), accuracy (0.938), F-1 score (0.933), and area under curve (AUC) (0.950). Our findings demonstrated the efficacy of machine learning models based on radiomics features and voxel-based anatomical locations in predicting the risk of epilepsy in supratentorial LGG. This model provides a highly accurate tool for distinguishing epilepsy-associated LGG in children, supporting precise treatment planning. Not applicable.