Elsevier

NeuroImage

Volume 74, 1 July 2013, Pages 254-265
NeuroImage

A novel in vivo atlas of human hippocampal subfields using high-resolution 3 T magnetic resonance imaging

https://doi.org/10.1016/j.neuroimage.2013.02.003Get rights and content

Abstract

The hippocampus is a neuroanatomical structure that has been widely studied in the context of learning, memory, stress, and neurodegeneration. Neuroanatomically, the hippocampus is subdivided into several subfields with intricate morphologies and complex three-dimensional relationships. Recent studies have demonstrated that the identification of different subfields is possible with high-resolution and -contrast image volumes acquired using ex vivo specimens in a small bore 9.4 T scanner and, more recently, in vivo, at 7 T. In these studies, the neuroanatomical definitions of boundaries between subfields are based upon salient differences in image contrast. Typically, the definition of subfields has not been possible using commonly available magnetic resonance (MR) scanners (i.e.: 1.5 or 3 T) due to resolution and contrast limitations. To overcome the limited availability of post-mortem specimens and expertise in state-of-the-art high-field imaging, we propose a coupling of MR acquisition and detailed segmentation techniques that allow for the reliable identification of hippocampal anatomy (including subfields). High-resolution and -contrast T1- and T2-weighted image volumes were acquired from 5 volunteers (2 male; 3 female; age range: 29–57, avg. 37) using a clinical research-grade 3 T scanner and have final super-sampled isotropic voxel dimensions of 0.3 mm. We demonstrate that by using these acquisition techniques, our data results in contrast-to-noise ratios that compare well with high-resolution images acquired with long scan times using post-mortem data at higher field strengths. For the subfields, the cornus ammonis (CA) 1, CA2/CA3, CA4/dentate gyrus, stratum radiatum/stratum lacunosum/stratum moleculare, and subiculum were all labeled as separate structures. Hippocampal volumes are reported for each of the substructures and the hippocampus as a whole (range for hippocampus: 2456.72–3325.02 mm3). Intra-rater reliability of our manual segmentation protocol demonstrates high reliability for the whole hippocampus (mean Dice Kappa of 0.91; range 0.90–0.92) and for each of the subfields (range of Dice Kappas: 0.64–0.83). We demonstrate that our reliability is better than the Dice Kappas produced by simulating the following errors: a translation by a single voxel in all cardinal directions and 1% volumetric shrinkage and expansion. The completed hippocampal atlases are available freely online (info2.camh.net/kf-tigr/index.php/Hippocampus) and can be coupled with novel computational neuroanatomy techniques that will allow for them to be customized to the unique neuroanatomy of different subjects, and ultimately be utilized in different analysis pipelines.

Graphical abstract

Manually segmented hippocampal subfields (CA1, CA2/CA3, CA4/DG, strata radiatum/lacunosum/moleculare, subiculum) on high-resolution MR images acquired on a 3 T scanner.

Segmentation protocol performs very well in reliability assessments.

Atlases available publicly online.

  1. Download : Download high-res image (128KB)
  2. Download : Download full-size image

Highlights

► High-resolution in vivo magnetic resonance images were acquired on a 3T scanner. ► Hippocampal subfields (CA1, CA2/CA3, CA4/DG, SR/SL/SM, subiculum) manually segmented. ► Segmentation protocol performs very well in reliability assessments. ► Full atlases available for download at info2.camh.net/kf-tigr/index.php/Hippocampus.

Introduction

The hippocampus is a brain structure of particular interest to several branches of neuroscientific, neurological, and psychiatric research. It is located in the medial temporal lobe, and is mainly associated with episodic memory and spatial reasoning (Milner, 1958, Scoville and Milner, 1957). The relationship between the hippocampus and different forms of cognitive abilities and brain dysfunction is too extensive to review comprehensively in this manuscript; however, well-known structural deficits in the hippocampus have been connected to several specific forms of brain dysfunction, such as Alzheimer's disease (Braak and Braak, 1991, Mouiha and Duchesne, 2011), schizophrenia (Altshuler et al., 1998, Bogerts et al., 1993, Heckers, 2001), and depression (Bremner et al., 2000, Campbell et al., 2004, Vakili et al., 2000). Therefore a high-resolution, accessible hippocampal atlas has the potential to be of significant value to a broad multidisciplinary community.

Classically, print atlases have been used to aid in the identification of hippocampal structure (Duvernoy, 2005, Mai et al., 2008, Talairach and Tournoux, 1988). The MR community represents the average hippocampal neuroanatomy of a population through the creation of MR-based atlases (Frey et al., 2011, Grabner et al., 2006, Mazziotta et al., 2001a, Mazziotta et al., 2001b, Talairach and Tournoux, 1988, Yushkevich et al., 2010). Such atlases are typically derived using group-wise registration techniques that capture both the neuroanatomical variability and commonalities in a group of subjects (Chakravarty et al., 2009). In order to obtain enhanced visualization not offered by MRI, some groups have developed methods for reconstructing and warping histological data in order to visualize both datasets interchangeably (Adler et al., 2012, Chakravarty et al., 2006, Toga et al., 2006, Yelnik et al., 2007).

The inter-subject morphometric variability in hippocampal shape and size and the partial volume, contrast, and resolution constraints of typical volumetric MR imaging protocols have led to the development of several manual (Frisoni and Jack, 2011, Mueller and Weiner, 2009, Pruessner et al., 2002, Van Leemput et al., 2009, Wisse et al., 2012, Yushkevich et al., 2009) and automated methods (Collins and Pruessner, 2010, Hammers et al., 2007, Leung et al., 2010, Van Leemput et al., 2009, Yushkevich et al., 2010) to facilitate identification of this structure. Typically, in order to properly achieve these goals, hippocampal neuroanatomy is greatly simplified, with little or no information regarding the subfields included (Collins and Pruessner, 2010).

Recent projects have utilized high-field MR technologies, post-mortem data, and long scan times to move past these more elementary definitions and successfully visualize and localize the hippocampal subfields (La Joie et al., 2010, Mueller et al., 2006, Van Leemput et al., 2009, Wisse et al., 2012, Yushkevich et al., 2009). Mueller and colleagues have demonstrated that hippocampal subfields can be identified on in vivo MR images acquired on a 4 T scanner with a resolution of 0.4 mm × 0.5 mm × 2 mm, while Van Leemput presents subfield-level segmentation using in vivo 3 T images with a 0.38 mm × 0.38 mm × 0.80 mm resolution (Mueller et al., 2006, Van Leemput et al., 2009). Neither, however, provides a segmentation protocol that includes the complete anterior–posterior hippocampal axis; Mueller segments only a few coronal slices, and Van Leemput excludes the hippocampal tail. Yushkevich and colleagues present 0.2 mm isotropic post-mortem images that offer greater anatomical detail, but rely on a small bore 9.4 T scanner and fixed medial temporal lobe specimens with a 15- or 63-hour scan time. Their work has been released to the public; however the applications of a protocol of this kind are limited given the lack of availability of post-mortem specimens, and the atlases are limited by the complexity of registering fixed images to in vivo data. Wisse et al. (2012) have recently attempted to overcome the latter limitation by scanning subjects in vivo at 7 T. Their T1- and T2-weighted data were acquired with significantly lower scan times, and their final image volumes have isotropic voxel dimensions of 0.35 mm. They have not released their atlases, however, and their technique necessitates access to a 7 T MR imaging system, which is not currently available at most academic research centers.

The purpose of this manuscript is threefold: first, to demonstrate that high-resolution and high-contrast MR data that allows for the reliable identification of the hippocampus and its subfields can be acquired on a clinical-grade 3 T scanner; second to describe in detail a protocol for segmenting the whole hippocampus and its subfields on data of this kind; and third, to make the complete atlas set available freely online for use by the neuroscientific community.

Each of the five complete atlases we present has final super-sampled isotropic voxel dimensions of 0.3 mm. Contrast in Signal intensity used to delineate the CA1, CA2/CA3, CA4/DG, stratum radiatum/stratum lacunosum/stratum moleculare, and subiculum of the hippocampal formation bilaterally in all five subjects, and the reliability of the protocol was assessed using a test–retest design along with multiple simulations that verified the accuracy and precision of our protocol. To assess the quality of our images, the contrast-to-noise ratio of specific subsections of the hippocampus was evaluated and compared to homologous regions in a hippocampal atlas derived from post-mortem data (Yushkevich et al., 2009). In this manuscript, we fully describe the protocol and include descriptions of the morphological features that were used to identify different subfields. All of these image volumes and their associated labels are publicly available in both MINC and NIfTI file formats (info2.camh.net/kf-tigr/index.php/Hippocampus).

Section snippets

Image acquisition

Subjects were recruited at the Centre for Addiction and Mental Health (CAMH) in Toronto, Canada. They were healthy volunteers, free of neurological and neuropsychiatric disorders and cases of severe head trauma (2 males and 3 females, ages 29–57, avg. 37). The study was conducted in keeping with the Declaration of Helsinki, was approved by the Centre for Addiction and Mental Health Research Ethics Board, and all subjects provided written, informed consent for data acquisition and sharing.

T1-

Results

Table 1 summarizes the results of our contrast-to-noise ratio (CNR) analyses. A comparison of the gray/white matter CNR of our T1- and T2-weighted data demonstrates that the T2-weighted image volumes have better contrast between CA1 gray matter and the adjacent temporal lobe white matter (T1: mean CNR = 5.05, range = 3.67–6.63; T2: mean CNR = 6.64, range = 5.38–7.51). The advantage of averaging the three single acquisitions is clear by comparing the CNR of a single acquisition (mean CNR = 4.45, range = 

Discussion

The main contribution of this manuscript is the development of five atlases of the human hippocampus derived from high-resolution T1- and T2-weighted in vivo MRI data acquired on a research-grade 3 T scanner. After super-sampling, rigid body alignment of the three scans, and voxel-by-voxel averaging, the final MR volumes have 0.3 mm isotropic voxel dimensions. We demonstrated that these final volumes provide comparable contrast-to-noise ratios compared to post-mortem data acquired on a small bore

Acknowledgements

We wish to thank Anusha Ravichandran for help with image acquisition, and we acknowledge support from the CAMH Foundation, thanks to Michael and Sonja Koerner, the Kimel Family, and the Paul E. Garfinkel New Investigator Catalyst Award. MMC is funded by the W. Garfield Weston Foundation and ANV is funded by the Canadian Institutes of Health Research, Ontario Mental Health Foundation, NARSAD, and the National Institute of Mental Health (R01MH099167).

References (53)

  • R. La Joie et al.

    Differential effect of age on hippocampal subfields assessed using a new high-resolution 3T MR sequence

    Neuroimage

    (2010)
  • K.K. Leung et al.

    Automated cross-sectional and longitudinal hippocampal volume measurement in mild cognitive impairment and Alzheimer's disease

    NeuroImage

    (2010)
  • A. Mouiha et al.

    Hippocampal atrophy rates in Alzheimer's disease: automated segmentation variability analysis

    Neurosci. Lett.

    (2011)
  • N. Schuff et al.

    Age-related metabolite changes and volume loss in the hippocampus by magnetic resonance spectroscopy and imaging

    Neurobiol. Aging

    (1999)
  • K. Vakili et al.

    Hippocampal volume in primary unipolar major depression: a magnetic resonance imaging study

    Biol. Psychiatry

    (2000)
  • L.E.M. Wisse et al.

    Subfields of the hippocampal formation at 7 T MRI: in vivo volumetric assessment

    NeuroImage

    (2012)
  • N.S. Yeghiazaryan et al.

    Temporal lobe epilepsy and hippocampal malrotation: is there a causal association?

    Epilepsy Behav.

    (2010)
  • J. Yelnik et al.

    A three-dimensional, histological and deformable atlas of the human basal ganglia. I. Atlas construction based on immunohistochemical and MRI data

    NeuroImage

    (2007)
  • P.A. Yushkevich et al.

    A high-resolution computational atlas of the human hippocampus from postmortem magnetic resonance imaging at 9.4 T

    NeuroImage

    (2009)
  • P.A. Yushkevich et al.

    Nearly automatic segmentation of hippocampal subfields in in vivo focal T2-weighted MRI

    NeuroImage

    (2010)
  • D.H. Adler et al.

    Reconstruction of the human hippocampus in 3D from histology and high-resolution ex-vivo MRI

    ISBI

    (2012)
  • L.L. Altshuler et al.

    Amygdala enlargement in bipolar disorder and hippocampal reduction in schizophrenia: an MRI study demonstrating neuroanatomic specificity

    Arch. Gen. Psychiatry

    (1998)
  • M. Boccardi et al.

    Survey of protocols for the manual segmentation of the hippocampus: preparatory steps towards a joint EADC-ADNI harmonized protocol

    J. Alzheimer's Dis.

    (2011)
  • Boccardi, M., Bocchetta, M., Ganzola, R., Robitaille, N., Redolfi, A., Duchesne, S., Jack, C. J., Frisoni, G. B., in...
  • H. Braak et al.

    Neuropathological stageing of Alzheimer-related changes

    Acta Neuropathol.

    (1991)
  • J.D. Bremner et al.

    Hippocampal volume reduction in major depression

    Am. J. Psychiatry

    (2000)
  • Cited by (201)

    • Perinatal maternal mental health and amygdala morphology in young adulthood

      2023, Progress in Neuro-Psychopharmacology and Biological Psychiatry
    View all citing articles on Scopus
    View full text