|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Special Report |
, PhD
1 From the Functional MRI Laboratory, Departments of Radiology (A.I.H.) and Neurosurgery (P.H.G.), Memorial Sloan Kettering Cancer Center, 1275 York Ave, New York, NY 10021; Department of Radiology, UMDNJNew Jersey Medical School, Newark, NJ (D.M.G.); and Department of Radiology, Weill Medical College of Cornell University, New York, NY (R.W., A.M.U.). From the 2003 RSNA Scientific Assembly. Received December 22, 2003; revision requested February 24, 2004; revision received May 6; accepted May 24. Address correspondence to A.I.H. (e-mail: holodnya@mskcc.org).
| ABSTRACT |
|---|
|
|
|---|
© RSNA, 2005
| INTRODUCTION |
|---|
|
|
|---|
Aside from the academic interest of deepening our understanding of neuroanatomy, the exact location and internal organization of the CST as it passes through the corona radiata and the PLIC has a number of important clinical applications. For example, such knowledge could be used for planning functional neurosurgery in patients with Parkinson disease and related disorders, for treatment of patients with stroke or lacunar infarcts in the vicinity of the PLIC, and, perhaps most important, for preoperative localization of the CST in patients with brain tumor (58). Inadvertent transection of the CST can have devastating consequences. Therefore, a method that could accurately depict the relationship of the CST to a tumor both preoperatively and intraoperatively would improve neurosurgical planning, as well as the actual resection (7). Since direct intraoperative white matter stimulation is much less reliable than cortical stimulation (9), it can be argued that correct preoperative identification of white matter tracts is perhaps even more important than identification of the eloquent cortex.
White matter tractography based on diffusion-tensor imaging is a magnetic resonance (MR) technique that can depict white matter tracts in the living human brain (1016). The goal of our study was to use diffusion-tensor imaging to define the normal location and organization of CSTs in the PLIC.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Imaging
Diffusion-tensor imaging data were obtained with a 3-T magnet (Eclipse; GE Medical Systems, Milwaukee, Wis). We used a spin-echo echo-planar imaging sequence (repetition time msec/echo time msec, 10 000/100; matrix, 128 x 128 x 60; voxel size, 1.88 x 1.88 x 2.5 mm; 26 gradient orientations; b values, 8151152 sec/mm2; six signals acquired without diffusion weighting). The diffusion tensor was calculated for each voxel by using single-value decomposition. Diagonalization of the tensor was used to calculate the eigenvalues (
1,
2,
3) and the eigenvector. The principal eigenvector, the vector corresponding to the largest eigenvalue, represents the direction in which water diffusion is greatest and is assumed to correspond to the predominant fiber orientation within each voxel.
In two of the eight volunteers, the hand and foot areas in the cortex were localized by using blood oxygen leveldependent functional MR imaging. Gradient-echo echo-planar images (3000/40; flip angle, 90°) were obtained in the transverse plane. The acquisition matrix was 64 x 64, with 30 sections acquired. The field of view for the functional acquisition was identical to that of the diffusion-tensor acquisition (240 x 240 x 150 mm) but with the voxel size doubled in the spatial dimension to 3.75 x 3.75 x 5 mm. The patients performed a finger-tapping paradigm to identify the motor cortex. The paradigm consisted of 30 seconds of self-paced finger tapping and 30 seconds of rest alternating with each other, for a total of five periods of activation and six periods of rest. This was repeated for toe tapping. The blood oxygen leveldependent data were analyzed by using Stimulate software (17), and functional maps were generated by using a cross-correlation technique for P < .03.
Tractography was performed in the following manner: Regions of interest were drawn by one of the authors (D.M.G.) to define the areas of the precentral gyrus known to control the movement of the foot and hand. These were designated as the "seed volumes." The areas of the precentral gyrus responsible for movement of the hand and foot were determined from the blood oxygen leveldependent functional MR data in the two volunteers who underwent this imaging study. In the other cases, the seed volumes were estimated by using known anatomic landmarks. A "destination volume" region of interest was placed to encompass the entire PLIC. Only tracts starting from the seed volumes and passing through the destination volumes were included in the trace of the CST.
A continuous tracking algorithm was used in which the path follows the principal eigenvector of the diffusion tensor on a subvoxel level until the voxel edge is met, at which point the direction abruptly changes to that of the new voxel (14,15). Tracking terminated when the relative anisotropy of the voxel decreased to below 0.15, which indicated that the directionality of the vector had become unreliable (15). The fibers of the CST were traced from the motor cortex through the PLIC and to the pons. The CSTs were viewed by using multiple oblique transverse projections to determine their actual and apparent relationships to other structures in the brain.
| RESULTS |
|---|
|
|
|---|
In 17 of the 20 tracts, the fibers of the CST were organized somatotopically in the PLIC, with the hand fibers lateral and slightly anterior to the foot fibers. In the remaining three tracts, the hand fibers were intermixed with the foot fibers. The three instances of intermixed fibers were seen in three different volunteers. In these three cases, the intermixing was seen only on the left side, not on the right side. One tumor displaced the hand fibers posteriorly but did not affect the internal organization of the CST (Figs 1, 2).
|
|
|
|
|
|
| DISCUSSION |
|---|
|
|
|---|
We did not see any evidence for the CST involving any other part of the PLIC. An important methodological point of our study is that we took the entire PLIC for the ending point. Therefore, the diffusion-tensor tractography method should have traced all of the white matter tracts from the precentral gyrus to any part of the PLIC. We were only able to demonstrate white matter tracts in a small area in the third quarter of the PLIC. In none of the cases did we see white matter tracts from the motor cortex traversing any other part of the PLIC.
CST Location in the PLIC
The location of the CST in the PLIC has been the subject of debate and has undergone a major revision during the past few decades (2,3). Classically, the CST was thought to lie in the anterior third of the PLIC (2224). This classic localization began to be challenged in the 1950s and 1960s, when a number of authors averred that the CST is located in the third quarter of the PLIC (18,20,21,25,26). The more posterior localization of the CST is supported by MR imaging investigations that did not involve diffusion-tensor tractography, such as that of Yagashita et al (27).
Notwithstanding the above-mentioned publications, the controversy continues (28,29). Two more recent anatomic studies that involved gross dissection of the brain (19,20) showed that the more cephalic part of the CST is in the midportion of the PLIC, while the more caudal portion is in the posterior aspect of the PLIC.
The apparent difference between these anatomic findings and the results of the current work can be explained by the different angulation of the transverse sections through the brain. Ross (19) and Kretschmann (20) transversely sectioned the brain perpendicular to the CST, so that the precentral gyrus appears to be at the apex of the brain. In most transverse radiologic studies, the angulation of the plane is more perpendicular to the long axis of the body, so the precentral gyrus is rarely seen at the apex; rather, it is located posteriorly. Because the CST is a relatively linear structure that originates in the precentral gyrus and traverses the PLIC, its "location" in the brain and in the cephalic part of the PLIC will also be influenced by the angulation of the transverse section. This is perhaps why the anatomists, who started their transverse dissection with the precentral gyrus in the midline, believed that the CST entered the PLIC near the midline, as well. Similarly, in studies that started with the precentral gyrus located relatively more posteriorly, the course of the CST was defined more posteriorly. However, it should be stressed that, irrespective of angulation, all recent results affirm that the CST traverses the PLIC at the level of the lower thalamus within the third quarter.
Internal Organization of CST in the PLIC
Given the difference of opinion regarding the location of the CST as it traverses the PLIC, it is not surprising that there would be disagreement regarding the internal organization of the CST in the PLIC. It is currently thought that the individual tracts that control movements of the hand, trunk, foot, and face are organized somatotopically in the PLIC, with the hand CST anterior and slightly medial to the foot CST along the long axis of the PLIC (2,21). The evidence for this somatotopic organization is "rather crude" (2).
The results of our study support a somatotopic organization for the separate tracts of the CST. However, on the basis of our results it appears that the fibers are organized along the left-to-right axis (the short axis of the PLIC) as opposed the anterior-posterior axis (the long axis), as is currently believed.
According to our results, the hand and foot CSTs are oriented exactly the same as the hand and foot homunculus in the motor strip: The hand CST is lateral and slightly anterior to the foot CST. In the prevailing model, however, the orientations of the hand and foot CSTs are rotated 90° from the orientation of the hand and foot homunculus in the precentral gyrus. Therefore, the organization of the CST in the PLIC that we have proposed appears to make more anatomic sense than does the prevailing model.
Most of what is currently believed about the internal organization of the CST in the PLIC is based on results from a careful study published in 1965 by Bertrand et al (21). The study by Bertrand et al suggested that the hand fibers were located more anteriorly and slightly medially to the foot fibers. However, Bertrand et al demonstrated a large overlap between the hand and foot fibers. Indeed, they were rather modest about their claims: "Electrical stimulation within what we have thought to be [emphasis added] the internal capsule itself has produced interesting information" (21). Unfortunately, the caution about their work was overlooked, and the results were translated into dogma in later publications, especially textbooks.
The Bertrand et al study had a number of limitations, which were inherent to the technology of 1965 and which they acknowledged: (a) They used a stereotactic system based on pneumoencephalography performed a few days prior to the procedure, and the system was correlated to orthogonal conventional radiographs during the procedure, after the head frame had been removed and reapplied. (b) The questions of cosine error on the radiographs, brain motion, and displacement of the brain by the actual needle were not addressed. (c) The axonal structure in the PLIC is strongly anisotropic. It would seem to follow that the impedance to current flow would also be similarly anisotropic. Therefore, current used to stimulate the fibers would travel differently in various directions, making anatomic localization less reliable. (d) The authors did not account for head size (even though one patient was a child) and did not account for the size of the third ventricle, which the authors suspected "may displace the internal capsule laterally" (21). (e) Stimulation of white matter tracts is known to be much less reliable than stimulation of the cortex (30). (f) They studied patients with Parkinson disease and other movement disorders, not healthy subjects.
Clearly, the limitations outlined above do not invalidate the results of the Bertrand et al (21) study. However, they leave room for the presentation of alternative results based on other modalities. An acknowledged limitation of diffusion-tensor imaging concerns crossing white matter fibers. If a voxel has two bundles of fibers that cross at an angle, it may not be possible at diffusion-tensor imaging to separate these two bundles. Instead, the two bundles may be combined into one large bundle that will have a direction intermediate between the two actual directions of the white matter bundles. Therefore, the white matter trace will follow the intermediate direction, which does not correspond to either of the true directions of the two white matter bundles.
Perhaps our method produced erroneous results due to this acknowledged limitation of diffusion-tensor imaging. We believe that this is unlikely, however, for several reasons: (a) We actually demonstrated intermingling of fibers in the left-to-right direction in three cases. (b) Crossing of fibers should cause a deflection of the trace. Such a deflection would have caused the white matter tracing to veer off course either anteriorly or posteriorly into the other parts of the PLIC. This did not occur. All of the CST fibers traversed the PLIC in the third quarter. (c) As far as we are aware, in anatomic terms there is no substantial crossing of white matter tracts in the PLIC in the anterior-posterior direction. (d) The fibers of the CST were traced from the motor cortex, through the PLIC, to their known anatomic location in the pons. As far as we are aware, the anatomic location of the CST in the pons is not in dispute. Therefore, to explain an error in the tracing of the internal organization of the CST through the PLIC, one would have to invoke a mechanism that would deflect the trace in one direction as the CST approached the PLIC and another mechanism that would deflect the trace in the opposite direction as one approached the pons, in order to put the CST back onto its correct trajectory. Since we did not detect traces of the CST in any other part of the PLIC, these two hypothetical mechanisms would have to counterbalance each other to a remarkable degree. From our understanding of human anatomy and the diffusion-tensor tractography mechanism, it appears unlikely that such mechanisms exist.
There are a number of other technical issues that can potentially limit diffusion-tensor imaging. These include field inhomogeneities, which are more prominent when echo-planar imaging sequences are used. Field inhomogeneities can cause distortions that lead to misregistration. However, for field inhomogeneities to account for the difference between the current results and the current model, one would have to propose a focal rotation of the diffusion-tensor images by 90° as the CST traversed the PLIC. Review of the source images made such a possibility unlikely.
The current study was also limited by the small number of people included. Perhaps more important, the diffusion-tensor imaging tracings of the CST (as well as our suppositions regarding the apparent differences between our results and previous findings at anatomic dissection) were not independently confirmed with any other method, such as anatomic dissection or intraoperative mapping.
In conclusion, we have noninvasively demonstrated the somatotopic organization of the CST in the PLIC. The somatotopic organization of the CST in the posterior limb of the internal capsule shows the hand fibers oriented anterolaterally to the foot fibers, not anteromedially as is currently believed. Diffusion-tensor imaging opens up a new avenue of research into the anatomic organization of the human brain, as well as possible new controversies.
| FOOTNOTES |
|---|
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, A.I.H., D.M.G.; study concepts and design, A.I.H., A.M.U., P.H.G.; literature research, A.I.H.; clinical and experimental studies, A.M.U., R.W.; data acquisition, R.W.; data analysis/interpretation, D.M.G., R.W.; statistical analysis, R.W., D.M.G.; manuscript preparation, definition of intellectual content, and final version approval, A.I.H.; manuscript editing and revision/review, A.I.H., P.H.G.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
A. Stadlbauer, E. Salomonowitz, G. Strunk, T. Hammen, and O. Ganslandt Age-related Degradation in the Central Nervous System: Assessment with Diffusion-Tensor Imaging and Quantitative Fiber Tracking Radiology, April 1, 2008; 247(1): 179 - 188. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Simonyan, F. Tovar-Moll, J. Ostuni, M. Hallett, V. F. Kalasinsky, M. R. Lewin-Smith, E. J. Rushing, A. O. Vortmeyer, and C. L. Ludlow Focal white matter changes in spasmodic dysphonia: a combined diffusion tensor imaging and neuropathological study Brain, February 1, 2008; 131(2): 447 - 459. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Smits, M.W. Vernooij, P.A. Wielopolski, A.J.P.E. Vincent, G.C. Houston, and A. van der Lugt Incorporating Functional MR Imaging into Diffusion Tensor Tractography in the Preoperative Assessment of the Corticospinal Tract in Patients with Brain Tumors AJNR Am. J. Neuroradiol., August 1, 2007; 28(7): 1354 - 1361. [Abstract] [Full Text] [PDF] |
||||
![]() |
S.K. Schimrigk, B. Bellenberg, M. Schluter, B. Stieltjes, R. Drescher, J. Rexilius, C. Lukas, H.K. Hahn, H. Przuntek, and O. Koster Diffusion Tensor Imaging-Based Fractional Anisotropy Quantification in the Corticospinal Tract of Patients with Amyotrophic Lateral Sclerosis Using a Probabilistic Mixture Model AJNR Am. J. Neuroradiol., April 1, 2007; 28(4): 724 - 730. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. M. Newton, N. S. Ward, G. J. M. Parker, R. Deichmann, D. C. Alexander, K. J. Friston, and R. S. J. Frackowiak Non-invasive mapping of corticofugal fibres from multiple motor areas--relevance to stroke recovery Brain, July 1, 2006; 129(7): 1844 - 1858. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. B. Wolbarst and W. R. Hendee Evolving and Experimental Technologies in Medical Imaging Radiology, January 1, 2006; 238(1): 16 - 39. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |