Sex-related variability of white matter tracts in the whole HCP cohort
Poster No:
2172
Submission Type:
Abstract Submission
Authors:
Bastien Herlin1,2, Ivy Uszynski1, Maëlig Chauvel3,1, Sophie Dupont2,4,5,6, Cyril Poupon1
Institutions:
1Neurospin, Gif-sur-Yvette, France, 2Rehabilitation Unit, AP-HP, Pitié-Salpêtrière Hospital, Paris, France, 3Max Planck Institute, Leipzig, Germany, 4Epileptology Unit, Reference Center for Rare Epilepsies, Department of Neurology, AP-HP, Pitié-Salpêtrière Hospital, Paris, France, 5Paris Brain Institute (ICM), Sorbonne-Université, Inserm U1127, CNRS 7225, Paris, France, 6Université Paris Sorbonne, Paris, France
First Author:
Bastien Herlin
Neurospin|Rehabilitation Unit, AP-HP, Pitié-Salpêtrière Hospital
Gif-sur-Yvette, France|Paris, France
Neurospin|Rehabilitation Unit, AP-HP, Pitié-Salpêtrière Hospital
Gif-sur-Yvette, France|Paris, France
Co-Author(s):
Sophie Dupont
Rehabilitation Unit, AP-HP, Pitié-Salpêtrière Hospital|Epileptology Unit, Reference Center for Rare Epilepsies, Department of Neurology, AP-HP, Pitié-Salpêtrière Hospital|Paris Brain Institute (ICM), Sorbonne-Université, Inserm U1127, CNRS 7225|Université Paris Sorbonne
Paris, France|Paris, France|Paris, France|Paris, France
Rehabilitation Unit, AP-HP, Pitié-Salpêtrière Hospital|Epileptology Unit, Reference Center for Rare Epilepsies, Department of Neurology, AP-HP, Pitié-Salpêtrière Hospital|Paris Brain Institute (ICM), Sorbonne-Université, Inserm U1127, CNRS 7225|Université Paris Sorbonne
Paris, France|Paris, France|Paris, France|Paris, France
Introduction:
Behavioral studies have shown that men and women have many similarities, but also some specificities (1–3). Whether these are due to biological or social differences, or a combination of both, remains unclear. Many studies have examined sex differences in cortical gray matter, but few have done so for white matter tracts. We therefore conducted a sex comparison of all deep white matter tracts in a large cohort of 1065 subjects.
Methods:
Cohort and imaging protocol
We used the dataset from the HCP cohort (4) of 1065 healthy subjects aged 22-35 years old. Each subject had a series of diffusion-weighted MRI (dMRI) sequences performed on a 3T Siemens MRI system using a 2D spin-echo single-shot multiband EPI sequence (1.25mm isotropic spatial resolution) over 3 shells along 90 diffusion sampling directions for each.
Data processing
For dMRI data processing, we used the Ginkgo toolbox developed by the CEA/NeuroSpin team, available at https://framagit.org/cpoupon/gkg. Our pipeline performed several successive steps: diffeomorphic registration to a common atlas space (the MNI ICBM 2009c nonlinear asymmetric template) with the Advanced Normalization Tools (ANTS) (5); computation of the Orientation Distribution Functions (ODF) with the analytical Q-ball model (6); computation of a whole-brain probabilistic tractography (7); and intra-subject fiber clustering (8) which groups fibers according to their geometric properties.
Group analysis
Cross-subject fiber clustering was performed on the entire cohort using the HDBscan algorithm (9). Deep white matter tracts were independently identified by two neuroanatomists using manual ROI selection. This allowed the construction of a deep white matter atlas of 79 tracts.
Statistical analysis
We measured the total brain volume (TBV) from the Freesurfer brain mask, and tract volumes from their density masks, which were then normalized to the subject's TBV. We computed several microstructural parameters: fractional anisotropy FA, mean diffusivity MD, parallel and transverse diffusivity (from the DTI model); generalized fractional anisotropy GFA (from the Q-ball model); neurite density index NDI, isotropic water volume fraction ISOVF, and orientation dispersion index ODI (from the NODDI model (10)). We applied Bonferroni correction for multiple comparisons (corrected threshold: p=7.0225*10-5), and estimated effect size with Cohen's d-test.
We used the dataset from the HCP cohort (4) of 1065 healthy subjects aged 22-35 years old. Each subject had a series of diffusion-weighted MRI (dMRI) sequences performed on a 3T Siemens MRI system using a 2D spin-echo single-shot multiband EPI sequence (1.25mm isotropic spatial resolution) over 3 shells along 90 diffusion sampling directions for each.
Data processing
For dMRI data processing, we used the Ginkgo toolbox developed by the CEA/NeuroSpin team, available at https://framagit.org/cpoupon/gkg. Our pipeline performed several successive steps: diffeomorphic registration to a common atlas space (the MNI ICBM 2009c nonlinear asymmetric template) with the Advanced Normalization Tools (ANTS) (5); computation of the Orientation Distribution Functions (ODF) with the analytical Q-ball model (6); computation of a whole-brain probabilistic tractography (7); and intra-subject fiber clustering (8) which groups fibers according to their geometric properties.
Group analysis
Cross-subject fiber clustering was performed on the entire cohort using the HDBscan algorithm (9). Deep white matter tracts were independently identified by two neuroanatomists using manual ROI selection. This allowed the construction of a deep white matter atlas of 79 tracts.
Statistical analysis
We measured the total brain volume (TBV) from the Freesurfer brain mask, and tract volumes from their density masks, which were then normalized to the subject's TBV. We computed several microstructural parameters: fractional anisotropy FA, mean diffusivity MD, parallel and transverse diffusivity (from the DTI model); generalized fractional anisotropy GFA (from the Q-ball model); neurite density index NDI, isotropic water volume fraction ISOVF, and orientation dispersion index ODI (from the NODDI model (10)). We applied Bonferroni correction for multiple comparisons (corrected threshold: p=7.0225*10-5), and estimated effect size with Cohen's d-test.
Results:
Volumetric analysis
Total brain volume was significantly different between men and women (mean+/-standard deviation: 1128+/-90 cm3 in women, 1290+/-102 cm3 in men; relative difference: 12.62%; p=1.30*10-127; d=1.7). 19 white matter tracts had a significant difference in normalized volume between men and women, two having a Cohen's d greater than 0.8: the corpus callosum genu and the parallel fibers (Fig 1), both larger in women (12.54% and 6.1%, respectively). Linear regression and ANCOVA confirmed a significant interaction between the normalized volume of these two tracts and sex (corpus callosum genu: p=3.49*10-65 ; parallel fibers: p=3.87*10-75).
Microstructural analysis
Many significant microstructural differences between sexes were found in several tracts with moderate or large effect sizes, from which a composite score was calculated (Fig 2). The most different tracts were associated with the motor system, most of which showed higher ODI, MD, parallel and transverse diffusivity in men; or with the limbic system, most of which showed higher FA, GFA, MD, and parallel diffusivity, and lower ODI in women.
Total brain volume was significantly different between men and women (mean+/-standard deviation: 1128+/-90 cm3 in women, 1290+/-102 cm3 in men; relative difference: 12.62%; p=1.30*10-127; d=1.7). 19 white matter tracts had a significant difference in normalized volume between men and women, two having a Cohen's d greater than 0.8: the corpus callosum genu and the parallel fibers (Fig 1), both larger in women (12.54% and 6.1%, respectively). Linear regression and ANCOVA confirmed a significant interaction between the normalized volume of these two tracts and sex (corpus callosum genu: p=3.49*10-65 ; parallel fibers: p=3.87*10-75).
Microstructural analysis
Many significant microstructural differences between sexes were found in several tracts with moderate or large effect sizes, from which a composite score was calculated (Fig 2). The most different tracts were associated with the motor system, most of which showed higher ODI, MD, parallel and transverse diffusivity in men; or with the limbic system, most of which showed higher FA, GFA, MD, and parallel diffusivity, and lower ODI in women.
Conclusions:
Our study revealed some differences in white matter tracts between men and women, both in their normalized volume and in their microstructure. Future research can build on our findings to further explore the complex relationship between brain connectivity and the cognitive and behavioral traits that differ between men and women.
Modeling and Analysis Methods:
Diffusion MRI Modeling and Analysis 2
Neuroanatomy, Physiology, Metabolism and Neurotransmission:
White Matter Anatomy, Fiber Pathways and Connectivity 1
Keywords:
MRI
Sexual Dimorphism
Tractography
White Matter
WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC
1|2Indicates the priority used for review
·Fig 1. Comparisons for total brain volume (top), normalized volume of white matter tracts (middle) and linear regression for the two most different tracts (bottom)
·Fig 2. Composite score of microstructural sex differences from the various parameters, keeping only significant differences with a d greater than 0.8 (2 points) or between 0.5-0.8 (1 point)
Provide references using author date format
2. Hyde JS. The gender similarities hypothesis. Am Psychol. 2005;60(6):581–92.
3. Giudice MD, Booth T, Irwing P. The Distance Between Mars and Venus: Measuring Global Sex Differences in Personality. PLOS ONE. 2012 Jan 4;7(1):e29265.
4. Van Essen DC, Smith SM, Barch DM, Behrens TEJ, Yacoub E, Ugurbil K, et al. The WU-Minn Human Connectome Project: an overview. NeuroImage. 2013 Oct 15;80:62–79.
5. Avants BB, Epstein CL, Grossman M, Gee JC. Symmetric diffeomorphic image registration with cross-correlation: evaluating automated labeling of elderly and neurodegenerative brain. Med Image Anal. 2008 Feb;12(1):26–41.
6. Descoteaux M, Angelino E, Fitzgibbons S, Deriche R. Regularized, fast, and robust analytical Q-ball imaging. Magn Reson Med. 2007 Sep;58(3):497–510.
7. Koch MA, Norris DG, Hund-Georgiadis M. An Investigation of Functional and Anatomical Connectivity Using Magnetic Resonance Imaging. NeuroImage. 2002 May 1;16(1):241–50.
8. Guevara P, Poupon C, Rivière D, Cointepas Y, Marrakchi-Kacem L, Descoteaux M, et al. Inference of a HARDI fiber bundle atlas using a two-level clustering strategy. Med Image Comput Comput-Assist Interv MICCAI Int Conf Med Image Comput Comput-Assist Interv. 2010 Jan 1;13:550–7.
9. Campello RJGB, Moulavi D, Sander J. Density-Based Clustering Based on Hierarchical Density Estimates. In: Pei J, Tseng VS, Cao L, Motoda H, Xu G, editors. Advances in Knowledge Discovery and Data Mining. Berlin, Heidelberg: Springer; 2013. p. 160–72. (Lecture Notes in Computer Science).
10. Zhang H, Schneider T, Wheeler-Kingshott CA, Alexander DC. NODDI: practical in vivo neurite orientation dispersion and density imaging of the human brain. Vol. 61, Neuroimage. 2012. p. 1000–16.