Russell Alan Poldrack

The paper describes the OpenNeuroPET initiative, which develops standards and open-source tools to standardize and share PET neuroimaging data, enhancing reproducibility and collaboration.

The paper shows that white matter tracts are positioned within the cortical hierarchy to support specific cognitive functions and connect regions with greater cognitive diversity.

This paper reviews how neighborhood-level factors relate to functional brain connectivity across the lifespan, finding associations between area-level disadvantage and altered brain network organization.

This study shows early adversity increases the size of the somato-cognitive action network in the developing brain, particularly in response to threat, and this change is linked to poorer cognition.

The paper shows that functional brain networks remain stable in healthy young adults over years, suggesting persistent individual brain organization.

This study evaluates deep learning models for predicting task activation from resting-state fMRI, finding that architectural changes improve efficiency without sacrificing accuracy.

The paper introduces a protocol for assessing and controlling MRI data quality using MRIQC software, focusing on anatomical, functional, and diffusion MRI scans of the human brain.

This paper validates data-driven factor models over RDoC-based models using fMRI and behavioral data, showing better reproducibility and clinical relevance across neural and behavioral domains.

The paper examines how head motion-related exclusions and differential retention affect the representativeness of longitudinal neuroimaging samples in the ABCD Study, highlighting disparities among racial groups.

This study identifies a twofold expansion of the salience network and contraction of the frontoparietal network in OCD, linked to symptom severity and dynamic connectivity changes over time.

The paper extends fMRIPrep to handle MRI data from infancy to old age, demonstrating high-quality preprocessing across a wide age range.

This study validates the RDoC framework using a latent variable approach, showing a bifactor model better fits fMRI data than the current framework and highlights underrepresentation of the arousal and regulatory systems domain.

The paper identifies unintended bias from collinearity reduction strategies in fMRI analysis and introduces a contrast-based variance inflation factor to assess their impact.

This study assesses how defacing structural MRI images affects quality assessments by humans and automated tools, finding significant human bias and minimal impact on MRIQC metrics.

The paper evaluates the impact of different preprocessing pipelines on functional MRI reproducibility and shows moderate inter-pipeline agreement, affecting brainwide association study consistency.

The paper shares a dataset of cognitive tasks, anatomical MRI, and functional MRI data from 103 healthy adults to study self-regulation.

hyve enables parametric, reusable visualizations for diverse brain geometries through a compositional framework.

This study evaluates how different analytic choices affect test-retest reliability in fMRI data using the Monetary Incentive Delay task across three independent samples.

The paper addresses the response time paradox in fMRI analyses by proposing a model that incorporates response times to separate condition effects from RT-related activation.

This paper demonstrates that global activation and pattern similarity are intertwined in neural data and shows that statistical methods to separate them are not always effective.

This paper introduces task-DyVA, a deep learning framework that models human response time sequences, capturing subject-specific behavioral differences and providing insights into cognitive theories.

TemplateFlow provides a FAIR framework for sharing and managing multi-scale, multi-species brain models to improve reproducibility in neuroimaging.

Neuroscout provides an automated platform for analyzing naturalistic fMRI data, enabling reproducible and generalizable research through machine learning and open tools.

OpenNeuro is an open archive for sharing brain imaging data following FAIR principles, hosting over 600 datasets from 20,000 participants across species and modalities, with over 150 publications citing its data.

This study compares massed and spaced training in reward-based learning, finding that spaced training reduces working memory contributions and isolates gradual learning mechanisms.

This paper identifies potential nootropic drug targets by integrating large-scale cognitive GWAS and transcriptomic data to find credible druggable genes.

The paper introduces a convolutional neural network to detect defaced facial features in MRI brain scans, achieving high accuracy and sensitivity using data augmentation.

The paper identifies design issues in the stop-signal task of the ABCD study and proposes solutions for improving data quality and analysis.

This study shows that intracranial electrical stimulation alters meso-scale network integration depending on the topological role of the stimulated brain regions.

This paper introduces fMRIPrep, a tool for preprocessing task-based fMRI data, and demonstrates its integration into a workflow for statistical analysis.