AutoMacq: an automatic pipeline to analyse macaque structural MRI data

Abstract

AbstractMRI scanning of rhesus macaques is a growing field due to their evolutionary proximity and similar neuroanatomy to humans. As such, there is a need for automatic macaque MRI processing pipelines. AutoMacq is a pipeline capable of processing rhesus macaque MRI data to produce both voxel-based and surface-based metrics. It involves minimal manual intervention and can be carried out without expert knowledge of macaque neuroanatomy. To test the quality of the pipeline, scans from 74 subjects across 8 different sites were processed. Results indicate that over 87% of tissue segmentations and surfaces were of satisfactory quality to not require additional manual correction. Hemispheric comparisons and analyses of scan-rescan data showed strong reliability of the volumetric and surface-based outputs. Finally, to illustrate potential applications of AutoMacq, the change in grey matter volume with ageing was investigated cross-sectionally using subjects aged 3-15 years (corresponding to adolescence until mid-adulthood). The analysis revealed a linear decrease in grey matter volume with age similar to what has been found in humans, reinforcing the value of rhesus macaques as a model of healthy human ageing.