Postdoctoral position in neuroimaging

Modeling heterogeneity and progression of cerebral small vessel disease and deciphering the role of glymphatic function using multi-modal neuroimaging

Are you interested in:

➢ Investigating the heterogeneity and progression of cerebral small vessel disease (cSVD)?
➢ Exploring the role of the brain’s glymphatic system in neurodegeneration?
➢ Working with large neuroimaging cohorts and developing novel MRI biomarkers?

General Description
Cerebral small vessel disease (cSVD) is a leading cause of stroke and dementia, yet no effective mechanistic treatments currently exist. A major clinical and scientific obstacle lies in the heterogeneity of its pathological manifestations. We hypothesize that the spatial distribution and co-occurrence patterns of the core MRI-visible cSVD markers, namely white matter hyperintensities (WMH), cerebral microbleeds (CMB), lacunes (LAC) and enlarged perivascular spaces (PVS), can be used to identify biologically meaningful subtypes and subtype-specific
disease progression.
Beyond vascular injury, recent studies suggest that failure in glymphatic function that clear metabolic waste from the brain may play a key role in the pathophysiology of cSVD. Two emerging markers of glymphatic function based on routinely acquired diffusion-weighted (DWI) and resting-state functional MRI (rs-fMRI) offer an opportunity to investigate the relevance of glymphatic dysfunction in cSVD. However, these markers remain poorly validated in large-scale studies, and have not been systematically linked to cSVD subtypes or stages. We hypothesize that glymphatic dysfunction plays a differential role in distinct cSVD trajectories. Investigating this relationship could uncover new mechanisms and open pathways for targeted diagnosis and
treatment.

Context
cSVD is a heterogeneous group of pathologies affecting small vessels of the brain, which can be detected on MRI with high prevalence in the aging population, often without clear clinical manifestations. Yet, the MRI-defined cSVD markers are major predictors of both ischemic and hemorrhagic stroke, dementia, as well as cognitive decline and mood disorders, thus representing a significant source of disability and societal burden in rapidly aging society. In parallel, the glymphatic system, a brain-wide fluid clearance network, has emerged as a key player in brain waste removal and vascular health. Advanced MRI techniques now allow non-invasive assessment of glymphatic function in humans, opening new avenues for investigating its role in cerebrovascular diseases. Despite these advances, the relationship between impaired glymphatic clearance and cSVD subtypes remains poorly understood.

Objectives
This project has two main objectives. First, we aim to apply a data-driven disease progression model (Subtype and Stage Inference: SuStaIn) to regional distributions of core MRI markers of cSVD in large-scale neuroimaging cohorts, in order to identify biologically distinct subtypes and stages of disease. This will enable the derivation of subtype-specific severity scores for precision phenotyping. Second, we aim to develop robust, standardized pipelines for extracting two promising MRI-based markers of glymphatic function: the DTI-ALPS (Diffusion Tensor Imaging analysis ALong the Perivascular Space) index from DWI, and the coupling between global blood-oxygenation-level-dependent (BOLD) signal and cerebrospinal fluid (gBOLD-CSF) from rs-fMRI. These markers will be evaluated in relation to cSVD subtypes and severity, with the goal of uncovering the contribution of glymphatic dysfunction to different cSVD trajectories.

Methods
This project will use multimodal neuroimaging datasets from two large cohort studies: First is the UK Biobank (UKB), accessible via an ongoing project on genetic and environmental determinants of cSVD (PI: Stephanie Debette), with neuroimaging data available from ~70K participants aged 40+. Second is the Bcube study (PI: Cecilia Samieri), a neuroimaging and biobank study in Bordeaux, with data from ~1000 volunteers aged 55+ already available. For Aim 1, two UKB subsamples (~5000 each) will serve as discovery and internal replication sets. T1w, FLAIR, and SWI scans will be processed with SHiVAi, a deep-learning-based pipeline for automated segmentation of core cSVD markers (WMH, PVS, CMB, and LAC). We will use a unified bullseye parcellation (Beyer et al, 2025) to quantify regional distributions. SuStaIn models will be trained on the discovery sample and tested on the replication sample. Longitudinal UKB data (~5000 subjects) will validate the inferred temporal progression. The Bcube cohort will serve as an external validation set. For Aim 2, we will develop standardized pipelines for two glymphatic markers: the DTI-ALPS index from DWI, and gBOLD-CSF from rs-fMRI. A refined ALPS pipeline addressing limitations of the original method will be implemented. Published gBOLD-CSF methods will be adapted for harmonization across cohorts. We will compute these markers in both cohorts and perform age-stratified analyses to assess their association with cSVD subtype and stage.

Expected results and impact
To our knowledge, this will be the first large-scale, data-driven attempt to integrate all core cSVD markers and their spatial patterns to identify biologically meaningful subtypes while modeling subtype-specific disease progression. This approach offers a major advance over currently available methods, which typically rely on summing visual rating scales or marker counts using arbitrary cut-offs. Although this project focuses on the potential role of glymphatic involvement on cSVD subtypes, the resulting data-driven subtyping and severity scoring framework will have a major impact on the broader research community by offering a novel, reproducible tool to stratify patients across the spectrum of cSVD. This framework can be directly applied in future studies to investigate genetic, environmental, and lifestyle determinants of specific cSVD trajectories. Further, the clarification of the role of glymphatic impairment in cSVD initiation and progression can shed new light on how brain clearance mechanisms may contribute to vascular brain aging.

VBHI Postdoctoral Fellowship (1 year)
PI: Ami Tsuchida, PhD  –