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


     

Published online before print February 21, 2008, 10.1148/radiol.2471070707
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2471070707v1
247/1/179    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 Stadlbauer, A.
Right arrow Articles by Ganslandt, O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Stadlbauer, A.
Right arrow Articles by Ganslandt, O.
(Radiology 2008;247:179-188.)
© RSNA, 2008

Neuroradiology

Age-related Degradation in the Central Nervous System: Assessment with Diffusion-Tensor Imaging and Quantitative Fiber Tracking1

Andreas Stadlbauer, PhD, Erich Salomonowitz, MD, Guido Strunk, PhD, Thilo Hammen, MD, and Oliver Ganslandt, MD

1 From the Department of Radiology, Landesklinikum St Poelten, Propst Fuehrer Strasse 4, A-3100 St Poelten, Austria. Received April 20, 2007; revision requested June 11; revision received July 27; final version accepted September 12. Address correspondence to E.S. (e-mail: erich.salomonowitz{at}stpoelten.lknoe.at).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Purpose: To prospectively quantify differences in age-related changes in the diffusivity parameters and fiber characteristics between association, callosal, and projection fibers.

Materials and Methods: This study was approved by the institutional review board, and informed consent was obtained. Diffusion-tensor imaging data with an isotropic voxel size of 1.9 mm3 were acquired at 3 T in 38 healthy volunteers (age range, 18–88 years; 18 women). Quantitative fiber tracking was used to calculate fractional anisotropy (FA) and mean diffusivity values, eigenvalues ({lambda}1, {lambda}2, and {lambda}3), the number of fiber projections, and the number of fiber projections per voxel for three-dimensional reconstructed association, callosal, projection, and total brain fibers. Bivariate linear regression models were used to analyze correlations. Significant differences between correlations were assessed with the Hotelling-Williams test.

Results: For FA, the strongest degradation in association fibers and no significant changes in projection fibers were observed. The difference in correlation was significant (P = .002). The number of fiber projections and the number of fiber projections per voxel showed strong to moderate negative correlations that were dependent on age (P < .001) in the three fiber structures and total brain fibers, with the exception of the number of fiber projections per voxel in projection fibers, which showed no significant correlation. The decrease in the number of fiber projections was significantly greater (P = .043) in projection fibers than in total brain fibers, whereas the decrease in the number of fiber projections per voxel was significantly weaker (P = .005). Association fibers showed the largest changes per decade of age for FA (–1.13%) and for the number of fiber projections per voxel (–4.7%), whereas callosal fibers showed the largest changes per decade of age for the number of fiber projections (–10.4%).

Conclusion: Quantitative fiber tracking enables identification of differences in diffusivity and fiber characteristics due to normal aging.

© RSNA, 2008


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
The relationship between normal aging and changes in the volume of gray and white brain matter has been investigated in several studies. There is consensus about the reduction in the volume of gray matter in relation to age (1). Data concerning the volume of white matter are less consistent. Some authors found no age-related changes (2–4), while others reported decreased white matter volume in parts of the brain (5,6) or global reduction of white matter volume (1,7). In all of these studies, tissue volumes were assessed on conventional magnetic resonance (MR) images. The temporal lobe, followed by the parietal and frontal lobes, had the strongest association with age, whereas the occipital lobe showed the weakest association with age. However, the structural organization of white matter cannot be assessed on conventional MR-based volumetric images.

The three major fiber structures of functional cortical regions are association, callosal, and projection fibers. Association fibers are cortex-to-cortex connections in the same hemisphere. Callosal fibers consist of an array of bundles that connect cortical areas of the hemispheres via the corpus callosum. Projection fibers connect the cortex with the thalamus, brainstem, and spinal cord (8,9). These highly organized formations of white matter affect the molecular movement of water, which is fastest when it is parallel to myelinated axonal fibers (10). Thus, the anisotropy of water diffusion reflects the microstructure of white matter tissue. Processes related to normal aging lead to changes in diffusion that result from loss of tissue organization and alterations in extracellular spaces (11,12).

Diffusion-tensor (DT) imaging enables in vivo detection of the anisotropy of water diffusion (13–16) with use of diffusion gradients in at least six directions. In general, DT data are evaluated by calculating parametric maps of fractional anisotropy (FA) and mean diffusivity (MD) that describe the directionality and magnitude of water diffusion, respectively. To determine potential regional differences, two-dimensional regions of interest (ROIs) can be defined manually within these maps. Fiber tracking enables reconstruction of white matter pathways in three dimensions with use of a tracking algorithm for comparison of orientations of water diffusion anisotropy on a voxel-by-voxel basis. Reconstructed fiber bundles may be evaluated visually or with qualitative assessment. Quantitative fiber tracking (17–20) is a technique used to determine diffusivity parameters (FA, MD, and eigenvalues) and fiber characteristics (number of fiber projections and mean number of fiber projections per voxel). The individual fiber bundles are defined by using a tracking algorithm for three-dimensional segmentation.

The purpose of our study was to prospectively quantify differences in age-related changes in diffusivity parameters and fiber characteristics between association, callosal, and projection fibers.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Participants and Power Analysis
All participants provided informed consent. The study protocol was approved by the institutional review board of the ethics committee of the University of Erlangen-Nuremberg, Erlangen, Germany. We wished to achieve a moderate bivariate correlation (r ≥ 0.5) between the diffusivity parameters (both diffusivity parameters and fiber characteristics) and age, and a power analysis (1 – β = 0.8, {alpha} = .1) revealed 36 subjects needed to be included in this study for it to have statistical significance. We examined 38 healthy volunteers (18 women, 20 men; mean age, 49.6 years ± 20.1 [standard deviation]; age range, 18–88 years). There were no significant differences in demographic variables between women and men. The mean ages of women and men were 51.6 years ± 22.1 and 47.8 years ± 18.6, respectively. The age difference between women and men was not significant (t = –0.569, df = 36, P = .573). Also, the age distribution of the participants, as assessed with the Kolmogorov-Smirnov z test, did not differ significantly (z = 0.855, P = .458).

An experienced radiologist (E.S., 32 years of experience in neuroradiology) interpreted conventional MR images to exclude morphologic abnormalities. Age-related white matter changes on MR images were not an exclusion criterion; however, neurologic or psychiatric disorders, head trauma, or any history of loss of consciousness was an exclusion criterion. Participants aged 50 years or older were screened for dementia by a neurologist (T.H., 13 years of experience in neurology) using the Mini-Mental State Examination (MMSE), as is usual for this age group (21–23). A minimum MMSE score of 27 points (maximum MMSE score, 30 points) was an inclusion criterion (24) (Table 1).


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

  		Table 1. Age, Sex, and Mini-Mental State Examination Scores of Participants

 
MR Imaging and White Matter Volumetry
MR imaging was performed with a 3-T whole-body imager (Magnetom Trio; Siemens, Erlangen, Germany) equipped with a receive-only eight-channel head coil. For DT imaging, a Stejskal-Tanner sequence with single-shot echo-planar imaging was used with the following parameters: 10.000/77 (repetition time msec/echo time msec), 128 x 128 matrix, 243 x 243-mm field of view, 1955 Hz/pixel bandwidth, and 0.82-msec echo spacing. Sixty transverse sections with no intersection gap and an isotropic voxel size of 1.9 x 1.9 x 1.9 mm were measured. Motion-probing gradients in six orientations with a b value of 700 sec/mm2 were applied after the acquisition of images with a b value of 0 sec/mm2. The sequence design was based on balanced diffusion gradients to minimize eddy current artifacts. For sufficient signal-to-noise ratio, four signals were acquired; a total DT data acquisition time of 5 minutes 40 seconds was required. An acceleration factor of two was applied by using the generalized autocalibrating partially parallel acquisition imaging technique to reduce echo train length and echo time. (Echo time would be 90 msec without use of this technique.) Section positioning and angulations were standardized. Each participant's head was fixed in a headrest and rendered immobile to minimize artifacts secondary to inadvertent motion.

Conventional MR imaging performed to define anatomic details consisted of (a) a coronal T2-weighted fast spin-echo sequence (7720/85, 5-mm section thickness), (b) a transverse fluid-attenuated inversion-recovery sequence (9000/133, 5-mm section thickness), (c) a sagittal T2-weighted fast spin-echo sequence (2800/83, 2-mm section thickness), and (d) a three-dimensional T1-weighted magnetization-prepared rapid acquisition gradient-echo sequence (1550/2.6, 0.47 x 0.47 x 0.94-mm voxel size, 176 sections).

We used T1-weighted magnetization-prepared rapid acquisition gradient-echo data sets and a region-growing process of the BrainVoyager QX software (version 1.7; Brain Innovation, Maastricht, the Netherlands) for volumetry of white matter (25). Talairach normalization was performed to transform subcortical brain structures to standardized spatial positions. This approach provided a priori positional knowledge, which was needed to segment white and gray brain matter from the ventricles and the surrounding head tissue, including the cerebellum.

To determine a threshold to automatically separate white matter from gray matter, a series of 10 signal intensity histograms across several transverse sections were analyzed. The curves showed two major peaks: The first peak corresponded to the intensity of gray matter, whereas the second peak corresponded to the intensity of white matter. The aforementioned software program was used to compute the threshold value that separated white matter from gray matter. This value was used for the segmentation procedure, which was used to perform a region-growing process to label white matter voxels.

Additionally, the region-growing process needed a seed voxel to start the growing process. The seed voxel was determined with an imaging process that started in the center of the brain and continued until a voxel that had an intensity higher than that of the determined white matter–gray matter threshold was found. From the determined seed voxel, neighboring voxels (ie, the six voxels in the direct neighborhood) are considered for inclusion in white matter segmentation. A neighboring voxel is added to the white matter segmentation only if its intensity is higher than the intensity threshold. The neighboring voxels of all included voxels were checked for inclusion with use of the same threshold. This process continued until no more neighbor voxels with intensities that exceeded the intensity threshold were found. After completion of the region-growing process, the gray matter voxels and remaining head tissue were set to zero; this resulted in a representation containing only the labeled white matter voxels. The labeled voxels were counted, and the white matter volume was calculated.

DT Data Processing and Quantitative Fiber Tracking
DT data were transferred offline to a workstation (Inspiron 8200; Dell, Round Rock, Tex) for analysis. One author (A.S., 6 years of experience in brain MR imaging) using DtiStudio software (version 2.4; H. Jiang, S. Mori, Department of Radiology, Johns Hopkins University, Baltimore, Md) (26) performed data processing. The DT was calculated for each voxel with a multivariate linear fitting method (27). Diagonalization of the DT was used to calculate three eigenvectors and three eigenvalues ({lambda}1, {lambda}2, {lambda}3) that corresponded to the main diffusion directions and associated diffusivities. The concept of diffusion ellipsoids enables three-dimensional representation of the diffusion distance in a given diffusion time. The eigenvalues are the axes of these ellipsoids and can be calculated for each image voxel (10,28,29). Parametric maps of FA and MD were calculated (30).

The three major white matter tracts that were investigated in this study were defined according to the classification of functional categories described by Wakana et al (27), as follows: (a) association fibers, including the superior longitudinal, inferior longitudinal, and inferior fronto-occipital fasciculi; (b) callosal fibers; and (c) projection fibers—including the corticobulbar and corticospinal tracts—and thalamic fibers—including anterior, superior, and posterior thalamic radiations.

The Fiber Assignment by Continuous Tracking, or FACT, algorithm was used for three-dimensional reconstruction of fiber tracts. With this algorithm, one starts tracking by tracking every voxel, and all voxels of the image volume are encompassed (16,31,32) so that each voxel has multiple fibers passing through it. The fibers of interest were selected by designating manually defined multiple ROIs and using the three logical operators AND, OR, and NOT (26). ROIs were determined by a radiologist (E.S.) and neurosurgeron (O.G., 13 years of experience in brain MR imaging) working in consensus and by using T2-weighted MR images or DT imaging–based color-coded maps (for the superior longitudinal fasciculus) according to the guidelines given by Wakana et al (27) and Mori et al (8). The tracking procedure was stopped when a track-turning angle greater than 60° was encountered. An FA threshold of 0.2 was used as the indicator for termination of tract elongation.

For visualization of the superior longitudinal fasciculus, one ROI was defined on a coronal section of the DT imaging–based color-coded maps at the posterior tip of the putamen. One ROI at the parieto-occipital sulcus, which was identified at the middle of the coronal section along the superior-inferior axis, was used for both the inferior longitudinal fasciculus and the inferior fronto-occipital fasciculus. For the inferior longitudinal fasciculus, an additional ROI was defined on a coronal section at the midtemporal lobe at the section level of the posterior tip of the putamen. For the inferior fronto-occipital fasciculus, an additional ROI was defined on a coronal section at the frontal lobe at the section level where the frontal and temporal lobes were separated. The ROIs for the inferior longitudinal fasciculus and the inferior fronto-occipital fasciculus were combined by using the operator AND (8).

The callosal fibers were reconstructed in three sections. The first ROI was placed in the corpus callosum at the midsagittal level. Two additional ROIs were combined with the operator AND and were placed on coronal sections in the corona radiata. One was placed anterior to the genu, and one was placed posterior to the splenium of the corpus callosum. One ROI was placed on transverse sections superior to the body of the corpus callosum. The three reconstructed fiber bundles were combined with the operator OR. The operator NOT was used to remove thalamic fibers, which were reconstructed occasionally (27).

The corticobulbar and corticospinal tracts were reconstructed by placing one ROI at the cerebral peduncle and one ROI at the internal capsule. Both ROIs were combined by using the operator AND. For reconstruction of the thalamic fibers, the entire thalamus was defined as the first ROI. The second ROI for the anterior and posterior thalamic radiations was placed on coronal sections to define the frontal lobe at the section level where the frontal and temporal lobes were separated and to define the occipital lobe at the section level of the posterior tip of the putamen, respectively. For superior thalamic radiation, the second ROI was defined on a transverse section above the corpus callosum, and it occupied the entire hemisphere. Again, the first and the second ROIs were combined by using the operator AND (8,27).

Quantitative fiber tracking of the reconstructed fibers was performed by using the tract statistics function of the DtiStudio software, which enables statistical evaluation of pixels occupied by reconstructed fibers. Diffusivity parameters, as well as fiber characteristics, were calculated for the selected fiber bundles. For total brain fibers, the diffusivity parameters (FA, MD, and eigenvalues) cannot be calculated with the tract statistics function of the DtiStudio software.

The number of fiber projections is the number of streamlines that were reconstructed and that penetrate the ROI. Hence, this parameter includes not only the fibers started from the points in the ROI but also the streamlines started from other points that pass through the ROI. Both the number of streamlines within each voxel of the reconstructed fiber structure and the number of voxels occupied by the fiber structure (measured in cubic millimeters) were counted and used to calculate the mean number of fiber projections per voxel. Therefore, the number of fiber projections per voxel was used to measure the fiber density of the selected fiber bundle. The number of fiber projections and the number of fiber projections per voxel are limited by the spatial resolution achieved and the technique applied in this study. Both parameters have to be interpreted as relative parameters and are not measures of the true number of axons passing through a voxel.

Statistical Methods
Data were analyzed by two authors (A.S., G.S.; each with 15 years of experience in statistics) by using statistical software (SPSS 12; SPSS, Chicago, Ill). A two-sided paired Student t test was used to compare differences between hemispheres. A Kolmogorov-Smirnov z test was used to compare age distributions between women and men. Bivariate linear regression models were used to analyze the influence of age on diffusivity parameters and fiber characteristics. The significance of differences between correlations was tested with the Hotelling-Williams test (33). Three authors (A.S., G.S., and O.G.) interpreted the correlation coefficients with consideration of the specifications for interpretation of correlation coefficients given by Zou et al (34). For all tests, the level of significance was set at P < .05.

To assess the influence of age and sex on diffusivity parameters and fiber characteristics, multiple linear regression models were used. Multiple linear correlations with sex as the control variable revealed that sex did not have a significant influence, and results were qualitatively similar to those obtained with the bivariate linear regression model. Thus, the authors thought it was justifiable to present the results for only the bivariate model.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Diffusivity Parameter
Association, callosal, and projection fibers were reconstructed for all 38 participants (Fig 1). Two-sided paired t tests revealed no significant differences between hemispheres for diffusivity parameters or fiber characteristics of association fibers (FA, P = .107; MD, P = .074; {lambda}1, P = .128; {lambda}2, P = .446; {lambda}3, P = .058; number of fiber projections, P = .349; and number of fiber projections per voxel, P = .073) and projection fibers (FA, P = .073; MD, P = .074; {lambda}1, P = .128; {lambda}2, P = .446; {lambda}3, P = .058; number of fiber projections, P = .349; and number of fiber projections per voxel, P = .073). For further statistical evaluation, we used (a) mean values from the left and right hemispheres for FA, MD, eigenvalues, and number of fiber projections per voxel and (b) the sum from the left and right hemispheres for the number of fiber projections.


Figure 1
View larger version (81K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Representative sets of fiber-tracking results obtained in, A, D, G, a 25-year-old man; B, E, H, a 55-year-old man; and, C, F, I, an 81-year-old woman and depicted on sagittal sections of the DT data set measured with a b value of 0 sec/mm2. A–C, Association fibers are divided into superior longitudinal (yellow), inferior fronto-occipital (orange), and inferior longitudinal (red) fasciculi. D–F, Callosal fibers (purple) and, G–I, projection fibers (blue) are also seen.

 
For FA, bivariate linear regression models revealed a strong negative correlation for association fibers (P < .001), a moderate negative correlation for callosal fibers (P = .002), and a weak negative correlation for projection fibers (P = .341) (Fig 2). The differences in correlation between association and projection fibers were significant (P = .002). There were no significant differences in correlation between callosal and projection fibers (P = .099) or between association and callosal fibers (P = .229) for FA.


Figure 2
View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: Scatterplots show correlation of FA and MD with age. A–C, FA versus age for, A, association, B, callosal, and C, projection fibers. D–E, MD versus age for, D, association, E, callosal, and F, projection fibers. The correlation coefficients and P values obtained with a bivariate linear regression model are included.

 
The results for MD differed from the results for FA. Similar moderate positive correlations were registered for all three fiber structures. No significant difference was found in the correlation of MD and age for the three fiber structures (association vs callosal fibers, P = .611; association vs projection fibers, P = .71; and callosal vs projection fibers, P = .869). The {lambda}1 of the three fiber structures showed moderately positive correlations that were dependent on age (association fibers: r = 0.349, P = .032; callosal fibers: r = 0.377, P = .02; and projection fibers: r = 0.39, P = .016). For {lambda}2, we found weakly positive correlations that were significant for association fibers (r = 0.315, P = .044) but not for callosal (r = 0.239, P = .148) or projection (r = 0.251, P = .128) fibers. The {lambda}3 showed moderately significant positive correlations that depended on age for all three fiber structures (association fibers: r = 0.317, P = .042; callosal fibers: r = 0.363, P = .025; and projection fibers: r = 0.378, P = .019). There were no significant differences in correlations of eigenvalues between the fiber structures. For {lambda}1, P values were as follows: .846 for association versus callosal fibers, .8 for association versus projection fibers, and .927 for callosal versus projection fibers. For {lambda}2, P values were as follows: .88 for association versus callosal fibers, .811 for association versus projection fibers, and .946 for callosal versus projection fibers. For {lambda}3, P values were as follows: .887 for association versus callosal fibers, .792 for association versus projection fibers, and .937 for callosal versus projection fibers.

Age-related Changes
Age-related changes in the number of fiber projections for association, callosal, projection, and total brain fibers are shown in Figure 3. For total brain fibers and all three fiber structures, a bivariate linear regression model showed there was a strong negative correlation between the number of fiber projections and age (P < .001). Only the difference between the correlations of projection fibers (r = –0.834) and total brain fibers (r = –0.688) was significant (P = .043) at Hotelling-Williams testing (ie, the number of fiber projections of the projection fibers decreased significantly more with age than did the number of fiber projections of the total brain fibers).


Figure 3
View larger version (17K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Scatterplots show correlation of number of fiber projections (NoF) with age in, A, association fibers, B, callosal fibers, C, projection fibers, and, D, total brain fibers. The correlation coefficients and P values obtained with a bivariate linear regression model are included.

 
The number of fiber projections per voxel showed similar results for association, callosal, and total brain fibers with regard to changes due to normal aging. The calculated correlation coefficients revealed moderately negative correlations for these two fiber structures and total brain fibers (P < .001). However, the correlation for projection fibers was not significant (P = .07) (Fig 4). Differences between the correlation of projection (r = –0.297) and association (r = –0.662) fibers, as well as between the correlation of projection and total brain fibers (r = –0.654), were significant (P = .005 and P = .016, respectively) at Hotelling-Williams testing.


Figure 4
View larger version (16K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: Scatterplots show correlation of mean number of fiber projections per voxel (FpV) with age in, A, association fibers, B, callosal fibers, C, projection fibers, and, D, total brain fibers. The correlation coefficients and P values obtained with a bivariate linear regression model are included.

 
Association fibers had the largest decrease both in FA per decade of age (–1.1%) and in the number of fiber projections per voxel per decade of age (–4.5%) (Table 2). We found callosal fibers had the largest increase in MD (2.2% per decade of age) and the largest relative change in the number of fiber projections (–10.2% per decade of age). Projection fibers had the largest absolute change in the number of fiber projections (–221 fibers per decade of age).


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

  		Table 2. Absolute and Relative Changes in Fiber and Diffusivity Parameters in Fiber Structures per Decade of Age

 
Correlation with White Matter Volumes
The absolute and relative changes in the white matter volume, obtained via segmentation and volumetry of T1-weighted magnetization-prepared rapid acquisition gradient-echo data, were 21.5 cm3 and –5.3%, respectively. A bivariate linear regression model revealed a moderate negative correlation with age (r = –0.503, P = .001) for white matter volume.

Hotelling-Williams testing revealed significant differences between correlations for the number of fiber projections for fiber structures and the white matter volume (P = .01 for association fibers, P = .039 for callosal fibers, P < .001 for projection fibers, and P < .001 for total brain fibers). However, the differences between correlations for the number of fiber projections per voxel for the fiber structures and white matter volume were not significant (P = .227 for association fibers, P = .792 for callosal fibers, P = .27 for projection fibers, and P = .192 for total brain fibers).

For diffusivity parameters, only FA in projection fibers had a significantly weaker correlation compared with the correlation of white matter volume (P = .027). All other correlations for diffusivity parameters were not significantly different from the correlation of white matter volume with age. For association fibers, P values were as follows: .303 for FA, .525 for MD, .4 for {lambda}1, .155 for {lambda}2, and .171 for {lambda}3. For callosal fibers, P values were as follows: .874 for FA, .975 for MD, .505 for {lambda}1, .196 for {lambda}2, and .439 for {lambda}3. For projection fibers, P values were as follows: .853 for MD, .545 for {lambda}1, .224 for {lambda}2, and .38 for {lambda}3.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Understanding patterns in age-related changes in neuronal connections is important if one wishes to interpret cognitive and behavioral changes throughout adulthood. To our knowledge, DT imaging is the only noninvasive method with which to identify fiber bundles. Evaluation of DT data with quantitative fiber tracking enables the segmentation of defined fiber structures as three-dimensional volumes of interest and the quantification of parameters for characterization of white matter changes.

Our results for the age-related changes in fiber characteristics in projection fibers—a large significant decrease in the number of fiber projections and a small insignificant decrease in the number of fiber projections per voxel—may be explained by the fact that an age-related decrease in the number of fiber projections may be associated with a decrease in the volume of the fiber structure that resulted in a smaller reduction in fiber density (the number of fiber projections per voxel) and FA. On the other hand, it appears that the volume of association fiber structures decreases the least compared with the volume of other white matter structures investigated in this study. The largest decrease was registered for the number of fiber projections per voxel. This finding is supported by the fact that the decrease in FA was largest in association fibers. This decrease was related to the largest and only significant increase in the perpendicular component of diffusion {lambda}2, and it may have been due to increased extracellular space (11,12).

Age-related changes in callosal fibers occupied an intermediate position: The most pronounced decrease in the number of fiber projections was associated with a large decrease in both the number of fiber projections per voxel and the FA, indicating a similar decrease in callosal volume. However, in our estimation, the most important parameter was the number of fiber projections, which was most affected by age-related degradation of fiber structures, including frontal white matter (callosal and projection fibers). Furthermore, there was a significantly larger decrease in the number of fiber projections in the three fiber structures and the total brain fibers compared with the decrease in total white matter volume, indicating additional age-related changes beyond the well-known brain atrophies.

In a number of studies, researchers used DT imaging and manually defined ROIs to determine age-related changes in the diffusivity of water in different regions of white matter. Study results have shown a significant age-related decrease in tissue volume in frontal white matter (35), the genu of the corpus callosum (36–38), and the posterior limb of the internal capsule (39,40). These results are in agreement with our findings, as we registered a significant correlation between FA and age in callosal fibers. The use of manually defined two-dimensional ROIs in the evaluation of DT data limits the detection of a three-dimensional process, such as that which occurs with aging in white matter structures of the brain. Advanced and interesting approaches for three-dimensional evaluation of age-related changes in DT data have been presented in two relatively recent studies (12,21).

Nusbaum et al (12) co-registered DT data to T1-weighted MR images and standardized maps of relative anisotropy to normal coordinates for statistical probability mapping. Pixels with which there was a significant correlation between relative anisotropy and age were calculated with a linear regression model. Significantly decreased anisotropy with increasing age was found in the frontal white matter, the genu and splenium of the corpus callosum, and the periventricular white matter. Significant increases in relative anisotropy were found in the internal capsules on both sides. Nusbaum et al (12) concluded that their technique may permit global assessment of changes in the organization of white matter pathways that occur with normal aging.

Pfefferbaum et al (21) presented an anteroposterior profile analysis of the integrity of the white matter microstructure across supratentorial regions by using DT data collected at 3 T. The DT imaging profile analysis revealed that the frontal distribution of low white matter FA was more robust in healthy older adults than in healthy younger adults. Sullivan et al (23) described the use of a quantitative fiber-tracking approach to examine age-related degradation in diffusivity parameters (FA and apparent diffusion coefficient) and the number of fiber projections in six regions of the corpus callosum. In the frontal callosal fibers particularly, older participants had lower FA values, higher apparent diffusion coefficient values (apparent diffusion coefficient is comparable to MD), and fewer fibers than did younger participants. Sullivan et al (23) concluded that the quantitative fiber-tracking approach enables one to confirm the selective vulnerability of frontal callosal fibers to normal aging as the mechanism underlying age-related decreases in cognition. The results of a study by Ota et al (41), in which the authors used quantitative fiber tracking, confirmed these results for age-related FA and MD changes in frontal and callosal fibers. However, unlike Sullivan et al (23) and us, Ota et al (41) did not evaluate fiber characteristics. The evaluation of fiber characteristics yields detailed information about the mechanisms of age-related changes in the fiber structures of white matter.

Our study was limited by the rather small number of participants examined. However, the number of participants in each age group was more or less equal. A further limitation might have been the fact that we did not include other fiber structures and did not investigate differences between the substructures of association, callosal, and projection fibers. These issues should be addressed in future studies (eg, for association fibers, projection fibers, or short-range U fibers in different lobes), with a view toward identifying potential differences between the substructures.

In conclusion, we found that quantitative evaluation of fiber-tracking results enabled us to identify differences in age-related changes in diffusivity parameters and fiber characteristics between different fiber structures. The degradation in the number of fiber projections was stronger compared with the decrease in total white matter volume, suggesting that a process more complex than the loss of white matter volume occurs in the normal aging brain.


   ADVANCE IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE
 References
 


   IMPLICATION FOR PATIENT CARE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
References
 


   FOOTNOTES
 

Abbreviations: DT = diffusion tensor • FA = fractional anisotropy • MD = mean diffusivity • ROI = region of interest

Author contributions: Guarantors of integrity of entire study, A.S., E.S., T.H., O.G.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, A.S., E.S., G.S., O.G.; clinical studies, A.S., E.S., T.H.; statistical analysis, A.S., E.S., G.S., O.G.; and manuscript editing, all authors

Authors stated no financial relationship to disclose.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCE IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 

  1. Walhovd KB, Fjell AM, Reinvang I, et al. Effects of age on volumes of cortex, white matter and subcortical structures. Neurobiol Aging 2005;26(9):1261–1270.[CrossRef][Medline]
  2. 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(1 pt 1):21–36.[CrossRef][Medline]
  3. Sullivan EV, Rosenbloom M, Serventi KL, Pfefferbaum A. Effects of age and sex on volumes of the thalamus, pons, and cortex. Neurobiol Aging 2004;25(2):185–192.[CrossRef][Medline]
  4. Kruggel F. MRI-based volumetry of head compartments: normative values of healthy adults. Neuroimage 2006;30(1):1–11.[CrossRef][Medline]
  5. Jernigan TL, Archibald SL, Fennema-Notestine C, et al. Effects of age on tissues and regions of the cerebrum and cerebellum. Neurobiol Aging 2001;22(4):581–594.[CrossRef][Medline]
  6. Allen JS, Bruss J, Brown CK, Damasio H. Normal neuroanatomical variation due to age: the major lobes and a parcellation of the temporal region. Neurobiol Aging 2005;26(9):1245–1260.[CrossRef][Medline]
  7. Benedetti B, Charil A, Rovaris M, et al. Influence of aging on brain gray and white matter changes assessed by conventional, MT, and DT MRI. Neurology 2006;66(4):535–539.[Abstract/Free Full Text]
  8. Mori S, Kaufmann WE, Davatzikos C, et al. Imaging cortical association tracts in the human brain using diffusion-tensor-based axonal tracking. Magn Reson Med 2002;47(2):215–223.[CrossRef][Medline]
  9. Holodny AI, Gor DM, Watts R, Gutin PH, Ulug AM. Diffusion-tensor MR tractography of somatotopic organization of corticospinal tracts in the internal capsule: initial anatomic results in contradistinction to prior reports. Radiology 2005;234(3):649–653.[Abstract/Free Full Text]
  10. Le Bihan D, Mangin JF, Poupon C, et al. Diffusion tensor imaging: concepts and applications. J Magn Reson Imaging 2001;13(4):534–546.[CrossRef][Medline]
  11. Meier-Ruge W, Ulrich J, Bruhlmann M, Meier E. Age-related white matter atrophy in the human brain. Ann N Y Acad Sci 1992;673:260–269.[Medline]
  12. 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(1):136–142.[Abstract/Free Full Text]
  13. Conturo TE, Lori NF, Cull TS, et al. Tracking neuronal fiber pathways in the living human brain. Proc Natl Acad Sci U S A 1999;96(18):10422–10427.[Abstract/Free Full Text]
  14. Poupon C, Clark CA, Frouin V, et al. Regularization of diffusion-based direction maps for the tracking of brain white matter fascicles. Neuroimage 2000;12(2):184–195.[CrossRef][Medline]
  15. Jones DK, Simmons A, Williams SC, Horsfield MA. Non-invasive assessment of axonal fiber connectivity in the human brain via diffusion tensor MRI. Magn Reson Med 1999;42(1):37–41.[CrossRef][Medline]
  16. Mori S, Crain BJ, Chacko VP, van Zijl PC. Three-dimensional tracking of axonal projections in the brain by magnetic resonance imaging. Ann Neurol 1999;45(2):265–269.[CrossRef][Medline]
  17. Concha L, Beaulieu C, Gross DW. Bilateral limbic diffusion abnormalities in unilateral temporal lobe epilepsy. Ann Neurol 2005;57(2):188–196.[CrossRef][Medline]
  18. Roberts TP, Liu F, Kassner A, Mori S, Guha A. Fiber density index correlates with reduced fractional anisotropy in white matter of patients with glioblastoma. AJNR Am J Neuroradiol 2005;26(9):2183–2186.[Abstract/Free Full Text]
  19. Schonberg T, Pianka P, Hendler T, Pasternak O, Assaf Y. Characterization of displaced white matter by brain tumors using combined DTI and fMRI. Neuroimage 2006;30(4):1100–1111.[CrossRef][Medline]
  20. Berman JI, Mukherjee P, Partridge SC, et al. Quantitative diffusion tensor MRI fiber tractography of sensorimotor white matter development in premature infants. Neuroimage 2005;27(4):862–871.[CrossRef][Medline]
  21. Pfefferbaum A, Adalsteinsson E, Sullivan EV. Frontal circuitry degradation marks healthy adult aging: evidence from diffusion tensor imaging. Neuroimage 2005;26(3):891–899.[CrossRef][Medline]
  22. Benson AD, Slavin MJ, Tran TT, Petrella JR, Doraiswamy PM. Screening for early Alzheimer's disease: is there still a role for the mini-mental state examination? Prim Care Companion J Clin Psychiatry 2005;7(2):62–69.[Medline]
  23. Sullivan EV, Adalsteinsson E, Pfefferbaum A. Selective age-related degradation of anterior callosal fiber bundles quantified in vivo with fiber tracking. Cereb Cortex 2006;16(7):1030–1039.[Abstract/Free Full Text]
  24. 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(3):189–198.[CrossRef][Medline]
  25. Kriegeskorte N, Goebel R. An efficient algorithm for topologically correct segmentation of the cortical sheet in anatomical MR volumes. Neuroimage 2001;14(2):329–346.[CrossRef][Medline]
  26. Jiang H, van Zijl PC, Kim J, Pearlson GD, Mori S. DtiStudio: resource program for diffusion tensor computation and fiber bundle tracking. Comput Methods Programs Biomed 2006;81(2):106–116.[CrossRef][Medline]
  27. Wakana S, Jiang H, Nagae-Poetscher LM, van Zijl PC, Mori S. Fiber tract–based atlas of human white matter anatomy. Radiology 2004;230(1):77–87.[Abstract/Free Full Text]
  28. Basser PJ, Mattiello J, LeBihan D. Estimation of the effective self-diffusion tensor from the NMR spin echo. J Magn Reson B 1994;103(3):247–254.[CrossRef][Medline]
  29. Basser PJ, Mattiello J, LeBihan D. MR diffusion tensor spectroscopy and imaging. Biophys J 1994;66(1):259–267.[Medline]
  30. Pierpaoli C, Basser PJ. Toward a quantitative assessment of diffusion anisotropy. Magn Reson Med 1996;36(6):893–906.[Medline]
  31. Mori S, van Zijl PC. Fiber tracking: principles and strategies—a technical review. NMR Biomed 2002;15(7-8):468–480.[CrossRef][Medline]
  32. Xue R, van Zijl PC, Crain BJ, Solaiyappan M, Mori S. In vivo three-dimensional reconstruction of rat brain axonal projections by diffusion tensor imaging. Magn Reson Med 1999;42(6):1123–1127.[CrossRef][Medline]
  33. Williams EJ. The comparison of regression variables. J R Stat Soc Series B 1959;21:395–399.
  34. Zou KH, Tuncali K, Silverman SG. Correlation and simple linear regression. Radiology 2003;227(3):617–622.[Abstract/Free Full Text]
  35. Madden DJ, Whiting WL, Huettel SA, White LE, MacFall JR, Provenzale JM. Diffusion tensor imaging of adult age differences in cerebral white matter: relation to response time. Neuroimage 2004;21(3):1174–1181.[CrossRef][Medline]
  36. Abe O, Aoki S, Hayashi N, et al. Normal aging in the central nervous system: quantitative MR diffusion-tensor analysis. Neurobiol Aging 2002;23(3):433–441.[CrossRef][Medline]
  37. 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(2):259–268.[CrossRef][Medline]
  38. Zhang YT, Zhang CY, Zhang J, Li W. Age-related changes of normal adult brain structure: analysed with diffusion tensor imaging. Chin Med J (Engl) 2005;118(13):1059–1065.[Medline]
  39. Salat DH, Tuch DS, Greve DN, et al. Age-related alterations in white matter microstructure measured by diffusion tensor imaging. Neurobiol Aging 2005;26(8):1215–1227.[CrossRef][Medline]
  40. Head D, Buckner RL, Shimony JS, et al. Differential vulnerability of anterior white matter in nondemented aging with minimal acceleration in dementia of the Alzheimer type: evidence from diffusion tensor imaging. Cereb Cortex 2004;14(4):410–423.[Abstract/Free Full Text]
  41. Ota M, Obata T, Akine Y, et al. Age-related degeneration of corpus callosum measured with diffusion tensor imaging. Neuroimage 2006;31(4):1445–1452.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Br. J. PsychiatryHome page
C. A. Chaddock, G. J. Barker, N. Marshall, K. Schulze, M. H. Hall, A. Fern, M. Walshe, E. Bramon, X. A. Chitnis, R. Murray, et al.
White matter microstructural impairments and genetic liability to familial bipolar I disorder
The British Journal of Psychiatry, June 1, 2009; 194(6): 527 - 534.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2471070707v1
247/1/179    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 Stadlbauer, A.
Right arrow Articles by Ganslandt, O.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Stadlbauer, A.
Right arrow Articles by Ganslandt, O.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE