Abstract
Recent in vivo MRI studies at 7.0 T have demonstrated extensive heterogeneity of T2* relaxation in white matter of the human brain. In order to study the origin of this heterogeneity, we performed T2* measurements at 1.5, 3.0, and 7.0 T in normal volunteers. Formalin-fixed brain tissue specimens were also studied using T2*-weighted MRI, histological staining, chemical analysis, and electron microscopy.
We found that T2* relaxation rate (R2*=1/ T2*) in white matter in living human brain is linearly dependent on the main magnetic field strength and the T2* heterogeneity in white matter observed at 7.0 T can also be detected, albeit weaker, at 1.5 and 3.0 T. The T2* heterogeneity exists also in white matter of the formalin fixed brain tissue specimens, with prominent differences between the major fiber bundles such as the cingulum and the superior corona radiada. The white matter specimen with substantial difference in T2*have no significant difference in the total iron content as determined by chemical analysis. On the other hand, evidence from histological staining and electron microscopy demonstrate these tissue specimen have apparent difference in myelin content and microstructure.
Keywords: T2*, Heterogeneity, Susceptibility, MRI contrast, Iron, Myelin, Histology
Introduction
Unlike T1- and T2-weighted MRI, which are routinely used for clinical examinations, the application of T2*-weighted MRI in clinical practice has been rather limited. Successful applications include venography (1), detection of hemorrhagic infarcts (2) and mapping areas of the brain with high iron content (3-6). Recent high-field studies at 7T and above have shown a dramatic increase in T2* contrast, allowing detection of laminar architecture of the human cerebral cortex and white matter microstructures (7,8).
In brain tissues, a number of sources can give rise to the magnetic susceptibility gradients underlying the T2* contrast (7,8). While iron, deoxy-hemoglobin, and myelin content, all may contribute to this contrast, their relative contributions remain unknown. For better understanding of the underlying mechanism of the extensive T2* heterogeneity observed in vivo in white matter, in this study, we experimentally investigated the B0 dependence of T2* heterogeneity among different white matter bundles and addressed the issue of whether the T2* variations among the different fiber bundles can also be detected at the clinically more relevant magnetic field of 1.5 and 3.0 T. We investigated also whether similar T2* contrast heterogeneity exist in formalin fixed brain tissue specimens in which additional measurements, such as, iron concentration, myelin content, and fiber microstructure were conducted in order to identify the sources contributing to the observed T2* heterogeneity.
Materials and Method
In vivo MRI Measurements of Normal Volunteers
MRI experiments were performed on three separate GE Signa whole-body MRI scanners (General Electric Medical Systems, Milwaukee, WI) with field strengths of 1.5, 3.0 and 7.0 T. Each scanner was equipped with a Twin-Speed gradient system capable of providing the maximum gradient amplitude of 40 mT/m and a slew rate of 150 T/m/s. Multi-channel (8-16 channels) phased-array detectors (Nova Medical, Wilmington, MA) were used for signal reception. Head motion was mitigated by inserting foam padding between the coil former and the subjects’ head.
Eight normal volunteers (female/male=3/5) with ages = 31 ± 5 years participated in the study and each completed three scan sessions at 1.5, 3.0 and 7.0 T, after having provided informed consents in accordance with MRI protocols approved by the Institutional Review Board of National Institute of Neurological Disorders and Stroke, NIH. At each field strength, the MRI protocol included the following scans: 1) a fast 3-plane localizer acquired using a spoiled gradient recalled echo pulse sequence; 2) whole-brain magnetic field shimming up to the 2nd order based on gradient recalled echo (GRE) spiral acquisitions at two different echo times; 3) T2*-mapping by multiple T *2-weighted scan sessions using a multiple-echo GRE pulse sequence; 4) a low resolution (64×64) coil sensitivity mapping using the standard GRE pulse sequence with TE/TR=7.3/400 ms and flip angle=90°; 5) a 3D T1-weighted scan using a pulse sequence based on magnetization prepared rapid gradient echo (MPRAGE) technique.
For T2* mapping, T2*-weighted images at four different echo times were acquired on each scanner. The acquisition parameters used at each magnetic field strength were kept as close as possible except for systematic adjustment of the echo times to adequately sample the signal decay curve at each field. In order to achieve similar SNR on the different scanners, the slice thickness/spacing, and number of signal average (NEX) were also adjusted. The main differences in data acquisition parameters for the in vivo T2*-mapping on the three scanners are detailed in Table 1. Other acquisition parameters were kept constant across the different scanners. These included the repetition time (TR) = 1000 ms, flip angle = 60°, bandwidth = 32 kHz, field of view (FOV) = 240 × 180 mm, matrix size 256×192, resulting in an in-plane resolution of 0.94×0.94 mm2. Sixteen axial slices were prescribed from top to bottom with the 5th slice through the cingulate cortex.
Table 1.
The main differences in data acquisitions for in vivo T2*-mapping on each scanner.
| B0 | 1.5 T | 3.0 T | 7.0 T |
|---|---|---|---|
| TE (ms) | 7.3 | 7.5 | 7.6 |
| 23.9 | 21.7 | 16.4 | |
| 40.5 | 35.9 | 23.2 | |
| 57.0 | 52.0 | 30 | |
| NEX | 4 | 2 | 1 |
| Slice thickness (mm) | 2 | 2 | 1 |
| Slice spacing (mm) | 1 | 1 | 2 |
In order to obtain a white matter mask for each T2* map, a 3D T1-weighted scan based on the MPRAGE technique was performed for each subject on all three scanners. Similarly, the acquisition parameters were kept as close as possible except for necessary adjustment of the inversion time (TI) to achieve the optimal contrast between gray and white matter. The TI values used at 1.5, 3.0, and 7.0T were 750, 900, 1200 ms, respectively. The other acquisition parameters for the MPRAGE scans were as followins: TR/TE=6.5/3.2 ms; flip angle = 8°; slice thickness = 1.5 mm; FOV = 240 × 180 mm, matrix size = 256×192, which results in the same in-plane resolution as the corresponding GRE images; and bandwidth = 32 kHz. For each subject the slab center of the 3D MPRAGE was carefully selected to match the center slice of the T2*-weighted scans. Therefore, the centers of slice locations between the MPRAGE and the T2*-weighted images overlapped and the white matter mask derived from the MPRAGE scans was directly applied to the T2* maps.
Analysis of the in vivo MRI data
Images were reconstructed off-line by using phase-sensitive combination of the individual coil data (9). The coil sensitivity reference was derived from the low-resolution GRE scans by doing a spatial low-pass-filtering to an in-plane resolution of 15 mm. T *2 maps were generated on a pixel-by-pixel basis by fitting a single exponential function to the T2*-weighted image intensities of different echo times using a least-square minimization algorithm in Matlab (the Mathworks Inc, Natick, MA). For better definition of the different tissue types and regions, the outlines of the matched MPRAGE images were first overlaid onto the corresponding T2* maps to segment gray and white matter. The T2* value for each region of interest (ROI) was evaluated as the average of multiple small ROIs each containing 9 voxels (3×3). For each white matter ROI, 6-8 small ROIs were manually selected by using a dedicated interface program based on Matlab. Care was taken in selecting each small ROI to avoid apparent large vessels and cerebrospinal fluid (CSF). The average T2* values for each subject at each magnetic field strength in seven white matter regions with relatively homogenous T2* values were evaluated. These white matter regions were: superior region of corona radiata (SCR); splenium of corpus callosum (SPL), genu of corpus callosum (GCC), tapetum (Tap), posterior region of coronal radiata (PCR), superior longitudinal fasciculus (SLF), and cingulum (CG).
Ex vivo Studies of Brain Tissue Specimen
In order to compare T2*-contrast between in-vivo and post-mortem brain, three formalin-fixed samples from males (age= 66-78) who died of non-neurological causes were also scanned at 7 T using a 2D gradient-recalled echo pulse sequence with similar acquisition parameters described for the in vivo measurements. To gain further insight into the underlying contrast mechanisms, a coronal slab containing the middle section of CG fiber bundle as well as the SCR was dissected from one of the post-mortem brain specimens and the slab was scanned at 7.0 Twith higher spatial resolution before it was sampled for further histological and chemical analysis. The relevant acquisition parameters used for the T2*-weighted MRI of the brain slab were as follows: TE/TR = 30/800 ms, flip angle= 30°, bandwidth = 32 kHz, FOV = 200 × 100 mm, matrix size = 1024 ×512, thickness = 1 mm, and NEX=4 giving rise to ~0.2 ×0.2 × 1mm3 voxel size. The same image data processing procedure as used in the in vivo data was performed.
After MRI scanning, six tissue specimens of approximate 2×2×2 mm3 were sampled in the CG and SCR structures for chemical analysis of the iron content. The tissue samples were sealed and sent to West Coast Analytical Service Inc (Santa Fe Springs, CA) for quantitative iron determination using inductively coupled plasma-mass spectrometry. The iron detection limit with this method is about 50 ppb (part per billion), therefore, a small sample in the order of 1 mg is sufficient.
Histological staining of the fixed brain tissues was performed at American Histolab Inc (Gaithersburg, MD) using the Luxol Fast Blue (10) and Bielschowsky methods (11).
Multiple regions of the CG and SCR structures were also examined using transmission electron miscopy (Jeol 1200 EX) with up to 10,000× direct amplification in order to compare the microstructure differences of the white matter fiber bundles. The micrographs were segmented using Matlab to estimate the area fraction of the myelinated structures and surface order parameter as defined for the nematic liquid crystals (12).
RESULTS
B0 dependence of T2* heterogeneity in white matter
Figure 1 shows a typical set of high-resolution T2*-weighted images with spatial resolution of 0.2×0.2×1 mm3 acquired at 7.0 T. It shows excellent contrast difference between the different white matter fiber bundles. A few white matte fiber pathways that can be readily identified are labeled in Fig. 1. These include superior region of corona radiata (SCR); Splenium of corpus callosum (SPL), genu of corpus callosum (GCC), external capsule (EC), superior frontal-occipital fasciculus (SFO), tapetum (Tap), posterior region of coronal radiata (PCR), superior longitudinal fasciculus (SLF), anterior commissar (AC), and cingulum (CG). As demonstrated previously at 7.0T, the intensity contrast observed in white matter in the T2*-weighted images is largely related to T2* heterogeneity. The proton density maps exhibited much less variations in white matter (8). This observation is further confirmed by the T2* mapping results.
FIG. 1.
A set of T2*-weighted images acquired at 7T from a normal human brain. The acquisition parameters were: TR/TE=800/30 ms, flip angle=30°, receiver bandwidth=32 kHz, matrix size 1024×768, FOV=220×165 mm2, slice thickness=1 mm. Substantial contrast difference is seen between gray matter, CSF, vessels, and white matter. Also remarkable is the contrast difference between the different fiber projections. The labeled white matter bundles include: superior region of corona radiata (SCR); Splenium of corpus callosum (SPL), genu of corpus callosum (GCC), external capsule (EC), superior frontal-occipital fasciculus (SFO), tapetum (Tap), posterior region of coronal radiata (PCR), superior longitudinal fasciculus (SLF), anterior commissar (AC), and cingulum (CG).
Table 2 shows the T2* values measured in 8 subjects at 3 different magnetic fields for 7 different white matter fiber bundles throughout the brain. The mean values for the different fiber bundles at 1.5, 3.0, and 7.0T are 69±8, 50±8, and 28±5ms, respectively. With the increase of the main magnetic field strength, not only is the T2* relaxation rate (R2*=1/T2*) increased, but also the T2* heterogeneity among the different fiber bundles is enhanced. The coefficients of variance for T2* at 1.5, 3.0, and 7.0T are 13, 16 and 19%, respectively. Using the T2* value for the SPL fiber bundle as a reference, at 7T the T2* values for all other measured fiber bundles are significantly different (p<0.02). Comparing 7.0T with 1.5 T, the T2* contrast difference between the different fiber bundles is enhanced by a factor of 4. The 3.0 T result is intermediate between the 7 T and 1.5 T results. The enhancement of T2* relaxation and its heterogeneity with the main magnetic field strength are more explicitly summarized in Fig. 2, which depicts T2* relaxation rate (R2* =1/T2*) for the different white matter fiber bundles as a function of the main magnetic field strength. There is nearly linear dependence, with the relaxation rate increasing proportional to the field strength.
Table 2.
T2*±std (ms) at three different field strengths for the different fiber bundles in individual subjects as well as their inter-subject averages. The least square fitting results of the linear equation, R2*=α+β*B0; to the data are listed at the end.
| Subject | B0 (Tesla) | T2* ± std (ms) for different white matter bundles |
||||||
|---|---|---|---|---|---|---|---|---|
| CG | SCR | GCC | SPL | SLF | PCR | Tap | ||
| 1 | 1.5 | 67±4 | 81±6 | 53±7 | 75±5 | 75±6 | 62±4 | 78±4 |
| 3.0 | 46±4 | 55±4 | 36±4 | 51±5 | 56±8 | 45±6 | 57±6 | |
| 7.0 | 26±3 | 31±4 | 24±3 | 29±6 | 27±3 | 20±1 | 32±4 | |
| 2 | 1.5 | 61±7 | 71±6 | 56±6 | 73±7 | 74±8 | 66±8 | 80±9 |
| 3.0 | 42±6 | 54±5 | 36±6 | 50±7 | 54±2 | 45±7 | 60±7 | |
| 7.0 | 21±3 | 31±2 | 20±3 | 26±3 | 30±4 | 26±3 | 35±5 | |
| 3 | 1.5 | 66±6 | 85±8 | 62±5 | 70±7 | 73±7 | 62±6 | 79±7 |
| 3.0 | 47±5 | 59±5 | 43±6 | 46±5 | 64±7 | 51±6 | 59±8 | |
| 7.0 | 25±3 | 32±3 | 26±3 | 33±3 | 35±4 | 29±6 | 36±6 | |
| 4 | 1.5 | 59±4 | 79±7 | 56±6 | 62±7 | 68±5 | 65±5 | 73±6 |
| 3.0 | 40±4 | 52±6 | 40±4 | 41±6 | 48±8 | 43±8 | 58±7 | |
| 7.0 | 19±1 | 33±2 | 18±3 | 27±2 | 32±4 | 24±3 | 34±4 | |
| 5 | 1.5 | 66±5 | 78±6 | 59±5 | 66±5 | 74±7 | 67±6 | 83±8 |
| 3.0 | 42±4 | 66±5 | 34±4 | 53±5 | 62±6 | 49±5 | 62±8 | |
| 7.0 | 22±3 | 35±4 | 22±4 | 28±3 | 34±3 | 26±4 | 35±4 | |
| 6 | 1.5 | 61±2 | 75±3 | 60±3 | 71±2 | 70±4 | 63±5 | 73±7 |
| 3.0 | 45±3 | 57±4 | 39±3 | 53±5 | 52±4 | 47±5 | 55±6 | |
| 7.0 | 22±3 | 32±3 | 21±2 | 28±2 | 31±4 | 25±3 | 35±5 | |
| 7 | 1.5 | 60±3 | 76±6 | 52±4 | 77±6 | 79±6 | 62±5 | 80±7 |
| 3.0 | 54±4 | 66±6 | 36±2 | 45±5 | 60±6 | 51±5 | 56±6 | |
| 7.0 | 19±2 | 33±2 | 18±3 | 29±3 | 34±4 | 22±4 | 33±6 | |
| 8 | 1.5 | 63±8 | 76±8 | 60±7 | 67±6 | 76±6 | 61±7 | 84±8 |
| 3.0 | 41±3 | 60±7 | 38±3 | 48±5 | 54±6 | 39±4 | 56±7 | |
| 7.0 | 22±2 | 33±3 | 21±2 | 29±2 | 33±5 | 29±4 | 34±5 | |
| Group | 1.5 | 63±3 | 78±4 | 57±4 | 70±5 | 74±3 | 64±2 | 79±4 |
| 3.0 | 45±5 | 59±5 | 36±4 | 48±4 | 56±5 | 46±4 | 58±2 | |
| 7.0 | 22±3 | 32±1 | 21±3 | 29±2 | 32±3 | 25±3 | 34±1 | |
| β (Hz/Tesla) | 5.45 | 3.29 | 5.33 | 3.72 | 3.24 | 4.41 | 3.00 | |
| α (Hz) | 7.04 | 7.63 | 9.90 | 9.01 | 8.40 | 8.85 | 8.26 | |
| CC | 0.998 | 0.999 | 0.999 | 0.999 | 0.999 | 0.999 | 1.000 | |
FIG. 2.
The T2* relaxation rate (R2* =1/T2*) for the different white matter fiber bundles as a function of the main magnetic field strength. The lines are least square fit of the equation R2*=α+β*B0. The fitting results for each fiber bundles are detailed in Table 2. Increasing the main magnetic field strength from 1.5T to 7.0T, the T2* relaxation contrast heterogeneity is significantly enhanced by more than 4 times.
Ex vivo investigation of formalin fixed brain tissue
Similar T2*-weighted contrast was observed in white matter in the formalin fixed brains, when the T2*-weighted images were acquired using similar acquisition parameters as used for the in vivo measurements As shown in Fig. 3a, the signal intensity for the SCR region is also substantially higher than that for CG (~30%). However, quantitative T2* comparison with the in vivo results may not be very informative, because T2* values measured in fixed brain tissues are affected by the fixation process and the level of dehydration during sample preparation. To better understand the underlying mechanisms for T2* heterogeneity in white matter, the ex vivo study was focused on the contrast difference between the CG and SCR fiber bundles.
FIG. 3.
(a) a T2*-weighted image acquired at 7.0 T in a coronal slab of formalin fixed brain tissue across the cingulate cortex. The acquisition parameters were: TR/TE=800/20 ms, flip angle=30°, receiver bandwidth=32 kHz, matrix size 1024×512, FOV=200×100 mm2, and slice thickness=1 mm, NEX=3. As detected in the living human brains, in the fixed brain tissue specimen there is also substantial contrast different between the superior region of corona radiata (SCR) and cingulum (CG). (b) myelin (Luxol fast blue) staining result for the SCR fiber bundle; (c) axonal staining (Bielschowsky's Silver Stain method) result for the SCR region. (d) myelin staining result for cingulum. (e) axonal staining result for the cingulum region. The myelin and axonal stains demonstrate that the white matter fiber bundles SCR and CG differ substantially in myelin content and axonal microstructure.
Table 3 details the measured iron contents in the CG and SCR regions. As shown, the average iron contents for the SCR and CG specimen are 176±13 and 182±9 μg/g, respectively. The iron contents between SCR and CG regions are not significantly different (p>0.5). The measured iron content from this study is consistent with previously reported literature results (4,13,14) for white matter.
Table 3.
The iron contents as determined in tissue specimen of the human brains1.
| Specimen | Iron (μg)1 | Dry weight (mg) |
|---|---|---|
| SCR_1 | 161 | 5.2 |
| SCR_2 | 179 | 6.8 |
| SCR_3 | 187 | 7.1 |
| Average ± std | 176±13 | 6.4±1.0 |
| CG_1 | 172 | 5.2 |
| CG_2 | 188 | 3.1 |
| CG_3 | 185 | 3.1 |
| Average ± std | 182±9 | 3.8±1.2 |
The iron contents were measured using inductively coupled plasma-mass spectrometry at West Coast Analytical Service Inc (Santa Fe Springs, CA). The detection limit for iron is about 50 ppb.
The myelin and axonal staining results for the SCR region are shown in Figs. 3b and 3c respectively. The myelin and axonal staining results for the CG region are shown in Figs. 3d and 3e respectively. The histological results demonstrate that the white matter fiber bundles SCR and CG differ substantially in myelin content and axonal microstructure. Compared with the fibers in the SCR region, the CG bundle is apparently composed of more compact and coherently oriented fibers. The difference is further confirmed by the EM micrographs shown in Fig. 4. The myelin areas estimated from the micrographs for the SCR (Fig. 4a) and CG (Fig. 4b) specimen were 24 and 32%, respectively. The corresponding order parameters were 0.28 and 0.35, respectively.
FIG. 4.
(a) Transmission electron micrograph of superior region of corona radiata (SCR) showing less compact myelinated axons. The orientations of the axons are also less coherent. The magnification is 10,000×. (b) Transmission electron micrograph of cingulum showing more compact and coherent oriented axons. The magnification is 10,000×.
Discussion
Mechanisms underlying T2* heterogeneity in white matter
There are a few potential sources that can give rise to local field inhomogeneity as summarized previously (7,8). The effect of iron and myelin content, which has been the focus of the present study, will be discussed further below.
Differences in iron concentration could lead to variations of both T2 and T2*. Such effects are expected to increase with increasing field strength. Previous MRI relaxation studies (15) of ferritin solutions have demonstrated that both R1 and R2 relaxation rates are linear functions of B0 and iron content. As mentioned above, previous in vivo measurements of R2 and R2’ in iron rich regions of the brain have also demonstrated a linear dependence on B0 and iron content (3,6,16). This leads us to speculate the linear dependence of T2* relaxation on B0 for the different white matter fiber bundles might be due to their difference in iron content. However, chemical analysis of the total iron content in brain tissue samples indicates that there is no significant difference in the total iron content between fiber bundles with substantial difference in T2*, such as CG and SCR (Table 3). An iron-mediated T2* contrast would be consistent with the observed dark appearance of U-fibers in T2*-weighted MRI (e.g. Fig. 1), which are known to have a high iron concentration (17). However, it seems contradictory to the relative dark appearance of the optical radiations, which reportedly have a relatively low iron concentration (17,18). Therefore, iron appears no to be the predominant contrast mechanism for T2* contrast between the major fiber bundles in normal brain. A complication is that the detailed microscopic distribution of iron and iron form can affect T2* under some conditions. Therefore, some of the observed differences in T2* relaxation rate may be due to variations in distribution and form of iron within these bundles.
Variations in myelin content are prevalent across cerebral white matter and can also significantly affect T2 and T2* values. Previous studies have shown that myelin can disproportionately enhance T2 (and T2*) relaxation and mediate sensitivity to MT (19,20). It has been suggested that the main contributor to these effects is galactorcerebrocide (21) which is disproportionately prevalent in myelin compared to other cell membranes (22). The galactose head group of galactorcerebrocide has 4 hydroxyl groups and extends outward from the membrane. This permits more extensive hydration of the hydroxyl groups than many other lipid molecules. Therefore, the bound water fraction associated with myelin is greatly increased. A previous study (21) based on multilamellar vesicle suspensions of different lipid compositions has shown that galactorcerebrocide contributes most significantly to the magnetization transfer and T2 shortening.
A higher fraction of bound water could lead to enhanced T2 relaxation and MT effect in heavily myelinated fibers through incoherent dephasing. Indeed, MT-weighted measurements suggest a high fraction of bound water in major fiber tracts and brain fasciculi (23), such as corpus callosum, cingulum, anterior commisure, and optical radiations. All of these fibers were found to have a relatively short T2* (see Tables 2). The rest of white matter has lower bound water fraction and relatively high T2*. Since T2 changes are included in T2* variations, the relaxation mechanisms giving rise to incoherent dephasing can become the dominant source behind the observed T2* heterogeneity. Both our histological and EM data collected in the CG and SCR regions support the notion that myelin content plays a major role in determining T2* relaxation of water in white mater fiber tracts.
A number of studies have reported on a general co-localization of brain iron and myelin (24-26). This co-localization derives from the fact that ferritin particles, which constitute the main storage pool of brain iron, are often located in the vicinity of myelin-generating oligodenrocytes. The current thinking is that this storage iron serves as a buffer for the iron demanding process of myelin generation and maintenance (27). Because of this general co-localization, iron is also expected to contribute more significantly to T2* relaxation in compact white mater fiber tracts. According to the empirical formula for iron content in fresh white matter (13), it can be estimated that the average iron content in white matter for the subject group (aged 31± 5 years) studied in this investigation is about 40 ppm (wet weight). Converting it into iron content per gram of dry tissue using a dry to wet brain tissue ratio of 0.22 (28,29), the result is 182 ppm, which agrees nearly perfectly with what was obtained from the chemical analysis of the fixed brain tissue samples (see table 3). In a recent B0 dependence study (6) of R2* in iron rich gray matter regions where iron related susceptibility is known to be the dominant source of the R2* relaxation mechanism, it has been shown that the slope of R2* dependence on B0 is approximately a linear function of the iron content (see fig. 8a in reference 6). Assuming the same linear correlation also applicable to the white matter regions studied here and using the above estimated iron content of 40 ppm, the expected slope of R2* dependence on B0 in the white matter tracts would be about 2.1 Hz/Tesla. However, as shown in Table 2, the experimentally measured values are in the range of 3.0 to 5.5 Hz/Tesla depending on the myelin content of the tracts. This indicates that myelin contribution to R2* is quite significant in the white matter regions and the myelin content is likely to be the main source underlying the R2* heterogeneity in white matter.
CONCLUSION
High-resolution T2*-weighted MRI at high field manifests a rich contrast offering a unique opportunity to study white matter architecture of the living human brain in fine detail. In particular in deep white matter, a strong contrast is apparent between the different fiber bundles. This contrast could potentially be exploited to study the microstructure differences of the major fiber pathways. Quantitative T2* measurements at three different magnetic fields confirm that the T2*-weighted contrast originates from microscopic susceptibility effects in the tissue. Preliminary results from studies of formalin fixed brain tissues indicate contrast differences between the major fiber bundles may primarily originate from differences in myelination.
Acknowledgements
The authors would like to acknowledge gratefully the experimental assistance from Drs. K. Matsuda and S. Chen. This work was supported by the intramural research program of the NIH, NINDS. A Matlab program for calculating the surface order parameter is available by contacting the corresponding author.
References
- 1.Reichenbach JR, Haacke EM. High-resolution BOLD venographic imaging: a window into brain function. NMR in biomedicine. 2001;14(7-8):453–467. doi: 10.1002/nbm.722. [DOI] [PubMed] [Google Scholar]
- 2.Scharf J, Brauherr E, Forsting M, Sartor K. Significance of haemorrhagic lacunes on MRI in patients with hypertensive cerebrovascular disease and intracerebral haemorrhage. Neuroradiology. 1994;36(7):504–508. doi: 10.1007/BF00593508. [DOI] [PubMed] [Google Scholar]
- 3.Gelman N, Gorell JM, Barker PB, Savage RM, Spickler EM, Windham JP, Knight RA. MR imaging of human brain at 3.0 T: preliminary report on transverse relaxation rates and relation to estimated iron content. Radiology. 1999;210(3):759–767. doi: 10.1148/radiology.210.3.r99fe41759. [DOI] [PubMed] [Google Scholar]
- 4.Haacke EM, Cheng NY, House MJ, Liu Q, Neelavalli J, Ogg RJ, Khan A, Ayaz M, Kirsch W, Obenaus A. Imaging iron stores in the brain using magnetic resonance imaging. Magnetic resonance imaging. 2005;23(1):1–25. doi: 10.1016/j.mri.2004.10.001. [DOI] [PubMed] [Google Scholar]
- 5.Waldvogel D, van Gelderen P, Hallett M. Increased iron in the dentate nucleus of patients with Friedrich's ataxia. Ann Neurol. 1999;46(1):123–125. doi: 10.1002/1531-8249(199907)46:1<123::aid-ana19>3.0.co;2-h. [DOI] [PubMed] [Google Scholar]
- 6.Yao B, Li TQ, Gelderen P, Shmueli K, de Zwart JA, Duyn JH. Susceptibility contrast in high field MRI of human brain as a function of tissue iron content. NeuroImage. 2009;44(4):1259–1266. doi: 10.1016/j.neuroimage.2008.10.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Duyn JH, van Gelderen P, Li TQ, de Zwart JA, Koretsky AP, Fukunaga M. High-field MRI of brain cortical substructure based on signal phase. Proceedings of the National Academy of Sciences of the United States of America. 2007;104(28):11796–11801. doi: 10.1073/pnas.0610821104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Li TQ, van Gelderen P, Merkle H, Talagala L, Koretsky AP, Duyn J. Extensive heterogeneity in white matter intensity in high-resolution T2*-weighted MRI of the human brain at 7.0 T. NeuroImage. 2006;32(3):1032–1040. doi: 10.1016/j.neuroimage.2006.05.053. [DOI] [PubMed] [Google Scholar]
- 9.de Zwart JA, Ledden PJ, Kellman P, van Gelderen P, Duyn JH. Design of a SENSE-optimized high-sensitivity MRI receive coil for brain imaging. Magn Reson Med. 2002;47(6):1218–1227. doi: 10.1002/mrm.10169. [DOI] [PubMed] [Google Scholar]
- 10.Kluver H, Barrera E. A method for the combined staining of cells and fibers in the nervous system. Journal of neuropathology and experimental neurology. 1953;12(4):400–403. doi: 10.1097/00005072-195312040-00008. [DOI] [PubMed] [Google Scholar]
- 11.Nauta WJ. Selective silver impregnation of degenerating axons in the central nervous system. Stain technology. 1952;27(3):175–179. doi: 10.3109/10520295209105080. [DOI] [PubMed] [Google Scholar]
- 12.Genes PGd. The Physics of Liquid Crystals. Oxford Press; 1974. [Google Scholar]
- 13.Hallgren B, Sourander P. The effect of age on the non-haemin iron in the human brain. Journal of neurochemistry. 1958;3(1):41–51. doi: 10.1111/j.1471-4159.1958.tb12607.x. [DOI] [PubMed] [Google Scholar]
- 14.Hebbrecht G, Maenhaut W, De Reuck J. Brain trace Elements and Aging. Nuclear Instruments and Methods in Physics Research B. 1999;150:208–213. [Google Scholar]
- 15.Vymazal J, Zak O, Bulte JW, Aisen P, Brooks RA. T1 and T2 of ferritin solutions: effect of loading factor. Magn Reson Med. 1996;36(1):61–65. doi: 10.1002/mrm.1910360111. [DOI] [PubMed] [Google Scholar]
- 16.Mitsumori F, Watanabe H, Takaya N, Garwood M. Apparent transverse relaxation rate in human brain varies linearly with tissue iron concentration at 4.7 T. Magn Reson Med. 2007;58(5):1054–1060. doi: 10.1002/mrm.21373. [DOI] [PubMed] [Google Scholar]
- 17.Drayer B, Burger P, Darwin R, Riederer S, Herfkens R, Johnson GA. MRI of brain iron. AJR Am J Roentgenol. 1986;147(1):103–110. doi: 10.2214/ajr.147.1.103. [DOI] [PubMed] [Google Scholar]
- 18.Chen JC, Hardy PA, Clauberg M, Joshi JG, Parravano J, Deck JH, Henkelman RM, Becker LE, Kucharczyk W. T2 values in the human brain: comparison with quantitative assays of iron and ferritin. Radiology. 1989;173(2):521–526. doi: 10.1148/radiology.173.2.2798884. [DOI] [PubMed] [Google Scholar]
- 19.Henkelman RM, Stanisz GJ, Kim JK, Bronskill MJ. Anisotropy of NMR properties of tissues. Magn Reson Med. 1994;32(5):592–601. doi: 10.1002/mrm.1910320508. [DOI] [PubMed] [Google Scholar]
- 20.Stanisz GJ, Kecojevic A, Bronskill MJ, Henkelman RM. Characterizing white matter with magnetization transfer and T(2). Magn Reson Med. 1999;42(6):1128–1136. doi: 10.1002/(sici)1522-2594(199912)42:6<1128::aid-mrm18>3.0.co;2-9. [DOI] [PubMed] [Google Scholar]
- 21.Kucharczyk W, Macdonald PM, Stanisz GJ, Henkelman RM. Relaxivity and magnetization transfer of white matter lipids at MR imaging: importance of cerebrosides and pH. Radiology. 1994;192(2):521–529. doi: 10.1148/radiology.192.2.8029426. [DOI] [PubMed] [Google Scholar]
- 22.O'Brien JS, Sampson EL. Lipid composition of the normal human brain: gray matter, white matter, and myelin. Journal of lipid research. 1965;6(4):537–544. [PubMed] [Google Scholar]
- 23.Yarnykh VL, Yuan C. Cross-relaxation imaging reveals detailed anatomy of white matter fiber tracts in the human brain. NeuroImage. 2004;23(1):409–424. doi: 10.1016/j.neuroimage.2004.04.029. [DOI] [PubMed] [Google Scholar]
- 24.Burdo JR, Martin J, Menzies SL, Dolan KG, Romano MA, Fletcher RJ, Garrick MD, Garrick LM, Connor JR. Cellular distribution of iron in the brain of the Belgrade rat. Neuroscience. 1999;93(3):1189–1196. doi: 10.1016/s0306-4522(99)00207-9. [DOI] [PubMed] [Google Scholar]
- 25.Gerber MR, Connor JR. Do oligodendrocytes mediate iron regulation in the human brain? Annals of neurology. 1989;26(1):95–98. doi: 10.1002/ana.410260115. [DOI] [PubMed] [Google Scholar]
- 26.Rajan KS, Colburn RW, Davis JM. Distribution of metal ions in the subcellular fractions of several rat brain areas. Life sciences. 1976;18(4):423–431. doi: 10.1016/0024-3205(76)90220-4. [DOI] [PubMed] [Google Scholar]
- 27.Todorich B, Pasquini JM, Garcia CI, Paez PM, Connor JR. Oligodendrocytes and myelination: the role of iron. Glia. 2009;57(5):467–478. doi: 10.1002/glia.20784. [DOI] [PubMed] [Google Scholar]
- 28.Markesbery WR, Ehmann WD, Hossain TI, Alauddin M, Goodin DT. Instrumental neutron activation analysis of brain aluminum in Alzheimer disease and aging. Annals of neurology. 1981;10(6):511–516. doi: 10.1002/ana.410100604. [DOI] [PubMed] [Google Scholar]
- 29.Stedman JD, Spyruo NM. Elemental analysis of the frontal lobe of “normal” brain tissue and that affected by Altzheimer's disease. Journal of Radioanalytical and Nuclear Chemistry. 1997;217(2):163–166. [Google Scholar]