Neuroinformatics

Neurodegenerative diseases such as Mild Cognitive Impairment (MCI), Multiple Sclerosis (MS), Parkinson s Disease (PD), and Amyotrophic Lateral Sclerosis (ALS) are becoming more prevalent. Each of these diseases, despite its specific pathophysiological mechanisms, leads to widespread reorganization of brain activity. However, the corresponding neurophysiological signatures of these changes have been elusive. As a consequence, to date, it is not possible to effectively distinguish these diseases from neurophysiological data alone. This work uses Magnetoencephalography (MEG) resting-state data, combined with interpretable machine learning techniques, to support differential diagnosis. We expand on previous work and design a Riemannian geometry-based classification pipeline. The pipeline is fed with typical connectivity metrics, such as covariance or correlation matrices. To maintain interpretability while reducing feature dimensionality, we introduce a classifier-independent feature selection procedure that uses effect sizes derived from the Kruskal-Wallis test. The ensemble classification pipeline, called REDDI, achieved a mean balanced accuracy of 0.81 (+/-0.04) across five folds, representing a 13% improvement over the state-of-the-art, while remaining clinically transparent. As such, our approach achieves reliable, interpretable, data-driven, operator-independent decision-support tools in Neurology.

Background: Developing generalizable neuroimaging models is often hindered by limited labeled data which has led to an increased interest in unsupervised inverse learning. Existing approaches often neglect geometric principles and struggle with diverse pathologies. We propose a symmetry-informed inverse learning foundation model to address these shortcomings for robust and efficient anomaly detection in brain MRI. Methods: Our framework employs a reconstruction-to-embedding pipeline, trained exclusively on healthy brain MRI slices. A 2D U-Net uses a novel, symmetry-aware masking strategy to reconstruct a disorder-free slice. Difference maps are embedded into a 1024-dimensional latent space via a Beta-VAE. Anomaly scoring is performed using Mahalanobis distance. We evaluated generalization by fine-tuning on external lesion datasets, BraTS Africa (SSA), and the ADNI-derived Alzheimer disease cohort (Alz). Results: On the source metastasis (Mets) dataset, the framework achieved high performance (AB1+MSE: 99.28% accuracy, 99.79% sensitivity). Generalization to the external lesion dataset (SSA) was robust, with the Symmetry ROC configuration achieving 91.93% accuracy. Transfer to the Alzheimer dataset (Alz) was more challenging, achieving a peak accuracy of 70.54% with a high false-positive rate, suggesting difficulty in separating subtle, diffuse changes. Conclusion: The symmetry-informed inverse learning framework establishes a robust foundation model for neuroimaging, showing strong performance for focal lesions and successful generalization under domain shift. Limitations in diffuse neurodegeneration underscore the necessity for richer representations and multimodal integration to improve future foundation models.

Menstrual cycles are major biological events with extensive effects on the brain and cognition, experienced by half of the human population. To develop a comprehensive account of human cognition, it is necessary to successfully integrate and characterize menstrual cycle effects in cognitive science research. However, current approaches to menstrual cycle analysis suffer from low data resolution and are not well-equipped to capture the highly variable, non-linear changes in outcomes of interest across the cycle. We present a validated standardized method remedying these issues, demonstrate its utility using hormonal, behavioral, and neuroimaging data, and provide an open-source toolkit to facilitate its use.

Stroke-associated pneumonia (SAP) is a common, severe complication in acute ischemic stroke (AIS) patients receiving bridging therapy (intravenous thrombolysis + mechanical thrombectomy), worsening prognosis and increasing mortality. Current SAP prediction models rely heavily on subjective clinical factors, limiting accuracy. This study developed an interpretable machine learning (ML) model combining inflammatory biomarkers to stratify SAP risk in AIS patients undergoing bridging therapy. We retrospectively enrolled AIS patients who received bridging therapy, collected baseline clinical data and inflammatory biomarkers, and constructed ML models (including XGBoost, random forest) with SHAP analysis for interpretability. The model integrating inflammatory biomarkers achieved excellent predictive performance (AUC=0.XX, 95%CI: XX-XX), outperforming traditional clinical models. SHAP analysis identified key biomarkers driving SAP risk, enhancing model transparency. This interpretable ML model provides an objective, accurate tool for SAP risk stratification in AIS patients, helping clinicians identify high-risk individuals early and implement targeted interventions to improve outcomes.

Background and Purpose: 2-[18F] fluoro-2-deoxy-D-glucose positron emission tomography (static PET) has mixed specificity and sensitivity in targeting epileptic zones in the noninvasive stage of epilepsy surgery evaluations. We compared the signal quality of static PET compared to a method of interictal dynamic PET (iD-PET). Materials and Methods: We calculated the signal quality of static PET and iD-PET obtained from a cohort of patients with focal epilepsy. We developed a Bayesian regional estimated signal quality (BRESQ) technique to objectively compare signal-to-noise ratios (SNRs) by region of interest (ROI) within subjects. Results: Adjusted for ROI size and neighboring regions, iDPET was superior to sPET with probability >95% in 8/36 regions; >90% in 21/36 regions; >80% in 29/36 regions. The top five regions with the largest adjusted SNR differences (greatest magnitude of iDPET superiority) were the Temporal Mesial (Left and Right), Occipital Lateral (Left and Right), and the Left Frontal Inferior Base. Conclusions: We found that iDPET yielded a superior SNR in most ROI. BRESQ offers a scalable and generalizable method to quantify signal quality between brain mapping modalities.

Background and Objectives: We report the first intraoperative deployment of a real-time machine vision system in neurosurgery, derived from our previous anatomical detection work, automatically identifying structures during endoscopic endonasal surgery. Existing systems demonstrate promising performance in offline anatomical recognition, yet so far none have been implemented during live operations. Methods: A real-time anatomy detection model was trained using the YOLOv8 architecture (Ultralytics). Following training completion in the PyTorch environment, the model was exported to ONNX format and further optimized using the NVIDIA TensorRT engine. Deployment was carried out using the NVIDIA Holoscan SDK, the system ran on an NVIDIA Clara AGX developer kit. We used the model for real-time recognition of intraoperative anatomical structures and compared it with the same video labelled manually as reference. Model performance was reported using the average precision at an intersection-over-union threshold of 0.5 (AP50). Furthermore, end-to-end delay from frame acquisition to the display of the annotated output was measured. Results: A mean AP50 of 0.56 was achieved. The model demonstrated reliable detection of the most relevant landmarks in the transsphenoidal corridor. The mean end-to-end latency of the model was 47.81 ms (median 46.57 ms). Conclusion: For the first time, we demonstrate that clinical-grade, real-time machine-vision assistance during neurosurgery is feasible and can provide continuous, automated anatomical guidance from the surgical field. This approach may enhance intraoperative orientation, reduce cognitive load, and offer a powerful tool for surgical training. These findings represent an initial step toward integrating real-time AI support into routine neurosurgical workflows.

Predicting response to induction chemotherapy (IC) and overall survival (OS) is critical for optimizing treatment in patients with locally advanced nasopharyngeal carcinoma (LANPC). This study aimed to develop and validate a multi-task deep learning model integrating pretreatment MRI and whole slide images (WSIs) to predict IC response and OS in LANPC. Pretreatment MRI and WSIs from 404 patients with LANPC were retrospectively collected to construct a multi-task model (MoEMIL) for the simultaneous prediction of early IC response and OS. MoEMIL employed multi-instance learning to process WSIs, PyRadiomics and a convolutional neural network (ResNet50) to extract MRI features, and fused multimodal features through a multi-gate mixture-of-experts architecture. Clustering-constrained attention multiple instance learning and gradient-weighted class activation mapping were applied for visualization and interpretation. MoEMIL effectively stratified patients into good and poor IC response groups, achieving areas under the curve of 0.917, 0.869, and 0.801 in the train, validation, and test sets, respectively, and outperformed the deep learning radiomics model, the pathomics model and TNM staging. The model also stratified patients into high- and low-risk OS groups (P < 0.05). MoEMIL shows promise as a decision-support tool for early IC response prediction and prognostication in LANPC. Author SummaryWe have developed a deep learning model that integrates two types of medical images, including magnetic resonance imaging (MRI) and digital pathological slices, to simultaneously predict response to induction chemotherapy and prognosis in patients with locally advanced nasopharyngeal carcinoma. Current treatment decisions primarily rely on traditional tumor staging (TNM), which often fails to comprehensively reflect the complexity of the disease. Our model, named MoEMIL, was trained and tested on data from 404 patients across two hospitals and consistently outperformed both single-model approaches and TNM staging methods. By identifying patients who exhibit poor response to induction chemotherapy or higher prognostic risk, our tool can assist clinicians in achieving personalized treatment, enabling intensified management for high-risk patients and avoiding unnecessary side effects for low-risk patients. Additionally, we visualize the models reasoning process through heat map generation, which highlights the image regions exerting the greatest influence on prediction outcomes. This work represents a step toward more precise treatment for nasopharyngeal carcinoma; however, larger-scale prospective studies are required before the model can be integrated into routine clinical practice.

Background and Objectives: Normal appearing white matter (NAWM) may already harbor subtle microstructural alterations not yet visible on conventional MRI. Quantitative Multi-Parametric Mapping (qMPM) such as Magnetization Transfer saturation (MTsat), longitudinal relaxation rate (R1), and Proton Density (PD) offer new possibilities for analyzing NAWM which are sensitive to demyelination, axonal loss, and edema. We aimed to characterize these alterations within white matter hyperintensities (WMH) and the perilesional NAWM (pNAWM), to gain insights into the underlying process of lesion progression. We also investigated their association with cerebrovascular risk factors (CVRF) and long-term cognitive performance. Methods: This investigation included the cerebral MRI data of 245 participants from the prospective Berlin Longterm Observation of Vascular Events (BeLOVE) study. Furthermore, 121 participants cognitive performance was evaluated at baseline and longitudinally at 2 years follow-up using Montreal Cognitive Assessment (MoCA). Regions of interest (ROIs) of WMH, pNAWM at 1, 2, 3 mm were assessed in comparison to the mirrored contralesional white matter (cWM). Linear mixed effects models were employed to demonstrate the pairwise comparisons between each region using estimated marginal means and the association of MPM metrics with CVRFs. Linear regression was used to assess the association with cognitive performance. Results: In 245 participants, (mean age 62 years, SD: 12 years; 29.8% females), MPM metrics demonstrated a clear spatial gradient of microstructural injury. MTsat and R1 values were lower in WMH compared to cWM (lower case Greek beta = -0.48 (-0.52 - -0.44) and lower case Greek beta = -0.07 (-0.08 - -0.06), p<0.001, respectively) and showed gradual recovery with increasing distance indicating a microstructural gradient in pNAWM. Conversely, PD values were higher in WMH and decreased peripherally (lower case Greek beta = 2.32 (2.05 - 2.61, p<0.001). No substantial associations were found between MPM parameters and CVRFs in our cohort. At baseline and 2-year follow-up, cognitive performance was associated with higher pNAWM R1 values, whereas MTsat were only moderately associated. Discussion: Quantitative MPM reliably detects microstructural alterations not only within WMH, but also in pNAWM, confirming the high sensitivity of qMPM to subtle tissue pathology and support its utility as a promising biomarker for longitudinal studies and monitoring therapeutic effects.

BackgroundDistinguishing epilepsy from functional/dissociative seizures (FDS) is an ongoing diagnostic challenge. Misdiagnosis delays appropriate treatment and puts patients at significant risk. Quantitative analyses of clinical EEG offer a potential avenue for developing decision-support tools in the diagnosis of seizure disorders. Recent work using univariate features demonstrated that reliably identifying diagnostic traits in the presence of confounding factors remains challenging. However, diagnostic information might be available in multivariate features such as network-based measures. Using a well-controlled dataset, we run the first diagnostic accuracy study assessing the potential of multivariate resting-state EEG markers to directly discriminate between a diagnosis of epilepsy and one of FDS at the time when a diagnosis is suspected and prior to treatment initiation. MethodsThe dataset, previously examined in a published study, includes 148 age- and sex-matched individuals with suspected seizure disorders who were later diagnosed with non-lesional epilepsy (n=75) or FDS (n=73). Eyes-closed, resting-state EEG data used for the analyses were normal on visual inspection, and acquired while participants were medication-free. Functional network measures in the 6-9 Hz range were extracted and machine learning implemented to assess their predictive potential; different model configurations (including varying model types, dimensionality reduction methods, and approaches to enhance feature stability) were tested to identify the most promising approach for future translational implementations. ResultsNetwork measures derived from resting-state EEG discriminate between conditions at levels significantly above chance (maximum balanced accuracy: 67.5%). Their sensitivity to epilepsy (81.8%) is consistently higher than their sensitivity to FDS (53.3%). A systematic assessment of model choices indicates that improving the temporal stability of network features through epoch-wise averaging improves classification accuracy (62.6% to 67.5%). Multiple nonlinear model types succeed on the classification problem, with the three-best performing assigning a consistent diagnostic label to 77.5% of the individuals; however, model choice remains a strong determinant of overall classification accuracy. Dimensionality reduction did not provide a significant advantage in our models. ConclusionWe establish evidence for the clinical validity of selected network-based markers to discriminate between a diagnosis of non-lesional epilepsy and FDS prior to treatment initiation, highlighting the measures potential to support post-test probability estimation in the clinic. Our models, configured to optimise balanced accuracy, classified people with epilepsy more accurately than people with FDS, indicating that these measures are specific to epilepsy and should not be interpreted as markers of a positive diagnosis of FDS.

Computer vision and deep learning techniques, including convolutional neural networks (CNNs) and transformers, have increased the performance of medical image classification systems. However, training deep learning models using medical images is a challenging task that necessitates a substantial amount of annotated data. In this paper, we implement data augmentation strategies to tackle dataset imbalance in the VinDr-SpineXR dataset, which has a lower number of spine abnormality X-ray images compared to normal spine X-ray images. Geometric transformations and synthetic image generation using Generative Adversarial Networks are explored and applied to the abnormal classes of the dataset, and classifier performance is validated using VGG-16 and InceptionNet to identify the most effective augmentation technique. Additionally, we introduce a hybrid augmentation technique that addresses class imbalance, reduces computational overhead relative to a GAN-only approach, and achieves ~99% validation accuracy with both classifiers across all three case studies. Keywords: Data augmentation, Generative Adversarial Network, VGG-16, InceptionNet, Class imbalance, Computer vision, Spine X-ray, Radiology.

Brain activity comprises both rhythmic (periodic) and arrhythmic (aperiodic) components. These signal elements vary across healthy aging, and disease, and may make distinct contributions to conscious perception. Despite pioneering techniques to parameterize rhythmic and arrhythmic neural components based on power spectra, the methodology for quantifying rhythmic activity remains in its infancy. Previous work has relied on parametric estimates of rhythmic power extracted from specparam, or estimates of rhythmic power obtained after detrending neural spectra. Variation in analytical choices for isolating brain rhythms from background arrhythmic activity makes interpreting findings across studies difficult. Whether these current approaches can accurately recover the independent contribution of these neural signal elements remains to be established. Here, using simulation and parameter recovery approaches, we show that power estimates obtained from detrended spectra conflate these two neurophysiological components, yielding spurious correlations between spectral model parameters. In contrast, modelled rhythmic power obtained from specparam, which detrends the power spectra and parametrizes brain rhythms, independently recovers the rhythmic and arrhythmic components in simulated neural time series, minimising spurious relationships. We validate these methods using resting-state recordings from a large cohort. Based on our findings, we recommend modelled rhythmic power estimates from specparam for the robust independent quantification of rhythmic and arrhythmic signal components for cognitive neuroscience.

Patients who participate in intracranial neuroscience research make invaluable contributions to our understanding of the brain, accelerating the development of neurotechnological interventions. Engagement of patients as part of this research presents unique challenges, where study goals can be distant from immediate clinical applications and require specialized domain knowledge. Yet methods for meaningfully integrating patient communities as part of these research efforts is essential, as intracranial neuroscience guides the application of artificial intelligence for understanding and enhancing human cognition. In order to identify what patients consider meaningful research engagement we interviewed individuals who participated in a study during their Deep Brain Stimulation (DBS) surgery and attended a group event where they interacted with our research team. Analysis of semi-structured interviews identified four main themes: interest in science and the future of clinical care, contributing to science to improve lives, connecting with others, and accessibility considerations. Based on these insights, we propose strategies for transformational participation of patient communities in intracranial neuroscience research with respect to engagement objectives, communication and scope. This approach offers a foundation for sustaining relationships between scientists and communities rooted in trust and transparency, to ensure that impacts of neurotechnology on human health and cognition are aligned with patient needs as well as desired public values.

Single photon emission computed tomography (SPECT) is a highly specialized imaging modality that enables measurement of regional cerebral perfusion and, in particular, resting cerebral blood flow (rCBF). Recent technological advances have improved SPECT quantification and reliability, making it increasingly useful for studying rCBF abnormalities and perfusion network alterations in psychiatric and neurological disorders. To characterize large scale functional organization in SPECT data, data driven decomposition methods such as independent component analysis (ICA) have been used to extract covarying perfusion patterns that map onto interpretable brain networks. Blind ICA provides a data driven approach to estimate these networks without strong prior assumptions. More recently, a hybrid approach that leverages spatial priors to guide a spatially constrained ICA (sc ICA) have been used to fully automate the ICA analysis while also providing participant-specific network estimates. While this has been reliably demonstrated in fMRI with the NeuroMark template, there is currently no comparable SPECT template. A SPECT template would enable automatic estimation of functional SPECT networks with participant-specific expressions that correspond across participants and studies. The current study introduces a new replicable NeuroMark SPECT template for estimating canonical perfusion covariance patterns (networks). We first identify replicable SPECT networks using blind ICA applied to two large sample SPECT datasets. We then demonstrate the use of the resulting template by applying sc-ICA to an independent schizophrenia dataset. In sum, this work presents and shares the first NeuroMark SPECT template and demonstrating its utility in an independent cohort, providing a scalable and robust framework for network-based analyses.

Saccadic eye movements are established biomarkers in neuroscience and clinical neurology, where video-oculography (VOG) remains the gold standard. However, VOG's high cost, bulky equipment, and poor portability restrict its clinical utility. Electrooculography (EOG) offers a promising alternative by detecting cornea-retinal potential changes during eye movements. To enable quantitative saccadic analysis using EOG as a VOG alternative, this study develops and validates a mathematical transformation model converting EOG data into VOG-equivalent values. A prospective observational study was conducted on 4 healthy adults without neurological or sleep disorders. Horizontal saccades were recorded simultaneously using EOG and VOG during controlled gaze shifts. EOG peak saccadic velocity was derived from voltage change rate, whereas VOG was calculated from angular displacement over time. A derivation dataset of fixed horizontal saccades ({+/-}20{degrees}) formulated the transformation model, achieving a strong correlation coefficient (r = 0.95 rightward, r = 0.93 leftward, p < 0.0001). Multiple filter settings were evaluated, and 0.3 Hz high-pass and 35 Hz low-pass filtering were identified as optimal. The fixed horizontal saccades derived model was applied to a validation dataset of random horizontal saccades, confirming robustness across saccades without significant differences from VOG measurements. These findings establish EOG's feasibility for quantitative analysis of horizontal saccades and provide a validated transformation model. By systematically optimizing filtering parameters, this approach enables EOG as a cost-effective VOG alternative while maintaining high-precision measurement accuracy.

In this work, the impact of outliers on the performance of machine learning and deep learning models is investigated, specifically for the case of histopathological images of colorectal cancer stained with Haematoxylin and Eosin. The evaluation of the impact is done through the systematic comparison of one machine learning model (Random Forests) and one deep learning model (ResNet-18). Both models were trained with the popular NCT-CRC-HE-VAL-100K dataset and tested on the CRC-HE-VAL-7K companion set. Then, a curation process was performed by analysing the divergence of patches based on chromatic, textural and topological features of the training set and removing outliers to repeat the training with a cleaned dataset. The results showed that machine learning models, can benefit more from improvements in the quality of data, than deep learning models. Further, the results suggest that deep learning models are more robust to outliers as, through the training process, the architectures can learn features other than those previously mentioned.

BACKGROUNDMild traumatic brain injury (mTBI) is a signature injury in civilian and military populations that remains invisible to detection by conventional radiological methods. Diffusion MRI has been identified as a potential clinical tool for revealing subtle microstructural alterations associated with mTBI. OBJECTIVEThis study evaluates whether a comprehensive and powerful diffusion MRI (dMRI) technique called mean apparent propagator (MAP) MRI can detect sequelae of mTBI. METHODSWe analyzed data from 417 participants of the GE/NFL prospective mTBI study which included 143 matched controls (mean age, 21.9 {+/-} 8.3 years; 76 women) and 274 patients with acute mTBI and GCS [&ge;]13 (mean age, 21.9 {+/-} 8.5 years; 131 women). All participants underwent MRI exams at up to four visits including structural high-resolution T1W, T2W, FLAIR-T2W, and dMRI, in addition to clinical assessments of post-concussive physical symptoms (RPQ-3), psychosocial functioning and lifestyle symptoms (RPQ-13), and postural stability (BESS). The dMRI data for each subject were co-registered across all visits and analyzed using the MAP-MRI framework to measure and map the distribution of net microscopic displacements of diffusing water molecules in tissue and ultimately compute the microstructural MAP-MRI tissue parameters including propagator anisotropy (PA), Non-Gaussianity (NG), return-to-origin probability (RTOP), return-to-axis probability (RTAP), and return-to-plane probability (RTPP). We quantified voxel-wise and region-of-interest (ROI)-based changes in these parameters across all four visits. RESULTSMAP-MRI parameter values were within the expected ranges and showed relatively little variation across visits. We found no significant differences in the longitudinal trajectories of these parameters between mTBI patients and controls. At acute post-injury timepoints, RPQ-3 and RPQ-13 scores were increased in mTBI patients relative to controls, while BESS scores were not significantly different between groups. Analysis of dMRI metrics and clinical mTBI markers showed significant correspondence between MAP-MRI metrics in cortical gray matter, caudate and pallidum and BESS scores. CONCLUSIONWe developed and tested a state-of-the-art quantitative image processing pipeline for sensitive analysis and detection of subtle tissue changes in longitudinal clinical diffusion MRI data. The absence of a significant statistical difference between populations in the dMRI parameters in this study suggests that the mTBI corresponded to acute post-injury clinical symptoms but that the injury was not severe enough to cause detectable microstructural damage/alterations, and that increased diffusion sensitization combined with improved analysis techniques may be needed. CLINICAL IMPACTThese findings suggest that acute mTBI (GCS[&ge;]13) may not be detectable with diffusion MRI. TRIAL REGISTRATIONClinicalTrials.gov NCT02556177

ImportanceEpilepsy is one of the most common neurological disorders globally. A significant proportion of patients fail to achieve effective seizure control with medication and ultimately develop drug-resistant epilepsy, particularly mesial temporal lobe epilepsy (MTLE). While surgical resection and laser interstitial thermal therapy (LITT) are effective treatments for drug-resistant MTLE, these procedures may be associated with severe adverse events. In contrast, allogeneic induced pluripotent stem cell (iPSC)-based therapy is expected to offer a novel, potentially safer therapeutic approach with fewer side effects for patients with drug-resistant MTLE. ObjectiveTo evaluate the safety and preliminary efficacy of a single intracranial injection of ALC05 (iPSC-derived GABAergic interneurons) in patients with unilateral MTLE, and to assess the therapeutic effects of different dosage levels. Design, Setting, and ParticipantsThis single-center, randomized, double-blind, Phase 1 clinical trial will enroll 12 subjects with unilateral MTLE. All subjects will be randomly assigned to either the low-dose or high-dose group in a 1:1 ratio. To minimize risks at each dose level, the first subject in each dose group will be monitored for safety for at least 3 months following ALC05 injection and must demonstrate acceptable safety and tolerability before the remaining subjects are enrolled. The primary outcome will be the incidence and severity of adverse events (AEs) and serious adverse events (SAEs). Secondary outcomes include cell engraftment and survival, responder rate, and seizure frequency. The follow-up period for this study is 1 year. After completing the follow-up period within this study, subjects will enter a 15-year long-term safety follow-up. DiscussionMTLE remains a significant challenge in neurology. The results of this study will provide critical data regarding the feasibility and preliminary efficacy of ALC05 in treating MTLE and may offer a transformative therapeutic option for this condition.

Background: Paediatric pneumonia is a leading cause of childhood morbidity and mortality worldwide. Chest X-rays (CXR) are an important diagnostic tool in the diagnosis of pneumonia, but shortages in specialist radiology services lead to clinically significant delays in CXR reporting. The ability to communicate findings both to clinicians and laypersons allows MLLMs to be deployed throughout clinical workflows, from image analysis to patient communication. However, MLLMs currently underperform state-of-the-art deep learning classifiers. Objective: To evaluate the diagnostic accuracy of ensemble strategies with MLLMs compared to the baseline average agent for paediatric radiological pneumonia detection. Methods: We conducted a retrospective cohort study using paediatric CXRs from two independent hospital datasets totalling 2300 CXRs. Fifteen MedGemma-4B-it agents independently classified each CXR into five pneumonia likelihood categories. Majority voting, soft voting, and GPTOSS-20B aggregation were compared against the average agent performance. The primary metric evaluated was OvR AUROC. Secondary metrics included accuracy, sensitivity, specificity, F1-score, Cohen's kappa, and OvO AUROC. Results: Soft voting achieved improvements in OvR AUROC (p_balanced = 0.0002, p_real-world = 0.0003), accuracy (p_balanced = 0.0008, p_real-world < 0.0001), Cohen's Kappa (p_balanced = 0.0006, p_real-world = 0.0054) and OvO AUROC (p_balanced < 0.0001, p_real-world = 0.0011) across both datasets, and a superior F1-value (pbalanced = 0.0028) for the balanced dataset. Conclusion: Soft voting enhances MedGemma's diagnostic discriminatory performance for paediatric radiological pneumonia detection. Our system enables privacy-preserving, near real-time clinical decision support with explainable outputs, having potential for integration into emergency departments. Our system's high specificity supports triage by flagging high-risk radiological pneumonia cases.

PurposePrevious research has demonstrated effects of head impact exposure on cortical neurophysiology, which may help with understanding variability in clinical sequelae. In separate lines of research, neurochemical and gene transcription markers of vulnerability to traumatic brain injury (TBI) have been established. The purpose of this study was to examine whether these cortical neurochemical and gene transcription gradients are spatially aligned with neurophysiological effects. Methods and MaterialsMagnetoencephalography (MEG) data was collected at a total of 278 pre- and post-season timepoints from 91 high school football players across up to four seasons of play. Of the 91 football players, 10 experienced a concussion, and of the remaining 81 non-concussed players, 71 met the criteria for complete imaging and kinematic data, with post-season evaluations less than six weeks after the end of the season. Head impacts were tracked over the course of the season with helmet-mounted sensors. MEG data underwent source-imaging, frequency-transformation, spectral parameterization, and linear modeling to examine the effects of concussive and non-concussive head impact exposure on pre-to-post-season changes in rhythmic and arrhythmic neurophysiological activity. To determine clinical effects, parent reported Post-Concussive Symptom Inventory scores related to cognitive symptoms were correlated with cortical neurophysiological changes. Multi-atlas data of neurochemical system densities from neuromaps and gene expression from the Allen Human Brain Atlas were examined for alignment with head impact-related alterations in neurophysiology via nonparametric spin-tests with autocorrelation-preserving null models (5,000 Hungarian spins; pFDR <.05). ResultsConcussion-related reductions in cortical excitability (i.e., aperiodic exponent slowing) were aligned with atlas-based norepinephrine transporter (NET) and alpha-4 beta-2 nicotinic receptor (4{beta}2) densities, and with apolipoprotein E (APOE) and brain-derived neurotrophic factor (BDNF) expression levels. More severe cognitive symptoms associated with concussion-related slowing of aperiodic neurophysiology were also aligned with atlas-based NET and 4{beta}2 receptor densities. Similar changes in cortical excitability related to non-concussive head impact exposure were colocalized with serotonin receptor (5-HT1A) density maps and APOE and BDNF expression. Rhythmic alpha activity was reduced by concussion and colocalized with histamine (H3) and mu-opioid (MOR) receptors, among others, as well as with gene transcription atlases of APOE and C-C chemokine receptor 5 (CCR5). ConclusionsThese findings extend our previous work to show that the effects of head impact exposure on neurophysiology are strongest in cortical areas with specific neurochemical and genetic profiles that are known to signal vulnerability to traumatic brain injury, and that these spatial alignments are also associated with self-reported symptom severity. Clinical Relevance / ApplicationChange in cortical excitability, as measured here by MEG, has potential value as a clinical tool for concussion diagnosis and prognosis. We provide genetic and neurochemical contextualization for these changes that may extend their clinical applications, for example to concussion risk assessment and pharmacotherapies.

Dysarthric speech severity assessment typically requires either trained clinicians or supervised machine learning models built from labelled pathological speech data, limiting scalability across languages and clinical settings. We present a training-free method (no supervised severity model is trained; feature directions are estimated from healthy control speech using a pretrained forced aligner) that quantifies dysarthria severity by measuring the degradation of phonological feature subspaces within frozen HuBERT representations. For each speaker, we extract phone-level embeddings via Montreal Forced Aligner, compute d scores along phonological contrast directions (nasality, voicing, stridency, sonorance, manner, and four vowel features) derived exclusively from healthy control speech, and construct a 12-dimensional phonological profile. Evaluating 890 speakers across10corpora, 5 languages for the full MFA pipeline (English, Spanish, Dutch, Mandarin, French) and 3 primary aetiologies (Parkinsons disease, cerebral palsy, amyotrophic lateral sclerosis), we find that all five consonant d features correlate significantly with clinical severity (random-effects meta-analysis rho = -0.50 to -0.56, p < 2 x 10^-4; pooled Spearman rho = -0.47 to -0.55 with bootstrap 95% CIs not crossing zero), with the effect replicating within individual corpora, surviving FDR correction, and remaining robust to leave-one-corpus-out removal and alignment quality controls. Nasality d decreases monotonically from control to severe in 6 of 7 severity-graded corpora. Mann-Whitney U tests confirm that all 12 features distinguish controls from severely dysarthric speakers (p < 0.001).The method requires no dysarthric training data and applies to any language with an existing MFA acoustic model (currently 29 languages) or a model trained from healthy speech alone. It produces clinically interpretable per-feature profiles. We release the full pipeline and phone feature configurations for six languages to support replication and clinical adoption. Author SummaryOne of the authors has lived with ALS for sixteen years. Bernard Muller, who built this entire analytical pipeline using only eye-tracking technology, has experienced the progression of the disease firsthand, including the dysarthric speech that comes with advancing ALS and the tracheostomy that followed. The problem this paper addresses is not abstract to him, and that shapes how the method was designed. We developed a method to measure how well a person with dysarthria can produce distinct speech sounds, without needing any recordings of disordered speech for training. Our approach works by analysing how a widely available AI speech model organises different sound categories -- such as nasal versus oral consonants, or voiced versus voiceless sounds -- and measuring whether those categories become harder to tell apart. We tested this on 890 speakers across 10 datasets in five languages, covering Parkinsons disease, cerebral palsy, and ALS. Because the method only needs healthy speech recordings to set up, it applies to any language with an existing acoustic model, currently covering 29 languages. The resulting profiles show clinicians which specific aspects of speech production are degrading, rather than providing a single opaque severity score. This could support remote monitoring of speech decline in neurodegenerative disease and enable screening in languages and settings where specialist assessment is unavailable.