Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print April 17, 2003, 10.1148/radiol.2273020721
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2273020721v1
227/3/731    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rovaris, M.
Right arrow Articles by Filippi, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rovaris, M.
Right arrow Articles by Filippi, M.
(Radiology 2003;227:731-738.)
© RSNA, 2003

Neuroradiology

Age-related Changes in Conventional, Magnetization Transfer, and Diffusion-Tensor MR Imaging Findings: Study with Whole-Brain Tissue Histogram Analysis1   

Marco Rovaris, MD, Giuseppe Iannucci, MD, Mara Cercignani, PhD, Maria Pia Sormani, PhD, Nicola De Stefano, MD, Simonetta Gerevini, MD, Giancarlo Comi, MD and Massimo Filippi, MD

1 From the Neuroimaging Research Unit (M.R., G.I., M.C., M.P.S., M.F.) and Clinical Trials Unit (G.C.) of the Department of Neuroscience, and Department of Neuroradiology (S.G.), Scientific Institute and University Ospedale San Raffaele, Via Olgettina 60, 20132 Milan, Italy; and Neurometabolic Unit, Institute of Neurological Sciences, University of Siena, Italy (N.D.S.). Received June 18, 2002; revision requested August 9; revision received September 24; accepted October 25. Address correspondence to M.F. (e-mail: filippi.massimo@hsr.it).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
PURPOSE: To investigate the influence of aging on conventional magnetic resonance (MR) imaging–, magnetization transfer MR imaging–, and diffusion-tensor MR imaging–derived measurements.

MATERIALS AND METHODS: Dual-echo T1-weighted magnetization transfer and diffusion-tensor MR images of the brain were obtained in 89 healthy subjects. Normalized brain parenchymal volume (NBV) was measured by using a fully automated technique. Magnetization transfer ratio (MTR), apparent diffusion coefficient (ADC), and fractional anisotropy (FA) histograms were created for the whole brain (MTR values) or for a large representative portion of it (ADC and FA values). Bivariate correlations were assessed by using the Spearman rank correlation coefficient. A stepwise selection procedure was used to identify the combination of variables that were most influenced by subject age in a multivariate regression model.

RESULTS: Significant correlations were found between subject age and the following variables: number of hyperintense areas in the brain at T2-weighted MR imaging (r = 0.63, P < .001), NBV (r = -0.79, P < .001), mean ADC (r = 0.34, P = .001), ADC peak height (r = -0.34, P = .001), and FA peak height (r = -0.57, P < .001). NBV correlated significantly with number of hyperintense areas (P < .001), MTR peak height (P < .001), mean ADC (P = .001), ADC peak height (P = .001), and FA peak height (P < .001). The final multivariable regression model included NBV and number of hyperintense areas at T2-weighted MR imaging as independent predictors of subject age.

CONCLUSION: In addition to the extent of T2-weighted MR imaging hyperintense areas and the measurement of NBV, diffusion-tensor MR imaging provides additional in vivo information about microstructural age-related brain tissue changes.

© RSNA, 2003

Index terms: Aging • Brain, atrophy, 10.142, 10.83 • Brain, MR, 10.121411, 10.121412, 10.121416, 10.121417, 10.12144, • Magnetic resonance (MR), diffusion tensor, 10.12144, • Magnetic resonance (MR), magnetization transfer, 10.121417


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Normal aging can be accompanied by the development of white matter lesions that are hyperintense on T2-weighted magnetic resonance (MR) images of the brain and by enlargement of the ventricles and cortical sulci that is reflective of nonfocal atrophy (1,2). However, the correct interpretation of these findings in elderly subjects may be challenging. Areas of hyperintensity in white matter seem to reflect zones of atrophic perivascular demyelination (3,4), but their presence and extent indicate only in part the occurrence and severity of age-related cognitive decline (5,6). MR imaging–measured brain volume decreases with increasing age (5,7–9), and this decrease correlates well with impaired cognitive function (5,6). However, the structural changes that lead to the development and progression of brain atrophy have not been fully clarified yet (5,9).

Magnetization transfer (10) and diffusion-weighted (11,12) MR imaging examinations have increased sensitivity in the detection of brain tissue damage that occurs at levels beyond the resolution capability of conventional imaging modalities. The magnetization transfer ratio (MTR), an index of magnetization transfer MR imaging, reflects the efficiency of the magnetization exchange between the protons of water inside tissue (relatively free) and the protons bound to the macromolecules, which decrease when brain tissue disorganization occurs owing to demyelination or axonal loss (13). Diffusion-weighted MR imaging enables quantitative measurements of other aspects of the brain tissue microstructure, which are obtained by exploiting the properties of water diffusion (12). Water molecules undergo a random Brownian motion that can be influenced by the characteristics of the surrounding medium because of the presence of partially permeable barriers and aligned structures. These characteristics are reflected by the magnitude and directionality of diffusion.

Because diffusion in solid tissues is an inherently three dimensional and anisotropic (ie, with different characteristics in different directions) process, diffusion-tensor MR imaging has been developed to fully characterize the various aspects of diffusion (14). From the tensor it is possible to calculate both the magnitude of diffusion, by using the apparent diffusion coefficient (ADC), and the degree of anisotropy, which is a measure of tissue organization that can be expressed by several indexes, including fractional anisotropy (FA), which has no dimension. The ADC is equal to one-third the trace of the diffusion tensor (ie, the average of the three diagonal elements), which represents isotropic diffusion averaged along three orthogonal directions. Tissue disruption, by removing structural barriers to water molecular motion, typically causes increased ADC values and decreased FA values (15,16).

One can postprocess both magnetization transfer MR imaging data and diffusion-tensor MR imaging data to obtain histograms of MTR, ADC, and FA values from large portions of the brain parenchyma (17,18). Although histogram analysis lacks spatial information when compared with a region-of-interest–based approach, it requires a reduced degree of human intervention and thus is less prone to the measurement variability caused by bias and uncertainty in the choice and location of regions of interest. Several studies (19–27) have been performed to investigate the effects of normal aging on magnetization transfer MR imaging– and diffusion-tensor MR imaging–derived quantities. However, these studies did not include assessment of either the relationship between the various quantities derived from magnetization transfer and diffusion-tensor MR images or the relationship between age-related brain volume decreases and observed changes in histogram parameters. The purpose of our study was to investigate the influences of aging on conventional MR imaging–, magnetization transfer MR imaging–, and diffusion-tensor MR imaging–derived measurements.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Subjects
Eighty-nine healthy subjects (50 female subjects and 39 male subjects; mean age, 43.6 years; age range, 11–76 years) volunteered to participate in this study after a written informed consent form was signed. The study was approved by the local ethics committee of the Scientific Institute and University Ospedale San Raffaele. All subjects were recruited from the general population. Five female subjects and three male subjects were aged 11–20 years; 10 women and five men, aged 21–30 years; 13 women and four men, aged 31–40 years; seven women and seven men, aged 41–50 years; seven women and 10 men, aged 51–60 years; seven women and six men, aged 61–70 years; and one woman and four men, aged 71–80 years. Given the predominance of female subjects in the younger age groups, the mean age of the male subjects (47.4 years) was slightly older than that of the female subjects (40.7 years). This difference, however, was not statistically significant.

None of the subjects had experienced episodes of neurologic dysfunction, and all had completely normal neurologic examination results, including age- and education-corrected Mini–Mental State scores (28) within the normal limits. Subjects who had cerebrovascular disease, epilepsy, migraines, hypertension, diabetes, and other disorders that potentially could affect the central nervous system, such as neoplasm, chronic alcohol intake, endocrine or metabolic disease, coagulopathy, or other thromboembolic disorders, were not eligible to participate in the study. These conditions led to the exclusion of 29 screened subjects.

Image Acquisition
All images were obtained with a 1.5-T MR imaging unit (Vision; Siemens, Erlangen, Germany). During a single examination, the following MR images were obtained without the subject moving from the magnet bore: (a) dual-echo turbo spin echo (repetition time, 3,300 msec; first-echo echo time, 16 msec; second-echo echo time [for T2-weighted sequence], 98 msec; echo train length of five; matrix, 256 x 256; field of view, 250 x 250 mm); (b) two-dimensional gradient echo (600/12 [repetition time msec/echo time msec]; flip angle, 20°; matrix, 256 x 256; field of view, 250 x 250 mm) with and without an off-resonance saturation pulse (offset frequency, 1.5 kHz; Gaussian envelope duration, 7.68 msec; flip angle, 500°); (c) conventional T1-weighted spin echo (768/15); and (d) pulsed gradient spin-echo echo planar (interecho spacing, 0.8; echo time, 123 msec; matrix, 128 x 128; field of view, 240 x 240 mm) with diffusion gradients applied in eight non-colinear directions that were chosen to uniformly cover a three-dimensional space (29).

The duration and maximum amplitude of the diffusion gradients were 25 msec and 21 mT/m, respectively, and resulted in a b factor of 1,044 sec/mm2 in each of the eight directions. To optimize the measurement of diffusion, we used only two b factors: a b1 of approximately 0 sec/mm2 and a b2 of 1,044 sec/mm2 (29). It has been shown (30) that when the trace of the diffusion tensor is known approximately, only one b factor (excluding the reference measurement at b = 0) should be used for a given acquisition time so that the signal can be attenuated to about one-third of its nonweighted intensity and the signal-to-noise ratio can be maximized. Fat saturation was performed by using a four–radio-frequency-pulse binomial pulse train to prevent chemical shift artifacts. A birdcage head coil of approximately 300 mm in diameter was used for radio-frequency transmission and signal reception.

For the dual-echo turbo spin-echo, T1-weighted spin-echo, and gradient-echo MR image acquisitions, 24 contiguous 5-mm-thick sections were obtained with interleaved excitation order. The sections were positioned to run parallel to a line that joins the most inferoanterior part of the corpus callosum to the most inferoposterior part of the corpus callosum. For the pulsed gradient spin-echo echo-planar MR image acquisitions, 10 transverse 5-mm-thick sections were obtained in the same orientation as the dual-echo sections. However, the second-to-last caudal section was positioned to match the central sections in the dual-echo image set.

Image Analysis
Two observers (G.I., S.G.), who had 5 years of experience in MR image reading, examined the film hard copies of the dual-echo MR images in consensus; they were blinded as to which subjects the images had been obtained in. The observers identified and counted the hyperintense lesions depicted on the T2-weighted MR images and reported the imaging characteristics of any other pathologic finding.

All quantitative image postprocessing was performed on a workstation (Sparcstation; Sun Microsystems, Mountain View, Calif) that was independent from the MR imaging unit. From the two gradient-echo MR acquisitions (with and without a magnetization transfer pulse), MTR maps were generated pixel-by-pixel according to the following formula: MTR = [(M0 - MS)/M0] x 100, where M0 is the signal intensity of the pixel on the image without the magnetization transfer pulse and MS is the signal intensity of the same pixel when the magnetization transfer pulse is applied. The pulsed gradient spin-echo echo-planar MR images were first corrected for the distortion induced by eddy currents by using an algorithm that maximizes the mutual information between the non–diffusion-weighted and diffusion-weighted images (31). This function is evaluated as follows:

where I is the signal intensity, p{m} is the probability of occurrence of the signal intensity m from the range of signal intensities (M) on the first image, and p{n} is the probability of occurrence of the signal intensity n from the range of signal intensities (N) on the second image. p{m,n} is the probability of occurrence of the pair of signal intensities {m,n} in the same voxel. By maximizing the mutual information, the information depicted on the combined image is minimized with respect to the information depicted separately on each of the two images, and, therefore, maximal homology between the two images is achieved.

Then, by assuming a monoexponential relationship between the signal intensity and the product of the b matrix (ie, a 3 x 3 matrix that expresses the relationship between the signal attenuation and the elements of the diffusion-tensor matrix) and the diffusion-tensor matrix components, we calculated the diffusion tensor for each pixel according to the following equation:

where M is the measured signal intensity; M0, the signal intensity at T2-weighted MR imaging; bij, the elements of the b matrix; and Dij, the elements of the diffusion-tensor matrix. The tensor was estimated by using a multivariate linear regression. After diagonalization of the estimated tensor matrix, the ADC and FA for every pixel were derived.

The creation of MTR, ADC, and FA maps is a fully automated process that takes about 15 minutes per map. The diffusion-tensor MR images were interpolated to the same image matrix size as the dual-echo MR images. Then, the MTR maps and the b = 0 area of the pulsed gradient spin-echo echo-planar MR images (T2-weighted, but not diffusion-weighted) were coregistered with the dual-echo T2-weighted images by using a three-dimensional rigid-body coregistration algorithm based on mutual information (31). The same transformation parameters were then used to coregister the ADC and FA images to the dual-echo images (Fig 1). Image interpolation and coregistration are fully automated processes that take about 10 minutes per MR image.



View larger version (69K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Transverse T2-weighted MR image (3,300/98) (A), MTR maps (B and E), ADC maps (C and F), and FA maps (D and G) of the brain obtained at the level of the lateral ventricles in a 45-year-old healthy man. MTR, ADC, and FA maps constructed before (B-D) and after (E-G) coregistration with the T2-weighted image (A) are shown.

 
The pixels containing cerebrospinal fluid and extracerebral tissue were carefully removed from the coregistered MTR and b = 0 area of the diffusion-tensor images by using a semiautomated technique based on local thresholding (32). Next, the corresponding pixels were removed from the FA and ADC maps. Human intervention is needed to remove pixels containing cerebrospinal fluid and extracerebral tissue; in the present study, the entire process took 5–10 minutes per map. Histograms were produced from these maps by using 10% wide bins. To correct for intersubject differences in brain size, each bin was normalized by using the total number of pixels contributing to the histogram. For each histogram (Fig 2), the following quantities were assessed: mean MTR, mean ADC, mean FA, histogram peak height, and histogram peak location. Histogram creation is fully automated, and the complete postprocessing of data from a single subject takes less than 1 minute. Additional information about the methodology that we used for histogram analysis is reported elsewhere (33,34).



View larger version (22K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a. Normalized (a) MTR, (b) ADC, and (c) FA histogram values from the brain parenchyma of a 38-year-old healthy woman that were obtained after removal of cerebrospinal fluid and extracerebral tissue pixels.

 


View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b. Normalized (a) MTR, (b) ADC, and (c) FA histogram values from the brain parenchyma of a 38-year-old healthy woman that were obtained after removal of cerebrospinal fluid and extracerebral tissue pixels.

 


View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c. Normalized (a) MTR, (b) ADC, and (c) FA histogram values from the brain parenchyma of a 38-year-old healthy woman that were obtained after removal of cerebrospinal fluid and extracerebral tissue pixels.

 
On the T1-weighted MR images, normalized volumes of the entire brain parenchyma were measured by using the fully automated software, SIENAX (Stephen Smith, PhD, and Paul Matthews, MD, PhD, Functional Magnetic Resonance Imaging of the Brain, Oxford University, England), which is the cross-sectional version of the structural image evaluation, using normalisation, of atrophy, or SIENA, software (35). First, SIENAX uses a brain extraction tool method to extract the pixels that form the brain and skull from MR images; this process is extensively described elsewhere (36). A tissue segmentation program (37) is then used to segment the extracted brain image into brain tissue, cerebrospinal fluid, and background sections and thus generate an estimate of the total brain tissue volume. Subsequently, the original MR images are registered to a canonical image in a standardized space (by using the image of the skull to provide the scaling cue); this procedure yields a spatial normalization scaling factor for each subject. The estimated brain tissue volume in the subject is then multiplied by the normalization factor to yield the normalized brain parenchymal volume (NBV). With use of the SIENAX software, the entire image postprocessing procedure for NBV calculation takes 40–60 minutes per MR image acquisition.

Statistical Analysis
Bivariate correlations among age, NBV, and histogram-derived measures were assessed by using the Spearman rank correlation coefficient. For this analysis, a Bonferroni correction for multiple comparisons was applied, and correlations with P <= .001 were considered to be significant at a 5% level after correction for 39 comparisons. MR imaging–derived variables with P < .002 at bivariate correlation analysis were included in a multivariate regression model with age as the dependent variable. All of the predictor variables included in the multivariate model were rank transformed to avoid violations of the assumptions of a linear relationship with age. The Mann-Whitney test was used for group comparisons.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
One or more hyperintense lesions were seen on the T2-weighted dual-echo MR images obtained in 36 subjects. Specifically, a single lesion was depicted at imaging in four subjects (in the 21–30-year, 41–50-year, 51–60-year, and 61–70-year age groups), two lesions were depicted in one subject (in the 31–40-year age group), and multiple patchy to confluent hyperintense lesions were depicted in the remaining 31 subjects (Table). The characteristics of the hyperintense lesions were always consistent with those of small nonspecific areas of hyperintensity in the white matter. No other major abnormalities such as infarct, vascular malformation, or tumor were found. Mean NBVs and mean MTR histogram–, ADC histogram–, and FA histogram–derived values for each age group and for the entire subject population are listed in the Table.


View this table:
[in this window]
[in a new window]

  		Conventional, Magnetization Transfer, and Diffusion-Tensor MR Imaging-derived Characteristics of Subjects in Different Age Groups

 
We observed significant correlations (Figs 3, 4) between subject age and each of the following parameters: number of hyperintense brain lesions at T2-weighted MR imaging (r = 0.63, P < .001), NBV (r = -0.79, P < .001), mean ADC (r = 0.34, P = .001), ADC peak height (r = -0.34, P = .001), and FA peak height (r = -0.57, P < .001).



View larger version (20K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a. Scatterplots of (a) number of hyperintense lesions at T2-weighted MR imaging and (b) NBV (in milliliters), both correlated with age in the entire subject population. Subject age correlated significantly with both NBV and number of hyperintense lesions.

 


View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b. Scatterplots of (a) number of hyperintense lesions at T2-weighted MR imaging and (b) NBV (in milliliters), both correlated with age in the entire subject population. Subject age correlated significantly with both NBV and number of hyperintense lesions.

 


View larger version (37K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Scatterplots of MTR histogram-derived (in percentages, top row), ADC histogram-derived (in mm2/sec x 10-3, middle row), and FA histogram-derived (bottom row) measurements correlated with age in the entire subject population. Mean ADC, ADC peak height, and FA peak height correlated significantly with subject age.

 
Number of hyperintense lesions at T2-weighted MR imaging correlated significantly with NBV (r = -0.54, P < .001), MTR peak height (r = -0.39, P < .001), and FA peak height (r = -0.39, P < .001). NBV correlated significantly with the following histogram-derived values: MTR peak height (r = 0.49, P < .001), mean ADC (r = -0.40, P = .001), ADC peak height (r = 0.37, P = .001), and FA peak height (r = 0.53, P < .001). There were no significant correlations between diffusion-tensor and magnetization transfer histogram indexes.

We analyzed number of hyperintense lesions at T2-weighted MR imaging, NBV, mean ADC, ADC peak height, and FA peak height by using a stepwise multivariate regression procedure, with age as the dependent variable. The final model included NBV (P < .001) and number of hyperintense lesions (P < .001) as independent predictors of subject age.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
REFERENCES
 
In the past few years, there has been growing interest in developing new MR imaging methods to investigate in vivo the dynamics of brain aging. In this context, histograms of MTR, ADC, and FA values from large portions of the brain parenchyma are promising tools for obtaining overall measurements of macro- and microstructural brain damage (17,18). Numerous studies (17,18,34,38–40) have focused on the potential role of quantitative histogram analysis in the study of central nervous system abnormalities. To our knowledge, however, there have been only a few studies on the effect of normal brain aging on MTR histogram–derived (22) and ADC histogram–derived (24,25) quantities in which magnetization transfer and diffusion-tensor MR imaging measures were found to be sensitive for the detection of age-related brain tissue changes and thus suitable reference parameters for comparison with neurodegenerative disease measurements. However, these previously performed studies did not address some important issues, such as the correlation between the age-related changes depicted at conventional MR imaging and the MTR histogram– and ADC histogram–derived measures or the relationship between magnetization transfer and diffusion-tensor MR imaging–derived findings.

About 40% (n = 36) of the subjects in our study had abnormalities at T2-weighted MR imaging, and a clear trend toward higher lesion burdens in older subjects was observed. Some of the subjects in the older age groups had a lesion burden that can be consistent with the lesion burdens associated with subclinical cerebrovascular disorders. The clustering of subjects in the older age groups who had hyperintense brain abnormalities at T2-weighted MR imaging also accounted for the moderate positive correlation between age and number of hyperintense lesions. These findings are similar to those of several previously performed studies (5,8,41,42) and suggest that although conventional MR imaging yields nonspecific information about brain tissue changes, the burden of hyperintense brain lesions at T2-weighted MR imaging does increase with normal aging. Quantitative estimates of hyperintense brain lesion load have been found to correlate, at least partially, with age-related changes in brain metabolism and declines in cognitive function (5).

Age-related decreases in regional (43,44) and global (5,8) brain tissue volume have been consistently reported in pathologic (6,9) and MR imaging studies (5,7,8,41–44). Although the results of previously performed studies are often conflicting, the majority of the available data indicate that the loss of tissue is higher in the white matter than in the gray matter, and most of the loss probably reflects both a decrease in the number of myelinated nerve fibers and an increase in the number of intercellular spaces. In our study, the use of normalized volume measurements led to an even stronger correlation between age and brain atrophy than the correlations previously reported (5,7,8), and this indicates that NBV is a sensitive MR imaging measure for the detection of age-related global brain changes. However, the occurrence of brain atrophy is an end-stage phenomenon; therefore, the assessment of this condition cannot yield information about the development and progression of earlier microstructural changes that precede the actual loss of brain parenchyma.

We found that some of the measures derived from brain ADC and FA histograms correlated significantly with subject age, although to a lesser degree compared with NBV and number of hyperintense lesions. Diffusion-tensor MR imaging–derived indexes reflect the microstructural integrity of brain tissue by enabling quantification of the extent and directionality of water diffusion. ADC is affected by cellular size, shape, and integrity and increases with decreases in structural barriers to water molecular motion (12). FA is an index of deviation from tissue anisotropy and can vary widely on a regional basis, even in normal brain structures, because it reflects the degree of alignment of cellular structures within fiber tracts, as well as the structural integrity of the cellular structures (19).

Considering these background data, our findings of a significant inverse relationship between ADC histogram peak height and subject age and between FA histogram peak height and subject age are consistent with an age-related disorganization of brain tissue. Similar histogram changes have been reported in other studies of normal brain aging (24,25) and can be observed in patients with organic brain disorders such as Alzheimer disease (40), as opposed to healthy age-matched control subjects. In addition, the results of the present study are consistent with the observation of decreasing levels of N-acetyl-aspartate (a marker of neuronal integrity) with increasing age in a sample of healthy subjects examined with proton MR spectroscopy (45).

The pathologic basis of age-related diffusion-tensor MR imaging changes in brain tissue remains unclear and needs to be clarified in correlative postmortem studies. Nevertheless, our study result of a decrease in the number of pixels with normal FA values in older subjects suggests a diffuse degeneration of white matter tracts, as well as an expansion of cerebrospinal fluid–filled spaces due to brain atrophy or microscopic changes at a cellular level. Although the latter two phenomena should affect MTR and ADC values equally, we found that among the various histographic values determined, those derived from FA histograms correlated more strictly with subject age than did those derived from ADC and MTR histograms. In addition, we observed no significant correlation between FA histogram–derived measurements and either MTR histogram– or ADC histogram–derived measurements; this finding suggests that measures of tissue anisotropy are independent from other measures that reflect brain tissue disorganization. All of these findings indicate that FA measurements can yield important information about the age-related microstructural deterioration of brain fiber coherence.

We recognize that variations in measurements derived from histograms for the entire brain might also reflect age-related changes in the gray matter–to–white matter ratio, given the different magnetization transfer and diffusion-tensor MR imaging characteristics of these two regions (20,46,47). In addition, ongoing maturational mechanisms might affect findings in younger subjects (48). However, whole-brain tissue histograms lack spatial information, and, thus, only those examinations performed after segmentation of the different brain tissue types might substantially help better address the issues of age-related changes in the gray matter–to–white matter ratio and the role of ongoing maturational mechanisms in younger subjects.

Another limitation of whole-brain histogram analysis is the presence of partial volume effects from cerebrospinal fluid–contaminated pixels at the edge of the brain. These effects can cause any concomitant decrease in brain parenchyma volume to have a potentially substantial influence on the histogram characteristics. In our study, because the diffusion-tensor MR images were acquired with a relatively low in-plane resolution, the risk of cerebrospinal fluid contamination was higher for the ADC and FA histograms than for the MTR histograms.

The strong influence of concomitant brain atrophy on changes in histogram-derived values was confirmed in our study population by the fact that NBV correlated significantly with mean ADC values and with MTR, ADC, and FA histogram peak heights. These data may help clarify the conflicting reports of highly variable correlations between brain volume measurements and magnetization transfer and diffusion-tensor MR imaging–derived values in previously performed studies (49–51). We believe that our study findings also help explain why, in a preliminary study of diffusion-tensor MR imaging (24), regional age-related changes in tissue anisotropy were substantial at the brain–cerebrospinal fluid interfaces.

In conclusion, of all of the magnetization transfer and diffusion-tensor MR imaging–derived parameters analyzed, FA histogram peak height was found to have the strongest correlation with subject age. Therefore, although number of hyperintense lesions at T2-weighted MR imaging and NBV seem to be the most sensitive measures for the detection of age-related changes in brain tissue, assessing changes at a microstructural level by using diffusion-tensor MR imaging–derived measures of tissue anisotropy can be useful in obtaining a better in vivo depiction of the dynamics of normal brain aging.


   ACKNOWLEDGMENTS
 
We thank Stephen Smith, PhD, and Paul Matthews, MD, PhD, of Functional Magnetic Resonance Imaging of the Brain, Oxford University, England, for kindly providing us with the SIENAX program.


   FOOTNOTES
 
Abbreviations: ADC = apparent diffusion coefficient, FA = fractional anisotropy, MTR = magnetization transfer ratio, NBV = normalized brain parenchymal volume

Author contributions: Guarantor of integrity of entire study, M.F.; study concepts, M.F., M.R.; study design, M.F.; literature research, M.R.; clinical studies, G.I., G.C.; data acquisition, G.I., S.G.; data analysis/interpretation, N.D.S., G.I., S.G., M.C.; statistical analysis, M.P.S., M.R.; manuscript preparation and definition of intellectual content, M.F., M.R.; manuscript editing, M.R.; manuscript revision/review, M.F., M.R.; manuscript final version approval, M.F., M.R., G.C., G.I., N.D.S.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 

  1. Drayer BP. Imaging of the aging brain. I: Normal findings. Radiology 1988; 166:785-796.[Abstract/Free Full Text]
  2. Drachman DA. Aging and the brain: a new frontier. Ann Neurol 1997; 42:819-828.[CrossRef][Medline]
  3. Braffman BH, Zimmerman RA, Trojanowski JQ, Gonatas NK, Hickey WF, Schlaepfer WW. Brain MR: pathologic correlation with gross histopathology. I. Lacunar infarction and Virchow-Robin spaces. AJR Am J Roentgenol 1988; 151:551-558.[Abstract/Free Full Text]
  4. van Swieten JC, van den Hout JHW, van Ketel BA, Hijdra A, Wokke JHJ, van Gijn J. Periventricular lesions in the white matter on magnetic resonance imaging in the elderly: a morphometric correlation with arteriolosclerosis and dilated perivascular spaces. Brain 1991; 114:761-774.[Abstract/Free Full Text]
  5. DeCarli C, Murphy DGM, Trahn M, et al. The effect of white matter hyperintensity volume on brain structure, cognitive performance and cerebral metabolism of glucose in 51 healthy adults. Neurology 1995; 45:2077-2084.[Abstract]
  6. Smith CD, Snowdon DA, Wang H, Markesberry WR. White matter volumes and periventricular white matter hyperintensities in aging and dementia. Neurology 2000; 54:838-842.[Abstract/Free Full Text]
  7. Coffey CE, Lucke JF, Saxton JA, et al. Sex differences in brain aging: a quantitative magnetic resonance imaging study. Arch Neurol 1998; 55:169-179.[Abstract/Free Full Text]
  8. Guttmann CRG, Jolesz FA, Kikinis R, et al. White matter changes with normal aging. Neurology 1998; 50:972-978.[Abstract/Free Full Text]
  9. Good CD, Johnsrude IS, Ashburner J, Henson RN, Friston KJ, Frackowiak RS. A voxel-based morphometric study of ageing in 465 normal adult human brains. Neuroimage 2001; 14:21-36.[CrossRef][Medline]
  10. Filippi M, Grossman RI, Comi G. Magnetization transfer in multiple sclerosis. Neurology 1999; 53(suppl 3).
  11. Le Bihan D, Turner R, Pekar J, Moonen CTW. Diffusion and perfusion imaging by gradient sensitization: design, strategy and significance. J Magn Reson Imaging 1991; 1:7-8.[Medline]
  12. LeBihan D, Breton E, Lallemand D, Grenier P. Cabanis E, Laval-Jeanter M. MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders. Radiology 1986; 161:401-407.[Abstract/Free Full Text]
  13. Brochet B, Dousset V. Pathological correlates of magnetization transfer imaging abnormalities in animal models and humans with multiple sclerosis. Neurology 1999; 53(suppl 3):S12-S17.
  14. Chenevert TL, Brunberg JA, Pipe JG. Anisotropic diffusion in human white matter: demonstration with MR techniques in vivo. Radiology 1990; 177:401-405.[Abstract/Free Full Text]
  15. Basser PJ, Pierpaoli C. Microstructural features measured using diffusion tensor imaging. J Magn Reson B 1996; 111:209-219.[CrossRef][Medline]
  16. Pierpaoli C, Jezzard P, Basser PJ, Blarnett A, Di Chiro G. Diffusion tensor MR imaging of the human brain. Radiology 1996; 201:637-648.[Abstract/Free Full Text]
  17. van Buchem MA, McGowan JC, Kolson DL, Polansky M, Grossman RI. Quantitative volumetric magnetization transfer analysis in multiple sclerosis: estimation of macroscopic and microscopic disease burden. Magn Reson Med 1996; 36:632-636.[Medline]
  18. Nusbaum AO, Tang CY, Wei TC, Buchsbaum MS, Atlas SW. Whole-brain diffusion MR histograms differ between MS subtypes. Neurology 2000; 54:1421-1426.[Abstract/Free Full Text]
  19. Gideon P, Thomsen C, Henriksen O. Increased self-diffusion of brain water in normal aging. Magn Reson Imaging 1994; 4:185-188.
  20. Mehta RC, Pike GB, Enzmann DR. Magnetization transfer MR of the normal adult brain. AJNR Am J Neuroradiol 1995; 16:2085-2091.[Abstract]
  21. Silver NC, Barker GJ, MacManus DG, Tofts PS, Miller DH. Magnetisation transfer ratio of normal brain white matter: a normative database spanning four decades of life. J Neurol Neurosurg Psychiatry 1997; 62:223-228.[Abstract]
  22. Hofman PAM, Kemerink GJ, Jolles J, Wilmink JT. Quantitative analysis of magnetization transfer images of the brain: effect of closed head injury, age and sex on white matter. Magn Reson Med 1999; 42:803-806.[CrossRef][Medline]
  23. Pfefferbaum A, Sullivan EV, Hedehus M, Lim KO, Adalsteinsson E, Moseley M. Age-related decline in brain white matter anisotropy measured with spatially corrected echo-planar diffusion tensor imaging. Magn Reson Med 2000; 44:259-268.[CrossRef][Medline]
  24. Nusbaum AO, Tang CY, Buchsbaum MS, Wei TC, Atlas SW. Regional and global changes in cerebral diffusion with normal aging. AJNR Am J Neuroradiol 2001; 22:136-142.[Abstract/Free Full Text]
  25. Chen ZG, Li TQ, Hindmarsh T. Diffusion tensor trace mapping in normal adult brain using single-shot EPI technique: a methodological study of the aging brain. Acta Radiol 2001; 42:447-458.[Medline]
  26. Sullivan EV, Adalsteinsson E, Hedehus M, et al. Equivalent disruption of regional white matter microstructure in ageing healthy men and women. Neuroreport 2001; 12:99-104.[CrossRef][Medline]
  27. Helenius J, Soinne L, Perkio J, et al. Diffusion-weighted MR imaging in normal human brains in various age groups. AJNR Am J Neuroradiol 2002; 23:194-199.[Abstract/Free Full Text]
  28. Folstein MF, Folstein SE, McHugh PR. Mini-Mental State: a practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 1975; 12:189-198.[CrossRef][Medline]
  29. Jones DK, Horsfield MA, Simmons A. Optimal strategies for measuring diffusion in anisotropic systems by magnetic resonance imaging. Magn Reson Med 1999; 42:515-525.[CrossRef][Medline]
  30. Bito Y, Hirata S, Yamamoto E. Optimal gradient factors for ADC measurements (abstr) Proceedings of the Third Meeting of the International Society for Magnetic Resonance in Medicine and the European Society of Magnetic Resonance in Medicine and Biology. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 1995; 913.
  31. Studholme C, Hill DLG, Hawkes DJ. Automated three-dimensional registration of magnetic resonance and positron emission tomography brain images by multiresolution optimization of voxel similarity measures. Med Phys 1996; 24:25-35.
  32. Rovaris M, Filippi M, Calori G, et al. Intra-observer reproducibility in measuring new putative MR markers of demyelination and axonal loss in multiple sclerosis: a comparison with conventional T2-weighted images. J Neurol 1997; 244:266-270.[CrossRef][Medline]
  33. Cercignani M, Iannucci G, Rocca MA, Comi G, Horsfield MA, Filippi M. Pathological damage in MS assessed by diffusion-weighted and magnetization transfer MRI. Neurology 2000; 54:1139-1144.[Abstract/Free Full Text]
  34. Filippi M, Iannucci G, Tortorella C, et al. Comparison of MS clinical phenotypes using conventional and magnetization transfer MRI. Neurology 1999; 52:588-594.[Abstract/Free Full Text]
  35. Smith SM, De Stefano N, Jenkinson M, Matthews P. Normalized accurate measurement of longitudinal brain change. J Comput Assist Tomogr 2001; 25:466-475.[CrossRef][Medline]
  36. Smith SM, Zhang Y, Jenkinson M, et al. Accurate, robust, and automated longitudinal and cross-sectional brain change analysis. Neuroimage 2002; 17:479-489.[CrossRef][Medline]
  37. Zhang Y, Brady M, Smith S. Segmentation of brain MR images through a hidden Markov random field model and the expectation maximization algorithm. IEEE Trans Med Imaging 2001; 20:45-57.[CrossRef][Medline]
  38. Filippi M, Rocca MA, Moiola L, et al. MRI and MTI changes in the brain and cervical cord from patients with Devic’s neuromyelitis optica. Neurology 1999; 53:1705-1710.[Abstract/Free Full Text]
  39. Rovaris M, Viti B, Ciboddo G, et al. Brain involvement in systemic immune-mediated diseases: magnetic resonance and magnetisation transfer imaging study. J Neurol Neurosurg Psychiatry 2000; 68:170-177.[Abstract/Free Full Text]
  40. Bozzali M, Franceschi M, Falini A, et al. Quantification of tissue damage in AD using diffusion tensor and magnetization transfer MRI. Neurology 2001; 57:1135-1137.[Abstract/Free Full Text]
  41. Hunt AL, Orrison WW, Yeo RA, et al. Clinical significance of MRI white matter lesions in the elderly. Neurology 1989; 39:1470-1474.[Abstract/Free Full Text]
  42. Yetkin FZ, Fischer ME, Papke RA, Haughton VM. Focal hyperintensities in cerebral white matter on MR images of asymptomatic volunteers: correlation with social and medical histories. AJR Am J Roentgenol 1993; 161:855-858.[Abstract/Free Full Text]
  43. Gur RC, Gunning-Dixon FM, Turetsky BI, Bilker WB, Gur RE. Brain region and sex differences in age association with brain volume: a quantitative MRI study of healthy young adults. Am J Geriatr Psychiatry 2002; 72-80.
  44. Xu J, Kobayashi S, Yamaguchi S, Iijima K, Okada K, Yamashita K. Gender effects on age-related changes in brain structure. AJNR Am J Neuroradiol 2000; 21:112-118.[Abstract/Free Full Text]
  45. Kadota T, Horinouchi T, Curoda C. Development and aging of the cerebrum: assessment with proton MR spectroscopy. AJNR Am J Neuroradiol 2001; 22:128-135.[Abstract/Free Full Text]
  46. Engelter ST, Provenzale JM, Petrella JR, DeLong DM, MacFall JR. The effect of aging on the apparent diffusion coefficient of normal-appearing white matter. AJR Am J Roentgenol 2000; 175:425-430.[Abstract/Free Full Text]
  47. Shimony JS, McKinstry RC, Akbudak E, et al. Quantitative diffusion-tensor anisotropy brain MR imaging: normative human data and anatomic analysis. Radiology 1999; 212:770-784.[Abstract/Free Full Text]
  48. Prayer D, Barkovich AJ, Kirschner DA, et al. Visualization of nonstructural changes in early white matter development on diffusion-weighted MR images: evidence supporting premyelination anisotropy. AJNR Am J Neuroradiol 2001; 22:1572-1576.[Abstract/Free Full Text]
  49. Miki Y, Grossman RI, Udupa JK, et al. Differences between relapsing-remitting and chronic progressive multiple sclerosis as determined with quantitative MR imaging. Radiology 1999; 210:769-774.[Abstract/Free Full Text]
  50. Wilson M, Morgan PS, Lin X, Turner BP, Blumhardt LD. Quantitative diffusion weighted magnetic resonance imaging, cerebral atrophy and disability in multiple sclerosis. J Neurol Neurosurg Psychiatry 2001; 70:318-322.[Abstract/Free Full Text]
  51. Iannucci G, Rovaris M, Giacomotti L, Comi G, Filippi M. Correlation of multiple sclerosis measures derived from T2-weighted, T1-weighted, magnetization transfer and diffusion tensor MR imaging. AJNR Am J Neuroradiol 2001; 22:1462-1467.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 StrokeHome page
A. Nitkunan, T. R. Barrick, R. A. Charlton, C. A. Clark, and H. S. Markus
Multimodal MRI in Cerebral Small Vessel Disease: Its Relationship With Cognition and Sensitivity to Change Over Time
Stroke, July 1, 2008; 39(7): 1999 - 2005.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Neuroradiol.Home page
C. Tessa, M. Giannelli, R. Della Nave, C. Lucetti, C. Berti, A. Ginestroni, U. Bonuccelli, and M. Mascalchi
A Whole-Brain Analysis in De Novo Parkinson Disease
AJNR Am. J. Neuroradiol., April 1, 2008; 29(4): 674 - 680.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
P. G. P. Nucifora, R. Verma, S.-K. Lee, and E. R. Melhem
Diffusion-Tensor MR Imaging and Tractography: Exploring Brain Microstructure and Connectivity
Radiology, November 1, 2007; 245(2): 367 - 384.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Neuroradiol.Home page
R. D. Nave, S. Foresti, A. Pratesi, A. Ginestroni, M. Inzitari, E. Salvadori, M. Giannelli, S. Diciotti, D. Inzitari, and M. Mascalchi
Whole-Brain Histogram and Voxel-Based Analyses of Diffusion Tensor Imaging in Patients with Leukoaraiosis: Correlation with Motor and Cognitive Impairment
AJNR Am. J. Neuroradiol., August 1, 2007; 28(7): 1313 - 1319.
[Abstract] [Full Text] [PDF]

Home page
 Mult SclerHome page
H. Vrenken, P.J.W. Pouwels, S. Ropele, D.L. Knol, J.J.G. Geurts, C.H. Polman, F. Barkhof, and J.A. Castelijns
Magnetization transfer ratio measurement in multiple sclerosis normal-appearing brain tissue: limited differences with controls but relationships with clinical and MR measures of disease
Multiple Sclerosis, July 1, 2007; 13(6): 708 - 716.
[Abstract] [PDF]

Home page
 Am. J. Neuroradiol.Home page
R. D. Nave, A. Magaudda, R. Michelucci, G. Capizzi, A. Calabro, L. Guerrini, C. Gavazzi, S. Diciotti, P. Riguzzi, O. Daniele, et al.
Whole-Brain Histogram and Voxel-Based Analyses of Apparent Diffusion Coefficient and Magnetization Transfer Ratio in Celiac Disease, Epilepsy, and Cerebral Calcifications Syndrome
AJNR Am. J. Neuroradiol., March 1, 2007; 28(3): 479 - 485.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Neuroradiol.Home page
A. Spilt, R. Goekoop, R.G.J. Westendorp, G.J. Blauw, A.J.M. de Craen, and M.A. van Buchem
Not all age-related white matter hyperintensities are the same: a magnetization transfer imaging study.
AJNR Am. J. Neuroradiol., October 1, 2006; 27(9): 1964 - 1968.
[Abstract] [Full Text] [PDF]

Home page
 NeurologyHome page
I. J. Deary, M. E. Bastin, A. Pattie, J. D. Clayden, L. J. Whalley, J. M. Starr, and J. M. Wardlaw
White matter integrity and cognition in childhood and old age
Neurology, February 28, 2006; 66(4): 505 - 512.
[Abstract] [Full Text] [PDF]

Home page
 NeurologyHome page
B. Benedetti, A. Charil, M. Rovaris, E. Judica, P. Valsasina, M. P. Sormani, and M. Filippi
Influence of aging on brain gray and white matter changes assessed by conventional, MT, and DT MRI
Neurology, February 28, 2006; 66(4): 535 - 539.
[Abstract] [Full Text] [PDF]

Home page
 NeurologyHome page
R. A. Charlton, T. R. Barrick, D. J. McIntyre, Y. Shen, M. O'Sullivan, F. A. Howe, C. A. Clark, R. G. Morris, and H. S. Markus
White matter damage on diffusion tensor imaging correlates with age-related cognitive decline
Neurology, January 24, 2006; 66(2): 217 - 222.
[Abstract] [Full Text] [PDF]

Home page
 BrainHome page
F. Fazekas, S. Ropele, C. Enzinger, F. Gorani, A. Seewann, K. Petrovic, and R. Schmidt
MTI of white matter hyperintensities
Brain, December 1, 2005; 128(12): 2926 - 2932.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
M. Cosottini, M. Giannelli, G. Siciliano, G. Lazzarotti, M. C. Michelassi, A. Del Corona, C. Bartolozzi, and L. Murri
Diffusion-Tensor MR Imaging of Corticospinal Tract in Amyotrophic Lateral Sclerosis and Progressive Muscular Atrophy
Radiology, October 1, 2005; 237(1): 258 - 264.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Neuroradiol.Home page
A. J. Fabiano, M. A. Horsfield, and R. Bakshi
Interhemispheric Asymmetry of Brain Diffusivity in Normal Individuals: A Diffusion-Weighted MR Imaging Study
AJNR Am. J. Neuroradiol., May 1, 2005; 26(5): 1089 - 1094.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Neuroradiol.Home page
A. Spilt, T. Geeraedts, A. J. M. de Craen, R. G. J. Westendorp, G. J. Blauw, and M. A. van Buchem
Age-Related Changes in Normal-Appearing Brain Tissue and White Matter Hyperintensities: More of the Same or Something Else?
AJNR Am. J. Neuroradiol., April 1, 2005; 26(4): 725 - 729.
[Abstract] [Full Text] [PDF]

Home page
 NeurologyHome page
A. F. Fotenos, A. Z. Snyder, L. E. Girton, J. C. Morris, and R. L. Buckner
Normative estimates of cross-sectional and longitudinal brain volume decline in aging and AD
Neurology, March 22, 2005; 64(6): 1032 - 1039.
[Abstract] [Full Text] [PDF]

Home page
 Mult SclerHome page
K. Schmierer, D. R Altmann, N. Kassim, H. Kitzler, C. M Kerskens, C. A Doege, O. Aktas, J. D Lunemann, D. H Miller, F. Zipp, et al.
Progressive change in primary progressive multiple sclerosis normal-appearing white matter: a serial diffusion magnetic resonance imaging study
Multiple Sclerosis, April 1, 2004; 10(2): 182 - 187.
[Abstract] [PDF]