Neuroinformatics

Neuroimaging research focusing on subcortical brain regions showed that they play a role in a wide range of cognitive functions and associated their structural alterations to various psychiatric disorders. Methods to analyse such structures were developed to observe changes, not simply in terms of volumes, but in terms of surfaces, which are able to detect subtle and local changes in shape and geometrical configuration. Such advanced analyses have however been largely restricted to command line environments with relatively limited access to Python users or researchers with low technical expertise. We present SubCortexMesh as a user-friendly toolbox which covers automated surface estimation from popular subcortical volume segmentations (FreeSurfer and the Functional Magnetic Resonance Imaging of the Brain Software Library (FSL)), computes shape-related vertex-wise metrics (thickness, surface area and curvature) for whole cohorts and includes statistical analyses. This is all done inside Python, automatically computing all subjects of a given preprocessed directory, with the explicit intent to minimise steps and manual coding required from the users. SubCortexMesh includes statistical tools which allow conventional random field theory-based cluster analyses on native subject metrics, standardised within common surface templates, thus containing a whole workflow necessary to run an up-to-date shape-wise subcortical analyses in one package.

Simplified spiking neuron models such as the Adaptive Exponential Integrate-and-Fire (AdEx) model reproduce the firing behaviour of diverse neuron types at a fraction of the cost of biophysically detailed models, making them attractive for large-scale brain simulation. Fitting their parameters to recordings, however, still relies on derivative-free search such as grid search or evolutionary algorithms, because the discrete spike-and-reset mechanism renders these models non-differentiable. Surrogate gradients, used in deep spiking networks, replace the spike derivative with a smooth approximation on the backward pass. Whether they also enable efficient gradient-based parameter estimation for single simplified neurons is untested. We implement a differentiable AdEx model with a surrogate-gradient variant in the Jaxley framework. The implementation is on par with target-compiled C++ Brian2 code while remaining end-to-end differentiable, vmap-batchable and GPU/TPU-portable. It is 200 times faster than Brian2's default runtime mode. Across a recovery-radius benchmark of 2,700 paired runs, gradient-based fitting never matched a gradient-free Nelder-Mead baseline. At sub- threshold levels, gradient descent instead beats Nelder-Mead. The obstacle is therefore the spike mechanism and the loss geometry it induces, not gradient descent itself.

Large-scale biophysically-detailed models of the brain are instrumental in unraveling how cellular, synaptic, and circuit-level mechanisms contribute to neural dynamics. Such models are built around large populations of single-neuron models whose morphologies must preserve physiological variability to reproduce integrative properties and firing patterns observed experimentally. Experimental reconstruction of morphologies is a time-consuming and resource-intensive process, unlikely to supply them at the required scale. Several data-driven methods for morphology synthesis have been developed, which nonetheless rely on extensive datasets for parameter estimation. To address this limitation, we developed NeuWalk, to our knowledge the first public Python library for morphology synthesis based on branching-and-annihilating, biased-and-correlated random walks. It estimates branching and annihilation rates from Sholl intersections and, optionally, bifurcation counts, while also ensuring their replication in synthetic morphologies; user-specified combinations of biases, implementing somatofugality, self-avoidance, spatial competition, and branching-angle control, sculpt morphological patterns characteristic of each neuron class. We then validated the method on mitral and middle tufted cells of the olfactory bulb, semilunar and pyramidal cells of the anterior piriform cortex, and neocortical pyramidal neurons: means and variances of Sholl intersections, bifurcation counts, and total dendritic length were statistically indistinguishable between synthetic and natural morphologies, with the single exception of its variance in apical dendrites of piriform cortex pyramidal neurons. As Sholl intersections and bifurcation counts are commonly reported in the literature, and biases can be inferred from published images or qualitative descriptions, NeuWalk requires no immediate access to experimental reconstructions, striking a trade-off between data-driven parameterization and empiricism without sacrificing morphological accuracy.

Biophysically detailed models of human cortical microcircuit provide key insights into brain function and disease biomarkers, yet their computational cost lengthens simulation runtime and limits scalability. To address this, we systematically evaluated recent simplification methods for morphological reduction and synapse merging in terms of preserving spiking and electroencephalographic (EEG) properties of human cortical microcircuit models. At the single- cell level, reduced compartmental models accurately reproduced spiking behaviors. However, at the microcircuit level, while achieving a 5-fold speed-up, morphological reduction failed to preserve spectral dynamics of spiking and EEG despite maintaining average baseline and response firing rates. Conversely, applying synapse merging alone on the full morphology successfully reproduced both spiking activity and EEG dynamics, when constrained to [≤] 50 m spatial intervals, and yielded a 2-fold speed-up. These results were consistent across both cortical layer 2/3 and layer 5 models, and in depression models with reduced somatostatin interneuron inhibition. Our study identifies model simplification methods that speed up cortical microcircuit simulations while preserving the accuracy of the simulated brain signals.

Stereotaxic intracranial microinjection in mice is a critical step in neuroscience workflows, enabled by robust hardware design, precise operation, and neuroanatomical expertise. Manual injections are prone to error and stereotax-mountable commercial automated injectors are extremely expensive. Further, the knowledge needed for performing accurate intracranial injections and surgeries is heterogeneous and disparate. Here we demonstrate the design and characterization of the UD Neuroinjector software-hardware-knowledge architecture, which serves to both reduce the cost of performing precisely controlled stereotaxic microinjections and provide a web platform for planning accurate targeting of specific mouse brain regions. We designed the UD Neuroinjector using low-cost hardware and 3D-printed parts, with a total build-cost of $300 --- less than 0.1x the cost of a commercial injector. Arduino-based firmware enables programmable flow rates, automated injection and extraction, and manual joystick control. Capillary-fluid measurements show stable, highly linear displacement (R2 > 0.93) at commonly used injection rates. In a head-to-head comparison with a commercial injector, in vivo DAPI injections into the mouse thalamus showed comparable spread and DAPI+ cell counts. For effective experimental design of intracranial injections, the web-based components of our architecture provide AI-enabled assistance and atlas-guided navigation. We used a hybrid retrieval-augmented generation (RAG) approach to ground AI-assisted protocol guidance for mouse intracranial surgeries using a corpus of 3,738 papers. Using the Common Coordinate Framework version 3 (CCFv3), we designed an interactive tool for identifying target coordinates. Together, the UD Neuroinjector architecture integrates and enables efficient knowledge discovery, experimental design, planning, and implementation of mouse intracranial surgeries.

Deep learning models for electroencephalography (EEG) pathology detection achieve competitive accuracy but offer limited interpretability. In this paper, we introduce concept bottleneck models (CBMs) for EEG pathology detection which use clinical EEG biomarkers computed analytically from the signal as a bottleneck to be used for prediction. Those concepts include band power, spectral ratios, alpha peak frequency and amplitude, hemispheric asymmetry and phase locking values. With the backbone based on shallow convolutional neural network (CNN), we compare a black box baseline with two variants of CBMs. The hybrid variant is a classifier receives a residual projection of the backbone features. The pure variant is a CBM where the classifier receives only the predicted concepts. Models are evaluated on abnormality detection and dementia classification followed by concept prediction quality and leakage analysis. Both CBMs approach the baseline on TUAB (82.7-82.9% vs 83.6%) and are 2-5 percentage points lower on CAUEEG (57.9-60.5% vs 62.7%). The pure CBM exhibits concept fidelity concern where the predicted concepts deviate from their ground truth values. Replacing them with their true values causes severe cost on accuracy by 26 and 22 points on TUAB and CAUEEG. The concept fidelity concern is mitigated by increasing the weight of the concept loss to match the ground truth concepts. the mitigation comes with small accuracy cost on TUAB but a relatively larger cost on CAUEEG. The code of the experiments and full logs are available at https://www.github.com/mhmdrdwn/cbm.

The choroid plexus (ChP) is a highly vascularized structure located in the brain ventricles that plays important roles in multiple central nervous system functions. ChP enlargement has been implicated in many neuropsychiatric conditions such as multiple sclerosis, depression, schizophrenia, and Alzheimer's disease. However, limitation in the accuracy of the current segmentation tools highlight the need for more reliable methods to quantify ChP volume in healthy and clinical human populations. The shape and size of the ChP vary drastically from person to person, and currently there are no large ChP datasets which allow comprehensive training of segmentation models. Here we introduce DataPlex, a collection of 700 structural MRI brain images where the lateral ventricle ChP has been manually labeled and checked by expert human raters. DataPlex is balanced and representative across a wide range of scanning parameters, demographics, and clinical profiles, and subject-level information will be available for each labeled image. DataPlex will be open source to support the creation of reliable segmentation tools for the human lateral ventricle ChP.

From a statistical point of view, clustering spike time-windows from multi-electrode arrays (MEAs) recordings is a challenging high-dimensional, unsupervised clustering task where stationarity, commonly assumed in standard time-series analysis, is often violated and for which a gold standard is currently unavailable. Here we aim at providing practical guidance on how to cluster this type of data by systematically comparing 108 clustering pipelines differing along three main dimensions: (i) the data feature space; (ii) the metrics used to quantify distance between data points; and (iii) the clustering algorithm. For our benchmark we used both labeled synthetic data, mimicking four physiologically-inspired classes of spike time-windows, and a set of real MEAs recordings from a two-dimensional in-vitro neural culture. The performance of the competing pipelines was evaluated in terms of balanced accuracy and computational time when analyzing the synthetic datasets, while the Silhouette score was used for the real dataset, where no ground-truth is available. Overall, our analysis shows that the best combination is formed by k-means with Euclidean distance applied after Principal Component Analysis (PCA) of the spike time-windows. Conversely, hierarchical clustering showed the highest computational burden, while Independent Component Analysis and kernel PCA provided less effective noise suppression.

Background. Automated segmentation of glioma subregions on multiparametric MRI can be improved by changing either the network architecture or the information presented to the network. These two design levers are usually studied together, which makes their individual contributions difficult to see. We present an integrated analysis of two completed, independent studies, each of which isolates one lever. Methods. In the architecture-level study (Slice), SIBA-UNet - a 2D U-Net combining depthwise separable convolution, inverted-bottleneck blocks and attention-guided skip fusion - was trained on 128 x 128 multimodal slices from BraTS 2020 with a patient-level split, and a sigmoid decision threshold was selected on validation data. In the representation-level study (Subtract), CMSR-3D augmented the four standard MRI sequences with four voxel-wise cross-modal subtraction maps (T1-T1CE, T2-FLAIR, FLAIR-T1CE, T2-T1CE) and evaluated 4-, 6- and 8-channel inputs with an identical 3D U-Net on 1,251 BraTS-GLI 2023 cases (1,001/125/125 split). Results. SIBA-UNet reached an overall validation Dice of 0.8601 at a validation-selected threshold of 0.51, with reported subregion Dice of 0.8795 (necrotic core), 0.8501 (edema) and 0.9125 (enhancing tumor); the source ablation was only partially reported. With the architecture held fixed, the 8-channel CMSR-3D configuration gave the highest mean Dice among the evaluated inputs (0.8699 vs. 0.8534 for the 4-channel baseline) and the lowest mean HD95 (5.9859 vs. 8.2308), whereas either two-map subset alone changed mean Dice by at most 0.0037. Conclusions. Architecture design and input-representation design are separable levers for brain tumor segmentation. The controlled representation ablation indicates that the complete set of subtraction maps was more informative than either subset under a fixed 3D U-Net. Both studies were evaluated internally on their respective datasets, and their numerical results are not directly compared because the datasets, dimensionality and evaluation protocols differ.

Independent component analysis (ICA) is widely used to remove artefacts from EEG and TMS-EEG recordings, but the components that correspond to artefacts are usually identified by manual inspection, which is subjective and difficult to reproduce. Existing automated classifiers are either designed for ordinary EEG and do not target the artefacts specific to concurrent transcranial magnetic stimulation, or are built on EEGLAB, so none is available for FieldTrip-based pipelines. We present TEFAR, a lightweight, configurable framework for semi-automatic artefact-component classification that applies the same detectors to ordinary EEG and TMS-EEG. TEFAR is built on FieldTrip and requires no additional MATLAB toolboxes. Each artefact class (line noise, blinks, lateral eye movements, cranial muscle, cardiac activity, and the TMS decay and recharge artefacts) is identified from the spectral, spatial, or temporal signature that its components are known to express. Detection thresholds use robust statistics (the median and the median absolute deviation rather than the mean and standard deviation), so that a single dominant artefact component cannot raise the threshold that is meant to detect it. In 30 ground-truth simulations, TEFAR reached a specificity of 1.000 and a sensitivity of 0.96-0.97 for both profiles. An application of TEFAR to real EEG and TMS-EEG recordings is the next step and will be reported separately.

Deep brain stimulation (DBS) is a well-established therapy for various neurological disorders. Clinical outcomes of DBS depend on electrode placement and stimulation programming, with growing evidence suggesting that targeting specific sites and networks can improve therapeutic efficacy. This targeting can be facilitated using high-accuracy, multi-modal biophysical models. In this article, we present the open-source platform OSS-DBS v2 that enables researchers and clinicians to build such biophys- ical models. Building upon previous work, the new version combines automated volume conductor generation, accurate and efficient electric field modelling and stimulation-driven neural simulations. OSS-DBS v2 provides an extensive library of intracranial electrodes, which can be modelled simul- taneously to test complex stimulation paradigms. Its application programming interface facilitates integration of external data workflows, such as impedance measurements. Furthermore, an integration with the Lead-DBS toolbox offered a straightforward graphical user interface for stimulation volume and pathway activation modelling, further complemented by an optimization framework. OSS-DBS v2 relies entirely on open-source software, supports all major operating systems and is designed for efficient execution on consumer-grade hardware.

The Hierarchical Gaussian Filter (HGF) is a hierarchical Bayesian model of learning under uncertainty and volatility, with a widely used MATLAB implementation. We present HGFX 1.0.0, a Python/JAX toolbox whose primary objective is validated behavioral and numerical compatibility with a frozen HGF Toolbox 8.2.0 reference, while removing MATLAB as a user-runtime dependency. Reimplementation is treated as a validation problem rather than source translation: configuration semantics, parameter transforms, forward trajectories, observation likelihoods, objectives, fitting and statistical surfaces, simulation workflows, official demo behavior, model selection, and backend agreement are compared against the pinned oracle under predeclared tolerances. Two official MATLAB demo workflows reproduce the reference at frozen trajectory tolerances, including a regime in which classic HGF encounters negative posterior precision while eHGF completes successfully. In paired model-selection validation, all 36 BIC winners agree between MATLAB and HGFX. Physical NVIDIA GPU applicability was confirmed on two Tesla T4 devices, with a maximum CPU-versus-GPU final-objective difference of 1.42e-14 against a frozen 1e-7 criterion. Historical parameter recovery results that did not meet predeclared criteria, together with exact MATLAB/HGFX limitation matches, are preserved and are not reclassified as scientific success. A prospectively gated comparison with pyhgf 0.3.2 shows binary64-scale agreement on mapped perceptual trajectories in one authorized three-level binary-HGF cell, while participant-response negative log-likelihood is retained as not directly comparable. HGFX therefore provides a MATLAB-independent Python implementation with an explicit evidence model that separates direct parity, matched reference limitations, backend applicability, and future performance claims.

Microstate analysis studies dynamic brain activity through a discrete representation of electroencephalographic (EEG) signals and their topography. While the body of literature describing findings in diverse experimental and clinical settings is growing rapidly, the effects of several core components of the microstate algorithm are still insufficiently understood. This study systematically investigates how two common temporal smoothing algorithms, parametric smoothing and GFP peak interpolation, influence a wide range of microstate sequence parameters, and how they shape the integrated picture of brain dynamics inferred from these metrics. We study eyes-closed resting-state EEG recordings from N=201 healthy subjects from a public EEG database. For the elementary temporal metrics, we find that widening smoothing windows markedly increase microstate duration, reduce occurrence rates, but leave coverage mostly unaffected. In terms of simple syntax properties, smoothing windows above 30 ms yield an inconsistent picture with a memoryless, first-order Markov structure at short time scales, but a positive Hurst phenomenon suggestive of long-range correlations at long time scales. Evidence of asymmetry and non-equilibrium dynamics are suppressed by smoothing. At intermediate time scales, microstate periodicity was only observed up to a window size of 28 ms. Smoothing had a strong impact on the complexity metrics entropy rate and excess entropy, decreasing the former and increasing the latter. For most metrics, the effects of GFP peak interpolation were comparable to those of smoothing windows in the range between 28 ms and 44 ms. In conclusion, elementary and higher-order microstate parameters must be interpreted in the context of preprocessing, as many parameters change continuously with smoothing intensity. This limits the comparability of microstate parameters obtained with different preprocessing settings. Strong smoothing with a window size above 30 ms or GFP peak interpolation yields a fundamentally different picture of resting state brain dynamics, i.e. that of a memoryless, aperiodic, time-reversible first-order Markov chain. Since insights from other neurophysiological modalities suggest that time series memory and time-irreversibility are markers of normal brain activity, our recommendation is to avoid intense smoothing to minimize alterations of the initial sequences.

Significance: There is potential for generative artificial intelligence (AI) to perform "full stack" academic publishing, including comprehensive manuscript fabrication, "peer"-review, and editorial decision making. This study evaluated the benefits and cautions in using AI for full stack academic publishing in clinical research about seizures. Methods: Three epileptologists each selected 1 published manuscript in Epilepsia and produced an AI-fabricated manuscript about seizures. AI tools were then used to perform "peer"-review and make editorial decisions with minimal human supervision. The rates of identifying AI-fabricated manuscripts and editorial decisions among the epileptologists was analyzed, and epileptologists' observations on how AI differed from typical human "peer"-review and editorial decisions was collected. Results: Four of 6 (67%) pairs of AI-fabricated and human-generated manuscripts were correctly identified. Only 1 AI-fabricated manuscript was editorially rejected, whereas another had a subtle indicator of fabrication. When compared to human content, AI generated content was comprehensive, detailed, and based on the highest standard of academic publication. However, generated content was more verbose, used more structured responses, and more critical feedback. Conclusions: These results demonstrate that the quality of AI-fabricated manuscripts about clinical research in seizures was sufficiently high that the current methods of peer and editorial review could not easily distinguish fabricated content compared to human-written content. Moreover, AI review and editorial decisions were high quality and appeared reasonable in many cases. AI tools have matured to the point that there should be thoughtful policies that permit AI assistance but prohibit insufficiently supervised autonomous scientific submissions, reviews, and editorial decisions.

Lesion Network Mapping (LNM) is a method used to map symptom-related circuits in the human brain. Recent work has raised concerns about the method and the specificity of LNM-derived networks, which often show high spatial overlap across unrelated conditions. Null models have been proposed to address the raised limitations1. Useful statistical tools in network neuroscience, they are reference procedures; in LNM, the normative connectome matrix C remains the source of the derived networks also under null modelling. Their broader role as a general solution for LNM therefore warrants careful evaluation. Here, we examine the proposed null models step-by-step and test them empirically across multiple datasets. We show that the proposed null-model correction effectively only replaces C with a transformed matrix C' (e.g., C' ~ C - mean(C)), one that is directly derived from C and retains much of its low-dimensional structure. Consequently, null-model-corrected LNM maps remain shaped by the same standard set of connectome-derived patterns and continue to show repetition and non-specificity across unrelated conditions. We discuss the scope and limitations of null models for LNM, showing that they can statistically benchmark LNMs, but that their application does not resolve the core limitation of non-specificity of LNM networks.

The dentate gyrus' principal neuron, the granule cell (GC), plays a key role in pattern separation, transforming similar inputs into dissimilar outputs. We used a new measure based on information theory to evaluate this transformation in an experimentally derived in silico GC model with realistic ion channel distributions across its dendrites. We found that model performance aligned well with current-clamp recordings from mouse brain slices. However, it was also possible to find artificial parameter combinations in the model that functionally outperformed the recorded cells while matching their action potential shapes. Rather than maximising computational power, the biologically realistic GC model balanced energy efficiency against increased functionality in a Pareto optimal way: No possible parameter combination was found to improve one of the objectives without impairing the other. Interestingly, immature GCs adopted a different weighted trade-off with a stronger emphasis on energy efficiency than pattern separation, but they remained Pareto optimal over their putative developmental trajectory. Our results support the idea that evolution favours neurons that can satisfy functional and energetic demands simultaneously.

There is increasing demand for quantitative measurements in neuroimaging research, including absolute quantification of regional myelin density. Fast whole-brain macromolecular proton fraction (MPF) imaging enables acquisition and reconstruction of quantitative maps of MPF, a biomarker of myelin, and has been used in many research applications. To date, no standard protocols have been published for parcellating MPF maps. FreeSurfer is a widely used software package for automated brain parcellation. It is designed for use with high-resolution T1-weighted anatomical images, typically acquired with an MPRAGE sequence. MPF maps inherently provide high contrast between white and gray matter and can potentially be used as input images for parcellation, enabling simultaneous quantitative tissue characterization and volumetric assessment. However, it is unclear how the use of MPF maps as input images with FreeSurfer software affect reliability and repeatability of regional MPF and volume estimates. We evaluated three FreeSurfer parcellation workflows for this purpose. The workflows differed in the amount of neuroimaging data required and computational intensity. MPF and MPRAGE data were acquired in two separate sessions for 11 adults. Reliability was evaluated using mean relative differences in estimates, Bland-Altman analysis, and intra-class correlation coefficients (ICCs). Repeatability was assessed using within-subject coefficients of variation (CVws). All three workflows produced similarly high within-subject repeatability across scans. However, reliability in regional MPF estimates was lower in gray matter than in white matter for workflows that used MPF maps as input. Workflows that use MPF maps as FreeSurfer input may be adequate for estimating mean MPF in GM and WM of cortical parcels and in subcortical regions. However, the MPRAGE-based workflow is recommended when reliable individual-level estimates or volume estimates are required.

For more than a century we have pictured the dendritic spine as the postsynaptic point where a single presynaptic bouton delivers its message. That picture is correct but incomplete, since it overlooks other neuronal elements surrounding each spine. Here we characterize that neighborhood for the first time by counting and measuring the proximity of the elements encircling each postsynaptic spine. We call this neighborhood the vicinome: the full set of neuronal and glial elements surrounding a dendritic spine. We used ground-truth three-dimensional meshes from the H01 petavoxel dataset, a reconstruction of the human cerebral cortex. We analyzed 4550 neighboring elements across 322 spines from 17 pyramidal neurons in cortical layers 2 through 5, with each spine classified as apical or basal. Neighbor number followed a depth gradient, falling from 17.2 {+/-} 4.2 elements per spine in layer 2 apical spines to 12.5 {+/-} 3.1 in layer 5 basal spines. A linear mixed model with a random intercept per neuron confirmed a strong effect of layer (p = 4.3 x 10-5) with large effect sizes (Cohen's d up to 1.45) and no effect of polarity. The minimum spine-to-neighbor distance stayed uniform, near 30 nm in every group, and although some depth-related contrasts were significant, all effect sizes were negligible (Cohen's d below 0.21). We conclude that cortical depth, and not pyramidal neuron polarity, was related to the number of neuronal elements neighboring the spine. We suggest that vicinome crowding could offer a measurable observable for comparison across regions, ages, species, and disease.

Electrical stimulation of brain activity has emerged as a key therapeutic intervention for neurological diseases. The success of these approaches relies on precisely tuned stimulation parameters. A major challenge is that these parameters should ideally be adjusted in real-time based on changes in brain activity. Closed-loop controllers address this challenge by monitoring brain activity and adapting stimulation according to ongoing changes in neural dynamics. While machine learning has been helpful in this context, closed-loop controllers typically do not adapt in real-time to fluctuations in brain activity. Here, we propose a novel framework based on reservoir computing to train a closed-loop, real-time neural controller. This approach, termed CLEAR-FORCE, trains the output weights of a recurrent network based on a combination of target signals and feedback from brain activity. Two series of simulations were performed, where the controller was successful at (1) inducing neural oscillations in a population of spiking neurons and (2) suppressing pathological oscillations. This framework provides a promising avenue for the implementation of neural controllers in low-powered, real-time, and adaptive hardware that harnesses machine learning for the personalized treatment of brain disorders.

An Ischemic stroke is one of the main causes of death and long-term disability around the world. There are very few treatment options available, especially for the cases where the short time period to give clot dissolving medicine has passed (Powers et al., 2019). While single-cell RNA sequencing (scRNA-seq) has emerged as a powerful tool for understanding cell-to-cell differences within at issue (cellular heterogeneity) (Tang et al., 2009). However translating these high dimensional datasets into clinically useful biomarkers remains challenging (Rajko-mar et al., 2019). A major problem with current biomarker studies is their reliance on a single machine learning model, where results may reflect model bias rather than true biological signals (Lopez-Rincon et al., 2020). Methods: We developed a multi-model consensus framework integrating eight different machine learning methods from three groups: traditional models (logistic regression, random forest, XGBoost), a hybrid deep learning model (CNN-XGBoost), and transformer-inspired models (scGPT, scBERT, sc-Former, Geneformer). The proposed framework was applied to scRNA-seq data from a mouse middle cerebral artery occlusion (MCAO) model (GSE174574; n=6 mice, 54,599 cells). To prevent data leakage, we implemented rigorous subject-wise train/test splitting, ensuring all cells from individual animals remained either in the training sets or in the testsets. Results: All eight models achieved excellent performance on the independent test data, with AUC scores between 0.9778 and 0.9975. From six models that allowed gene level importance extraction, we identified 63 consensus genes (threshold: [≥] 4/6 models), with 17 genes selected by all 6 models. Pathway enrichment analysis showed significant enrichment of interferon signaling, with a novel dual-interferon pattern. Twelve of the consensus genes respond to interferon, and five of them (Ifitm6, Ifitm1, Gbp2, Igtp, Tgtp1) respond to both Type I (IFN-/{beta}) and Type II (IFN-{gamma}) interferons. Notably the consensus included Il1rn (interleukin-1 receptor antagonist), the target of Anakinra -- an FDA-approved drug that has completed Phase II stroke trials with promising safety profiles (Smith et al., 2018; Emsley et al., 2005). Conclusions: The Multi-model consensus framework, proposed in this paper, provides a robust and reliable approach to biomarker discovery that goes beyond the limitations of any single method and reduces individual algorithm bias. The convergence of classical, ensemble, and deep learning methods on interferon and IL-1 pathway genes supports their biological validity as stroke biomarkers. Our findings provide a prioritized gene list for further clinical validation and highlight interferon signaling as a possible treatment strategy warranting further investigation. Keywords: ischemic stroke; machine learning; single-cell RNA sequencing; biomarker discovery; consensus analysis; interferon signaling; neuroinflammation