Processing pipeline for atlas-based imaging data analysis of structural and functional mouse brain MRI (AIDAmri)

Abstract

Magnetic resonance imaging (MRI) is a key technology in multimodal animal studies of brain connectivity and disease pathology. In vivo MRI provides non-invasive, whole brain macroscopic images containing structural and functional information, thereby complementing invasive in vivo high-resolution microscopy and ex vivo molecular techniques. Brain mapping, the correlation of corresponding regions between multiple brains in a standard brain atlas system, is widely used in human MRI. For small animal MRI, however, there is no scientific consensus on pre-processing strategies and atlas-based neuroinformatics. Thus, it remains difficult to compare and validate results from different pre-clinical studies which were processed using custom-made code or individual adjustments of clinical MRI software and without a standard brain reference atlas. Here, we describe AIDAmri, a novel Atlas-based Imaging Data Analysis pipeline to process structural and functional mouse brain data including anatomical MRI, fiber tracking using diffusion tensor imaging (DTI) and functional connectivity analysis using resting-state functional MRI (rs-fMRI). The AIDAmri pipeline includes automated pre-processing steps, such as raw data conversion, skull-stripping and bias-field correction as well as image registration with the Allen Mouse Brain Reference Atlas (ARA). Following a modular structure developed in Python scripting language, the pipeline integrates established and newly developed algorithms. Each processing step was optimized for efficient data processing requiring minimal user-input and user programming skills. The raw data is analyzed and results transferred to the ARA coordinate system in order to allow an efficient and highly-accurate region-based analysis. AIDAmri is intended to fill the gap of a missing open-access and cross-platform toolbox for the most relevant mouse brain MRI sequences thereby facilitating data processing in large cohorts and multi-center studies.