Effects of Compensation, Connectivity and Tau in a Computational Model of Alzheimer's Disease

TL;DR

This paper enhances a computational model of Alzheimer's Disease by incorporating tau pathology and small-world connectivity, revealing how these factors influence network capacity, robustness, and damage progression.

Contribution

It introduces a novel simulation of tau pathology and examines the effects of small-world connectivity on neural network behavior in AD models.

Findings

Small-world connectivity affects capacity, retrieval time, and robustness.

Using remote memories for compensation accelerates network damage.

Tau pathology simulation results in more severe network damage.

Abstract

This work updates an existing, simplistic computational model of Alzheimer's Disease (AD) to investigate the behaviour of synaptic compensatory mechanisms in neural networks with small-world connectivity, and varying methods of calculating compensation. It additionally introduces a method for simulating tau neurofibrillary pathology, resulting in a more dramatic damage profile. Small-world connectivity is shown to have contrasting effects on capacity, retrieval time, and robustness to damage, whilst the use of more easily-obtained remote memories rather than recent memories for synaptic compensation is found to lead to rapid network damage.