2775
Diffusion magnetic resonance imaging in the human hippocampal subfields using super-resolution HYDI1Department of Radiology and Imaging Sciences, Indiana University School of Medicine, Indiana University, Indianapolis, IN, United States, 2Indiana Alzheimer Disease Center, Indianapolis, IN, United States, 3University of Bordeaux, LaBRI, UMR 5800, PICTURA, F-33400 Talence, France, 4CNRS, LaBRI, UMR 5800, PICTURA, F-33400 Talence, France
Synopsis
The hippocampal atrophy is known to
be the most validated biomarker of Alzheimer’s disease. Accordingly, in this study, we develop a method to enable the
structural connectivity mapping through tractography of the hippocampal
subfields using super-resolution diffusion data.
Introduction
The hippocampal atrophy is known to be the most validated biomarker of Alzheimer’s disease 1,2. Over the last few years, high-resolution anatomical magnetic resonance imaging (MRI) such as T1-weighted (T1W) or T2-weighted (T2W) imaging enabled the feasibility of submillimeter in-plane resolution of the hippocampus, which has allowed the automatic segmentation of hippocampal subfields 3,4. Using anatomical MRI, most of the previous studies focused on the volumetric changes in the hippocampal subfields. However, changes to specific hippocampal fiber pathways within the brain’s structural networks have not been closely investigated, largely due to the limited spatial resolution in diffusion MRI. With an adequate spatial resolution, diffusion MRI could provide unbiased microstructural measurements and the associated tractography for the in-vivo human hippocampal subfields. Accordingly, we developed a sub-millimeter super-resolution hybrid diffusion imaging (HYDI) 5,6 technique using the well-established collaborative patch-based super-resolution method 7. This enables the characterization of the structural connectivity of the hippocampal subfields and the delineation of the fiber tracts towards the cortex.Methods
MR acquisition
High-resolution T1W images were acquired with the MPRAGE sequence at 0.8 x 0.8 x 0.8 mm3 resolution with TR= 2400 ms, TE= 2.22 ms, TI=1000 ms, flip-angle=8°, 256 mm field-of-view, 208 slices, and GRAPPA acceleration iPAT= 2. In addition, we acquired a high-resolution Turbo-Spin-Echo T2W images with high in-plane resolution of 0.4 x 0.4 mm in the oblique plane perpendicular to the main axis of the hippocampus, a slice thickness of 2 mm, TR= 8310 ms, TE= 50 ms, flip-angle=122°, 175 mm field-of-view, 32 slices, and GRAPPA acceleration iPAT= 2.
HYDI 5,6 was performed on a healthy volunteer. The diffusion images were acquired on a Siemens Prisma scanner using a single-shot spin-echo EPI with a multiband factor of 3, TE=74.2 ms, and TR=4164 ms, 220 mm field of view, 114 slices, isotropic resolution of 1.25mm, and 10:51 (min:sec) acquisition time. A four-shell diffusion imaging with monopolar diffusion scheme was used with b-values 500, 800, 1600, 2600 s/mm2, 134 diffusion directions, and 8 non-diffusion weighted volumes. Two sets of data were acquired with reversed phase-encode blips.
Preprocessing
All the DW images were denoised from Rician noise using overcomplete local Principal Component Analysis as proposed in 8. FSL-topup and FSL-eddy 9, part of the FSL package version 5.0.11 (FMRIB, Oxford, UK), were used to correct motion, susceptibility and eddy current distortions. To achieve super resolution, a collaborative patch-based method 7 was then used to upsample the 1.25-cubic-mm resolution HYDI data to a submillimeter 0.625-mm-cubic resolution.
Post-processing
The high-resolution T1W and T2W acquisitions were used for segmenting the hippocampal subfields with the Automatic Segmentation of Hippocampal Subfields (ASHS) method 4 (Figure 1). HYDI data was used to compute DTI metrics and compartment modeling (results not shown). The hippocampal subfields were transformed from the T2W imaging space to the diffusion space using ANTs 10. In addition, DSI studio 11 was used to process the HYDI multishell data for tractography with a generalized q-sampling imaging (GQI) approach 12.
Results and Discussion
The zoomed-in box in Figure 2a demonstrates DTI eigenvectors in each hippocampal subfield. The means and standard deviations of DTI fractional anisotropy (FA) and mean diffusivity (MD) for each hippocampal subfield are shown in Figure 2b. While sulcus, a cell-free region, had lowest FA and highest MD, CA1, CA2, and subiculum had relatively high FA and low MD. White-matter connection as in tractography streamlines from the hippocampus to cortical gray matter are shown in Figure 3. It appears that the hippocampus has a direct connection to the temporal lobe, cingulum gyrus, and fornix. Tractography was performed between the hippocampal subfields to generating subfield connectivity (Figure 4). The strongest connections (as in streamline count) appear to be between the Brodmann areas 35 and 36 and between the CA1 and Subiculum. Similar information can be appreciated using the count connectogram generated using 13 (Figure 5).Conclusion
We presented a feasibility study of diffusion analyses and tractography within the hippocampal subfields using super-resolution HYDI. Future directions are to refine the pipelines for clinical applications in the aging brain and Alzheimer’s disease.Acknowledgements
The work is supported by grant NIH NIA R01 AG053993.References
1. Dubois, B. et al., Research criteria for the diagnosis of Alzheimer’s disease: revising the NINCDS–ADRDA criteria. Lancet Neurol, 734–746 (2007).
2. Frisoni, G., Fox, N., Jack, C., Sheltens, P. & Thompson, P., The clinical use of structural MRI in Alzheimer disease. Nat. Rev. Neurol., 67–77 (2010).
3. Van Leemput, K. et al., Automated segmentation of hippocampal subfields from ultra‐high resolution in vivo MRI. Hippocampus 19, 549-557 (2009).
4. Yushkevich, P. et al., Automated volumetry and regional thickness analysis of hippocampal subfields and medial temporal cortical structures in mild cognitive impairment. Hum Brain Mapp. 36, 258-287 (2015).
5. Wu, Y.-C. & Alexander, A. L., Hybrid diffusion imaging. NeuroImage 36, 617-629 (2007).
6. Wu, Y.-C., Field, A. S. & Alexander, A. L., Computation of diffusion function measures in q-space using magnetic resonance hybrid diffusion imaging. IEEE Trans Med Imaging 27 (6), 858-865 (2008).
7. Coupé, P., Manjón, J., Chamberland, M., Descoteaux, M. & Hiba, B., Collaborative patch-based super-resolution for diffusion-weighted images. NeuroImage 83, 245-261 (2013).
8. Manjón, J., Coupé, P., Concha, L., Buades, A. & Collins, D., Diffusion weighted image denoising using overcomplete local PCA. PLoS ONE 8 (9) (2013).
9. Andersson, J., Graham, M., Zsoldos, E. & Sotiropoulos, S., Incorporating outlier detection and replacement into a non-parametric framework for movement and distortion correction of diffusion MR images. NeuroImage 141, 556-572 (2016).
10. Avants, B. et al., A reproducible evaluation of ANTs similarity metric performance in brain image registration. Neuroimage 54, 2033-44 (2011).
11. http://dsi-studio.labsolver.org.
12. Yeh, F., Wedeen, V. & Tseng, W., Generalized q-sampling imaging. IEEE Trans Med Imaging 29, 1626-1635 (2010).
13. Krzywinski, M. et al., Circos: an Information Aesthetic for Comparative Genomics. Genome Res. 19, 1639-1645 (2009).
Figures
