Thomas E. Nichols

Diagnose Harmonisation (DHARM): An open-source tool for the harmonisation and analysis of batch effects in MRI-derived measures.

DHARM is an open-source tool that diagnoses, harmonizes, and evaluates batch effects in MRI data using standardized workflows and multivariate diagnostics.

Spurious model comparisons are widespread in biomedical artificial intelligence

97% of 184 high-impact biomedical AI studies compare cross-validated models with tests assuming fold independence; in 69% of re-analyzable studies a significant comparison does not survive correction.

ComCat: Combating Covariate Effects in Brain Analysis

ComCat improves brain analysis by reducing unwanted variability from heterogeneous datasets, outperforming existing methods in brain age prediction across multiple datasets.

MIDFA: Scalable Bayesian Factor Analysis for Mixed and Incomplete Data

The paper introduces a scalable Bayesian factor analysis framework for handling mixed and incomplete data, demonstrating its effectiveness on MS datasets.

FEMA-Long: Modeling unstructured covariances for discovery of time-dependent effects in large-scale longitudinal datasets

FEMA-Long enables efficient modeling of time-dependent effects in longitudinal data, discovering genetic variants with time-dependent effects on infant growth traits.

The Hidden Landscape of Missed Effects in Human Functional Neuroimaging

The paper identifies inflated effect sizes in functional neuroimaging studies and proposes a correction method to improve study planning and detection of brain effects.

Connectome-based spatial statistics enabling large-scale population analyses of human connectome across cohorts

This paper introduces CBSS, a framework for analyzing structural connectomes across multiple cohorts, showing its reliability and utility in cross-cohort diffusion MRI studies.

The methodological foundations of lesion network mapping remain sound

The paper defends lesion network mapping by showing that it provides meaningful differences through specificity testing, refuting criticisms that it reflects normative connectome properties.

Structural Brain Indicators of Cognitive Performance in Middle and Late Adulthood: The Human Connectome Project in Aging/Aging Adult Brain Connectome Cohort

This study identifies structural brain differences associated with cognitive performance in middle and late adulthood using data from the Human Connectome Project.

Excessive Censoring Degrades Individual-Specific Cortical Parcellations and Personalized TMS Targets

The paper evaluates how different censoring strategies affect cortical parcellations and TMS targets in rs-fMRI, finding that lenient censoring yields better results than strict censoring.

Age associations with cortical and subcortical brain structure in adolescents age 9-17

The paper uses an extended mixed-effects model to analyze age-related changes in brain structure across a large adolescent cohort, finding distinct developmental patterns in cortical and subcortical regions.

Longer scans boost prediction and cut costs in brain-wide association studies

The paper shows longer scans improve prediction accuracy in brain-wide association studies while reducing costs, with 30-minute scans being the most cost-effective.

NiCLIP: Neuroimaging contrastive language-image pretraining model for predicting text from brain activation images

NiCLIP predicts cognitive tasks and domains from brain activation maps using a contrastive language-image model trained on neuroscientific articles.

Convergent and Divergent Brain-Cognition Development

The paper examines how stable and changing brain network features predict cognitive development in adolescents using rs-fMRI and cognitive data from the ABCD Study.

BARTharm: MRI Harmonization Using Image Quality Metrics and Bayesian Non-parametric

BARTharm improves MRI data harmonization by using image quality metrics and Bayesian non-parametric models, outperforming ComBat in removing scanner-induced bias while preserving biological signals.

Advancing image-based meta-analysis for fMRI: A framework for leveraging NeuroVault data

This paper develops a framework for image-based meta-analysis using NeuroVault data, showing that manual selection provides precise results while heuristic methods are effective for heterogeneous domains.

Characterising ongoing brain aging and baseline effects from cross-sectional data

The paper disambiguates ongoing brain aging from baseline effects using cross-sectional data and validates findings with longitudinal data from UK Biobank.

Genome-wide analysis identifies 66 variants underlying anatomical variation in human neuroendocrine structures and reveals links to testosterone

This study identifies 66 genetic variants linked to neuroendocrine structure volumes and links some to testosterone levels using UK Biobank data.

Partitioning variance in cortical morphometry into genetic, environmental, and subject-specific components

The paper estimates genetic, environmental, and subject-specific variance components in cortical morphometry using longitudinal data from the ABCD Study.

Leveraging the Adolescent Brain Cognitive Development Study to improve behavioral prediction from neuroimaging in smaller replication samples

The paper uses the ABCD Study to determine optimal replication sample sizes for reproducible brain-behavior associations using neuroimaging data.

Two common and distinct forms of variation in human functional brain networks

This paper identifies two distinct forms of variation in human functional brain networks: border shifts and ectopic intrusions, which differ in location, network associations, and behavioral predictions.

Decentralized Mixed Effects Modeling in COINSTAC

This paper introduces a decentralized linear mixed-effects model for neuroimaging analysis, enabling large-scale collaboration without data pooling, and shows it achieves similar performance to centralized models.

FEMA: Fast and efficient mixed-effects algorithm for large sample whole-brain imaging data

FEMA enables fast and efficient mixed-effects analyses of large-scale whole-brain imaging data, revealing distinct spatial patterns of brain maturation in early adolescence.

Brain-wide functional connectome analysis of 40,000 individuals reveals brain networks that show aging effects in older adults

The paper analyzes brain functional connectomes in 40,000 individuals to identify aging-related network changes, finding two motor-related subnetworks with age-related declines in functional connectivity.

A scalable approach for continuous time Markov models with covariates

The paper introduces a scalable method for fitting continuous time Markov models with covariates using stochastic gradient descent and Pade approximation, showing improved performance on a large-scale dataset.

BLMM: Parallelised Computing for Big Linear Mixed Models

BLMM addresses scalability and missing data issues in large-scale fMRI analyses using a parallelized linear mixed models approach.

Application of a convolutional neural network to the quality control of MRI defacing

The paper develops a convolutional neural network to automate MRI defacing quality control, reducing manual checks and improving efficiency.

Improved Interpretability of Brain-Behavior CCA with Domain-driven Dimension Reduction

This paper improves the interpretability of brain-behavior CCA by using domain-driven dimension reduction, showing more stable and interpretable results compared to standard PCA.

Causal Inference on Neuroimaging Data with Mendelian Randomisation

This paper introduces Mendelian randomisation methods for causal inference in neuroimaging data, using UK Biobank imaging phenotypes to demonstrate their application and limitations.

Isolating the Sources of Pipeline-Variability in Group-Level Task-fMRI results

The paper identifies key stages in a task-fMRI analysis pipeline where different software packages produce divergent results, finding that variability is concentrated in specific stages and varies across datasets.