Nighres: a python toolbox for high-resolution neuroimaging
Poster No:
1900
Submission Type:
Abstract Submission
Authors:
Pierre-Louis Bazin1, Julia Huntenburg2, Julia Huck3, Leevi Kerkela4, Hoang Dung Do3, Tristan Glatard3, Christopher Steele3
Institutions:
1Universiteit van Amsterdam, Amsterdam, North Holland, 2Systems Neuroscience Lab, Champalimaud Research, Lisbon, n/a, 3Concordia University, Montreal, Quebec, 4UCL Great Ormond Street Institute of Child Health, University College London, London, n/a
First Author:
Co-Author(s):
Leevi Kerkela
UCL Great Ormond Street Institute of Child Health, University College London
London, n/a
UCL Great Ormond Street Institute of Child Health, University College London
London, n/a
Introduction:
With the recent advances of ultra-high field MRI into the neuroscientific and clinical domains, high resolution and multi-dimensional MRI data is more common. Yet, moving into sub-millimeter resolutions and handling quantitative contrasts are major challenges for many image analysis toolboxes, initially designed for conventional 1.5T and 3T imaging data. With this new toolbox, we address specifically these issues. The tools gathered in it cover the main steps of structural image processing, from quantitative parameter reconstruction to laminar cortical depth modeling. It focuses not only on the cerebral cortex, but also provides new tools to investigate the subcortex and the cerebellum. The methods in this toolbox scale well with image resolution, and can handle data at 400 µm routinely.
Methods:
The toolbox, Nighres (for "neuroimaging at high resolution"; Huntenburg et al., 2019; https://nighres.readthedocs.io/en/latest/), is a Python library developed collaboratively between multiple centers over the open source Github platform. It incorporates Java-based algorithms from the CBSTools (https://github.com/piloubazin/cbstools-public) and the IMCN toolkit (https://github.com/IMCN-UvA/imcn-imaging) for computational efficiency. The Java algorithms are largely standalone, and wrapped with the JCC utility. Documentation is generated automatically from the Python functions and the code is automatically tested through Travis, a continuous integration engine. Nighres has also been released as a Docker environment and adapted for parallel computations with Dask (https://github.com/dohoangdzung/neuroimg_pipelines). Example processing pipelines with test data are included for the major analysis tools.
Because of its decentralized and modular architecture, Nighres includes a variety of tools, some very specialized to given imaging sequences such as the MP2RAGE (Marques et al., 2010) and MP2RAGEME (Caan et al., 2018) and others more generic such as cortical extraction or topology correction. Currently, the toolbox includes tools for quantitative MRI reconstruction (estimation of T1 and T2* from MP2RAGE and multi-echo sequences), high-resolution image denoising (Local-Complex PCA; Bazin et al., 2019), general brain segmentation (skull stripping, whole brain segmentation, cortical reconstruction; Bazin et al., 2014), cortical depth modeling (volumetric layering, cortical profile sampling, laminar averaging; Waehnert et al., 2016), vascular modeling (Huck et al., 2019), white matter tract labeling (Bazin et al., 2011), subcortical parcellation, topology correction, cortical surface inflation, and tissue classification. Additional utilities to handle multiple registration steps between images, average shapes, handle level-set and mesh representations of surfaces, unwrap phase data or perform simple statistics are also included.
Because of its decentralized and modular architecture, Nighres includes a variety of tools, some very specialized to given imaging sequences such as the MP2RAGE (Marques et al., 2010) and MP2RAGEME (Caan et al., 2018) and others more generic such as cortical extraction or topology correction. Currently, the toolbox includes tools for quantitative MRI reconstruction (estimation of T1 and T2* from MP2RAGE and multi-echo sequences), high-resolution image denoising (Local-Complex PCA; Bazin et al., 2019), general brain segmentation (skull stripping, whole brain segmentation, cortical reconstruction; Bazin et al., 2014), cortical depth modeling (volumetric layering, cortical profile sampling, laminar averaging; Waehnert et al., 2016), vascular modeling (Huck et al., 2019), white matter tract labeling (Bazin et al., 2011), subcortical parcellation, topology correction, cortical surface inflation, and tissue classification. Additional utilities to handle multiple registration steps between images, average shapes, handle level-set and mesh representations of surfaces, unwrap phase data or perform simple statistics are also included.
Results:
Examples of processing made possible by Nighres are given in Fig.1. Because the algorithms are designed to handle high-resolution data, the main requirement to run is adequate amounts of memory available, which scales linearly with data size. Most included methods range in complexity from O(N) to O(NlogN), resulting in reasonable computation times on a single processor, a few minutes to a few hours for 7T image analysis, depending on algorithm and image resolution. For instance the complete segmentation pipeline including skull stripping, brain segmentation, cortical reconstruction and depth modeling takes under 12 minutes for 0.7 mm resolution data on a modern laptop.
Conclusions:
With the Nighres toolbox, we provide a set of carefully curated tools tailored to high resolution MRI, integrating new MR sequences and new analysis techniques for spatially detailed neuroimaging analysis. The toolbox is fully open, compatible with the vast majority of neuroimaging tools, supported across multiple centers, and complemented by examples and utilities to help researchers integrate it into their analysis procedures.
Modeling and Analysis Methods:
Image Registration and Computational Anatomy
Motion Correction and Preprocessing
Segmentation and Parcellation
Neuroinformatics and Data Sharing:
Workflows 1
Novel Imaging Acquisition Methods:
Anatomical MRI 2
Keywords:
Cerebellum
Computing
Cortical Layers
HIGH FIELD MR
Informatics
Segmentation
Spatial Normalization
STRUCTURAL MRI
Sub-Cortical
White Matter
1|2Indicates the priority used for review
My abstract is being submitted as a Software Demonstration.
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Was any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using author date format
Bazin, P.-L., Weiss, M., Dinse, J., Schäfer, A., Trampel, R., Turner, R., 2014. A computational framework for ultra-high resolution cortical segmentation at 7Tesla. NeuroImage 93, 201–209. https://doi.org/10.1016/j.neuroimage.2013.03.077
Bazin, P.-L., Alkemade, A., van der Zwaag, W., Caan, M., Mulder, M., Forstmann, B.U., 2019. Denoising High-Field Multi-Dimensional MRI With Local Complex PCA. Frontiers in Neuroscience 13. https://doi.org/10.3389/fnins.2019.01066
Caan, M.W.A., Bazin, P., Marques, J.P., Hollander, G., Dumoulin, S.O., Zwaag, W., 2019. MP2RAGEME: T 1 , T 2 * , and QSM mapping in one sequence at 7 tesla. Human Brain Mapping 40, 1786–1798. https://doi.org/10.1002/hbm.24490
Huntenburg, J.M., Steele, C.J., Bazin, P.-L., 2018. Nighres: processing tools for high-resolution neuroimaging. GigaScience 7. https://doi.org/10.1093/gigascience/giy082
Huck, J., Wanner, Y., Fan, A.P., Jäger, A.-T., Grahl, S., Schneider, U., Villringer, A., Steele, C.J., Tardif, C.L., Bazin, P.-L., Gauthier, C.J., 2019. High resolution atlas of the venous brain vasculature from 7 T quantitative susceptibility maps. Brain Structure and Function 224, 2467–2485. https://doi.org/10.1007/s00429-019-01919-4
Marques, J.P., Kober, T., Krueger, G., van der Zwaag, W., Van de Moortele, P.-F., Gruetter, R., 2010. MP2RAGE, a self bias-field corrected sequence for improved segmentation and T1-mapping at high field. NeuroImage 49, 1271–1281. https://doi.org/10.1016/j.neuroimage.2009.10.002
Waehnert, M.D., Dinse, J., Schäfer, A., Geyer, S., Bazin, P.-L., Turner, R., Tardif, C.L., 2016. A subject-specific framework for in vivo myeloarchitectonic analysis using high resolution quantitative MRI. NeuroImage 125, 94–107. https://doi.org/10.1016/j.neuroimage.2015.10.001