Vladimir Fonov

This paper improves brain age prediction models by reducing generalization error through preprocessing, data augmentation, and regularization.

QRATER is a web-based QC tool for neuroimaging that enables rating of QC images, showing varying time and agreement across tasks and rater experience.

This paper evaluates automated methods for hippocampal segmentation and HVR calculation, finding that a CNN-based method outperforms others in accuracy and reliability.

This study uses MRI and deformation-based morphometry to show that nucleus basalis of Meynert degeneration occurs earlier than entorhinal cortex or hippocampus degeneration in Alzheimer's disease.

The paper introduces a multi-contrast neonatal brain template and atlases to improve image processing and spatial normalization in neonatal neuroimaging.

This study assesses white matter hyperintensity severity using multimodal MRI in Alzheimer's Disease by analyzing signal measures from different MRI contrasts and their associations with clinical variables.

The paper develops a functional parcellation method for both cortical and subcortical brain surfaces, showing improved homogeneity and alignment with known subnuclei compared to arbitrary parcellations.

The paper introduces disease-specific neuroimaging templates for frontotemporal dementia (FTD) to improve registration accuracy and reduce biases in analyses of affected populations.

DARQ is a deep learning method for automated quality control of stereotaxic MRI registration, achieving high accuracy comparable to manual assessment.

The paper uses MRI and beta-galactosidase expression to study Capn15 knockout mice, revealing brain size reductions and expression patterns in adult mice.

The paper introduces CerebrA, an atlas registered to the MNI-ICBM2009c template, with manual label correction for improved anatomical accuracy in brain imaging.

The paper predicts Alzheimer’s dementia onset 2-7 years before clinical diagnosis using combined cognitive scores and MRI biomarkers from a single baseline visit.

This paper develops an unsupervised domain adaptation method to improve automated neuroanatomy segmentation in MRI across different domains without target domain annotations.

The paper shows that combining cortical and subcortical white/gray contrast improves age prediction and reveals cognitive performance differences through residual analysis.

This study compares hippocampal volume and SNIPE score to predict dementia onset in cognitively normal individuals, finding SNIPE score has a larger effect size in distinguishing progressors.

The paper creates age-appropriate structural brain atlases for Chinese children using MRI data and shows they improve brain age prediction compared to Western-based atlases.

The paper introduces a deep learning method for automatic quality control of stereotaxic MRI registration, achieving higher accuracy than manual assessment.

The paper presents a deep learning method for tissue classification in infant MRI scans and applies it to the iSeg-2017 challenge data.

This study compares diffeomorphic and non-diffeomorphic methods for creating average anatomical templates and finds that diffeomorphic frameworks do not necessarily improve accuracy.

This study uses voxel-wise T2 relaxometry to analyze brain development in healthy infants and toddlers, providing maps of T2 evolution and comparing them to manual estimates.

The paper develops a defacing algorithm for MRI data to anonymize faces while preserving data integrity for subsequent processing.

The paper shows that cortical thinning in early Parkinson’s Disease correlates with connectivity to subcortical disease reservoirs, suggesting disease spread along neural networks.