ACID: A Comprehensive Toolbox for Image Processing and Modeling of Brain, Spinal Cord, and Post-mortem Diffusion MRI Data

Gergely Dávid, Björn Fricke, Jan Malte Oeschger, Lars Ruthotto, Francisco J. Fritz, Ora Ohana, Laurin Mordhorst, Thomas Sauvigny, Patrick Freund, Karsten Tabelow, Siawoosh MohammadibioRxiv 2023 · DOI

TL;DR

This paper introduces ACID, an open-source MATLAB toolbox that extends the SPM software for comprehensive processing and modeling of diffusion MRI (dMRI) data from the brain, spinal cord, and ex vivo specimens. ACID addresses key challenges in dMRI analysis, including artifact correction, advanced diffusion modeling, and biophysical parameter estimation, making it a valuable tool for neuroscience research.

What Problem Does This Paper Address?

Diffusion MRI (dMRI) is a powerful technique for imaging tissue microstructure, with growing applications beyond the brain to the spinal cord and ex vivo samples. However, analyzing dMRI data is complex due to artifacts (motion, eddy currents, susceptibility) and the need for sophisticated models to interpret the signal. Existing toolboxes often focus on brain data, lack comprehensive artifact correction, or don't integrate advanced biophysical models. This paper addresses the need for a unified, user-friendly, and extensible toolbox that can handle the unique challenges of dMRI analysis across different tissue types and experimental conditions, while also adhering to data standardization principles like BIDS.

Methods

ACID is built upon the Statistical Parametric Mapping (SPM) framework, leveraging its robust image processing and statistical analysis capabilities. Key methodological contributions include:

  • Artifact Correction: State-of-the-art algorithms for eddy current and motion correction (ECMOCO), adaptive denoising (msPOAS), Rician bias correction (RBC), and susceptibility artifact correction (HySCO). Notably, ECMOCO offers slice-wise registration for improved spinal cord alignment.
  • Diffusion Modeling: Implementation of Diffusion Tensor Imaging (DTI) and Diffusion Kurtosis Imaging (DKI) with various fitting algorithms (OLS, WLS, robust fitting, NLLS).
  • Biophysical Modeling: Inclusion of models like WMTI-Watson and NODDI-DTI to estimate microstructural properties such as axonal water fraction and orientation dispersion.
  • BIDS Compliance: Adherence to the Brain Imaging Data Structure (BIDS) standard for organized data output, facilitating sharing and compatibility with other tools.
  • Modularity and Extensibility: A modular design allows for seamless integration of external tools and customization of analysis pipelines.

Datasets

The paper describes the use of three example datasets for demonstrating ACID's capabilities:

  • An in vivo brain dMRI dataset.
  • An in vivo spinal cord dMRI dataset.
  • An ex vivo temporal lobe specimen dMRI dataset.

Details on the scan parameters for these datasets are provided in Table 4.

Key Insights and Ideas

  • Spinal Cord dMRI Challenges: Analyzing spinal cord dMRI is significantly more challenging than brain dMRI due to lower SNR, complex motion (respiration, pulsation), and different artifact profiles, requiring specialized correction strategies like slice-wise registration.
  • Biophysical Models for Microstructure: Biophysical models offer a more direct link between dMRI signal and tissue microstructure, providing parameters like axonal water fraction and orientation dispersion, which are crucial for understanding tissue health and disease.
  • BIDS Standardization: Adopting BIDS is essential for reproducible research, data sharing, and interoperability with the growing ecosystem of neuroimaging tools.
  • SPM Integration: Leveraging SPM's established framework provides ACID users with access to a wide range of powerful tools for segmentation, spatial normalization, and statistical analysis, creating a unified analysis environment.

Key Results

  • ACID provides a comprehensive pipeline from raw DICOM data to group-level statistics.
  • The toolbox successfully processes dMRI data from brain, spinal cord, and ex vivo specimens.
  • Example pipelines demonstrate the effectiveness of artifact correction and model fitting for these different data types.
  • Parallel computing is implemented to accelerate processing, with computation times provided for selected modules.
  • ACID has been used in various clinical and neuroscience research studies, including those on multiple sclerosis, Parkinson's disease, gliomas, and spinal cord injury.

Limitations and Future Work

  • Parameter Optimization: Performance can be highly dependent on parameter settings, necessitating systematic exploration for optimal results on specific datasets.
  • Quantitative Comparison: A dedicated future study is warranted for a thorough quantitative comparison of ACID with other toolboxes.
  • Ongoing Development: The authors acknowledge ongoing developments and plan to integrate more advanced biophysical models and estimators (e.g., Rician maximum likelihood estimator).
  • Clinical Use: ACID is intended for academic research purposes and is not approved for clinical diagnosis.

Summary updated

Abstract

Diffusion MRI (dMRI) has become a crucial imaging technique in the field of neuroscience, with a growing number of clinical applications. Although most studies still focus on the brain, there is a growing interest in utilizing dMRI to investigate the healthy or injured spinal cord. The past decade has also seen the development of biophysical models that link MR-based diffusion measures to underlying microscopic tissue characteristics, which necessitates validation through ex vivo dMRI measurements. Building upon 13 years of research and development, we present an open-source, MATLAB-based academic software toolkit dubbed ACID: A Comprehensive Toolbox for Image Processing and Modeling of Brain, Spinal Cord, and Ex Vivo Diffusion MRI Data. ACID is an extension to the Statistical Parametric Mapping (SPM) software, designed to process and model dMRI data of the brain, spinal cord, and ex vivo specimens by incorporating state-of-the-art artifact correction tools, diffusion and kurtosis tensor imaging, and biophysical models that enable the estimation of microstructural properties in white matter. Additionally, the software includes an array of linear and non-linear fitting algorithms for accurate diffusion parameter estimation. By adhering to the Brain Imaging Data Structure (BIDS) data organization principles, ACID facilitates standardized analysis, ensures compatibility with other BIDS-compliant software, and aligns with the growing availability of large databases utilizing the BIDS format. Furthermore, being integrated into the popular SPM framework, ACID benefits from a wide range of segmentation, spatial processing, and statistical analysis tools as well as a large and growing number of SPM extensions. As such, this comprehensive toolbox covers the entire processing chain from raw DICOM data to group-level statistics, all within a single software package.
BibTeX
@misc{david2023acid,
  author = {Gergely Dávid and Björn Fricke and Jan Malte Oeschger and Lars Ruthotto and Francisco J. Fritz and Ora Ohana and Laurin Mordhorst and Thomas Sauvigny and Patrick Freund and Karsten Tabelow and Siawoosh Mohammadi},
  title = {{ACID: A Comprehensive Toolbox for Image Processing and Modeling of Brain, Spinal Cord, and Post-mortem Diffusion MRI Data}},
  year = {2023},
  doi = {10.1101/2023.10.13.562027},
  howpublished = {bioRxiv},
}
Figure