Assessing substantia nigra pars compacta connections with 3T and 7T multi-shell diffusion MRI
Poster No:
1604
Submission Type:
Abstract Submission
Authors:
Federica Franza1, Giuseppina Caiazzo1, Giovanni Cirillo1, Michele Papa1, Mario Cirillo1, Fabrizio Esposito1
Institutions:
1University of Campania "Luigi Vanvitelli", Naples, Italy
First Author:
Co-Author(s):
Introduction:
Structural connections through the substantia nigra pars compacta (SNc) could be disrupted at the early stages of Parkison's disease [1]. High angular resolution diffusion imaging (HARDI) tractography from multi-shell (MS) HARDI MRI series allows investigating whole-brain structural connections non-invasively in humans [2], [3] but its application to the quantitative analysis of subcortico-subcortical connections is still limited [4]. Here we compared the relative percentages of connections through SNpc terminating in the striatum or the thalamus across three different data sets, acquired at 3T and 7T in healthy human adults, using 3-shell protocols derived from ADNI (https://adni.loni.usc.edu) and HCP (https://www.humanconnectome.org) initiatives.
Methods:
Two MS-HARDI MRI data sets were serially acquired on a 3 T scanner at 1.5 mm isotropic resolution using ADNI (MS1) and HCP (MS2) protocols in 10 healthy human subjects (5 males, mean age ± std. 23.8 ± 3.65 years, age range: 25-30 years). Additional HCP 7T data sets from 10 age- and sex-matched healthy subjects, acquired at 1.0 mm isotropic resolution, were retrieved from the public HCP database [5]. Whole-brain tractograms were reconstructed via MRtrix (www.mrtrix.org) using the inverse fiber orientation distribution algorithm ver. 2 [6] within the anatomical constrained tractography framework [7]. Binary masks for subcortical regions of inclusion/exclusion were taken from validated anatomical atlases. Starting from the full set of streamlines through SNc, only those terminating in the striatum or thalamus were selected, after excluding those encompassing nearby structures according to previous knowledge [4]. The streamline counts were normalized to the total number of SNc streamlines, obtaining a relative connection index (%). Wilcoxon signed rank and rank sum tests were used to compare median estimates from MS1 vs. MS2 data sets (same subjects) and from MS1/MS2 vs. HCP-7T data sets (different subjects).
Results:
For the left (right) nigrostriatal pathway (figure 1), the median connection index (± IQR) was 7.20 ± 1.26 (6.56% ± 1.21) for MS1, 5.79% ± 1.35 (4.95% ± 1.98) for MS2, and 4.91% ± 1.91 (7.47% ± 3.28) for HCP (figure 2a). For the left (right) nigrothalamic pathway, the median connection index (± IQR) was 13.17% ± 5.74 (9.05% ± 3.62) for MS1, 10.99% ± 3.26 (7.86% ± 2.89) for MS2, and 9.66% ± 3.45 (6.12% ± 3.27) for HCP (figure 2b). The median index of the SNc-striatum connection was significantly higher for MS1 (ADNI-3T), compared to MS2 (HCP-3T) (left: p = 0.0020, right: p = 0.0039) and, for the left hemisphere, to HCP-7T (p = 0.0073), data sets. The median index of SNc-Thalamus connection was not significant different between MS1 and MS2 (3T) data sets and between MS1/MS2 (3T) and HCP 7T datasets.
·Figure 1. 3D visualization of the streamlines between SNc and the striatum for one representative subject acquired with MS1 (ADNI-3T).
·Figure 2. Relative connection index estimated from MS1 (ADNI-3T) , MS2 (HCP-3T) and HCP-7T datasets for both hemispheres: (a) target region: striatum; (b) target region: thalamus.
Conclusions:
MS-HARDI protocols derived from ADNI and HCP initiatives identified non-zero putative nigro-striatal and nigro-thalamic connections. A significantly higher SNc-Striatum connectivity was obtained from MS1 (ADNI-3T), compared to MS2 (HCP-3T) data sets, on the same subjects, despite MS1 having less directions. Most likely, the shorter acquisition time of MS1 data sets might have reduced the impact of motion artifacts. MS1 (3T) data sets also revealed relatively more SNc-Striatum connections than HCP (7T) data sets in the left hemisphere despite the lower spatial resolution. A larger intersubject variability was observed for the putative SNc-Thalamus connections with no significant differences between datasets. Nevertheless, these estimates should be taken more carefully since there is no current evidence in humans for direct nigro-thalamic pathways. Furthermore, the estimates did not expressly account for volumes of, or distances between, the regions, making it not appropriate to compare estimates between different target regions [8].
Modeling and Analysis Methods:
Diffusion MRI Modeling and Analysis 1
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
Subcortical Structures 2
White Matter Anatomy, Fiber Pathways and Connectivity
Keywords:
MRI
STRUCTURAL MRI
Sub-Cortical
Thalamus
Tractography
White Matter
WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC
1|2Indicates the priority used for review
Provide references using author date format
[2] F. Trojsi et al., «High angular resolution diffusion imaging abnormalities in the early stages of amyotrophic lateral sclerosis», Journal of the Neurological Sciences, vol. 380, pp. 215–222, set. 2017, doi: 10.1016/j.jns.2017.07.039.
[3] B. Jeurissen et al., «Multi-tissue constrained spherical deconvolution for improved analysis of multi-shell diffusion MRI data», NeuroImage, vol. 103, pp. 411–426, dic. 2014, doi: 10.1016/j.neuroimage.2014.07.061.
[4] A. Quartarone et al., «New insights into cortico-basal-cerebellar connectome: clinical and physiological considerations.», Brain, vol. 143, fasc. 2, pp. 396–406, feb. 2020, doi: 10.1093/brain/awz310.
[5] A. T. Vu et al., «High resolution whole brain diffusion imaging at 7T for the Human Connectome Project.», Neuroimage, vol. 122, pp. 318–331, nov. 2015, doi: 10.1016/j.neuroimage.2015.08.004.
[6] J.-D. Tournier et al., «Improved probabilistic streamlines tractography by 2nd order integration over fibre orientation distributions», Proc. Intl. Soc. Mag. Reson. Med. (ISMRM), vol. 18, gen. 2010.
[7] R. E. Smith et al., «Anatomically-constrained tractography: improved diffusion MRI streamlines tractography through effective use of anatomical information.», Neuroimage, vol. 62, fasc. 3, pp. 1924–1938, set. 2012, doi: 10.1016/j.neuroimage.2012.06.005.
[8] Y.-N. Chang et al., «Distance-dependent distribution thresholding in probabilistic tractography.», Hum Brain Mapp, vol. 44, fasc. 10, pp. 4064–4076, lug. 2023, doi: 10.1002/hbm.26330.