Reaction-Diffusion Model of Cortical Atrophy Spread during Early Stages of Alzheimer's Disease
Kulason, S.; Miller, M.; Trouve, A.
Show abstract
1.This study introduces a reaction-diffusion model of atrophy spread across the rhinal cortex during early stages of Alzheimers disease. Our finite elements model of atrophy spread is motivated by histological evidence of a spatio-temporally specific pattern of neurofibrillary tau accumulation, and evidence of grey matter atrophy correlating with sites of neurofibrillary tau accumulation. The goal is to estimate disease-related parameters such as the origin of atrophy, the speed at which atrophy spreads, and the stage of the disease. We solve a constrained optimization problem using the adjoint state method and gradient descent to match modeled cortical thickness to observed cortical thickness as calculated from 3T MRI scans. Simulation testing shows that disease-related parameters can be estimated accurately with as little as 2 years of annual observations, depending on the stage of the disease. Case studies of 3 subjects suggests that we can pinpoint the origin of atrophy to the anterior transentorhinal cortex, and that the speed of atrophy spread is less than 1 mm per year. In the future, this type of modeling could be useful to stage the progression of the disease prior to the onset of clinical symptoms. 2. Author SummaryMisfolded tau proteins are associated with Alzheimers disease. They are known to accumulate and spread across the rhinal cortex, which is an area of the temporal lobe. Recent imaging studies suggest that we can detect grey matter thinning that occurs in pattern similar to tau spread. In this study, we introduce a model of disease spread to examine where thinning begins, how fast it spreads, and the stage of the disease. The results show that the origin of thinning corresponds with the earliest known location of tau accumulation, and spreads at a rate of less than 1 mm per year. Future work may focus on staging the progression of the disease using this type of model.
Matching journals
The top 6 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
- ReMiND: Recovery of Missing Neuroimaging using Diffusion Models with Application to Alzheimer’s Disease 94%
- Feature attention graph neural network for estimating brain age and identifying important neural connections in mouse models of genetic risk for Alzheimer's disease 94%
- Leveraging multimodal neuroimaging and GWAS for identifying modality-level causal pathways to Alzheimer’s disease 94%
Similar papers in this journal
- c-Triadem: A constrained, explainable deep learning model to identify novel biomarkers in Alzheimer’s disease 93%
- Eigenvector alignment: assessing functional network changes in amnestic mild cognitive impairment and Alzheimer's disease 93%
- Associations between regional blood-brain barrier disruption, aging, and Alzheimers disease biomarkers in cognitively normal older adults 93%
Similar papers in this journal
- Establishing a baseline for human cortical folding morphological variables: a multicenter study 94%
- The impact of genetic risk for Alzheimers disease on the structural brain networks of young adults 93%
- Cortical thickness estimation in individuals with cerebral small vessel disease, focal atrophy, and chronic stroke lesions. 92%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.