Abstract
Neurodegenerative diseases are devastating conditions characterized by progressive cognitive decline with few available treatments. Neurodegeneration can be quantified in vertebrate and invertebrate models of disease by analysis of vacuolation - the formation of empty spaces within brain tissue. Previous approaches for quantifying this phenotype have required time-consuming methods such as manual counting and measuring of vacuole dimensions, which can be subjective. Here we describe VacQuant, a novel application that can be paired with existing machine learning software to automatically measure the area of vacuolation in brain tissue. Using Drosophila brain sections from tauopathy model flies, a well-described model of dementia-related neurodegeneration, we quantified a significant increase in brain vacuolation at several timepoints in adult flies with the aid of VacQuant. When compared with quantification by five blinded volunteers, the machine learning method positively correlated with their group average, confirming its accuracy and functionality. This automated method developed with VacQuant removes human bias and measurement variation, providing a consistent threshold for all brain sections and experiments. This automated pipeline will be particularly useful for high-throughput screening for genetic modifiers or therapeutic compounds in animal models of neurodegeneration.
