Simulating neurodegeneration through longitudinal population analysis of structural and diffusion weighted MRI data

Area of Science:

• Neuroimaging
• Computational Neuroscience
• Biomarker Discovery

Background:

• Neuroimaging biomarkers are crucial for diagnosing and tracking neurodegenerative diseases.
• Evaluating the accuracy of neuroimaging analysis methods is challenging due to the absence of a realistic ground truth.
• Existing methods for extracting disease-specific markers lack robust validation due to limited ground truth data.

Purpose of the Study:

• To propose a proof-of-concept patient- and disease-specific brain neurodegeneration simulator.
• To generate realistic simulated longitudinal neuroimaging data for method validation.
• To assess the simulator's ability to reproduce known patterns of neurodegeneration.

Main Methods:

• Development of a neurodegeneration simulator based on longitudinal, multi-modal brain imaging data.
• Application of the simulator to populations including normal controls, Alzheimer's disease (AD), and frontotemporal dementia (FTD) patients.
• Simulation of follow-up brain scans from baseline data and comparison with actual repeat scans; generation of simulated volume change maps for comparison with real data.

Main Results:

• The proposed simulator successfully generated realistic patient-specific patterns of longitudinal brain changes.
• Simulated follow-up images closely matched real repeat images for the studied populations.
• Generated simulated volume change maps showed comparability with those estimated from real longitudinal data.

Conclusions:

• The developed brain neurodegeneration simulator provides a biologically realistic, patient-specific ground truth for validating neuroimaging biomarkers.
• This tool can significantly aid in the development and evaluation of robust diagnostic and progression-tracking methods for neurodegenerative diseases.
• The simulator's ability to reproduce AD and FTD specific patterns highlights its potential for advancing neuroimaging research in these conditions.