Resomapper: A Versatile Pipeline for Multiparametric MRI Processing. A Demonstrative Application in Mouse Brain Imaging

Abstract

ObjectivesMagnetic resonance imaging (MRI) is essential in both research and clinical settings, with quantitative MRI (qMRI) enhancing reproducibility and sensitivity. However, qMRI processing can be complex, especially for users with limited coding experience. We introduce Resomapper, an open-source, cross-platform tool that integrates established processing libraries into a unified, user-friendly workflow, simplifying qMRI analysis while promoting accessibility, reproducibility, and data sharing. Materials and MethodsResomapper is a Python-based pipeline designed for intuitive multiparametric MRI processing. It supports T1, T2, and T2* relaxometry, magnetization transfer imaging (MTI) and diffusion tensor imaging (DTI) model fitting. The software includes advanced preprocessing options such as denoising, Gibbs artifact removal, and bias field correction. Users can process data interactively through a sequential pipeline or via an automated JSON-configured workflow, both executable through simple command-line instructions. Resomapper ensures compatibility by converting raw MRI data from different formats into the standardized NIfTI format within a BIDS-like structure, enhancing reproducibility, scalability, and data management efficiency. To demonstrate its application, we present a brain MRI study carried out on healthy, adult C57BL/6J mice, both sexes. MRI acquisitions were conducted on a Bruker Biospec 7T system using a multiparametric MRI protocol that included anatomical T2W images, T2 and T2* maps, MTI and DTI. The data were processed with Resomapper and co-registered using ANTsPy. Finally, a region of interest (ROI)-based analysis was performed to examine differences between sexes and brain areas, focusing on the cortex (Cx), hippocampus, (HPC), thalamus (Thal), and hypothalamus (HTH). ResultsDifferences in all MRI parameters were found across brain regions, as expected. Additionally, a small significant sex difference in T2* was observed, with higher values in the thalamus and hypothalamus of female mice compared to males. This may reflect sex-specific responses to anesthesia. Moreover, this study also serves as a reference for standardized multiparametric qMRI studies in mice using Resomapper. ConclusionsBy integrating multiple processing tools into a single, accessible framework, Resomapper streamlines qMRI workflows and enables reproducible, high-quality image analysis. Thanks to its ability to handle diverse preprocessing techniques, multiple imaging modalities, and standardized data formats, the software proves to be a valuable resource for researchers with varying levels of programming expertise.