Musculoskeletal ImagingFree Access

Diffusion Tensor Imaging in Musculoskeletal Disorders

Abstract

Diffusion tensor (DT) imaging is an emerging magnetic resonance (MR) imaging technique for evaluating the microstructure of well-organized biologic tissues such as muscles and nerves. DT imaging provides information about tissue microstructure by producing three-dimensional maps of water molecule movements. The two main parameters of measurement at DT imaging, fractional anisotropy and the apparent diffusion coefficient, allow quantitation of architectural changes occurring in tissue. These parameters are modified in the presence of cervical spondylotic myelopathy, cervical spine trauma, carpal tunnel syndrome, lumbar nerve compression, peripheral nerve tumors, and muscle ischemia. Their alteration may be observed at DT imaging even when no abnormality is seen at conventional MR imaging, a fact that suggests that DT imaging allows the detection of abnormalities at an earlier stage of injury. Experimental studies in animals have shown that DT imaging consistently allows identification of pathophysiologic alterations in tissue that correlate with histologic findings. Tractographic images accurately depict both normal and abnormal diffusion in anatomic structures such as the thigh and pelvic muscles, cervical spine, and lumbar nerves. Patients with chronic diseases also may benefit from follow-up evaluation with DT imaging, although DT imaging sequences must be further adapted to improve the evaluation of specific anatomic regions by reducing artifacts, optimizing spatial resolution, and minimizing acquisition time. Given its proven potential for use in identifying abnormalities that are otherwise identifiable only with electrophysiologic and histopathologic studies, and with future technical improvements, DT imaging could soon become a standard method for early diagnosis, management, and follow-up of disease in the spine, muscles, and peripheral nerves.

©RSNA, 2014

SA-CME LEARNING OBJECTIVES

After completing this journal-based SA-CME activity, participants will be able to:

■ Explain the roles of diffusion and anisotropy in DT imaging.
■ Describe the physical principles underlying MR tractography.
■ Discuss the current state of research in the use of DT imaging for musculoskeletal evaluations.

Introduction

Medical imaging is focused mainly on the morphologic identification of normal anatomic structures and the description of their appearances in the presence of various pathologic conditions. Most disease processes encountered in daily medical practice can be managed on the basis of adequate knowledge of signs and symptoms. Yet the information provided by morphologic imaging is limited. Musculoskeletal imaging specialists and general radiologists routinely encounter cases in which patients with authentic radicular pain in the lumbar region have no visible abnormalities in the lumbar spine at computed tomography (CT) or magnetic resonance (MR) imaging. Conversely, radiologic images obtained in patients with no reported radicular pain sometimes depict severe lumbar disk arthrosis or impingement of lumbar nerve roots. Such discrepancies between the imaging findings and the clinical situation suggest that although accurate morphologic information is often provided by standard imaging studies, functional information is missing.

However, recent technical developments in radiologic imaging allow us to go beyond the morphologic characterization of anatomic structures and explore the inner structure of biologic tissues. For example, MR or CT perfusion studies may be used to evaluate tissue blood supply; MR spectroscopy, to quantify different tissue components; diffusion imaging, to globally assess the motion of water molecules; and diffusion tensor (DT) imaging, to infer microstructural information from the three-dimensional (3D) motion of water molecules in tissue.

The article briefly describes the main concepts underlying DT imaging and MR tractography, including basic principles of physics, and reviews the findings from recent studies of the use of DT imaging for musculoskeletal evaluation. The literature includes dedicated studies in which the technique is described in detail and its principles are defined by mathematical formulas (1,2). Here, we focus instead on providing a glimpse into the potential advantages of DT imaging as reflected in the results of studies showing its feasibility for evaluating specific parts of the anatomy, anatomic studies demonstrating the capability of tractography for the 3D representation of muscles and nerves, experimental studies confirming the ability of DT imaging to depict specific physiologic or pathologic processes, and studies describing how the parameters measured by DT imaging are altered by certain diseases in the spinal cord and peripheral nerves.

Conceptual Basis of DT Imaging

Water Diffusion and Anisotropy

Water molecules move randomly in biologic tissues. This diffusion of water molecules, referred to as Brownian movement, causes tiny disturbances in the magnetic field that are not measurable with conventional MR imaging sequences. However, by applying additional pulses in the gradient echo sequence, the effect of these movements on the MR signal can be observed; this is the principle underlying diffusion imaging (3). The specific parameters that can be used to characterize this motion are the apparent diffusion coefficient (ADC) and mean diffusivity. The ADC represents the extent to which the movement of water molecules is unrestrained. The freer the motion, the higher the ADC. In other words,

the ADC and mean diffusivity express the existence of obstacles to the motion of water molecules among the cells, whether these obstacles are physiologic or induced by a pathologic process.

For example, imagine yourself walking down an empty street. With no other pedestrian or vehicle impeding your movement, your ADC (reflecting the facility of your movement down the street) would be high. On the contrary, on an overcrowded street, other people would hinder your movements, lowering your ADC. In physiologic conditions, the cellular wall is one such obstacle. The extent to which the walls of cells impede diffusion depends on the microstructural organization of the tissue (Fig 1a). In pathologic conditions, phenomena such as cytotoxic edema (ie, a swelling of the cells that reduces the extracellular compartment) can restrict diffusion; this occurs in the presence of cerebral ischemia (4) (Fig 1b). Hypercellular tumors also reduce diffusion (5) (Fig 1c). By contrast, if the cellular walls are destroyed by a pathologic process, diffusion may be increased (Fig 1d). Therefore, pathologic conditions can lead to increased or decreased ADC (6).

Figure 1a Schematics show the diffusion of water molecules (dark blue ovals) in biologic tissue (light blue background). Yellow rings represent cell walls, and red arrows represent the potential movement of the water molecules. **(a)** Diffusion in normal tissue is relatively unrestricted. **(b, c)** Diffusion in the presence of cellular edema **(b)** or hypercellularity **(c)** is restricted. **(d)** In necrotic tissue, the cell walls are destroyed, and diffusion is increased.

Figure 1b Schematics show the diffusion of water molecules (dark blue ovals) in biologic tissue (light blue background). Yellow rings represent cell walls, and red arrows represent the potential movement of the water molecules. **(a)** Diffusion in normal tissue is relatively unrestricted. **(b, c)** Diffusion in the presence of cellular edema **(b)** or hypercellularity **(c)** is restricted. **(d)** In necrotic tissue, the cell walls are destroyed, and diffusion is increased.

Figure 1c Schematics show the diffusion of water molecules (dark blue ovals) in biologic tissue (light blue background). Yellow rings represent cell walls, and red arrows represent the potential movement of the water molecules. **(a)** Diffusion in normal tissue is relatively unrestricted. **(b, c)** Diffusion in the presence of cellular edema **(b)** or hypercellularity **(c)** is restricted. **(d)** In necrotic tissue, the cell walls are destroyed, and diffusion is increased.

Figure 1d Schematics show the diffusion of water molecules (dark blue ovals) in biologic tissue (light blue background). Yellow rings represent cell walls, and red arrows represent the potential movement of the water molecules. **(a)** Diffusion in normal tissue is relatively unrestricted. **(b, c)** Diffusion in the presence of cellular edema **(b)** or hypercellularity **(c)** is restricted. **(d)** In necrotic tissue, the cell walls are destroyed, and diffusion is increased.

DT imaging is based on the ability of water molecules to move differentially in various directions through biologic tissue (6). The diffusion of water molecules within biologic tissues is anisotropic. Anisotropic diffusion occurs preferentially in one direction instead of homogeneously in all directions

(1). Diffusion in the other directions is impeded by the walls of the cells constituting the tissue.

Let us consider another example: A swimmer in the open ocean can move around freely without any nearby obstacles to disturb his movements; because his movement capacity is the same in any direction, this environment is said to be isotropic (Fig 2a). On the other hand, the movement probability of someone standing in the middle of a narrow street is not the same in every direction: it is likely that he or she will move down the street. This environment is said to be anisotropic (Fig 2c).

Figure 2a **(a, b)** Photograph and schematic represent an isotropic diffusion environment, in which the probability of water molecule movement (green arrows) is the same in all directions. The geometric model of this environment is a sphere. **(c, d)** Photograph and schematic show an anisotropic environment, in which the probability of water molecule movement (green arrows) is highest in one direction (represented by the longest arrow). The geometric model is an ellipsoid defined by three main vectors, named eigenvectors, each with a determined length or eigenvalue: λ¹, λ², λ³. **(e, f)** Photographs of whole and broken noodles illustrate the striking contrast between healthy tissue composed of well-organized parallel fibers conducive to anisotropic diffusion **(e)** and pathologic disruption of tissue fiber alignment leading to a loss of anisotropy **(f)**.

Figure 2b **(a, b)** Photograph and schematic represent an isotropic diffusion environment, in which the probability of water molecule movement (green arrows) is the same in all directions. The geometric model of this environment is a sphere. **(c, d)** Photograph and schematic show an anisotropic environment, in which the probability of water molecule movement (green arrows) is highest in one direction (represented by the longest arrow). The geometric model is an ellipsoid defined by three main vectors, named eigenvectors, each with a determined length or eigenvalue: λ¹, λ², λ³. **(e, f)** Photographs of whole and broken noodles illustrate the striking contrast between healthy tissue composed of well-organized parallel fibers conducive to anisotropic diffusion **(e)** and pathologic disruption of tissue fiber alignment leading to a loss of anisotropy **(f)**.

Figure 2c **(a, b)** Photograph and schematic represent an isotropic diffusion environment, in which the probability of water molecule movement (green arrows) is the same in all directions. The geometric model of this environment is a sphere. **(c, d)** Photograph and schematic show an anisotropic environment, in which the probability of water molecule movement (green arrows) is highest in one direction (represented by the longest arrow). The geometric model is an ellipsoid defined by three main vectors, named eigenvectors, each with a determined length or eigenvalue: λ¹, λ², λ³. **(e, f)** Photographs of whole and broken noodles illustrate the striking contrast between healthy tissue composed of well-organized parallel fibers conducive to anisotropic diffusion **(e)** and pathologic disruption of tissue fiber alignment leading to a loss of anisotropy **(f)**.

Figure 2d **(a, b)** Photograph and schematic represent an isotropic diffusion environment, in which the probability of water molecule movement (green arrows) is the same in all directions. The geometric model of this environment is a sphere. **(c, d)** Photograph and schematic show an anisotropic environment, in which the probability of water molecule movement (green arrows) is highest in one direction (represented by the longest arrow). The geometric model is an ellipsoid defined by three main vectors, named eigenvectors, each with a determined length or eigenvalue: λ¹, λ², λ³. **(e, f)** Photographs of whole and broken noodles illustrate the striking contrast between healthy tissue composed of well-organized parallel fibers conducive to anisotropic diffusion **(e)** and pathologic disruption of tissue fiber alignment leading to a loss of anisotropy **(f)**.

Figure 2e **(a, b)** Photograph and schematic represent an isotropic diffusion environment, in which the probability of water molecule movement (green arrows) is the same in all directions. The geometric model of this environment is a sphere. **(c, d)** Photograph and schematic show an anisotropic environment, in which the probability of water molecule movement (green arrows) is highest in one direction (represented by the longest arrow). The geometric model is an ellipsoid defined by three main vectors, named eigenvectors, each with a determined length or eigenvalue: λ¹, λ², λ³. **(e, f)** Photographs of whole and broken noodles illustrate the striking contrast between healthy tissue composed of well-organized parallel fibers conducive to anisotropic diffusion **(e)** and pathologic disruption of tissue fiber alignment leading to a loss of anisotropy **(f)**.

Figure 2f **(a, b)** Photograph and schematic represent an isotropic diffusion environment, in which the probability of water molecule movement (green arrows) is the same in all directions. The geometric model of this environment is a sphere. **(c, d)** Photograph and schematic show an anisotropic environment, in which the probability of water molecule movement (green arrows) is highest in one direction (represented by the longest arrow). The geometric model is an ellipsoid defined by three main vectors, named eigenvectors, each with a determined length or eigenvalue: λ¹, λ², λ³. **(e, f)** Photographs of whole and broken noodles illustrate the striking contrast between healthy tissue composed of well-organized parallel fibers conducive to anisotropic diffusion **(e)** and pathologic disruption of tissue fiber alignment leading to a loss of anisotropy **(f)**.

It is common to use a mathematical model to represent the probability of water molecules’ movement with arrows called eigenvectors (7). In the theoretical case of a perfectly isotropic tissue, the probability of movement is exactly the same in any direction; every arrow has the same length (eigenvalue) (7). The geometric model representing this probability is a sphere (Fig 2b). In an anisotropic tissue such as nerves or muscles, the probability of water molecule movement is much greater along the fibers than perpendicular to them. In our model, one arrow is longer than the other ones; that is, its eigenvalue (λ¹) is higher than those of the other arrows (λ²and λ³). The model is no longer a sphere; it is now an ellipsoid (6) (Fig 2d). Fractional anisotropy (FA) is the main parameter reflecting the degree of a tissue’s anisotropy. This value varies from 0 (perfect isotropy) to 1 (total anisotropy).

Tissues such as peripheral nerves and muscle fibers are highly anisotropic in comparison with cerebrospinal fluid or subcutaneous fat because they are well organized.

If a highly anisotropic tissue becomes less anisotropic (that is, if its FA decreases), it is likely that its well-organized cellular architecture has been disrupted by a physiologic or pathologic process

(6) (Figs 2e, 2f).

Measurement of Diffusion Parameters

Several parameters can be measured with DT imaging by using software developed specifically for this purpose (8). Eigenvalues (λ¹, λ², and λ³) can be calculated for the three main directions of the ellipsoid (7) (Fig 2d) and used to characterize the relative probability of movement of water molecules in each direction. However, other indices (eg, mean diffusivity, FA) derived from these “raw” values are far more frequently used. As described earlier, mean diffusivity and ADC represent the degree to which the diffusion of water molecules is unrestricted. FA reflects the extent to which the organization of the tissue determines a preferential directionality of movement of water molecules (6).

It is important to keep in mind that these parameters of measurement are not absolute. They vary according to the magnetic field strength as well as the voxel size, number of diffusion gradients, and signal-to-noise ratio (9). Variability in the use of imaging hardware (same hardware, different operators) also may be responsible for variations in measurement (10,11), especially if different gradient schemes or different gradient directions are used (12). However, the reproducibility of DT imaging measurements in longitudinal studies performed with the same imaging system and in the same location is good (13,14). Preliminary studies are needed to assess potential bias before a multicenter study of DT imaging can be performed (15).

DT Images

The DT image is generally a color-coded map of the FA values in each voxel (Fig 3a). The color of the voxels depicted on an FA map corresponds to the direction of the main eigenvector: Blue signifies x-axis movement along the patient’s body (in a craniocaudal or caudocranial direction), red stands for y-axis (laterolateral) movement, and green signifies z-axis (anteroposterior or posteroanterior) motion. The luminosity of each pixel indicates the degree of anisotropy; the higher the FA, the greater the luminosity (16). To identify the boundaries of the anatomic structure undergoing evaluation, the fusion of FA maps with morphologic images is often necessary and is usually achieved by using T2-weighted images (Fig 3b). Navigation within the T2-weighted image dataset makes it possible to find the voxels corresponding to those depicted on the FA map. However, distortions on the images may hinder accurate matching of pixels. Before superimposing the images, one can correct these distortions by performing a step called coregistration. Image fusion and coregistration are performed by using specific software (8).

Figure 3a **(a)** Axial FA map shows values measured in the nerves within the dural sac at the level of the L5-S1 intervertebral disk. The nerve fibers where the L5 roots exit the intervertebral foramen (arrows) are obliquely oriented in both the craniocaudal and laterolateral planes, which produces their purplish-pink coloration on the map because of the mixing of blue with red. (Blue = orientation along the x-axis, red = orientation along the y-axis, green = orientation along the z-axis.) The nearly perfect craniocaudal orientation of the sacral nerves (arrowheads) within the dural sac results in their deep blue coloration on the FA map. **(b)** Axial T2-weighted MR image obtained for anatomic comparison shows the lumbar nerve roots (arrows) and sacral nerves (arrowheads).

Figure 3b **(a)** Axial FA map shows values measured in the nerves within the dural sac at the level of the L5-S1 intervertebral disk. The nerve fibers where the L5 roots exit the intervertebral foramen (arrows) are obliquely oriented in both the craniocaudal and laterolateral planes, which produces their purplish-pink coloration on the map because of the mixing of blue with red. (Blue = orientation along the x-axis, red = orientation along the y-axis, green = orientation along the z-axis.) The nearly perfect craniocaudal orientation of the sacral nerves (arrowheads) within the dural sac results in their deep blue coloration on the FA map. **(b)** Axial T2-weighted MR image obtained for anatomic comparison shows the lumbar nerve roots (arrows) and sacral nerves (arrowheads).

Tractographic Images

The tractographic image is a 3D representation of an anisotropic structure. In the DT acquisition, the main eigenvector represents the orientation of the dominant fiber tracts. Hence, the imaging volume may be envisioned as a constellation of many arrows (7), each with its own orientation (eigenvector) and with a length (eigenvalue) proportional to the degree of anisotropy in each voxel. Various techniques for the propagation of neural tracts on tractographic images, each with different pros and cons, have been described in the literature (17).

Postprocessing software connects voxels together (Fig 4) by taking into account both the orientation of the vector and the FA (17). Adjacent voxels with similar FA and main vector orientation are assumed to be connected.

Thus, the threshold values for both parameters must be carefully selected (17). The operator chooses a low FA limit to exclude structures that are insufficiently anisotropic (6). If the FA is low in a voxel, this voxel probably does not belong to the anisotropic structure adjacent to the fiber tract, even if the direction of the main eigenvector is about the same. However, while a low FA threshold correctly excludes many aberrant fibers, it may also lead the radiologist to overlook authentic pathologic fibers (6) that have a lower FA than normal fibers. The operator also sets the limit for angular deviation so as to exclude voxels in which the main eigenvector is oriented differently (6). If the direction of the main eigenvector of two proximate voxels diverges markedly, they probably belong to different structures, even if they have similar FA values. This operation results in groups of lines corresponding to the anisotropic structures, making the 3D depiction of nerve bundles or muscle fibers possible. It is important to remember that the fiber tracts are mathematical constructs based on the presumed connectedness of voxels with similar diffusion properties: they do not correspond to axons or muscle fibers (6), structures that are much smaller than the voxels on tractographic images.

Figure 4 Schematic represents the voxels of an anisotropic structure (orange voxels) on an FA map. All three eigenvectors are oriented in the same direction; thus, the eigenvalues are high (long arrows). To ensure accurate tracking of voxels with similar FA values and similar primary eigenvector orientation, the radiologist must carefully define the limits for these two parameters in the postprocessing software. The FA threshold must be high enough so that adjacent structures with lower FA values (purple voxels) relative to those of the fiber are correctly excluded; however, if the FA threshold is set too high, decreased FA (short arrows) due to a pathologic process (yellow voxels) in part of the fiber may lead to erroneous exclusion of the affected voxels, with resultant inaccuracy in the estimation of FA. Likewise, if the allowed angular deviation is too low, a bifurcation of the fiber such as that depicted in the upper right corner of the schematic might not be taken into consideration, whereas if it is too high, an unrelated anisotropic structure (blue voxels) might be incorrectly included.

However, DT imaging makes the quantitative analysis of individual fibers or bundles of fibers possible (8). Measures of FA and ADC can be obtained in selected fibers by positioning a region of interest (ROI) directly on the FA or ADC map (18–20). Correlation of DT maps with standard anatomic MR images can be helpful for avoiding partial volume effects. Another option is to measure the DT parameters directly on tractographic images (18,19,21–23). To obtain the FA of an ROI in a selected fiber on tractographic images, the operator crops the image to focus on the structure of interest after the tractography software performs the calculations on the basis of the entire volume. This method can help lessen the subjectivity involved in manual positioning of the ROI and may minimize the risk of including voxels that do not belong to the nerve. With either method, however, the risk of including or excluding voxels because of faulty threshold settings is the same. The DT parameters measured at tractography can be exported for statistical analysis. The key steps in tractography performed at our institutions are summarized in Figure 5 .

Figure 5a Screen captures show MR tractographic images of the lumbar nerves obtained with open-source postprocessing software (MedINRIA, version 1.9.0). The spheroid objects represent the limits of the cropping boxes. **(a)** Superior view of the entire “raw” image data set shows many structures with marginal anisotropy, most of which are attributable to image noise. To remove these structures, the FA threshold was increased and the cropping tool was used to delimit the volume of interest for measurement of the DT imaging parameters. **(b–d)** The 3D navigation tool allowed the radiologist to scroll through the cropped tractographic image volume from the initial superior view **(b)** to an anterior view **(c)**. Each lumbar nerve was isolated for independent analysis as shown in d, an anterior view of the left L5 root. FA or mean diffusivity values were extracted along each fiber and reported on corresponding histograms to allow reproducible level-by-level comparisons. Last, open-source software (FiberViewer 1.2.3) was used to fuse the tractographic image data set with a T2-weighted image data set obtained at the level of the L5 nerve roots. **(e)** Screen capture shows the resultant composite display. Fusion image (top center) and left and right L5 tractographic images (bottom right and bottom left, respectively) show a mass effect on the right L5 nerve root (thick white arrow), a finding that results from compression by a herniated disk. A corresponding decrease in FA is seen on the histogram for the right nerve root (*). No similar abnormalities are seen on either the tractographic image of the left L5 nerve root (black arrow) or on the corresponding histogram.

Figure 5b Screen captures show MR tractographic images of the lumbar nerves obtained with open-source postprocessing software (MedINRIA, version 1.9.0). The spheroid objects represent the limits of the cropping boxes. **(a)** Superior view of the entire “raw” image data set shows many structures with marginal anisotropy, most of which are attributable to image noise. To remove these structures, the FA threshold was increased and the cropping tool was used to delimit the volume of interest for measurement of the DT imaging parameters. **(b–d)** The 3D navigation tool allowed the radiologist to scroll through the cropped tractographic image volume from the initial superior view **(b)** to an anterior view **(c)**. Each lumbar nerve was isolated for independent analysis as shown in d, an anterior view of the left L5 root. FA or mean diffusivity values were extracted along each fiber and reported on corresponding histograms to allow reproducible level-by-level comparisons. Last, open-source software (FiberViewer 1.2.3) was used to fuse the tractographic image data set with a T2-weighted image data set obtained at the level of the L5 nerve roots. **(e)** Screen capture shows the resultant composite display. Fusion image (top center) and left and right L5 tractographic images (bottom right and bottom left, respectively) show a mass effect on the right L5 nerve root (thick white arrow), a finding that results from compression by a herniated disk. A corresponding decrease in FA is seen on the histogram for the right nerve root (*). No similar abnormalities are seen on either the tractographic image of the left L5 nerve root (black arrow) or on the corresponding histogram.

Figure 5c Screen captures show MR tractographic images of the lumbar nerves obtained with open-source postprocessing software (MedINRIA, version 1.9.0). The spheroid objects represent the limits of the cropping boxes. **(a)** Superior view of the entire “raw” image data set shows many structures with marginal anisotropy, most of which are attributable to image noise. To remove these structures, the FA threshold was increased and the cropping tool was used to delimit the volume of interest for measurement of the DT imaging parameters. **(b–d)** The 3D navigation tool allowed the radiologist to scroll through the cropped tractographic image volume from the initial superior view **(b)** to an anterior view **(c)**. Each lumbar nerve was isolated for independent analysis as shown in d, an anterior view of the left L5 root. FA or mean diffusivity values were extracted along each fiber and reported on corresponding histograms to allow reproducible level-by-level comparisons. Last, open-source software (FiberViewer 1.2.3) was used to fuse the tractographic image data set with a T2-weighted image data set obtained at the level of the L5 nerve roots. **(e)** Screen capture shows the resultant composite display. Fusion image (top center) and left and right L5 tractographic images (bottom right and bottom left, respectively) show a mass effect on the right L5 nerve root (thick white arrow), a finding that results from compression by a herniated disk. A corresponding decrease in FA is seen on the histogram for the right nerve root (*). No similar abnormalities are seen on either the tractographic image of the left L5 nerve root (black arrow) or on the corresponding histogram.

Figure 5d Screen captures show MR tractographic images of the lumbar nerves obtained with open-source postprocessing software (MedINRIA, version 1.9.0). The spheroid objects represent the limits of the cropping boxes. **(a)** Superior view of the entire “raw” image data set shows many structures with marginal anisotropy, most of which are attributable to image noise. To remove these structures, the FA threshold was increased and the cropping tool was used to delimit the volume of interest for measurement of the DT imaging parameters. **(b–d)** The 3D navigation tool allowed the radiologist to scroll through the cropped tractographic image volume from the initial superior view **(b)** to an anterior view **(c)**. Each lumbar nerve was isolated for independent analysis as shown in d, an anterior view of the left L5 root. FA or mean diffusivity values were extracted along each fiber and reported on corresponding histograms to allow reproducible level-by-level comparisons. Last, open-source software (FiberViewer 1.2.3) was used to fuse the tractographic image data set with a T2-weighted image data set obtained at the level of the L5 nerve roots. **(e)** Screen capture shows the resultant composite display. Fusion image (top center) and left and right L5 tractographic images (bottom right and bottom left, respectively) show a mass effect on the right L5 nerve root (thick white arrow), a finding that results from compression by a herniated disk. A corresponding decrease in FA is seen on the histogram for the right nerve root (*). No similar abnormalities are seen on either the tractographic image of the left L5 nerve root (black arrow) or on the corresponding histogram.

Figure 5e Screen captures show MR tractographic images of the lumbar nerves obtained with open-source postprocessing software (MedINRIA, version 1.9.0). The spheroid objects represent the limits of the cropping boxes. **(a)** Superior view of the entire “raw” image data set shows many structures with marginal anisotropy, most of which are attributable to image noise. To remove these structures, the FA threshold was increased and the cropping tool was used to delimit the volume of interest for measurement of the DT imaging parameters. **(b–d)** The 3D navigation tool allowed the radiologist to scroll through the cropped tractographic image volume from the initial superior view **(b)** to an anterior view **(c)**. Each lumbar nerve was isolated for independent analysis as shown in d, an anterior view of the left L5 root. FA or mean diffusivity values were extracted along each fiber and reported on corresponding histograms to allow reproducible level-by-level comparisons. Last, open-source software (FiberViewer 1.2.3) was used to fuse the tractographic image data set with a T2-weighted image data set obtained at the level of the L5 nerve roots. **(e)** Screen capture shows the resultant composite display. Fusion image (top center) and left and right L5 tractographic images (bottom right and bottom left, respectively) show a mass effect on the right L5 nerve root (thick white arrow), a finding that results from compression by a herniated disk. A corresponding decrease in FA is seen on the histogram for the right nerve root (*). No similar abnormalities are seen on either the tractographic image of the left L5 nerve root (black arrow) or on the corresponding histogram.

Potential Uses of DT Imaging

Many interesting studies have been conducted to determine the capabilities of DT imaging for evaluating the muscles and nerves, both in humans and animals. The following sections review the findings in these studies as reported in the literature.

Depiction of Normal Anatomy

The results of several feasibility studies showed that the muscles and nerves can be imaged with tractography and that the resultant images correspond to anatomic reality. In a DT imaging study of thigh muscles (16), the appearance of the muscles on 3D tractographic images obtained in living volunteers was well correlated with the findings in dissected cadavers (Figs 6 , 7 ). All fibers were positioned within the expected muscle boundaries depicted on corresponding anatomic T2-weighted images. The fibers were followed to the site of their insertion on the aponeurosis. Yet, the authors could not precisely identify the corresponding sites of origin and insertion of the whole muscle, as the tendons did not appear on the images. They hypothesized that the small number of water molecules present in the tendons relative to the number of water molecules in the muscle fibers was probably the reason for the limitations of this technique.

Figure 6a **(6a)** MR tractographic image depicts the narrow, ribbonlike sartorius muscle *(S)* in a medial view: Where the muscle crosses the thigh obliquely in an anteroposterior direction, it is mainly green and blue; its proximal and distal portions, which follow a more vertical course, are blue. The gracilis muscle *(G)*, which follows a more vertical course along the thigh, appears more uniformly blue. **(6b)** Pho-tograph provides a medial view of the same muscles in a dissected human cadaver, allowing satisfactory correlation with the features seen in a. RF = rectus femoris, VM = vastus medialis.

Figure 6b **(6a)** MR tractographic image depicts the narrow, ribbonlike sartorius muscle *(S)* in a medial view: Where the muscle crosses the thigh obliquely in an anteroposterior direction, it is mainly green and blue; its proximal and distal portions, which follow a more vertical course, are blue. The gracilis muscle *(G)*, which follows a more vertical course along the thigh, appears more uniformly blue. **(6b)** Pho-tograph provides a medial view of the same muscles in a dissected human cadaver, allowing satisfactory correlation with the features seen in a. RF = rectus femoris, VM = vastus medialis.

Figure 7a **(7a)** Anterior MR tractographic image shows the bipennate organization of the superficial fibers of the rectus femoris muscle. **(7b)** Photograph provides an anterior view of the rectus femoris *(RF)* muscle in a dissected specimen. S = sartorius, VL = vastus lateralis, VM = vastus medialis. (Figs 6 and 7 reprinted, with permission, from reference 16.)

Figure 7b **(7a)** Anterior MR tractographic image shows the bipennate organization of the superficial fibers of the rectus femoris muscle. **(7b)** Photograph provides an anterior view of the rectus femoris *(RF)* muscle in a dissected specimen. S = sartorius, VL = vastus lateralis, VM = vastus medialis. (Figs 6 and 7 reprinted, with permission, from reference 16.)

In another study, the complex anatomy of the overlapping tongue muscles was depicted with tractography (24). This was possible because the color encoding scheme clearly separated the muscles according to their different spatial orientations. A recent work concerning the pelvic floor showed that the pubovisceral and internal obturator muscles were easily identified according to their different orientation (25). The circular pattern of muscle fibers surrounding the anal canal and urethral sphincter complexes were also well depicted on tractographic images (25).

Depiction of Normal Physiologic Alterations in Muscle

Effects of Aging.—DT imaging has high sensitivity for the depiction of age-related changes in muscle, as demonstrated by Galban et al (26). DT parameters reflecting anisotropy were altered in the anterior and posterior muscles of the leg in elderly human subjects compared with young adults. A loss of anisotropy was the primary change noted between the two groups. Although the authors could offer no definitive explanation for these differences, they hypothesized that the deformation and loss of muscle fibers, increase in noncontractile tissue mass (eg, intramuscular accumulation of fat), and possible shifts in fiber type could result in a looser fiber network and contribute to a reduction in the transverse anisotropy of water molecule movement.

Effects of Muscle Contraction and Exercise.—Okamoto et al (27) performed DT imaging in the calves of two volunteers immediately, 24 hours, 48 hours, and 1 week after a regimen of unilateral exercises involving repeated flexion and extension of the ankle with loading. Immediately after exercise, a decrease in FA values was observed in the exercised gastrocnemius and soleus muscles in comparison with FA values in the contralateral calf muscles, which remained at rest. This alteration in FA was reversed within a few days. The FA decrease in the volunteer who performed the exercises for 60 minutes was more marked than that observed in the volunteer who performed the exercises for 30 minutes. Although the sample was too small to allow definitive conclusions, this study showed alterations in the DT parameter that were consistent with physiologic changes occurring during muscular contraction.

The same authors compared DT parameters in calf muscles of athletes with those in nonathletes and found significant differences (28). Physical training leads to well-described microstructural changes in muscle: Contractile filaments of protein (eg, actin and myosin) increase in size, number, and density. Skeletal muscle hypertrophy induces an increase in the thickness of endomysium and sarcoplasmic reticulum. Both intracellular and extracellular transformations may influence the properties of water diffusion.

In another study (29), opposing alterations in DT parameters were observed during active contraction or passive muscle elongation. DT images of the calf were acquired at rest, during dorsal flexion of the foot, and during plantar flexion of the foot. During dorsal flexion, the three eigenvalues and the ADC of the tibialis anterior muscle increased substantially, whereas those in the medial gastrocnemius muscle decreased slightly. A reverse pattern of alterations in the ADC and three eigenvalues was observed during plantar flexion, with a substantial increase in the medial gastrocnemius and a slight decrease in the tibialis anterior. Similar results were obtained in a more recent study (30). The precise relations between DT imaging parameter alterations and known physiologic phenomena such as fiber shortening, modification of the pennation angles, and changes in local temperature and venous pressure have yet to be established.

Depiction of Pathophysiologic Conditions in Muscle

Ischemia.—In an experimental study published in Radiology, Heemskerk et al (31) induced muscular ischemia by performing unilateral ligation of the femoral artery in mice. Follow-up DT imaging studies of both thighs and histologic analyses were performed immediately after surgery as well as on postoperative days 3, 10, and 21. Increased ADC and decreased FA were observed in the ischemic muscles relative to the values measured in the contralateral normal muscles. The DT parameters had changed the most on day 3, when the ischemic muscle damage seen at histologic analysis was greatest: most of the myocytes were round and swollen, almost no nuclei were visible, and signs of inflammation were observed at the periphery of the muscles. By day 10, the DT parameters had begun to normalize in the periphery of the muscles; this was consistent with centripetal recovery seen in the histologic study. FA and ADC returned to normal levels by day 21, contemporaneously with tissue recovery. Thus, the diffusion parameters changed dynamically, while remaining consistently associated with the severity and spatial location of ischemic muscle damage.

Trauma.—Zeng et al (32) conducted an experimental study of DT imaging in four rabbits with directly induced trauma to skeletal muscle in the thighs. Histopathologic analysis was performed for comparison. Decreased FA and increased ADC seen in the damaged skeletal muscle of the rabbits corresponded to areas of edema, injury, and tearing of muscle fibers. The most interesting fact was that the abnormalities in DT parameters were seen even in muscle that appeared normal on conventional MR images. The results of this study suggest that DT imaging might have higher sensitivity than conventional MR imaging for the depiction of traumatic muscle injury.

Myositis.—We performed a study with DT imaging of the thigh in patients with polymyositis and in healthy volunteers (Fig 8 ). The number of muscle fibers seen decreased with increasing severity of muscle lesions. Although this finding is insufficient to validate the use of DT imaging for evaluating muscle abnormalities, it does suggests that DT imaging can demonstrate tissue changes that are related to the presence and extent of lesions. Further studies are needed to test this hypothesis.

Figure 8a Posterior MR tractographic images show the semitendinous muscle in a healthy volunteer **(a)** and in two patients with polymyositis **(b, c)**. Conventional T1-weighted MR imaging showed partial fatty replacement of the muscle fibers in the same patient as in b and nearly complete fatty replacement of the muscle fibers in the same patient as in c.

Figure 8b Posterior MR tractographic images show the semitendinous muscle in a healthy volunteer **(a)** and in two patients with polymyositis **(b, c)**. Conventional T1-weighted MR imaging showed partial fatty replacement of the muscle fibers in the same patient as in b and nearly complete fatty replacement of the muscle fibers in the same patient as in c.

Figure 8c Posterior MR tractographic images show the semitendinous muscle in a healthy volunteer **(a)** and in two patients with polymyositis **(b, c)**. Conventional T1-weighted MR imaging showed partial fatty replacement of the muscle fibers in the same patient as in b and nearly complete fatty replacement of the muscle fibers in the same patient as in c.

Depiction of Peripheral Neuropathies

Wallerian Degeneration.—The term wallerian degeneration is used to describe the series of events that follows traumatic nerve injury (33). Nerve bundles in the peripheral nervous system are mainly composed of axons, Schwann cells that form a myelin sheath around each axon, and fibroblasts scattered between nerve fibers. Traumatic injury to peripheral nerves is produced by tissue damage at a site of physical trauma with consequent impairment of neural functions. Nerve segments at distal locations that did not sustain direct injury subsequently undergo cellular changes due to wallerian degeneration (33). This process begins within 24–36 hours after the traumatic event. The part of the axon that is separated from the neuronal cell body degenerates distal to the site of injury: the axonal skeleton disintegrates, and axonal membranes break apart. Deterioration of the myelin sheath follows. Eventually, macrophages clear up the debris, and repair is achieved through the regeneration of damaged axons and reinnervation of tissues. Within 4 days after the initial traumatic nerve injury, the distal end of the proximal nerve fiber sends out sprouts toward the hollow axonal tube consisting of Schwann cells. Growing at a rate of up to 1 mm per day, these sprouts soon enter the tube and extend along it to reinnervate the tissue at the site of injury (34). Because of multiple clinical factors, however, peripheral axonal regrowth is often delayed and seldom completed (33,35).

Takagi et al (36) performed surgical contusion of the sciatic nerve in 120 rats. They followed the evolution of neural damage with repeated DT imaging examinations in which intact (preinjured) excised nerves were compared with injured nerves excised at 3 hours, 1 day, 4 days, and seven other intervals 1–12 weeks after the crushing injury (n = 10 nerves at each time-point) (Fig 9 ). Histologic analysis of each excised nerve was performed for correlation. The pattern of change in DT parameters was consistent with the modifications seen at histologic analysis: degeneration of the nerve was followed by regeneration, beginning at the site of contusion and extending distally. The DT parameters returned to normal levels as muscle strength was recovered. Restoration of the FA value at the lesion epicenter was strongly correlated with clinical parameters of motor and sensory functional recovery (ie, leg muscle contraction test, rotarod test, and von Frey filament test). The FA values measured in the peripheral nerves were more strongly correlated with axonal parameters (eg, density and diameter) than with myelin sheath–related parameters (eg, density and thickness), suggesting that axonal membranes play a major role in anisotropic water diffusion and that myelination modulates the degree of anisotropy. The alteration in FA values within different segments of the nerve was correlated with both histologic and functional changes associated with wallerian degeneration and regeneration of peripheral nerves. These results are corroborated by the results of other studies of nerve regeneration (37,38).

Figure 9a **(a, b)** FA maps from DT imaging **(a)** and MR tractography **(b)** in a rat show intact *(Proximal, Distal)* and injured *(Epicenter)* portions of a sciatic nerve before *(pre)* and at various time intervals (hours, days, and weeks) after surgical contusion. **(c)** Histograms show the temporal evolution of FA values at and adjacent to the site of nerve injury (A, proximal to injury; B, epicenter of injury; C, distal to injury). Distal nerve degeneration due to contusion is clearly identifiable by the downward trend of the curves in B and C, followed by the gradual upward trend of tissue recovery over the ensuing hours, days, and weeks. (Reprinted, with permission, from reference 36.)

Figure 9b **(a, b)** FA maps from DT imaging **(a)** and MR tractography **(b)** in a rat show intact *(Proximal, Distal)* and injured *(Epicenter)* portions of a sciatic nerve before *(pre)* and at various time intervals (hours, days, and weeks) after surgical contusion. **(c)** Histograms show the temporal evolution of FA values at and adjacent to the site of nerve injury (A, proximal to injury; B, epicenter of injury; C, distal to injury). Distal nerve degeneration due to contusion is clearly identifiable by the downward trend of the curves in B and C, followed by the gradual upward trend of tissue recovery over the ensuing hours, days, and weeks. (Reprinted, with permission, from reference 36.)

Figure 9c **(a, b)** FA maps from DT imaging **(a)** and MR tractography **(b)** in a rat show intact *(Proximal, Distal)* and injured *(Epicenter)* portions of a sciatic nerve before *(pre)* and at various time intervals (hours, days, and weeks) after surgical contusion. **(c)** Histograms show the temporal evolution of FA values at and adjacent to the site of nerve injury (A, proximal to injury; B, epicenter of injury; C, distal to injury). Distal nerve degeneration due to contusion is clearly identifiable by the downward trend of the curves in B and C, followed by the gradual upward trend of tissue recovery over the ensuing hours, days, and weeks. (Reprinted, with permission, from reference 36.)

DT imaging was used to assess wallerian degeneration in another experimental study carried out by Zhang et al (39). The authors followed the DT imaging changes occurring in the spinal cord in rats after unilateral L2-L4 dorsal root axotomy. They noticed alterations of DT parameters (parallel and perpendicular diffusivity) on day 3 in the ipsilateral dorsal column, from the cervical to the lumbar area, in comparison with diffusivity on the contralateral unharmed side. Contemporaneous histologic changes included loss in and alteration of neurofilaments as well as formation of myelin ovoids, without any statistically significant decrease in the quantity of myelin. Although parallel diffusivity remained low over 30 days, the authors noticed a slow but statistically significant increase in perpendicular diffusivity between day 3 and day 30, a finding that correlated with a gradual clearance of myelin, without any statistically significant change in the accumulation of nonphosphorylated neurofilaments. The results of this study established that DT imaging has high sensitivity for the detection of early axonal changes and later myelin changes in wallerian degeneration.

Lumbar Nerve Abnormalities.—Balbi et al (23) demonstrated the feasibility of DT imaging of the lumbar nerve roots. Tractography enabled the identification of L5 and S1 nerve roots in the lateral recess and foramen, the areas most often affected by lumbar disk herniation (Fig 10a). The authors included 19 patients with unilateral sciatica (S1 or L5) related to unilateral disk herniation seen on CT or MR images and matched them with 19 healthy volunteers. Lumbar root compression sites were clearly identified on the tractographic images (Fig 10b). Statistically significant changes in diffusion parameters (decreased FA and increased mean diffusivity) were found in the compressed lumbar nerves, in comparison with the contralateral nerve root or the corresponding nerve roots in volunteers. FA changes were also found in a similar study published recently (40). Kitamura et al (41) reported that DT imaging was helpful in a case of sciatica resulting from extraforaminal compression. These results show that DT imaging holds promise not only for ascertaining whether a nerve is compressed but also for objectively establishing the presence and extent of neural damage, a capability not provided by current clinical imaging techniques. Further studies are required to determine whether DT parameters are directly correlated with clinical symptoms and whether they may be prognostic of postoperative outcome.

Figure 10a Fusion tractographic–T2-weighted MR images obtained in a patient with unilateral sciatica show the L5 and S1 nerves in anterior **(a)** and left lateral **(b)** views. The orientation of the nerve fibers is signaled by their different colors (blue = craniocaudal, purple = oblique craniocaudal-laterolateral, red = laterolateral). Mass effect of a herniated disk (arrow) on the lumbar nerve root is clearly visible in b.

Figure 10b Fusion tractographic–T2-weighted MR images obtained in a patient with unilateral sciatica show the L5 and S1 nerves in anterior **(a)** and left lateral **(b)** views. The orientation of the nerve fibers is signaled by their different colors (blue = craniocaudal, purple = oblique craniocaudal-laterolateral, red = laterolateral). Mass effect of a herniated disk (arrow) on the lumbar nerve root is clearly visible in b.

Median Nerve Abnormalities.—In a study of DT imaging in patients with carpal tunnel syndrome, Khalil et al (21) showed that mean FA of the median nerve at DT imaging in symptomatic patients with carpal tunnel syndrome was lower than that in controls even when conventional imaging sequences showed no morphologic abnormalities of the nerve. The authors hypothesized that these FA changes resulted from a combination of segmental demyelination, wallerian degeneration, and axonal damage, classic findings in patients with focal neuropathies. Khalil and colleagues proposed that changes in connective tissue may also be involved, as postmortem studies of the median nerve in the presence of severe carpal tunnel syndrome have shown a marked increase in the proportion of endoneural and perineural connective tissue located deep and proximal to the flexor retinaculum. Differences in the DT parameters were corroborated in further studies, including a recent study by Wang et al (20), in which the authors demonstrated that decreased FA and increased ADC within the compressed nerve were correlated with electrophysiologic abnormalities. A significant correlation was found between the degree of FA and ADC alteration and the severity of carpal tunnel syndrome.

In a recent study, Hiltunen et al (42) compared DT parameters in pre- and postoperative patients with carpal tunnel syndrome with those in both young and age-matched control groups. By contrast with results from previous studies, the only parameter that was statistically different between preoperative patients and age-matched controls was mean diffusivity measured in the distal nerve. Postoperative clinical improvement was reflected in diffusivity but not in anisotropy. However, Hiltunen and colleagues also demonstrated that aging alone caused changes similar to those seen in carpal tunnel syndrome. Young control subjects had higher anisotropy and lower diffusivity than the other groups. Considerable inter- and intraindividual variations in mean diffusivity and FA also were reported. The authors thus inferred that age and inter- and intraindividual variability can bias group comparisons (42).

Guggenberger et al (43) corroborated these results, confirming that FA decreases and ADC increases with increasing age, as well as in the presence of carpal tunnel syndrome. They also proposed specific FA and ADC threshold values for use in the diagnosis of this disease.

Other Peripheral Neuropathies.—The literature on DT imaging of peripheral nerves other than the median nerve is limited. The feasibility of DT imaging of the ulnar and median nerves in the forearm was recently reported (18,44). Kakuda et al (45) studied the tibial nerves in patients with chronic inflammatory demyelinating polyradiculoneuropathy, comparing them to tibial nerves in sex- and age-matched healthy volunteers. A significant correlation was found between FA values and amplitude of action potential in the nerves at electrophysiologic examination. The use of DT imaging to evaluate the peroneal and tibial nerves around the knee also has been reported (46).

To our knowledge, no study of DT imaging of nerves in the ankle, elbow, or shoulder has been reported. However, patients in whom an entrapment syndrome in one of those areas is suspected might well benefit from DT imaging, since nerve abnormalities are not always identifiable on conventional MR images (47). In a recent study of the usefulness of neurography for diagnosis of ulnar neuropathy at the elbow, DT imaging was acknowledged to improve diagnostic accuracy (48).

Peripheral Nerve Tumors.—DT imaging is used mainly for preoperative assessment in patients with brain tumors (49,50). DT imaging–based descriptions of peripheral nerve tumors are scarce, but a few studies have shown different neural fiber patterns within tumors. Some nerves have a globally swollen aspect on tractographic images (51), whereas in others the lesion appears to spread the fibers apart (52,53). Although only case reports and preliminary study results are available, these suggest a need for further investigation.

In a case report of an intramural sciatic nerve perineuroma in a 7-year-old girl, tractographic images depicted global swelling of the nerve and its extension to the sacral roots (51). In another study, Viallon et al (52) reported a case of a schwannoma of the brachial plexus. On DT images, the fibers clearly spread out around the tumor, whereas conventional MR images obtained with a 3D short inversion time inversion-recovery sequence showed a mass that could not be distinguished from the nerve bundles. In Vargas et al (53), DT imaging depicted white-matter tracts spreading around an ependymoma of the cervical spinal cord. Although DT imaging–based signs of nerve tumors have yet to be defined, the features described in these case studies are consistent with the known histopathologic aspects of the lesions depicted.

Depiction of Spinal Neuropathies

Cervical Spondylotic Myelopathy.—In a recent study of DT imaging in which the findings in patients with cervical spondylotic myelopathy were compared with those in healthy volunteers (22), a significant correlation was found between FA values and clinical scores according to the Japanese Orthopedic Association Cervical Myelopathy Evaluation Questionnaire (JOACMEQ). However, findings of focal high signal intensity in the spinal cord on T2-weighted images were not correlated with either the DT parameters or the clinical scores (Fig 11 ). These results were corroborated by Jones et al (54) in their study of 30 patients with cervical spondylotic myelopathy, in which the FA values correlated well with the clinical scores obtained by using the modified Japanese Orthopedic Association scoring system. Moreover, higher FA values on preoperative DT images obtained in 15 patients who subsequently underwent decompressive surgery were correlated with functional recovery after surgery, as demonstrated by improvement in the Neck Disability Index score. Kara et al (55) found FA and ADC changes in patients with neurologic signs and symptoms of cervical spondylotic myelopathy but without signal hyperintensity in the spinal cord on T2-weighted MR images.

Figure 11a Spinal injuries. **(a–c)** Sagittal T2-weighted MR image **(a)**, fused tractographic–T2-weighted image **(b)**, and magnified 3D tractographic image **(c)** obtained in a patient with poor JOACMEQ clinical scores (16/100 for the upper extremities and 9/100 for the lower extremities) show severe spinal stenosis (arrow), a finding consistent with the low FA values seen in b and c and the low clinical scores. **(d, e)** Sagittal T2-weighted MR image **(d)** and fused tractographic–T2-weighted image **(e)** obtained in a patient with intermediate JOACMEQ clinical scores suggestive of moderate impairment (of 95/100 for the upper limbs and 59/100 for the lower limbs) show a focal region with increased signal intensity (arrow in d) but high mean FA (area between the two horizontal lines in e) suggestive of preservation of the cord microstructure. The imaging findings correlated well with the results of clinical assessment. **(f, g)** Sagittal T2-weighted MR image **(f)** and fused tractographic–T2-weighted image **(g)** obtained in a patient with spinal cord compression show normal signal intensity but decreased FA values in the compressed segment, a finding that corresponded to the patient’s low JOACMEQ scores (68/100 and 55/100 for the upper and lower limbs, respectively). (Reprinted, with permission, from reference 22.)

Figure 11b Spinal injuries. **(a–c)** Sagittal T2-weighted MR image **(a)**, fused tractographic–T2-weighted image **(b)**, and magnified 3D tractographic image **(c)** obtained in a patient with poor JOACMEQ clinical scores (16/100 for the upper extremities and 9/100 for the lower extremities) show severe spinal stenosis (arrow), a finding consistent with the low FA values seen in b and c and the low clinical scores. **(d, e)** Sagittal T2-weighted MR image **(d)** and fused tractographic–T2-weighted image **(e)** obtained in a patient with intermediate JOACMEQ clinical scores suggestive of moderate impairment (of 95/100 for the upper limbs and 59/100 for the lower limbs) show a focal region with increased signal intensity (arrow in d) but high mean FA (area between the two horizontal lines in e) suggestive of preservation of the cord microstructure. The imaging findings correlated well with the results of clinical assessment. **(f, g)** Sagittal T2-weighted MR image **(f)** and fused tractographic–T2-weighted image **(g)** obtained in a patient with spinal cord compression show normal signal intensity but decreased FA values in the compressed segment, a finding that corresponded to the patient’s low JOACMEQ scores (68/100 and 55/100 for the upper and lower limbs, respectively). (Reprinted, with permission, from reference 22.)

Figure 11c Spinal injuries. **(a–c)** Sagittal T2-weighted MR image **(a)**, fused tractographic–T2-weighted image **(b)**, and magnified 3D tractographic image **(c)** obtained in a patient with poor JOACMEQ clinical scores (16/100 for the upper extremities and 9/100 for the lower extremities) show severe spinal stenosis (arrow), a finding consistent with the low FA values seen in b and c and the low clinical scores. **(d, e)** Sagittal T2-weighted MR image **(d)** and fused tractographic–T2-weighted image **(e)** obtained in a patient with intermediate JOACMEQ clinical scores suggestive of moderate impairment (of 95/100 for the upper limbs and 59/100 for the lower limbs) show a focal region with increased signal intensity (arrow in d) but high mean FA (area between the two horizontal lines in e) suggestive of preservation of the cord microstructure. The imaging findings correlated well with the results of clinical assessment. **(f, g)** Sagittal T2-weighted MR image **(f)** and fused tractographic–T2-weighted image **(g)** obtained in a patient with spinal cord compression show normal signal intensity but decreased FA values in the compressed segment, a finding that corresponded to the patient’s low JOACMEQ scores (68/100 and 55/100 for the upper and lower limbs, respectively). (Reprinted, with permission, from reference 22.)

Figure 11d Spinal injuries. **(a–c)** Sagittal T2-weighted MR image **(a)**, fused tractographic–T2-weighted image **(b)**, and magnified 3D tractographic image **(c)** obtained in a patient with poor JOACMEQ clinical scores (16/100 for the upper extremities and 9/100 for the lower extremities) show severe spinal stenosis (arrow), a finding consistent with the low FA values seen in b and c and the low clinical scores. **(d, e)** Sagittal T2-weighted MR image **(d)** and fused tractographic–T2-weighted image **(e)** obtained in a patient with intermediate JOACMEQ clinical scores suggestive of moderate impairment (of 95/100 for the upper limbs and 59/100 for the lower limbs) show a focal region with increased signal intensity (arrow in d) but high mean FA (area between the two horizontal lines in e) suggestive of preservation of the cord microstructure. The imaging findings correlated well with the results of clinical assessment. **(f, g)** Sagittal T2-weighted MR image **(f)** and fused tractographic–T2-weighted image **(g)** obtained in a patient with spinal cord compression show normal signal intensity but decreased FA values in the compressed segment, a finding that corresponded to the patient’s low JOACMEQ scores (68/100 and 55/100 for the upper and lower limbs, respectively). (Reprinted, with permission, from reference 22.)

Figure 11e Spinal injuries. **(a–c)** Sagittal T2-weighted MR image **(a)**, fused tractographic–T2-weighted image **(b)**, and magnified 3D tractographic image **(c)** obtained in a patient with poor JOACMEQ clinical scores (16/100 for the upper extremities and 9/100 for the lower extremities) show severe spinal stenosis (arrow), a finding consistent with the low FA values seen in b and c and the low clinical scores. **(d, e)** Sagittal T2-weighted MR image **(d)** and fused tractographic–T2-weighted image **(e)** obtained in a patient with intermediate JOACMEQ clinical scores suggestive of moderate impairment (of 95/100 for the upper limbs and 59/100 for the lower limbs) show a focal region with increased signal intensity (arrow in d) but high mean FA (area between the two horizontal lines in e) suggestive of preservation of the cord microstructure. The imaging findings correlated well with the results of clinical assessment. **(f, g)** Sagittal T2-weighted MR image **(f)** and fused tractographic–T2-weighted image **(g)** obtained in a patient with spinal cord compression show normal signal intensity but decreased FA values in the compressed segment, a finding that corresponded to the patient’s low JOACMEQ scores (68/100 and 55/100 for the upper and lower limbs, respectively). (Reprinted, with permission, from reference 22.)

Figure 11f Spinal injuries. **(a–c)** Sagittal T2-weighted MR image **(a)**, fused tractographic–T2-weighted image **(b)**, and magnified 3D tractographic image **(c)** obtained in a patient with poor JOACMEQ clinical scores (16/100 for the upper extremities and 9/100 for the lower extremities) show severe spinal stenosis (arrow), a finding consistent with the low FA values seen in b and c and the low clinical scores. **(d, e)** Sagittal T2-weighted MR image **(d)** and fused tractographic–T2-weighted image **(e)** obtained in a patient with intermediate JOACMEQ clinical scores suggestive of moderate impairment (of 95/100 for the upper limbs and 59/100 for the lower limbs) show a focal region with increased signal intensity (arrow in d) but high mean FA (area between the two horizontal lines in e) suggestive of preservation of the cord microstructure. The imaging findings correlated well with the results of clinical assessment. **(f, g)** Sagittal T2-weighted MR image **(f)** and fused tractographic–T2-weighted image **(g)** obtained in a patient with spinal cord compression show normal signal intensity but decreased FA values in the compressed segment, a finding that corresponded to the patient’s low JOACMEQ scores (68/100 and 55/100 for the upper and lower limbs, respectively). (Reprinted, with permission, from reference 22.)

Figure 11g Spinal injuries. **(a–c)** Sagittal T2-weighted MR image **(a)**, fused tractographic–T2-weighted image **(b)**, and magnified 3D tractographic image **(c)** obtained in a patient with poor JOACMEQ clinical scores (16/100 for the upper extremities and 9/100 for the lower extremities) show severe spinal stenosis (arrow), a finding consistent with the low FA values seen in b and c and the low clinical scores. **(d, e)** Sagittal T2-weighted MR image **(d)** and fused tractographic–T2-weighted image **(e)** obtained in a patient with intermediate JOACMEQ clinical scores suggestive of moderate impairment (of 95/100 for the upper limbs and 59/100 for the lower limbs) show a focal region with increased signal intensity (arrow in d) but high mean FA (area between the two horizontal lines in e) suggestive of preservation of the cord microstructure. The imaging findings correlated well with the results of clinical assessment. **(f, g)** Sagittal T2-weighted MR image **(f)** and fused tractographic–T2-weighted image **(g)** obtained in a patient with spinal cord compression show normal signal intensity but decreased FA values in the compressed segment, a finding that corresponded to the patient’s low JOACMEQ scores (68/100 and 55/100 for the upper and lower limbs, respectively). (Reprinted, with permission, from reference 22.)

Kerkovský et al (56) correlated clinical and electrophysiologic data with DT parameters. A group of patients in whom spondylotic cervical cord encroachment was identified on MR images was divided into two subgroups: asymptomatic patients and patients with clinically diagnosed cervical spondylotic myelopathy. No further clinical evaluation was done. Kerkovský and colleagues found that DT parameters in the compressed spinal cord differed significantly between patients and controls as well as between the asymptomatic and symptomatic subgroups of patients with a diagnosis of myelopathy. Electrophysiologic abnormalities were more frequent in symptomatic patients than in asymptomatic patients, but the difference was without statistical significance. Furthermore, there was no difference in any of the DT parameters between the patient subset with electrophysiologic abnormalities and the subset without such abnormalities. Kerkovský et al concluded that the accuracy of DT imaging for discriminating between the two clinical subgroups of patients was higher than that of either conventional MR imaging or electrophysiologic study (56).

These studies demonstrated that FA abnormalities exist in discrete clinical states of cervical spondylotic myelopathy, are correlated with clinical scores, and have higher accuracy for diagnosis than T2 signal hyperintensities (22,55) or abnormalities found in electrophysiologic studies (56).

DT imaging is capable of depicting microstructural alterations that remain occult at conventional MR imaging.

The variation in DT parameters is correlated with functional outcome after surgery (54). These features suggest that DT imaging may have a role to play in surgical decision-making models.

Cervical Trauma.—In a study that included 10 patients with chronic injuries of the cervical spine, Ellingson et al (19) demonstrated that the magnitude of overall diffusion within the spinal cord was higher at the injury site but significantly reduced in the upper cervical segments, suggesting that regions remote from the injury site may be affected by the pathophysiologic processes associated with long-term recovery from spinal injury. This is consistent with known phenomena such as wallerian degeneration and regeneration. These findings were corroborated in a more recent study by Cheran et al (57) in which patients suffering from hemorrhagic or nonhemorrhagic acute injury of the cervical spine were compared with controls. Notably, the American Spinal Injury Association clinical injury motor scores were significantly correlated with four DT parameters (FA, mean diffusivity, radial diffusivity, and longitudinal diffusivity) at the injury site in the group of patients with nonhemorrhagic injury, whereas no correlation was found in the group with hemorrhagic injury. This finding is in agreement with the hypothesis that DT parameters reflect microstructural damage in tissue. There is indeed no reason to expect that DT parameters measured in a hematoma must reflect any neural damage. Moreover, although the subject of study was acute cervical trauma, Cheran and colleagues found a statistically significant reduction in FA at the whole-cord level between the injured (both hemorrhagic and nonhemorrhagic) and control groups (57). This finding suggests that DT imaging can reliably depict microstructural modifications that occur in the spinal cord outside the injury site, even when no abnormality is visible on conventional MR images.

Future Prospects for DT Imaging

Technical Developments

DT imaging sequences will surely undergo further improvement (58). Imaging techniques that are used to evaluate the muscles and nerves must be adapted to overcome the special challenges presented by different anatomic locations. The depiction of delicate neural structures requires high spatial resolution within the plane of imaging as well as along the z-axis (23), particularly in areas such as the lumbar spine, where these anisotropic structures are surrounded by cerebrospinal fluid. Partial volume effects can have dramatic consequences for FA measurements. Yet, the loss of signal intensity that results from the use of smaller voxel sizes requires that the need to avoid partial volume effects must be weighed against the necessity of maintaining a high signal-to-noise ratio so as to obtain accurate measurements of diffusion parameters. Moreover, acquisition time must be kept as short as possible because patients who are experiencing nerve pain may move during the imaging examination.

Recent studies demonstrated benefits from the use of a reduced field of view for DT imaging of the spinal cord and peripheral nerves (59–61). With this technique, only the signal coming from a predefined area of interest is conserved. The advantages include the reduction of foldover artifacts and decreased readout time with resultant reduction in chemical shift and susceptibility artifacts. Superior results were demonstrated with the use of a reduced field of view, in comparison with a full field of view, at diffusion imaging of the spinal cord (59) and at DT imaging of lumbar nerve roots (60). Further studies are needed to assess the benefits of this technique in other anatomic locations.

Methodological Considerations

The pertinence of normative values must be assessed: Should normative values be identified for each nerve or muscle group? Under what conditions would such values be valid? In other words, would they be valid only when using the same scanner and the same sequence, or only when evaluating subjects of the same age group or the same sex? Considering the variability of DT imaging measurements in some studies (9,10,15), multiple “control” measurements obtained in different anatomic locations in the same patient could well be used to determine whether an abnormality is present (eg, in the assessment of unilateral radicular pain) (23). This approach, however, would be useless for establishing whether a systemic disease such as polymyositis is present.

Clinical Effectiveness

The vast majority of DT imaging studies of the muscles and nerves have been performed in small patient samples (22,23,40,42,45). Cohort studies of the effectiveness of DT imaging in comparison with other imaging methods for establishing specific clinical diagnoses, or of DT imaging in various patient groups and in controls, have yet to be conducted. Likewise, studies assessing the relative benefits of DT imaging in individual cases are still needed. Longitudinal studies are also mandatory to determine the prognostic value of DT imaging measurements, especially for surgical outcomes.

Conclusion

DT imaging is an MR imaging technique that allows noninvasive assessment of the biologic microstructures of muscles and nerves. Three-dimensional depiction of those structures is possible with tractography. The degree of FA and the mean diffusivity, parameters that can be objectively measured at DT imaging, represent properties of biologic tissues that may be altered by various physiologic states and pathologic conditions. Thus, DT imaging shows promise for the detection of early abnormalities in muscles and nerves before they become apparent at conventional MR imaging. However, the clinical utility of DT imaging for musculoskeletal evaluations is still under assessment, and much work remains to be done before its role may be established.

Recipient of a Magna Cum Laude award for an education exhibit at the 2011 RSNA Annual Meeting.

For this journal-based SA-CME activity, the authors, editor, and reviewers have no relevant relationships to disclose.

References

1. 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, Google Scholar
2. Basser PJ, Jones DK. Diffusion-tensor MRI: theory, experimental design and data analysis—a technical review. NMR Biomed 2002;15(7-8):456–467. Crossref, Medline, Google Scholar
3. Le Bihan D, Breton E, Lallemand D, Grenier P, Cabanis E, Laval-Jeantet M. MR imaging of intravoxel incoherent motions: application to diffusion and perfusion in neurologic disorders. Radiology 1986;161(2):401–407. Link, Google Scholar
4. Loubinoux I, Volk A, Borredon J, et al. Spreading of vasogenic edema and cytotoxic edema assessed by quantitative diffusion and T2 magnetic resonance imaging. Stroke 1997;28(2):419–426; discussion 426–427. Crossref, Medline, Google Scholar
5. Lee SK. Diffusion tensor and perfusion imaging of brain tumors in high-field MR imaging. Neuroimaging Clin N Am 2012;22(2):123–134, ix. Crossref, Medline, Google Scholar
6. Oppenheim C, Ducreux D, Rodrigo S, et al. Diffusion tensor imaging and tractography of the brain and spinal cord [in French]. J Radiol 2007;88(3 pt 2):510–520. Crossref, Medline, Google Scholar
7. Bammer R, Acar B, Moseley ME. In vivo MR tractography using diffusion imaging. Eur J Radiol 2003;45(3):223–234. Crossref, Medline, Google Scholar
8. Hasan KM, Walimuni IS, Abid H, Hahn KR. A review of diffusion tensor magnetic resonance imaging computational methods and software tools. Comput Biol Med 2011;41(12):1062–1072. Crossref, Medline, Google Scholar
9. Santarelli X, Garbin G, Ukmar M, Longo R. Dependence of the fractional anisotropy in cervical spine from the number of diffusion gradients, repeated acquisition and voxel size. Magn Reson Imaging 2010;28(1):70–76. Crossref, Medline, Google Scholar
10. Takao H, Hayashi N, Ohtomo K. Effect of scanner in asymmetry studies using diffusion tensor imaging. Neuroimage 2011;54(2):1053–1062. Crossref, Medline, Google Scholar
11. Vollmar C, O’Muircheartaigh J, Barker GJ, et al. Identical, but not the same: intra-site and inter-site reproducibility of fractional anisotropy measures on two 3.0T scanners. Neuroimage 2010;51(4):1384–1394. Crossref, Medline, Google Scholar
12. Saotome K, Ishimori Y, Isobe T, et al. Comparison of diffusion tensor imaging-derived fractional anisotropy in multiple centers for identical human subjects [in Japanese]. Nippon Hoshasen Gijutsu Gakkai Zasshi 2012;68(9):1242–1249. Crossref, Medline, Google Scholar
13. Mulcahey MJ, Samdani A, Gaughan J, et al. Diffusion tensor imaging in pediatric spinal cord injury: preliminary examination of reliability and clinical correlation. Spine 2012;37(13):E797–E803. Crossref, Medline, Google Scholar
14. Danielian LE, Iwata NK, Thomasson DM, Floeter MK. Reliability of fiber tracking measurements in diffusion tensor imaging for longitudinal study. Neuroimage 2010;49(2):1572–1580. Crossref, Medline, Google Scholar
15. Teipel SJ, Reuter S, Stieltjes B, et al. Multicenter stability of diffusion tensor imaging measures: a European clinical and physical phantom study. Psychiatry Res 2011;194(3):363–371. Crossref, Medline, Google Scholar
16. Budzik JF, Le Thuc V, Demondion X, Morel M, Chechin D, Cotten A. In vivo MR tractography of thigh muscles using diffusion imaging: initial results. Eur Radiol 2007;17(12):3079–3085. Crossref, Medline, Google Scholar
17. Mori S, van Zijl PC. Fiber tracking: principles and strategies—a technical review. NMR Biomed 2002; 15(7-8):468–480. Crossref, Medline, Google Scholar
18. Jambawalikar S, Baum J, Button T, Li H, Geronimo V, Gould ES. Diffusion tensor imaging of peripheral nerves. Skeletal Radiol 2010;39(11):1073–1079. Crossref, Medline, Google Scholar
19. Ellingson BM, Ulmer JL, Kurpad SN, Schmit BD. Diffusion tensor MR imaging in chronic spinal cord injury. AJNR Am J Neuroradiol 2008;29(10): 1976–1982. Crossref, Medline, Google Scholar
20. Wang CK, Jou IM, Huang HW, et al. Carpal tunnel syndrome assessed with diffusion tensor imaging: comparison with electrophysiological studies of patients and healthy volunteers. Eur J Radiol 2012;81(11):3378–3383. Crossref, Medline, Google Scholar
21. Khalil C, Hancart C, Le Thuc V, Chantelot C, Chechin D, Cotten A. Diffusion tensor imaging and tractography of the median nerve in carpal tunnel syndrome: preliminary results. Eur Radiol 2008;18(10):2283–2291. Crossref, Medline, Google Scholar
22. Budzik JF, Balbi V, Le Thuc V, Duhamel A, Assaker R, Cotten A. Diffusion tensor imaging and fibre tracking in cervical spondylotic myelopathy. Eur Radiol 2011;21(2):426–433. Crossref, Medline, Google Scholar
23. Balbi V, Budzik JF, Duhamel A, Bera-Louville A, Le Thuc V, Cotten A. Tractography of lumbar nerve roots: initial results. Eur Radiol 2011;21(6):1153–1159. Crossref, Medline, Google Scholar
24. Gaige TA, Benner T, Wang R, Wedeen VJ, Gilbert RJ. Three dimensional myoarchitecture of the human tongue determined in vivo by diffusion tensor imaging with tractography. J Magn Reson Imaging 2007; 26(3):654–661. Crossref, Medline, Google Scholar
25. Zijta FM, Froeling M, van der Paardt MP, et al. Feasibility of diffusion tensor imaging (DTI) with fibre tractography of the normal female pelvic floor. Eur Radiol 2011;21(6):1243–1249. Crossref, Medline, Google Scholar
26. Galbán CJ, Maderwald S, Stock F, Ladd ME. Age-related changes in skeletal muscle as detected by diffusion tensor magnetic resonance imaging. J Gerontol A Biol Sci Med Sci 2007;62(4):453–458. Crossref, Medline, Google Scholar
27. Okamoto Y, Kunimatsu A, Miki S, Shindo M, Niitsu M, Minami M. Fractional anisotropy values of calf muscles in normative state after exercise: preliminary results. Magn Reson Med Sci 2008;7(3):157–162. Crossref, Medline, Google Scholar
28. Okamoto Y, Mori S, Kujiraoka Y, Nasu K, Hirano Y, Minami M. Diffusion property differences of the lower leg musculature between athletes and non-athletes using 1.5T MRI. MAGMA 2012;25(4): 277–284. Crossref, Medline, Google Scholar
29. Deux JF, Malzy P, Paragios N, et al. Assessment of calf muscle contraction by diffusion tensor imaging. Eur Radiol 2008;18(10):2303–2310. Crossref, Medline, Google Scholar
30. Okamoto Y, Kunimatsu A, Kono T, Nasu K, Sonobe J, Minami M. Changes in MR diffusion properties during active muscle contraction in the calf. Magn Reson Med Sci 2010;9(1):1–8. Crossref, Medline, Google Scholar
31. Heemskerk AM, Strijkers GJ, Drost MR, van Bochove GS, Nicolay K. Skeletal muscle degeneration and regeneration after femoral artery ligation in mice: monitoring with diffusion MR imaging. Radiology 2007;243(2):413–421. Link, Google Scholar
32. Zeng H, Zheng JH, Zhang JE, et al. Grading of rabbit skeletal muscle trauma by diffusion tensor imaging and tractography on magnetic resonance imaging. Chin Med Sci J 2006;21(4):276–280. Medline, Google Scholar
33. Rotshenker S. Wallerian degeneration: the innate-immune response to traumatic nerve injury. J Neuroinflammation 2011;8:109. Crossref, Medline, Google Scholar
34. Delmotte A, Rigoard S, Buffenoir K, et al. Physiology of the injured peripheral nerve [in French]. Neurochirurgie 2009;55(suppl 1):S13–S21. Crossref, Medline, Google Scholar
35. Zochodne DW. The challenges and beauty of peripheral nerve regrowth. J Peripher Nerv Syst 2012; 17(1):1–18. Crossref, Medline, Google Scholar
36. Takagi T, Nakamura M, Yamada M, et al. Visualization of peripheral nerve degeneration and regeneration: monitoring with diffusion tensor tractography. Neuroimage 2009;44(3):884–892. Crossref, Medline, Google Scholar
37. Morisaki S, Kawai Y, Umeda M, et al. In vivo assessment of peripheral nerve regeneration by diffusion tensor imaging. J Magn Reson Imaging 2011; 33(3):535–542. Crossref, Medline, Google Scholar
38. Lehmann HC, Zhang J, Mori S, Sheikh KA. Diffusion tensor imaging to assess axonal regeneration in peripheral nerves. Exp Neurol 2010;223(1):238–244. Crossref, Medline, Google Scholar
39. Zhang J, Jones M, DeBoy CA, et al. Diffusion tensor magnetic resonance imaging of Wallerian degeneration in rat spinal cord after dorsal root axotomy. J Neurosci 2009;29(10):3160–3171. Crossref, Medline, Google Scholar
40. Eguchi Y, Ohtori S, Orita S, et al. Quantitative evaluation and visualization of lumbar foraminal nerve root entrapment by using diffusion tensor imaging: preliminary results. AJNR Am J Neuroradiol 2011; 32(10):1824–1829. Crossref, Medline, Google Scholar
41. Kitamura M, Eguchi Y, Inoue G, et al. A case of symptomatic extra-foraminal lumbosacral stenosis (“far-out syndrome”) diagnosed by diffusion tensor imaging. Spine 2012;37(14):E854–E857. Crossref, Medline, Google Scholar
42. Hiltunen J, Kirveskari E, Numminen J, Lindfors N, Göransson H, Hari R. Pre- and post-operative diffusion tensor imaging of the median nerve in carpal tunnel syndrome. Eur Radiol 2012;22(6):1310–1319. Crossref, Medline, Google Scholar
43. Guggenberger R, Markovic D, Eppenberger P, et al. Assessment of median nerve with MR neurography by using diffusion-tensor imaging: normative and pathologic diffusion values. Radiology 2012;265(1): 194–203. Link, Google Scholar
44. Zhou Y, Kumaravel M, Patel VS, Sheikh KA, Narayana PA. Diffusion tensor imaging of forearm nerves in humans. J Magn Reson Imaging 2012;36(4): 920–927. Crossref, Medline, Google Scholar
45. Kakuda T, Fukuda H, Tanitame K, et al. Diffusion tensor imaging of peripheral nerve in patients with chronic inflammatory demyelinating polyradiculoneuropathy: a feasibility study. Neuroradiology 2011;53(12):955–960. Crossref, Medline, Google Scholar
46. Hiltunen J, Suortti T, Arvela S, Seppä M, Joensuu R, Hari R. Diffusion tensor imaging and tractography of distal peripheral nerves at 3 T. Clin Neurophysiol 2005;116(10):2315–2323. Crossref, Medline, Google Scholar
47. Budzik JF, Wavreille G, Pansini V, Moraux A, Demondion X, Cotten A. Entrapment neuropathies of the shoulder. Magn Reson Imaging Clin N Am 2012;20(2):373–391, xii. Crossref, Medline, Google Scholar
48. Bäumer P, Dombert T, Staub F, et al. Ulnar neuropathy at the elbow: MR neurography—nerve T2 signal increase and caliber. Radiology 2011;260(1): 199–206. Link, Google Scholar
49. Bick AS, Mayer A, Levin N. From research to clinical practice: implementation of functional magnetic imaging and white matter tractography in the clinical environment. J Neurol Sci 2012;312(1-2): 158–165. Crossref, Medline, Google Scholar
50. Leclercq D, Delmaire C, de Champfleur NM, Chiras J, Lehéricy S. Diffusion tractography: methods, validation and applications in patients with neurosurgical lesions. Neurosurg Clin N Am 2011;22(2): 253–268, ix. Crossref, Medline, Google Scholar
51. Merlini L, Viallon M, De Coulon G, Lobrinus JA, Vargas MI. MRI neurography and diffusion tensor imaging of a sciatic perineuroma in a child. Pediatr Radiol 2008;38(9):1009–1012. Crossref, Medline, Google Scholar
52. Viallon M, Vargas MI, Jlassi H, Lövblad KO, Delavelle J. High-resolution and functional magnetic resonance imaging of the brachial plexus using an isotropic 3D T2 STIR (Short Term Inversion Recovery) SPACE sequence and diffusion tensor imaging. Eur Radiol 2008;18(5):1018–1023. Crossref, Medline, Google Scholar
53. Vargas MI, Delavelle J, Jlassi H, et al. Clinical applications of diffusion tensor tractography of the spinal cord. Neuroradiology 2008;50(1):25–29. Crossref, Medline, Google Scholar
54. Jones JG, Cen SY, Lebel RM, Hsieh PC, Law M. Diffusion tensor imaging correlates with the clinical assessment of disease severity in cervical spondylotic myelopathy and predicts outcome following surgery. AJNR Am J Neuroradiol 2013;34(2):471–478. Crossref, Medline, Google Scholar
55. Kara B, Celik A, Karadereler S, et al. The role of DTI in early detection of cervical spondylotic myelopathy: a preliminary study with 3-T MRI. Neuroradiology 2011;53(8):609–616. Crossref, Medline, Google Scholar
56. Kerkovský M, Bednarík J, Dušek L, et al. Magnetic resonance diffusion tensor imaging in patients with cervical spondylotic spinal cord compression: correlations between clinical and electrophysiological findings. Spine 2012;37(1):48–56. Crossref, Medline, Google Scholar
57. Cheran S, Shanmuganathan K, Zhuo J, et al. Correlation of MR diffusion tensor imaging parameters with ASIA motor scores in hemorrhagic and nonhemorrhagic acute spinal cord injury. J Neurotrauma 2011;28(9):1881–1892. Crossref, Medline, Google Scholar
58. Khalil C, Budzik JF, Kermarrec E, Balbi V, Le Thuc V, Cotten A. Tractography of peripheral nerves and skeletal muscles. Eur J Radiol 2010;76(3):391–397. Crossref, Medline, Google Scholar
59. Zaharchuk G, Saritas EU, Andre JB, et al. Reduced field-of-view diffusion imaging of the human spinal cord: comparison with conventional single-shot echo-planar imaging. AJNR Am J Neuroradiol 2011; 32(5):813–820. Crossref, Medline, Google Scholar
60. Budzik JF, Verclytte S, Lefebvre G, Monnet A, Forzy G, Cotten A. Assessment of reduced field of view in diffusion tensor imaging of the lumbar nerve roots at 3 T. Eur Radiol 2013;23(5):1361–1366. Crossref, Medline, Google Scholar
61. Barakat N, Mohamed FB, Hunter LN, et al. Diffusion tensor imaging of the normal pediatric spinal cord using an inner field of view echo-planar imaging sequence. AJNR Am J Neuroradiol 2012;33(6): 1127–1133. Crossref, Medline, Google Scholar

Article History

Received: Apr 11 2012
Revision requested: Aug 29 2012
Revision received: July 1 2013
Accepted: July 1 2013
Published online: May 2014
Published in print: May 2014

Vol. 34, No. 3

Abbreviations

Abbreviations:
ADC	apparent diffusion coefficient
DT	diffusion tensor
FA	fractional anisotropy
ROI	region of interest
3D	three-dimensional

Metrics

Downloaded 4,695 times

Altmetric Score