|
|
||||||||
Musculoskeletal Imaging |
1 Department of Diagnostic Radiology, Academic Medical Center, Meibergdreef 9/G1-231, 1105 AZ Amsterdam, the Netherlands.
| Abstract |
|---|
|
|
|---|
MATERIALS AND METHODS: Two-point Dixon chemical shift imaging was performed in 31 consecutive patients clinically suspected to have bone marrow disease. In some patients, Dixon studies were performed before and after the intravenous administration of gadopentetate dimeglumine, and in some patients follow-up examinations were performed, for a total of 64 studies. Areas of interest were the hand, wrist, foot, ankle, and lower leg. There was a special interest in the neuropathic foot and osteomyelitis. The uniformity of fat suppression in the entire field of view, the frequency of displacement artifacts, and the applicability of the technique in routine patient treatment were evaluated.
RESULTS: In 64 (100%) Dixon studies, uniform fat suppression was achieved. In 59 (92%) studies, there were no displacement artifacts. In five (8%) studies, displacement artifacts occurred; however, in only one (2%) study did they severely hamper the reading. Thus, in 63 (98%) studies, adequate diagnostic quality was obtained.
CONCLUSION: Two-point Dixon chemical shift imaging is a good technique for achieving uniform fat suppression in the distal parts of the extremities. Because the frequency of displacement artifacts is low, the technique is applicable in a routine clinical setting.
Index terms: Bone marrow, MR, 40.121414, 40.121415 Bones, infection, 40.20 Bones, neuropathic disorder, 40.821 Diabetes mellitus, complications, 40.20 Foot, MR, 46.21 Hand, MR, 43.21 Magnetic resonance (MR), chemical shift, 43.121414, 46.121414 Magnetic resonance (MR), fat suppression, 43.121415, 46.121415
| Introduction |
|---|
|
|
|---|
A good alternative technique for suppression of the fat signal is the short inversion time inversion-recovery (STIR) sequence. STIR is based on the rapid T1 recovery of fat and is, therefore, not substantially affected by field inhomogeneities. STIR is considered the most sensitive technique for identifying subtle bone marrow disease (1). Unfortunately, STIR suppresses all short T1 species, including tissues that have absorbed gadolinium. Hence, it is not possible to use STIR fat suppression to improve the detection of contrast medium enhancement (2). For the detection of osteomyelitis, this is an important disadvantage, because previous studies (7,8) have suggested that MR imaging with both fat suppression and gadolinium enhancement may be the imaging method of choice in clinically complicated situations.
An alternative method for fat suppression is the phase-contrast method, described by Dixon in 1984 (9), that is based on the chemical shift phenomenon. This technique does not require a high field homogeneity to achieve adequate fat suppression, thus field inhomogeneity is not a limiting factor (9). This technique has been applied successfully in MR imaging of adrenal masses and bone marrow (1013). A potential problem in using this chemical shift technique is the occurrence of displacement artifacts that hamper its application (3,4). The necessary postprocessing requires that the patient remain in the same position with absolutely no shift of the imaged structures between the in-phase and opposed-phase series (4,14).
In our hospital (Academic Medical Center, Amsterdam, the Netherlands), there is a relatively large population of patients with neuropathic foot disease, usually due to diabetes mellitus or leprosy. In this group, MR imaging is requested when infection is suspected. With our MR imager, the frequency-selective presaturation fat suppression of the distal extremities is inhomogeneous. Because much of this problem is caused by physical properties of the patient, it is largely independent of the equipment used (1,6), so we looked for an alternative technique. In a pilot study, we found the two-point Dixon chemical shift MR imaging technique in combination with the administration of gadopentetate dimeglumine promising (15). Subsequently, the present study was conducted to answer the following questions: (a) Is two-point Dixon chemical shift imaging useful in obtaining uniform fat suppression in the hands and feet? (b) Is this technique applicable in a clinical setting for the evaluation of bone marrow disease in the extremities? (c) At what frequency do displacement artifacts occur?
| MATERIALS AND METHODS |
|---|
|
|
|---|
Two-point Dixon chemical shift imaging was performed as an extension of the routine clinical MR imaging examination. In four patients, follow-up two-point Dixon chemical shift imaging examinations were performed. In a previous study (15), the image quality of two-point Dixon chemical shift imaging was tested in four healthy volunteers.
Two-Point Dixon Chemical Shift Imaging Technique
All MR examinations were performed at our institution (Academic Medical Center, Amsterdam, the Netherlands) on a 1.5-T Magnetom 63SP/4000 imager (Siemens, Erlangen, Germany). The hand or foot of interest was placed inside the circularly polarized head coil. This coil was used because it provided the best signal-to-noise ratio; the quality of our dedicated extremity coil at that time was inferior to the quality of the head coil.
The technique we used is based on the chemical shift technique described by Dixon (9). A Dixon study consists of two spin-echo series. The first one is a conventional spin-echo series. The second series is identical to the first one, except that the 180° inverting radio-frequency pulse is shifted by a time
, such that in the center of the readout gradient (
x = 0) the signals of water and fat have opposed phases (9,16). The value of
depends on the field strength (1.5 T) and on the frequency difference between fat (CH2) and water (H2O), which is approximately 3.4 ppm. Because of the presence of other resonances in the fat molecules, the optimal value for
varies a little with the chemical composition of the lipids (17). For the extremities, we use an optimal
of 1.17 msec.
In the conventional spin-echo series, the water and fat signals are in phase, yielding images in which water and fat signal intensities add up. The opposed-phase series yields images in which the water and fat signal intensities are subtracted. By combining the two sets of images with a proper postprocessing procedure, the water and fat signals are separated, which leads to images in which there is only a water signal, so-called water-only images, and therefore fat suppression, and images in which there is only a fat signal, so-called fat-only images. The postprocessing was performed on a workstation (SPARC 20-51; Sun Microsystems, Mountain View, Calif) by using a region-growing algorithm developed in our department (14).
In the present examinations, we used T1-weighted Dixon studies (570620/17 [repetition time msec/echo time msec]; section thickness, 3 mm; number of signals acquired, two to three; matrix, 256 x 256; field of view, 150240 mm). The acquisition time was 57 minutes for one spin-echo series and therefore was 1014 minutes for a Dixon study. The series were performed in succession, because it was not possible to image the in-phase and opposed-phase series interleaved.
Administration of Gadopentetate Dimeglumine
As mentioned in the introduction, the use of gadopentetate dimeglumine with STIR is not appropriate. Gadolinium produces a shortening of the T1 in water, and because STIR suppresses all short T1 species, it suppresses fat and gadolinium-containing water alike. The Dixon technique, on the other hand, like the conventional fat-saturation method, is based on the frequency difference between fat and water. So, T1-weighted Dixon water-only images will show gadolinium enhancement just as conventional T1-weighted images do.
Gadopentetate dimeglumine (0.1 millimole per kilogram of body weight; Magnevist; Schering, Berlin, Germany) was administered intravenously when desirable, mainly for the demarcation of abscesses or the detection of osteomyelitis. We routinely performed one Dixon study before and one after the administration of gadopentetate dimeglumine.
Analysis
In retrospect, two musculoskeletal radiologists (M.M., P.F.D.) analyzed the consecutive Dixon studies by consensus and considered the following parameters: (a) Uniformity was defined as overall uniform fat suppression in the entire field of view. For this purpose, water-only images were evaluated. Uniformity was scored as "Yes" or "No." (b) Displacement artifacts were scored by interpreting the water-only images as showing (i) "no displacement" or good image quality; (ii) "little displacement" or acceptable, still diagnostic image quality; or (iii) "much displacement" or unacceptable, nondiagnostic image quality.
The fat-only images were not used for diagnostic purposes.
| RESULTS |
|---|
|
|
|---|
|
Analysis
Uniformity.In 64 (100%) Dixon studies, uniform fat suppression in the entire field of view was attained. This was evident to both radiologists. There were no artifacts in the fat-water separation. The water-only images showed no areas of persistent fat signal intensity (Fig 1). Also, in the phalanges, there was complete fat suppression.
|
|
|
| DISCUSSION |
|---|
|
|
|---|
Because chemical shift imaging, as described by Dixon (9), in combination with an appropriate postprocessing technique, such as that developed in our department (14,15), is a potentially good alternative for obtaining fat-suppressed images, we examined its use in a clinical setting on a routine basis.
We were able to obtain uniform fat suppression in the entire field of view in every study. However, we did not perform another fat-suppression technique at the same time, which renders a comparative study of several techniques impossible.
A potential weakness of the Dixon technique is the shift of position between the in-phase and opposed-phase series, which leads to displacement artifacts (1,3). Because the amount of displacement perpendicular to the imaging plane cannot be estimated from the image data, and because nonuniform displacement (eg, the bending of a toe) is even harder to quantify, we chose a qualitative analysis of this phenomenon. Displacement artifacts were encountered in only five (8%) Dixon studies. Only one (2%) Dixon study was inconclusive because of too much displacement between in-phase and opposed-phase images. To prevent displacement artifacts, we seek to develop a sequence in which both the in-phase and opposed-phase series can be acquired interleaved in one sequence. Thus, our clinical results with this Dixon technique are both applicable and promising.
Diagnosing osteomyelitis is a well-known challenge in diagnostic radiology (19,20). In neuropathic foot disease, the discrimination between neuropathic osteoarthropathy and osteomyelitis is an additional problem. Because of inhomogeneous suppression of the high marrow signal on both T1-weighted spin-echo and fast T2-weighted spin-echo images, some parts of the foot cannot be judged adequately and are prone to misconception (1). Therefore, homogeneous fat suppression in the entire field of view both before and after the intravenous administration of gadopentetate dimeglumine is mandatory. The results of the present study indicate that our two-point Dixon chemical shift imaging technique leads to good uniform fat suppression. Its potential to help obtain images both before and after the intravenous administration of gadopentetate dimeglumine is an important development and may be of benefit in diagnosing osteomyelitis in complex situations and in the distal parts of the extremities, such as the neuropathic foot (Fig 3). Further studies in this field are necessary and will be performed.
|
|
| Acknowledgments |
|---|
| Footnotes |
|---|
Abbreviation: STIR = short inversion time inversion recovery
Author contributions: Guarantor of integrity of entire study, M.M.; study concepts and design, M.M., P.F.D.; definition of intellectual content, M.M., P.F.D., E.M.A.; literature research, M.M., E.M.A.; clinical studies, M.M., P.F.D., E.M.A.; experimental studies, M.M., P.F.D., E.M.A.; data acquisition, M.M., P.F.D., E.M.A.; data analysis, M.M., P.F.D.; manuscript preparation and review, M.M., P.F.D., E.M.A.; manuscript editing, M.M., E.M.A.
Received December 9, 1997;
revision requested February 27, 1998; revision received June 8, 1998;
accepted August 20, 1998.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. A. W. Verhagen, M. Maas, M. G. W. Dijkgraaf, J. L. Tol, R. Krips, and C. N. van Dijk Prospective study on diagnostic strategies in osteochondral lesions of the talus: IS MRI SUPERIOR TO HELICAL CT? J Bone Joint Surg Br, January 1, 2005; 87-B(1): 41 - 46. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. A. Bus, M. Maas, P. R. Cavanagh, R. P. J. Michels, and M. Levi Plantar Fat-Pad Displacement in Neuropathic Diabetic Patients With Toe Deformity: A magnetic resonance imaging study Diabetes Care, October 1, 2004; 27(10): 2376 - 2381. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Maas, C. E. M. Hollak, E. M. Akkerman, J. M. F. G. Aerts, J. Stoker, and G. J. Den Heeten Quantification of Skeletal Involvement in Adults with Type I Gaucher's Disease: Fat Fraction Measured by Dixon Quantitative Chemical Shift Imaging as a Valid Parameter Am. J. Roentgenol., October 1, 2002; 179(4): 961 - 965. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. J. Rybicki, T. Chung, J. Reid, D. Jaramillo, R. V. Mulkern, and J. Ma Fast Three-Point Dixon MR Imaging Using Low-Resolution Images for Phase Correction: A Comparison with Chemical Shift Selective Fat Suppression for Pediatric Musculoskeletal Imaging Am. J. Roentgenol., November 1, 2001; 177(5): 1019 - 1023. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |