Ashish Raj

The paper develops a biophysically realistic model of tau progression that integrates exosome-mediated transport and microtubule dynamics to simulate macroscopic spread in the brain.

The paper uses gene gradients to infer directed structural connectivity across species and applies it to human data to reveal source-sink network organization.

This study uses resting-state fMRI to stratify chronic low back pain patients into phenotypes based on brain activity and sensory abnormalities, revealing distinct connectivity patterns and predicting pain locations.

This paper reframes Global Signal Removal (GSR) as graph spatial filtering, introducing variants with different effects on network connectivity.

MAMBAxBrain is a neural framework that models brain functional dynamics across multiple tasks, achieving high accuracy in brain fingerprinting, cognitive decoding, reaction time prediction, and schizophrenia classification.

The paper develops a computational model integrating tau propagation and neuroinflammation to explore their roles in Alzheimer's disease, finding that inflammation significantly influences tau spread.

The paper introduces HONeD-in, a method to deconvolve passive diffusion effects from functional MRI signals, revealing active brain activity patterns.

The paper identifies presynaptic and postsynaptic risk genes and key connectomes involved in pathological α-synuclein spreading using mathematical models and gene regulatory network analysis.

The paper uses a network diffusion model to analyze how genetic factors and brain connectivity influence selective vulnerability or resilience to Alzheimer's tauopathy.

This study shows that directional bias in tau transmission along fiber tracts explains divergent spatiotemporal progression of tauopathies in mouse models.

This study uses mathematical models to analyze amyloid and tau interactions in Alzheimer's Disease, finding that a one-way amyloid-to-tau aggregation model best explains their spatiotemporal progression.

The paper introduces a framework to unify integration and segregation in the brain using structural connectivity harmonics and the gap-spectrum, organizing them into three regimes with distinct properties.

This paper introduces a biophysical, connectivity-based generative model for rs-fMRI that integrates spectral and structural information to quantify structure-function relationships.

This study identifies dynamic functional connectivity features in MEG data that distinguish Alzheimer's disease from healthy controls, showing higher spatial complexity and lower temporal complexity in AD patients.

This study shows how local brain manipulations affect global brain dynamics using simulations and experimental data, revealing specific network changes and localization of effects.

This paper develops a Bayesian procedure for inferring spectral graph model parameters using simulation-based inference, showing comparable performance to annealing algorithms while enabling uncertainty quantification.

The paper identifies cell types and classes that confer vulnerability or resilience to tau pathology in Alzheimer's disease using single-cell RNA sequencing data and a cell-type mapping pipeline.

This paper introduces a computational framework to reconstruct functional connectivity from structural connectomes using joint subspace eigenmodes, showing competitive performance and better interpretability.

The paper analyzes the stability and dynamics of a spectral graph model of brain oscillations, showing it can capture alpha band oscillations without mesoscopic oscillations and estimate time-varying parameters from MEG data.

The paper examines associations between neuronal subpopulation parameters and tau/A{beta} in AD patients using MEG and PET data.

The paper finds that high neuronal activity and high functional connectivity are mutually exclusive in both zebrafish and human brains using calcium imaging and fMRI data.

The paper introduces a computational pipeline called MISS to map diverse cell types in the murine brain using single-cell RNAseq and in situ hybridization data, showing improved accuracy through gene subset selection.

This study uses a Network Diffusion Model to show that structural brain networks drive pathological spread in ALS, with seed regions in basal ganglia, thalamus, and insula, not the primary motor cortex.

This paper introduces an improved spectral graph theory-based model that captures MEG frequency spectra and spatial patterns of alpha and beta bands from resting healthy subjects.

The paper develops a model to predict future tau-PET burden in Alzheimer's disease using network diffusion and baseline data, achieving moderate predictive accuracy in MCI and AD patients.

This paper develops a multi-task deep learning model to predict current and future cognitive scores in Alzheimer's patients using MRI and demographic data, achieving high accuracy and interpretability.

The paper develops a computational model of tau-modified axonal transport to show how directional biases in tau spread arise from transport feedback and conformation-specific dysregulation.

This paper shows how functional brain networks can emerge from the structural connectome by using a complex-valued graph Laplacian that outperforms real-valued models.

The paper uses a mathematical model to show that AD-like mouse models show retrograde tau spread biases, while non-AD-like models do not.

This paper introduces a spectral graph model to predict brain oscillations from structural connectomes, showing that certain oscillations emerge from network topology.