DIVE: A spatiotemporal progression model of brain pathology in neurodegenerative disorders

Abstract

Current models of progression in neurodegenerative diseases use neuroimaging measures that are averaged across pre-defined regions of interest (ROIs). Such models are unable to recover fine details of atrophy patterns; they tend to impose an assumption of strong spatial correlation within each ROI and no correlation among ROIs. Such assumptions may be violated by the influence of underlying brain network connectivity on pathology propagationa strong hypothesis e.g. in Alzheimer's Disease. Here we present DIVE: Data-driven Inference of Vertexwise Evolution. DIVE is an image-based disease progression model with single-vertex resolution, designed to reconstruct long-term patterns of brain pathology from shortterm longitudinal data sets. DIVE clusters vertex-wise (i.e. point-wise) biomarker measurements on the cortical surface that have similar temporal dynamics across a patient population, and concurrently estimates an average trajectory of vertex measurements in each cluster. DIVE uniquely outputs a parcellation of the cortex into areas with common progression patterns, leading to a new signature for individual diseases. DIVE further estimates the disease stage and progression speed for every visit of every subject, potentially enhancing stratification for clinical trials or management. On simulated data, DIVE can recover ground truth clusters and their underlying trajectory, provided the average trajectories are sufficiently different between clusters. We demonstrate DIVE on data from two cohorts: the Alzheimer's Disease Neuroimaging Initiative (ADNI) and the Dementia Research Centre (DRC), UK. The DRC cohort contains patients with Posterior Cortical Atrophy (PCA) as well as typical Alzheimer's disease (tAD). DIVE finds similar spatial patterns of atrophy arXiv:1901.03553v1 [cs.CV] 11 Jan 2019 for tAD subjects in the two independent datasets (ADNI and DRC), and further reveals distinct patterns of pathology in different diseases (tAD vs PCA) and for distinct types of biomarker datacortical thickness from Magnetic Resonance Imaging (MRI) vs amyloid load from Positron Emission Tomography (PET). We demonstrate that DIVE stages have potential clinical relevance, despite being based only on imaging data, by showing that the stages correlate with cognitive test scores. Finally, DIVE can be used to estimate a fine-grained spatial distribution of pathology in the brain using any kind of voxelwise or vertexwise measures including Jacobian compression maps, fractional anisotropy (FA) maps from diffusion tensor imaging (DTI) or other PET measures.DIVE availability: DIVE source code, written in Python3, is available at https://github.com/mrazvan22/dive and can be easily applied on any registered voxelwise images or images processed with the Freesurfer software. ADNI data can be downloaded from the Laboratory of NeuroImaging at the University of Southern California.

Fully Bayesian longitudinal unsupervised learning for the assessment and visualization of AD heterogeneity and progression

“…Hence, the data-driven algorithm provides explicit information on whether a subject has a distinct atrophy pattern or a mixture of patterns through the estimation of subject component probabilities. The proposed framework clusters subjects of a cohort into groups (provides probability of subjects to belong in any of the clusters) and not patterns of atrophy into groups for a cohort (clusters of regions/vertices) as in the study of Marinsecu and colleagues (Marinescu et al, 2019).…”

“…In the AD research field, many studies have focused on the unbiased identification of cortical and subcortical patterns of atrophy with structural MRI (sMRI). One recent study utilizes longitudinal atrophy markers to find sets of brain regions with common progression patterns (Marinescu et al, 2019). However, to date no cluster-based study has included longitudinal atrophy data in their method scheme, in order to identify groups of individuals with similar atrophy trajectories and our current study intends to meet this necessity.…”

Monotonic Gaussian Process for spatio-temporal disease progression modeling in brain imaging data

“…This allowed us to demonstrate the advantages of building more complex approaches such as MGPA for the problem we tackle in this work. Concerning the comparison with the state of the art in disease progression modelling, to the best of our knowledge the two closest approaches are [27] and [21]. However, these two methods are specifically designed for modelling data defined on brain surfaces.…”

“…On the contrary, our method aims at progression modeling using full 3D volumetric information. The data dimension we tackle is thus an order of magnitude greater than the one of [27] and [21], preventing these methods to scale to the spatial geometry of our data.…”

Data-driven detection of latent atrophy factors related to phenotypical variants of posterior cortical atrophy

“…Posterior cortical atrophy is a clinical-radiological syndrome defined by progressive loss of higher-order visual functions, and atrophy that markedly affects posterior brain regions such as the parietal and occipital cortices (Benson et al, 1988;Whitwell et al, 2007;Koedam et al, 2011, Lehmann et al, 2011bCrutch et al, 2012;Alves et al, 2013, Ossenkoppele et al, 2015bFirth et al, 2019;Marinescu et al, 2019). While multiple pathologies may underlie the posterior cortical atrophy syndrome, the most common biological substrate is Alzheimer's disease, accounting for ~80% of the cases (Renner et al, 2004;Tang-Wai et al, 2004;Montembeault et al, 2018).…”