Skip to main content
Advertisement

Main menu

  • Home
  • Content
    • Current Issue
    • Publication Preview--Ahead of Print
    • Past Issue Archive
    • Case of the Week Archive
    • Classic Case Archive
    • Case of the Month Archive
    • COVID-19 Content and Resources
  • For Authors
  • About Us
    • About AJNR
    • Editors
    • American Society of Neuroradiology
  • Submit a Manuscript
  • Podcasts
    • Subscribe on iTunes
    • Subscribe on Stitcher
  • More
    • Subscribers
    • Permissions
    • Advertisers
    • Alerts
    • Feedback
  • Other Publications
    • ajnr

User menu

  • Subscribe
  • Alerts
  • Log in
  • Log out

Search

  • Advanced search
American Journal of Neuroradiology
American Journal of Neuroradiology

American Journal of Neuroradiology

  • Subscribe
  • Alerts
  • Log in
  • Log out

Advanced Search

  • Home
  • Content
    • Current Issue
    • Publication Preview--Ahead of Print
    • Past Issue Archive
    • Case of the Week Archive
    • Classic Case Archive
    • Case of the Month Archive
    • COVID-19 Content and Resources
  • For Authors
  • About Us
    • About AJNR
    • Editors
    • American Society of Neuroradiology
  • Submit a Manuscript
  • Podcasts
    • Subscribe on iTunes
    • Subscribe on Stitcher
  • More
    • Subscribers
    • Permissions
    • Advertisers
    • Alerts
    • Feedback
  • Follow AJNR on Twitter
  • Visit AJNR on Facebook
  • Follow AJNR on Instagram
  • Join AJNR on LinkedIn
  • RSS Feeds
OtherBRAIN
Open Access

Fully-Automated Quantification of Regional Brain Volumes for Improved Detection of Focal Atrophy in Alzheimer Disease

J.B. Brewer, S. Magda, C. Airriess and M.E. Smith
American Journal of Neuroradiology March 2009, 30 (3) 578-580; DOI: https://doi.org/10.3174/ajnr.A1402
J.B. Brewer
  • Find this author on Google Scholar
  • Find this author on PubMed
  • Search for this author on this site
S. Magda
  • Find this author on Google Scholar
  • Find this author on PubMed
  • Search for this author on this site
C. Airriess
  • Find this author on Google Scholar
  • Find this author on PubMed
  • Search for this author on this site
M.E. Smith
  • Find this author on Google Scholar
  • Find this author on PubMed
  • Search for this author on this site
  • Article
  • Figures & Data
  • Supplemental
  • Info & Metrics
  • References
  • PDF
Loading

Abstract

SUMMARY: Volumetric analysis of structural MR images of the brain may provide quantitative evidence of neurodegeneration and help identify patients at risk for rapid clinical deterioration. This note describes tests of a fully automated MR imaging postprocessing system for volumetric analysis of structures (such as the hippocampus) known to be affected in early Alzheimer disease (AD). The system yielded results that correlated highly with independent computer-aided manual segmentation and were sensitive to the anatomic atrophy characteristic of mild AD.

It is well established that Alzheimer disease (AD) and related neurodegenerative disorders are associated with atrophy in the brain that can be reliably detected with quantitative analysis of MR images. Consistent with histologic findings,1 quantification of structural MR images can reveal significant atrophy in medial temporal lobe structures.2 Such loss of tissue is typically accompanied by increases in ventricular volume, particularly in the inferior portion of the lateral ventricle. Quantitative MR imaging may also aid in the detection of AD-related neuropathologic disorders at the prodromal stage.3 Patients with moderately poor performance on 1 or more standardized tests of cognitive function are often classified as having mild cognitive impairment (MCI).4 When memory is one of the cognitive domains involved, individuals are at increased risk of converting to AD. Medial temporal atrophy has been found to be predictive of whether a patient with MCI will show rapid clinical decline or remain stable.5,6

Such findings imply a need for improved tools for MR imaging volumetric analysis that are suitable for routine clinical use. Furthermore, although many efforts are underway to develop treatments that could prevent AD or slow its progression, these efforts are also hindered in part by the lack of practical yet sensitive in vivo biomarkers that may help to demonstrate disease-modifying effects of potential treatments7 and to identify those patients most likely to benefit from aggressive intervention. Both of these needs might be better addressed by efficient methods for the quantitative measurement of atrophy in the living brain. This technical note describes a practical, fully automated method to derive regional brain volumes. The system's validity with respect to results obtained with conventional manual segmentation is examined, as is its sensitivity to the neurodegenerative characteristic of AD.

Technique

No human subjects were recruited specifically for this study; data included MR images from a group of elderly subjects (n=40) randomly selected from the Open Access Series of Imaging Studies8 but balanced for sex, age, and diagnosis. Half had been diagnosed with mild probable AD (Clinical Dementia Rating [CDR] scale, 0.5 or 1), and half were healthy control (HC) subjects (CDR, 0). The average age was 77 years in each group.

For each subject, a set of T1-weighted magnetization-prepared rapid acquisition of gradient echo sequence anatomic images was available that was obtained with use of a 1.5T Vision system (Siemens, Erlangen, Germany). These images were processed by the NeuroQuant software package (CorTechs Labs, La Jolla, Calif), a tool that has received US Food and Drug Administration (FDA) 510K clearance for clinical use to measure volumes of brain structures in MR images. The algorithm used by this software package includes 1) a quality-checking step to determine that the MR imaging sequence conforms to the specifications required to perform automated segmentations, 2) correction for gradient nonlinearity9 and B1 field inhomogeneity,10 and 3) skull stripping followed by automated segmentation of anatomic structures of interest. An active contour, or “snake,” model is used for skull stripping. Discrete cosine transform nonlinear registration aligns volumes to a probabilistic atlas customized for labeling anatomy in elderly subjects. Output of the software includes numeric volumes and images that have been annotated with graphic color overlays, with each color representing a specific segmental structure (Fig 1).

Fig 1.
  • Download figure
  • Open in new tab
  • Download powerpoint
Fig 1.

Example coronal sections from automatically segmented MR images from a healthy brain (left) and a patient with AD (right). Arrows indicate medial temporal lobe structures. Colors indicate segmented structures as listed in the Table.

The automated segmentation methods used by the software build on widely used semiautomated methods11,12 and rely on probabilistic atlas-based methods to provide volumetric analysis of segmental structures (including those listed in the Table). The segmentation procedure assigns a neuroanatomic label to each voxel on the basis of probabilistic information automatically estimated from an atlas. The labeling of each point in space is achieved by finding the segmentation for each dataset that maximizes the probability of input given the previous probabilities from the atlas. The probability of a class at each point is iteratively computed as the probability that the given class appeared at that location in a training set, modulated by the probability of the surrounding configuration of labels in the 6 cardinal directions, times the likelihood of obtaining the subject-specific measured image intensity value from that class. The NeuroQuant processing procedure takes approximately 8 minutes to complete on a conventional desktop computer and does not require any user input, aside from selecting the study to be segmented. The software interacts directly with a PACS server or can be configured as a remotely hosted server accessed via a Web interface. A report is generated, which returns volumes of structures in cubic centimeters and percentage of intracranial volume. For the hippocampus, lateral ventricle, and temporal horn of the lateral ventricle, a normative range based on previous segmentations of healthy subjects aged 50 to 95 years is provided for reference purposes. An example of full-volume segmentation and 2 example reports are provided as supplementary material.

View this table:
  • View inline
  • View popup

Manual vs automated segmentation

For comparison purposes, segmentation was also performed independently by a third-party commercial entity (Neuromorphometrics, Somerville, Mass) that was blinded to the diagnosis of patients and to the results of automated segmentation. This entity was paid for its services but had no financial interest in the outcome of the comparison. This segmentation process was conducted with computer-assisted semimanual methods by using the open source software package NVM (http://neuromorphometrics.org), which includes features that let the user efficiently delineate and label anatomy in 3 dimensions. Contour drawing and editing were performed by an expert neuroanatomist who used structure boundary definitions as described by the Center for Morphometric Analysis (http://www.cma.mgh.harvard.edu/).

We completed statistical comparisons by using SPSS for Windows, Rel. 14. 2005 (SPSS, Chicago, Ill). To evaluate the accuracy of results produced by the automated process relative to segmentation performed in a computer-assisted manual fashion, we computed the intraclass correlation coefficient (ICC) by using Cronbach alpha for each segmented structure obtained (Table 1). ICCs ranged from a low of r = 0.61 for the nucleus accumbens region (1-tailed P < .002) to r = 0.99 for the lateral ventricle (1-tailed P < .001). A high level of agreement between methods was observed in medial temporal regions that are known to atrophy in the early stages of AD (inferior lateral ventricle ICC, 0.92; hippocampal ICC, 0.93; P < .001 for both). A very similar pattern of agreement between methods was observed with the Pearson product-moment correlation coefficient.

To evaluate sensitivity of the automated method to the neuropathologic characteristic of AD, we performed comparisons between volumetric measures in the AD and HC subsets of the overall sample, after correcting for differences in total intracranial volume (ICV). In particular, for each subject the volume of each labeled structure was converted to a percentage of total ICV by dividing it by the sum of the volumes for all labeled structures. Highly significant differences between the 2 groups were observed in medial temporal lobe regions that would be expected to be affected in early stage AD, including expansion of the inferior lateral ventricle (AD > HC, t(38) = 3.2; P < .003), and atrophy of the hippocampus (AD < HC, t(38) = −4.7; P < .001) and amygdala (AD < HC, t(38) = −4.4; P < .001). In contrast, no significant differences in volume were observed in structures that would not be expected to be affected in early AD such as the thalamus (t(38) = 0.64; P = .53, n.s.) or cerebellum (t(38) = 0.55; P = .58, n.s.).

Discussion

Several studies have shown that neurodegenerative disorders are reliably associated with patterns of progressive neural atrophy that can be identified with quantitative analysis of MR images. Volumetric analysis of regional brain structures may thus help characterize a patient's condition by providing objective quantitative evidence of a disease process. Such measures may also have prognostic use, helping to identify patients with the greatest risk for rapid clinical deterioration. However, existing methods to derive quantitative volumetric information from MR images have not been suitable for use outside of small-scale research studies.

The results described above indicate that a practical FDA-cleared medical device postacquisition MR imaging analysis system for whole-brain anatomic segmentation and for quantification of the volumes of brain structures known to be affected in AD yielded volumetric results that correlated highly with results obtained by independent, expert computer-aided manual anatomic segmentation, and were sensitive to differences in volume between age-matched healthy control subjects and patients with early AD. Normative data from younger subjects will help extend these results to other diseases, such as traumatic brain injury, in which measurement of focal brain atrophy may help predict clinical outcome.13 However, it should be noted that the presence of large hemorrhage or other gross pathologic processes would likely interfere with automated segmentation of adjacent regions. Nevertheless, the results suggest that a fully automated system accurately measures brain segmental volumes in patients with AD, and this procedure can be applied to T1-weighted volumes from common clinical platforms regardless of field strength. Technology for volumetric MR imaging measurements may now be sufficiently mature to incorporate objective metrics of brain atrophy into everyday clinical practice.

Footnotes

  • This study was supported by grants to M.E.S. from the US National Institute on Aging and the US National Institute on Neurological Disorders and Stroke.

  • Some results of this study previously presented at: International Conference on Alzheimer's Disease, July 26–31, 2008, Chicago, Ill.

References

  1. ↵
    Braak H, Braak E. Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol 1991;82:239–59
    CrossRefPubMed
  2. ↵
    Jack CR Jr, Petersen RC, Xu YC, et al. Medial temporal atrophy on MRI in normal aging and very mild Alzheimer's disease. Neurology 1997;49:786–94
    Abstract/FREE Full Text
  3. ↵
    Hampel H, Burger K, Teipel SJ, et al. Core candidate neurochemical and imaging biomarkers of Alzheimer's disease. Alzheimers Dement 2008;4:38–48
    CrossRefPubMed
  4. ↵
    Petersen RC. Mild cognitive impairment as a diagnostic entity. J Intern Med 2004;256:183–94
    CrossRefPubMed
  5. ↵
    Jack CR Jr, Shiung MM, Weigand SD, et al. Brain atrophy rates predict subsequent clinical conversion in normal elderly and amnestic MCI. Neurology 2005;65:1227–31
    Abstract/FREE Full Text
  6. ↵
    Devanand DP, Pradhaban G, Liu X, et al. Hippocampal and entorhinal atrophy in mild cognitive impairment: prediction of Alzheimer disease. Neurology 2007;68:828–36
    Abstract/FREE Full Text
  7. ↵
    Salloway S, Mintzer J, Weiner MF, et al. Disease-modifying therapies in Alzheimer's disease. Alzheimers Dement 2008;4:65–79
    CrossRefPubMed
  8. ↵
    Marcus DS, Wang TH, Parker J, et al. Open Access Series of Imaging Studies (OASIS): cross-sectional MRI data in young, middle aged, nondemented, and demented older adults. J Cogn Neurosci 2007;19:1498–507
    CrossRefPubMed
  9. ↵
    Jovicich J, Czanner S, Greve D, et al. Reliability in multi-site structural MRI studies: effects of gradient non-linearity correction on phantom and human data. Neuroimage 2006;30:436–43
    CrossRefPubMed
  10. ↵
    Sled JG, Zijdenbos AP, Evans AC. A nonparametric method for automatic correction of intensity nonuniformity in MRI data. IEEE Trans Med Imaging 1998;17:87–97
    CrossRefPubMed
  11. ↵
    Dale AM, Liu AK, Fischl BR, et al. Dynamic statistical parametric mapping: combining fMRI and MEG for high-resolution imaging of cortical activity. Neuron 2000;26:55–67
    CrossRefPubMed
  12. ↵
    Fischl BR, Salat DH, Busa E, et al. Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron 2002;33:341–55
    CrossRefPubMed
  13. ↵
    Adamson DM, Burnam MA, Burns RM, et al. Invisible Wounds of War: Psychological and Cognitive Injuries, Their Consequences, and Services to Assist Recovery. Santa Monica: RAND Corp.;2008
  • Received July 10, 2008.
  • Accepted after revision October 12, 2008.
  • Copyright © American Society of Neuroradiology
View Abstract
PreviousNext
Back to top

In this issue

American Journal of Neuroradiology: 30 (3)
American Journal of Neuroradiology
Vol. 30, Issue 3
March 2009
  • Table of Contents
  • Index by author
Advertisement
Print
Download PDF
Email Article

Thank you for your interest in spreading the word on American Journal of Neuroradiology.

NOTE: We only request your email address so that the person you are recommending the page to knows that you wanted them to see it, and that it is not junk mail. We do not capture any email address.

Enter multiple addresses on separate lines or separate them with commas.
Fully-Automated Quantification of Regional Brain Volumes for Improved Detection of Focal Atrophy in Alzheimer Disease
(Your Name) has sent you a message from American Journal of Neuroradiology
(Your Name) thought you would like to see the American Journal of Neuroradiology web site.
CAPTCHA
This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.
Citation Tools
Fully-Automated Quantification of Regional Brain Volumes for Improved Detection of Focal Atrophy in Alzheimer Disease
J.B. Brewer, S. Magda, C. Airriess, M.E. Smith
American Journal of Neuroradiology Mar 2009, 30 (3) 578-580; DOI: 10.3174/ajnr.A1402

Citation Manager Formats

  • BibTeX
  • Bookends
  • EasyBib
  • EndNote (tagged)
  • EndNote 8 (xml)
  • Medlars
  • Mendeley
  • Papers
  • RefWorks Tagged
  • Ref Manager
  • RIS
  • Zotero
Share
Fully-Automated Quantification of Regional Brain Volumes for Improved Detection of Focal Atrophy in Alzheimer Disease
J.B. Brewer, S. Magda, C. Airriess, M.E. Smith
American Journal of Neuroradiology Mar 2009, 30 (3) 578-580; DOI: 10.3174/ajnr.A1402
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo
  • Tweet Widget
  • Facebook Like
  • Google Plus One
Purchase

Jump to section

  • Article
    • Abstract
    • Technique
    • Discussion
    • Footnotes
    • References
  • Figures & Data
  • Supplemental
  • Info & Metrics
  • References
  • PDF

Related Articles

  • PubMed
  • Google Scholar

Cited By...

  • A Large-scale Comparison of Cortical and Subcortical Structural Segmentation Methods in Alzheimers Disease: a Statistical Approach
  • Predictive Utility of Marketed Volumetric Software Tools in Subjects at Risk for Alzheimer Disease: Do Regions Outside the Hippocampus Matter?
  • Automated brain volumetrics in multiple sclerosis: a step closer to clinical application
  • A randomized, double-blind, placebo-controlled trial of resveratrol for Alzheimer disease
  • Mesial Temporal Sclerosis: Accuracy of NeuroQuant versus Neuroradiologist
  • An Expanded Role for Neuroimaging in the Evaluation of Memory Impairment
  • Evaluation of Automatic Measurement of the Intracranial Volume Based on Quantitative MR Imaging
  • Predicting MCI outcome with clinically available MRI and CSF biomarkers
  • Crossref
  • Google Scholar

This article has not yet been cited by articles in journals that are participating in Crossref Cited-by Linking.

More in this TOC Section

  • Usefulness of Quantitative Susceptibility Mapping for the Diagnosis of Parkinson Disease
  • White Matter Alterations in the Brains of Patients with Active, Remitted, and Cured Cushing Syndrome: A DTI Study
  • Qualitative and Quantitative Analysis of MR Imaging Findings in Patients with Middle Cerebral Artery Stroke Implanted with Mesenchymal Stem Cells
Show more BRAIN

Similar Articles

Advertisement

News and Updates

  • Lucien Levy Best Research Article Award
  • Thanks to our 2022 Distinguished Reviewers
  • Press Releases

Resources

  • Evidence-Based Medicine Level Guide
  • How to Participate in a Tweet Chat
  • AJNR Podcast Archive
  • Ideas for Publicizing Your Research
  • Librarian Resources
  • Terms and Conditions

Opportunities

  • Share Your Art in Perspectives
  • Get Peer Review Credit from Publons
  • Moderate a Tweet Chat

American Society of Neuroradiology

  • Neurographics
  • ASNR Annual Meeting
  • Fellowship Portal
  • Position Statements

© 2023 by the American Society of Neuroradiology | Print ISSN: 0195-6108 Online ISSN: 1936-959X

Powered by HighWire