Characterizing Behavioral Phenotypes in the 5xFAD Alzheimer’s Disease Mouse Model Using Computer Vision-based Monitoring

von Eckartsberg, Alex

Full Metadata

Overview

Year:: 2024
Contributor:: von Eckartsberg, Alex (creator); Jong, Yong (Reader); Fleischmann, Alexander (Reader); Brown University. Biology and Medicine: Biotechnology (sponsor)
Genre:: theses
Subject:: Alzheimer's disease; Artificial intelligence; Automated Continuous Behavioral Monitoring; Mouse behavior; Nervous system--Degeneration; Deep learning (Machine learning); Alzheimer's disease--Treatment; Alzheimer's disease--Prevention; Neural computers; Neuromuscular diseases
Extent:: vi, 38 p.

Files

Description

Abstract:: By 2017, only four approved drugs to treat symptoms of AD existed, while a resounding 146 drug candidates had failed to be approved by the FDA. Today, this theme remains, with only a single disease modifying therapy (DMT) in lecanumab (Leqembi ®) gaining approval while showing underwhelming results in slowing the progression of the disease. While characterizing the neuropathological changes that occur in testing these therapies on mice and other animals is critical for preclinical development and ultimately FDA approval, behavioral assessments that capture changes in motor function over time are limited. Historically, neuromotor behavioral testing in mouse models includes open field analysis, grip strength analysis, rotarod, and wire hang, among others. While these traditional-style tests provide useful information, they can be prone to human error and inconsistencies. In this study, we use a common Alzheimer’s Disease (AD) mouse model and assess motor dysfunction and behavior using an unbiased, automated, AI-powered recording model entitled Automated Continuous Behavioral Monitoring (ACBM). The 5xFAD transgenic mouse is most conducive to testing motor dysfunction due to excessive amyloid plaque build-up. At 8 months, a novel behavioral phenotype of “early activity onset” was identified showing 5xFAD mice becoming highly active approximately one hour earlier than WT mice during early morning hours. Additionally, phenotypes confirming historical trends in 5xFAD behavior were identified in both Walk, Hang, and Sniff behaviors showing significant increases hyperactivity, along with decreased body composition compared to the WT mice. Importantly, our novel comprehensive analysis of groom activity in the 5xFAD mouse is only preceded by one other study monitoring the behavior for just 5 minutes during open field testing (Ullah et al., 2020). Our data suggest that not only does ACBM confirm previous behavior summarized by traditional behavior testing, but that it can deliver consistent monitoring while identifying potentially new behaviors or deficits powered by computer vision that would otherwise not be known. Our lab continues to utilize ACBM to monitor additional ages of 5xFAD mice and others to form a comprehensive dataset with the ultimate goal of testing new DMTs/anti-amyloids, such as lecanumab, to observe any behavioral changes.
Notes:: Thesis (Sc. M.)--Brown University, 2024

Content

Citation

von Eckartsberg, Alex, "Characterizing Behavioral Phenotypes in the 5xFAD Alzheimer’s Disease Mouse Model Using Computer Vision-based Monitoring" (2024). Brown Digital Repository. Brown University Library. https://repository.library.brown.edu/studio/item/bdr:gjun5yrw/