Functional connectivity analyses have given considerable insights into human brain function and organization. As research moves towards clinical application, test-retest reliability has become a main focus of the field. So far, the majority of studies have relied on resting-state paradigms to examine brain connectivity, based on its low demand and ease of implementation. However, the reliability of resting-state measures is mostly poor to fair, potentially due to its unconstrained nature. Recently, naturalistic viewing paradigms have gained popularity because they probe the human brain under more ecologically valid conditions, possibly increasing reliability. We here compared the reliability of graph metrics extracted from resting-state and naturalistic viewing in functional networks, across two sessions. We show that naturalistic viewing can increase reliability over resting-state, but that its effect varies between stimuli and networks. Furthermore, we demonstrate that the effect of naturalistic viewing differs between two cohorts with Asian and European cultural backgrounds. Taken together, our study encourages the use of naturalistic viewing to increase reliability, but emphasizes the need to carefully select the appropriate stimulus for the network at hand.

Objective:The treatment of carotid artery stenosis (CAS) for stroke prevention is a matter of debate due to conflicting data, missing recent data, and advances in medical treatment options but also in interventional techniques and surgery. Therefore, the establishment of an easily available marker for brain damage might be a key tool in this patient group to guide treatment. Methods:A retrospective cross-sectional study was conducted leveraging a vascular surgery biobank of 95 patients aged 60 to 80 years. Serum neurofilament light chain (sNfL) and serum glial fibrillary acid protein (sGFAP) were evaluated using highly sensitive electrochemiluminescence immunoassays and z-scores. Discriminatory performance was assessed to differentiate between 19 symptomatic and 76 asymptomatic patients with CAS and their correlation with the degree of stenosis according to ultrasound-based North American Symptomatic Carotid Endarterectomy Trial (NASCET) criteria. Results:SNfL levels were markedly elevated in symptomatic compared with asymptomatic patients (median 17.4 vs 3.8 pg/mL; P < .001). sNfL robustly discriminated between these patients (area under the curve = 0.83; 95% confidence interval, 0.72-0.94), as did the NfL z-score (area under the curve = 0.83; 95% confidence interval, 0.71-0.95). Interestingly, within the asymptomatic cohort, sNfL levels demonstrated a significant, positive correlation with the degree of stenosis (Spearman's ρ = 0.24; P = .036). Serum levels of SGFAP were also associated with symptomatic status, albeit with a P-value >.05 (0.1 vs 0.1 pg/mL; (P = .057). Conclusions:The study provides evidence of increased sNfL in symptomatic vs asymptomatic CAS and, of note, of ongoing neuronal or glial damage in some patients with clinically asymptomatic CAS, with a positive correlation between sNfL and the degree of CAS. sNfL is a promising and already accessible blood biomarker that may guide therapeutic decisions in this patient population. The role of sGFAP remains elusive and must be evaluated in larger studies. Clinical Relevance:The CREST-2 trial has recently highlighted the high efficacy of intensive medical management in asymptomatic carotid artery stenosis (CAS), making the selection of patients for revascularization increasingly complex. Our study addresses this challenge by evaluating serum neurofilament light chain (sNfL) and serum glial fibrillary acid protein (sGFAP) as an objective biomarker for neuronal injury. Using individualized z-scores, we demonstrate that sNfL effectively differentiates symptomatic status (area under the curve = 0.828) and significantly correlates with the degree of stenosis in asymptomatic cohorts. These results suggest that sNfL can detect subclinical "silent" damage, providing a valuable biological tool to pinpoint high-risk patients who may require intervention beyond medical therapy alone. This represents a significant step toward personalized stroke prevention and refined risk stratification in the highly debated field of CAS management.

Background:Timely reperfusion offers the greatest benefit in acute ischaemic stroke within the first hour after onset. However, geographic disparities in stroke care access persist across Germany. Despite the potential of telemedicine and mobile stroke units, nationwide data that quantify existing care gaps or systematically investigate the benefit of early imaging with subsequent thrombolysis in locally accessible, CT-equipped hospitals with telemedicine are lacking. This study modelled nationwide access and compared direct transfer to specialised hospitals with a hub-and-spoke strategy (nearest CT plus telemedicine) for early thrombolysis. Methods:We performed a cross-sectional geospatial analysis combining national facility registries and 2023 hospital quality reports (data collected February 1st-July 23rd, 2025). We mapped German CT-equipped hospitals (n = 1475), stroke-ready hospitals (≥100 annual cases of "complex neurological treatment of acute stroke", n = 463) and certified stroke units (n = 349). For these facilities we modelled driving-time access up to 60 min in 5-min intervals using a local installation of openrouteservice, overlaying these on population and settlement grids. Additional scenarios simulated variations in ambulance speed (default: standard openrouteservice vehicle speed) or additional in-hospital delays when using a hub-and-spoke scenario. We then compared hub-and-spoke versus direct transfer to specialised hospitals on a national, state- and county-level. Findings:Within 30 min, nearly all residents (82,484,915/83,420,000; 98.9%) could reach a CT-equipped hospital, 90.0% (75,051,793) a stroke-ready hospital, but only 85.0% (70,875,055) a certified stroke unit. Compared with direct transfer to a stroke unit, a hub-and-spoke pathway would let 36.4% of inhabitants start imaging ≥10 min sooner (assuming normal speed, other scenarios varying from 4.5% to 40.0%) and for 14.2% it could save ≥20 min (varying from 1.1% to 18.2% across scenarios). Estimated benefits of a hub-and-spoke model depended on assumed driving speed and declined with simulated CT delays. Rural regions had particularly pronounced access gaps to stroke care, as evidenced by lower levels of urbanisation in regions with a higher hub-and-spoke benefit potential. State-level analysis illustrated heterogeneity, with 48.6% of inhabitants potentially benefitting from a hub-and-spoke model in Saxony-Anhalt, but <5% in city-states (assuming normal driving speed and 10-min delay). Interpretation:Marked intra-national inequities in rapid accessibility to stroke care persist. Leveraging CT-equipped hospitals with telemedicine could enable timelier thrombolysis within a hub-and-spoke model. Open questions regarding implementation and economics should be investigated. Funding:University Hospital Düsseldorf, the B. Braun Foundation, and the Ministry of Economic Affairs, Innovation, Digitalization and Energy of the State of North Rhine-Westphalia (funding number 005-2008-0055).

Wilson disease (WD) is a genetic disorder of copper metabolism that leads to progressive brain damage, yet the microstructural mechanisms underlying white matter alterations remain insufficiently understood. Diffusion tensor imaging provides limited biological specificity and is sensitive to free water contamination. Therefore, we applied neurite orientation dispersion and density imaging to characterize white matter microstructure and its clinical relevance in 30 patients with WD, including neurological and hepatic phenotypes, and 30 matched healthy controls. We demonstrate widespread, phenotype-specific microstructural alterations, with reductions in neurite density and orientation dispersion accompanied by increased extracellular volume fraction and diffusivity metrics (mean, axial, and radial diffusivity) in neurological WD and isolated increased isotropic volume fraction without abnormalities in conventional diffusion metrics in hepatic WD across major white matter tracts. These patterns provide a coherent explanation for previously inconsistent fractional anisotropy findings. Reduced neurite density was associated with greater neurological impairment and lower cognitive performance, particularly in processing speed and visual attention. Together, these findings highlight neurite density as a potential marker of clinically relevant white matter disruption and reveal distinct microstructural signatures across WD phenotypes, consistent with differential underlying mechanisms and a potential progression from free water-related alterations to axonal degeneration. Advanced diffusion imaging thus enables a more specific characterization of white matter pathology and supports linking microstructural alterations to clinical outcomes in neurological disorders.

Background: Neurite orientation dispersion and density imaging (NODDI) shows promise in providing specific insights into the neurite morphology underlying white matter (WM) damage in neurodegenerative diseases. This study aimed to advance the currently limited knowledge by characterizing NODDI-derived microstructural WM alterations in Wilson disease (WD) and examining their relationships with clinical symptoms. Methods: 30 WD patients, including 19 with predominant neurological involvement (neuro-WD) and 11 with hepatic manifestation (hep-WD), and 30 matched healthy controls underwent multi-shell diffusion-weighted magnetic resonance imaging. NODDI metrics, including neurite density index (NDI), orientation dispersion index (ODI), and isotropic volume fraction (ISOVF), and diffusion tensor imaging-based fractional anisotropy (FA) were estimated. Group differences in diffusion parameters across the WM skeleton were determined using tract-based spatial statistics. Additionally, voxel-wise correlations with neurological and cognitive scores were investigated. Results: We observed widespread NDI and ODI reductions in neuro-WD patients and ISOVF increases in hep-WD patients compared with healthy controls, particularly involving the corpus callosum, corona radiata, superior longitudinal fasciculus, external and internal capsule, and superior fronto-occipital fasciculus. A comparable yet more subtle pattern was found when comparing phenotypes. Distinct NDI and ODI constellations were identified as the microstructural determinants of FA alterations. Decreased NDI in the aforementioned fibers were correlated with neurological impairment, processing speed, and visual attention. Conclusions: Phenotype-specific microstructural WM alterations were identified, characterized by globally reduced axonal density and fiber organization in neuro-WD and excess free water in hep-WD. NODDI could be useful as an imaging biomarker for forecasting conversion to neurological WD manifestations and monitoring of disease progression. ### Competing Interest Statement Unrelated to this study, Alfons Schnitzler has served as a consultant for Abbott, Medtronic, and Zambon, and has received speaker honoraria from AbbVie, Abbott, Alexion, bsh medical communication, GE Healthcare, and Novartis. Julian Caspers reports no conflicts of interest related to this study; he has received honoraria for lecturing and travel expenses from Allergan and Pfizer. Christian Johannes Hartmann received honoraria and travel support from AbbVie, Abbott, Alexion Pharma Germany GmbH, Ideogen, Orphalan SA, and Univar Solutions B. V. outside the submitted work. The other authors declare no conflicts of interest related or unrelated to the present work. ### Funding Statement This research did not receive any funding. Julian Caspers work is supported by the German Federal Ministry of Education and Research (BMBF; 01GP2113C) and the German Research Foundation (DFG; KI2434/4-1). ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: The ethics committee of the Medical Faculty at the Heinrich-Heine-University Duesseldorf Germany gave ethical approval for this work (reference number 2019-470_4; 7th of February 2022). I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes Study data is not publicly available due to privacy reasons but anonymized data may be requested from the corresponding author upon reasonable request.

To evaluate the feasibility of a locally deployable large language model (LLM) system for automated MRI protocol selection addressing data privacy, annotation burden, and scalability limitations. This retrospective study included 598 German-language MRI order entries from three neuroradiology domains (brain, head/neck, spine) between June 2018 and January 2023. A radiologist labeled entries for 27 protocol classes based on institutional standard operating procedures (SOP). An SOP-grounded AI system using MedGemma 27B was developed to predict the MRI protocol from the order entry. The system was optimized using Stochastic Introspective Mini-Batch Ascent (SIMBA), a self-reflective prompt optimization algorithm, and compared with a hierarchical system that first classified the body region and then the MRI protocol. Data efficiency was evaluated using training subsets of 10–119 examples across 3 optimization runs per subset size. The flat zero-shot model achieved 73.07

Artificial intelligence (AI) has become an increasingly prominent force in medicine, driven by rapid technical advances and a growing number of clinical applications. As the field matures, it now increasingly moves from experimental development toward a phase of broader implementation. However, debates surrounding medical AI are often shaped by exaggerated risks and overly optimistic expectations. This paper seeks to contribute to a more balanced and realistic discussion. Rather than framing AI as either an existential threat or a universal solution, we advocate for an open-minded, evidence-based understanding of AI as a tool to support healthcare and discuss current and emerging challenges related to clinical validation, human–AI interaction, bias and discrimination, education, agentic AI, and the development and maintenance of trust. Question Discussions about AI in medicine continue to be dominated by exaggerated risks and overly optimistic expectations. Findings We provide a more realistic evaluation of the current opportunities of AI in medicine and want to highlight some genuine challenges that lie ahead. Clinical relevance AI is here to stay in medicine. Focusing on real and present challenges, rather than being distracted by exaggerated risks, as well as responsible expectation management, will be key to its success.

Background: Multiple sclerosis (MS) is influenced by age-related brain alterations and affects cellular aging mechanisms. Machine-learning models can estimate brain-predicted age from magnetic resonance imaging (MRI) to quantify these aging-related changes. Objectives: This study examines whether the difference between predicted and chronological age (BrainAGE) relates to clinical disability and biomarkers of neuro-axonal injury in MS. Design: This study analyzed brain-predicted age from structural 3D T1-weighted MRI in 82 patients with relapsing MS enrolled in three prospective clinical trials and 30 healthy controls. Methods: BrainAGE, calculated as MRI-predicted minus chronological age, was correlated with the Expanded Disability Status Scale (EDSS), MS Functional Composite subtests, and serum neurofilament light chain and glial fibrillary acidic protein. Results: The mean chronological age of patients and healthy controls included in this study was 39.2 and 40.9 years, respectively. Patients with MS ( n = 82) showed a higher BrainAGE (6.48 ± 6.83 years) than controls ( n = 30; 0.69 ± 6.5 years; p = 0.0002). BrainAGE increased stepwise from controls to patients with EDSS < 3 and EDSS ⩾3 ( p < 0.0001). Higher BrainAGE correlated with worse 9-Hole Peg Test (9HPT, ρ = 0.34, p = 0.002) and Timed 25-Foot Walk performance (T25FW, ρ = 0.23, p = 0.043), but not with serum neurofilament light chain ( p = 0.68) or glial fibrillary acidic protein ( p = 0.33). In multivariable regression models adjusting for chronological age, sex, disease duration, and disease-modifying therapy, BrainAGE remained significantly associated with EDSS, 9HPT, and T25FW performance. sNfL and sGFAP remained nonsignificant after adjustment. Conclusion: Our findings suggest that BrainAGE and serum biomarkers capture complementary aspects of MS pathology, supporting a multimodal approach to assess disease progression. Trial registration: ClinicalTrials.gov ID: SATURATE: NCT05701423, 360PMS: NCT06501950, SAFEGUIDE-MS: NCT06461481.

In recent years, artificial intelligence (AI) has become one of the most dynamic areas of innovation in medi-cine. In radiological diagnostics in particular, which relies heavily on the processing of complex image data, AI-based methods are being discussed as promising tools to support medical work. Applications range from the optimization of imaging procedures to the detection and classification of abnormalities to support in triage and prioritization decisions. At the same time, the use of such systems raises fundamental medical, ethical, and legal questions that go far beyond technical performance parameters. This title is also available as Open Access.

Progressive supranuclear palsy (PSP) shows a characteristic but incompletely defined pattern of neurodegeneration, in part because prior imaging studies have been limited by small and heterogeneous cohorts. Here, we consolidated evidence for PSP-related gray matter (GM) loss using a coordinate-based meta-analysis and interpreted the resulting atrophy pattern in a network and molecular framework to infer disease-relevant mechanisms. We conducted an Anatomical Likelihood Estimation (ALE) meta-analysis of whole-brain morphometry studies investigating atrophy in PSP, followed by functional decoding to evaluate the functions recruiting the atrophied regions and meta-analytic connectivity to delineate co-activation-based connectivity profiles. Finally, we explored potential neurochemical underpinnings by correlating the atrophy map with PET-derived neurotransmitter density distributions. ALE meta-analysis identified clusters of robust gray matter (GM) atrophy in PSP in the bilateral thalamus & midbrain, left anterior-dorsal insula as well as bilateral caudate nucleus. These regions were shown to be functionally associated with language, body perception, somatosensation and emotion processing. Connectivity analyses indicated coupling with fronto-insular salience-network circuitry and with key subcortical nodes, consistent with a distributed systems-level disturbance. At the molecular level, PSP-related atrophy aligned with higher densities of dopaminergic, serotonergic, and-most prominently-cholinergic markers, suggesting multi-transmitter-system vulnerability with a critical role of the cholinergic architecture. Together, these findings identify an interconnected set of cortical-subcortical targets in PSP whose functional and molecular profiles map onto core clinical features, supporting a network-based view of PSP neurodegeneration beyond isolated local atrophy with critical roles of the insula and the cholinergic system.

Ventriculoperitoneal shunts (VPS) are an essential part of the treatment of hydrocephalus, with numerous valve models available with different ways of indicating pressure levels. The model types often need to be identified on X‑rays to assess pressure levels using a matching template. Artificial intelligence (AI), in particular deep learning, is ideally suited to automate repetitive tasks such as identifying different VPS valve models. The aim of this work was to investigate whether AI, in particular deep learning, allows the identification of VPS models in cranial X‑rays. 959 cranial X‑rays of patients with a VPS were included and reviewed for image quality and complete visualization of VPS valves. The images included four VPS model types: Codman Hakim (n = 774, 81

Removal of nuisance signals (such as motion) from the BOLD time series is an important aspect of preprocessing to obtain meaningful resting-state functional connectivity (rs-FC). The nuisance signals are commonly removed using denoising procedures at the finest resolution, that is the voxel time series. Typically, the voxel-wise time series are then aggregated into predefined regions or parcels to obtain an rs-FC matrix as the correlation between pairs of regional time series. Computational efficiency can be improved by denoising the aggregated regional time series instead of the voxel time series. However, a comprehensive comparison of the effects of denoising on these two resolutions is missing. In this study, we systematically investigate the effects of denoising at different time series resolutions (voxel-level and region-level) in 370 unrelated subjects from the HCP-YA dataset. Alongside the time series resolution, we considered additional factors such as aggregation method (Mean and first eigenvariate [EV]) and parcellation granularity (100, 400, and 1000 regions). To assess the effect of those choices on the utility of the resulting whole-brain rs-FC, we evaluated the individual specificity (fingerprinting) and the capacity to predict age and three cognitive scores. Our findings show generally equal or better performance for region-level denoising with notable differences depending on the aggregation method. Using Mean aggregation yielded equal individual specificity and prediction performance for voxel-level and region-level denoising. When EV was employed for aggregation, the individual specificity of voxel-level denoising was reduced compared to region-level denoising. Increasing parcellation granularity generally improved individual specificity. For the prediction of age and cognitive test scores, only fluid intelligence indicated worse performance for voxel-level denoising in the case of aggregating with the EV. Based on these results, we recommend the adoption of region-level denoising for brain-behavior investigations when using Mean aggregation. This approach offers equal individual specificity and prediction capacity with reduced computational resources for the analysis of rs-FC patterns.

Selection of appropriate imaging sequences protocols for cranial magnetic resonance imaging (MRI) is crucial to address the medical question and adequately support patient care. Inappropriate protocol selection can compromise diagnostic accuracy, extend scan duration, and increase the risk of misdiagnosis. Typically, radiologists determine scanning protocols based on their expertise, a process that can be time-consuming and subject to variability. Language models offer the potential to streamline this process. This study investigates the capability of bidirectional encoder representations from transformers (BERT)-based models to suggest appropriate MRI protocols based on referral information.A total of 410 anonymized electronic referrals for cranial MRI from a local order-entry system were categorized into nine protocol classes by an experienced neuroradiologist. A locally hosted instance of four different, pre-trained BERT-based classifiers (BERT, ModernBERT, GottBERT, and medBERT.de) were trained to classify protocols based on referral entries, including preliminary diagnoses, prior treatment history, and clinical questions. Each model was additionally fine-tuned for local language on a large dataset of electronic referrals.The model based on medBERT.de with local language fine-tuning was the best-performing model and correctly predicted 81% of all protocols, achieving a macro-F1 score of 0.71, macro-precision and macro-recall values of 0.73 and 0.71, respectively. Moreover, we were able to show that local language fine-tuning led to performance improvements across all models.These results demonstrate the potential of language models to predict MRI protocols, even with limited training data. This approach could accelerate and standardize radiological protocol selection, offering significant benefits for clinical workflows.

To investigate the feasibility of Retrieval-augmented Generation (RAG)-enhanced Large Language Models (LLMs) in answering questions about two German neurovascular guidelines. Four LLMs (GPT-4o-mini, Llama 3.1 405B Instruct Turbo, Mixtral 8 × 22B Instruct, and Claude 3.5 Sonnet) with RAG as well as GPT-4o-mini without RAG were evaluated for generating answers about two German neurovascular guidelines (“S3 Guideline for Diagnosis, Treatment, and Follow-up of Extracranial Carotid Stenosis” and “S2e Guideline for Acute Therapy of Ischemic Stroke”). The answers were classified as “correct”, “inaccurate”, or “incorrect” by two neurovascular experts in consensus. Additionally, retrieval performance of five retrieval strategies was analyzed on a synthetic dataset of 384 questions. Claude Sonnet 3.5 achieved the highest answer correctness (70.6

Purpose Modern endovascular techniques enable the treatment of various neurovascular diseases. Given the complexity of these interventions, a certification system was introduced to ensure standardized care at specialized treatment centers. We used a driving-time-based isochrone approach to identify care gaps in different German Society of Interventional Radiology (DeGIR) certified neurovascular treatment centers. Methods DeGIR-certified neurovascular centers for minimally invasive stroke care (module E), neurovascular vessel anomalies (module F), and neurovascular therapy (module EF) were geocoded and driving-time-based isochrones were calculated for 30, 60, 90, and 120 min. The resulting contours were aggregated and combined with the 2025 population estimates from the Global Human Settlement Layer to estimate residents’ access. Results The analysis identified gaps in under-60-minute reachability, notably in northeastern Germany and parts of Rhineland-Palatinate, Saarland, and the southwest, with modules EF and F most affected, while module E fared better. Within 120 min, coverage was nearly complete across all modules. On a population level, 59.4% of residents lived within 30 min, 92.81% within 60 min, and 99.98% within 120 min of a module E center. Module F reached 45.8%, 84.26%, and 99.73%, respectively, with module EF showing intermediate accessibility. Discussion The driving-time-based isochrone approach identifies regions where access to specialized neurovascular care is limited—a critical issue in emergencies like stroke or aneurysm hemorrhage. Although immediate stroke care is generally more accessible than care for neurovascular anomalies, thrombectomy within an acceptable timeframe is not available to the entire population. These findings can guide strategies to enhance neurovascular care across Germany.

This study investigates the influence of carotid artery elongation on neurovascular intervention and outcome in acute stroke treatments proposing an easily assessable imaging marker for carotid elongation. 118 patients who underwent mechanical thrombectomy for middle cerebral artery occlusions were included. The carotid elongation ratio (CER), center-line artery length to scan’s Z-axis, was measured on the affected side in CT-angiographies. Full and partial correlations of CER with periprocedural times, complications and outcome were computed. Multivariate logistic regression, including comorbidities, for prediction of dichotomized mRS outcome after 3 months was performed. CER showed no significant correlation with recanalization success. Weak, outlier-driven correlation was found with recanalization time (p = 0.021, cor = 0.2). Weak correlations were found with improvement of NIHSS score at discharge and mRS score after 3 months (p = 0.023 and p = 0.031, each rho=-0.2). There was moderate correlation with NIHSS score at discharge (p = 0.001, rho = 0.3). Patients with favorable outcomes (mRS 0–2) exhibited lower CER (p = 0.012). Partial correlations of CER with favorable outcomes were observed after correcting for age, sex and cardiovascular risk factors (cor = 0.2, p = 0.048). Multivariate analysis (Nagelkerke’s R2 = 0.42) identified NIHSS score at admission, diabetes, hypertension and intervention time as significant factors for predicting outcome at 3 month, while CER showed the highest log Odd’s (2.97). Correlations between CER and clinical improvement suggest that carotid elongation might be a risk factor for poorer outcome without relevant effect on endovascular treatment and should not guide treatment decisions. Further studies should consider carotid elongation as an individual neurovascular risk factor, independent of hypertension. •The study investigates the influence of carotid elongation on endovascular stroke treatment and outcome. •There is correlation between carotid elongation and clinical improvement. However, no relevant effect on endovascular treatment was found. •Carotid elongation should not dictate acute stroke treatment. Rather it should be considered as an individual neurovascular risk factor.