The Human Connectome Project and beyond: initial applications of 300 mT/m gradients

Abstract

The engineering of a 3 T human MRI scanner equipped with 300 mT/m gradients - the strongest gradients ever built for an in vivo human MRI scanner - was a major component of the NIH Blueprint Human Connectome Project (HCP). This effort was motivated by the HCP's goal of mapping, as completely as possible, the macroscopic structural connections of the in vivo healthy, adult human brain using diffusion tractography. Yet, the 300 mT/m gradient system is well suited to many additional types of diffusion measurements. Here, we present three initial applications of the 300 mT/m gradients that fall outside the immediate scope of the HCP. These include: 1) diffusion tractography to study the anatomy of consciousness and the mechanisms of brain recovery following traumatic coma; 2) q-space measurements of axon diameter distributions in the in vivo human brain and 3) postmortem diffusion tractography as an adjunct to standard histopathological analysis. We show that the improved sensitivity and diffusion-resolution provided by the gradients are rapidly enabling human applications of techniques that were previously possible only for in vitro and animal models on small-bore scanners, thereby creating novel opportunities to map the microstructure of the human brain in health and disease.

Keywords: Axon diameter; Consciousness; Corpus callosum; Diffusion MRI; Human connectome; In vivo; Postmortem; Tractography; Traumatic coma.

Copyright © 2013 Elsevier Inc. All rights reserved.

Figures

FIG. 1

Regions of interest and deterministic tractography results for ARAS connectivity analysis in a patient who recovered from traumatic coma. a) Zoomed axial view of a susceptibility-weighted image at the level of the traumatic microhemorrhage (red arrow) in the dorsolateral midbrain. The hemorrhage is seen overlapping with the left pedunculopontine nucleus (PPN) at this level (turquoise outlines). b) Posterior view of regions of interest (ROIs) for the PPN (turquoise), and the thalamic target nuclei: centromedian/parafascicular nucleus (CEM/Pf, pink), central lateral nucleus (CL, blue) and the reticular nucleus (Ret, purple). All ROIs are superimposed on an axial diffusion-weighted image (DWI) at the level of the caudal midbrain and a coronal DWI at the level of the mid-thalamus. The microhemorrhage shown in panel a) is color-coded red and indicated by the red arrow. This hemorrhage involves the dorsal aspect of the left PPN nucleus. c) Inferior view of deterministic streamlines passing through the PPN at the level of the caudal midbrain. Streamlines are color-coded according to the thalamic nucleus with which they connect: pink, PPN-CEM/Pf; blue, PPN-CL; purple, PPN-Ret. The microhemorrhage is again shown in red and indicated by the red arrow. d) Zoomed posterior view from panel b) demonstrating streamlines connecting PPN to CEM/Pf, CL, and Ret. The hemorrhage is indicated by the red arrow. In panels c) and d), fewer streamlines are seen passing through the left PPN, particularly in the region of the hemorrhage, as compared to the right PPN. Neuroanatomic landmarks: 3V, third ventricle; Aq, cerebral aqueduct; CP, cerebral peduncle; Hippo, hippocampus.

FIG. 2

Probabilistic tractography results for the four control subjects and the patient who recovered from traumatic coma. Probabilistic data are superimposed on a coronal greyscale fractional anisotropy map at the level of the mid-thalamus. Within each voxel, the primary diffusion vector is shown in red: medial-lateral vectors are seen in the body of the corpus callosum at the superior margin of each image, and superior-inferior vectors are seen in the posterior limbs of the internal capsules at the lateral margins of each image. Probabilistic data for each analysis of pedunculopontine nucleus (PPN) connectivity are color-coded according to the thalamic nucleus with which the streamlines are connecting: red, centromedian/parafascicular nucleus; blue, central lateral nucleus; yellow, reticular nucleus. The intensity of the color in each voxel represents the number of streamlines passing through that voxel (i.e. brighter voxels represent a higher streamline count than darker voxels). For the purpose of clarity and to exclude streamlines with low probability, streamline results are thresholded in eachanalysis to show voxels with between 1000 and 10,000 streamlines passing through them.

FIG. 3

A comparison of probabilistic streamline connectivity index (PSCI) measurements for the pathways in the four healthy volunteers (C1-C4) and one patient (P) who recovered from traumatic coma. Light colors represent the left hemisphere. Darker colors represent the right hemisphere. The left-right difference for the PPN-Ret pathway in the patient was determined with a t-test to be statistically significantly different from the left-right differences in the controls (p-value=0.0031). Each pathway was seeded in the pedunculopontine nucleus (PPN). The target ROIs were the following thalamic nuclei: centromedian/parafascicular nucleus (CEM/Pf), central lateral nucleus (CL) and reticular nucleus (Ret).

FIG. 4

Sagittal diffusion-weighted images from one healthy volunteer demonstrate the quality of images used in the AxCaliber analysis. Here we show 18 different b-values, all acquired with δ = 15.3 ms and Δ=46.3 ms. Each image represents 12 averages.

FIG. 5

Mean (n=4) signal decay curves for ROIs drawn in the a) genu, b) body and c) splenium of the corpus callosum. Green, magenta, white, red and blue curves represent 123 ms, 87 ms, 57 ms, 47 ms and 38 ms diffusion times. Error bars represent the standard deviation across 4 subjects.d) Axon diameter distributions for the genu (red), body (blue) and splenium (yellow) found from the data shown in a),b) and c).

FIG. 6

a-d) ROI delineation (top) and axon diameter distributions (bottom) for subjects 1-4.

FIG. 7

Pixel-wise estimates in the corpus callosum of subjects 1-4 (left-to-right) for axon diameter (a-d) and axonal density (e-h). The pixel-wise axon diameter value represents the peak of the estimated gamma distribution. The axonal density represents the restricted fraction (fr) estimated in the AxCaliber signal model.

FIG. 8

Ex vivo diffusion imaging of the human brainstem; a) Ventral view of a dissected brainstem-diencephalon specimen (consisting of pons, midbrain, thalamus, hypothalamus, and basal forebrain) from a 53-year-old woman who died of non-neurological causes. b) Right lateral view of a whole-brain specimen from a 58-year-old woman who died of non-neurological causes. c) Sagittal view of the directionally-encoded color(DEC) map for the specimen in a). d) Sagittal view of the DEC map for the specimen in b). e)Axial view of the DEC map for the specimen in a) at the level of the rostral pos. f)Axial view of the DEC map for the specimen in b) at the level of the rostral pons. The crosshairs are located in the same voxel in c) and e), as well as in d) and f). The diffusion sequence parameters that were used in each analysis are summarized at the bottom of the figure. The connectome analysis of a whole-brain specimen provides equal spatial resolution (0.6 mm isotropic) and superior angular resolution (i.e. higher b value with similar number of diffusion directions), without sacrificing the signal-to-noise properties and without having to dissect a brain specimen for scanning on a small-bore scanner. Neuroanatomic landmarks: CST, corticospinal tract; MCP, middle cerebellar peduncle; PC, pontine crossing fibers; SCP, superior cerebellar peduncle.

FIG. 9

Orientation distribution function (ODF) analysis of voxels within the cuneiform/subcuneiform nucleus(CSC; mesencephalic reticular formation) in the dorsolateral aspect of the rostral midbrain. (a) The CSC region of interest is outlined on an axial greyscale fractional anisotropy map. (b) Zoomed view of ODFs within the 5 × 5 voxel CSC region of interest. Voxel size is 0.6 mm in each plane, and the ODFs are color-coded according to the direction of diffusion: blue, superior-inferior; red, medial-lateral; green, anterior-posterior. The ODFs reveal at least three different directional vectors of water diffusion within the CSC. Although the precise neuroanatomic trajectories and targets of each vector cannot be definitively determined, known connectivity from animal and human studies suggests that the blue superior-inferior pathways are ascending to the thalamus via the medial and lateral dorsal tegmental tracts (DTTM and DTTL), the green anterior-posterior pathways are projecting ventrally to the hypothalamus via the caudal ventral tegmental tract (VTTC), and the red medial-lateral pathways are projecting across the posterior commissure to the contralateral midbrain tegmentum. Nomenclature and patterns of connectivity for DTTM, DTTL, and VTTC were first proposed by Shute and Lewis in rats [Shute and Lewis, 1963] and recently refined by Edlow et al. in the human brain [Edlow et al., 2012]. Neuroanatomic landmarks: Aq, cerebral aqueduct; CP, cerebral peduncle; ML, medial lemniscus; MLF, medial longitudinal fasciculus; PAG, periaqueductal grey matter; RN, red nucleus; SC, superior colliculus; SN, substantia nigra.

FIG. 10

Orientation distribution function (ODF) analysis of voxels within the basis pontis of the brainstem. (a) A 5 × 5 voxel region of interest is outlined on an axial non-diffusion-weighted (b0) image at the level of the rostral pons. (b) Zoomed view of ODFs within the basis pontis region of interest. Voxel size is 0.6 mm in each plane, and the ODFs are color-coded according to the direction of diffusion: blue, superior-inferior; red, medial-lateral; green, anterior-posterior. The ODFs reveal multiple directional vectors of water diffusion within the basis pontis. Although the precise neuroanatomic trajectories and targets of each vector cannot be definitively determined, known connectivity from animal and human studies suggests that the blue superior-inferior pathways are ascending fibers of the corticospinal tracts (CST) and the red medial-lateral pathways are pontocerebellar crossing fibers (PCF). The green anterior-posterior pathways are of uncertain connectivity. Neuroanatomic landmarks: MCP, middle cerebellar peduncle; ML, medial lemniscus; MLF, medial longitudinal fasciculus; SCP, superior cerebellar peduncle.