| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Medical and Health Science Center, University of
Debrecen, Nagyerdei krt. 98, H-4012 Debrecen, Hungary.
2 Department of Ophthalmology, Medical and Health Science Center, University of
Debrecen, H-4012 Debrecen, Hungary.
3 Department of Nuclear Medicine, Medical and Health Science Center, University
of Debrecen, H-4012 Debrecen, Hungary.
4 Endocrine Section, Department of Medicine, Washington Hospital Center, 110
Irving St., N.W., Washington, DC 20010-2975.
5 Department of Internal Medicine, Medical and Health Science Center, University
of Debrecen, H-4012 Debrecen, Hungary.
Received January 16, 2002;
accepted after revision April 4, 2002.
Address correspondence to Z. Szucs-Farkas.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. The four rectus muscles in 110 orbits (35 patients and 20 control subjects) were evaluated with MR imaging. The diameter at the largest extent of the muscle belly, as well as the long and the short diameters and the cross-sectional areas in a preselected coronal scan, were measured for each muscle and were compared with the corresponding muscle volume measured on contiguous T1-weighted transverse slices.
RESULTS. The measured coronal area correlated well with the volume of the superior (r = 0.694, p < 0.0001) and inferior (r = 0.783, p < 0.0001) recti, and the largest transverse diameter showed strong correlation with the volume of the lateral (r = 0.868, p < 0.0001) and medial (r = 0.869, p < 0.0001) recti. For the latter muscle, the coronal area also exhibited a good correlation with the volume (r = 0.838, p < 0.0001). Coronal cross-sectional areas can be well estimated by measuring both the short and long coronal muscle diameters (r values were between 0.914 and 0.966; p < 0.0001).
CONCLUSION. In Graves' ophthalmopathy, the volume of three of the rectus muscles can be well estimated by simple measurements on a single coronal slice, and the largest transverse diameter of the lateral rectus is suitable for the same purpose.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The judgment of therapeutic efficacy in Graves' ophthalmopathy rests to a large extent on improvement in the patients' clinical status. Unfortunately, clinical activity scores based on symptoms fail to consistently provide reliable follow-up data for therapy monitoring, and their use is limited [2]. Decreasing edema and volume of the orbital components is a more reliable sign of successful therapy. Although the water content of the muscles can easily be deduced from T2 relaxation times, monitoring muscle volume changes is not an easy task. It requires special computer analysis, including applicable software and hardware, and radiologists' valuable time [3,4,5,6]. Other techniques for muscle volume assessment with MR imaging are fairly cumbersome [7].
Until the development of a simple, accurate method for automated volume measurement, clinicians need a parameter that can be measured quickly and that reflects changes in muscle volume. The usual approach to the enlarged eye muscles on MR imaging or CT is that the examiner evaluates one or two diameters of each muscle, and consecutive measurements of the same diameter(s) are performed during therapy for follow-up [2, 8]. No consensus exists among radiologists concerning in which plane these diameters should be measured. Gorman [2] reported that cross-sectional areas might estimate muscle volume more precisely. However, it has not yet been verified whether muscle diameters or even cross-sectional areas in any plane can actually substitute for the volume.
We analyzed the correlation between diameters, coronal areas, and volumes of the extraocular muscles. We also investigated the possibility of reducing the time of image evaluation by volume estimation of every muscle in the same time on a standard preselected slice.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Contiguous unenhanced T1-weighted transverse, coronal, and sagittal slices were performed with a conventional spin-echo sequence (TR/TE, 500/15; field of view, 200 x 200 mm; matrix, 512 x 512; slice thickness, 3 mm). Transverse and sagittal scans were obtained on a course parallel to the optic nerve. Coronal slices were perpendicular to the midsagittal plane. The subjects were asked to maintain a forward gaze and gentle eye closure to eliminate any measurement error from asymmetric extraocular muscle contraction. All scans and raw data were transferred to a central server computer and saved on compact discs.
Image Analysis
For image analysis, we used a computer program developed especially for
postprocessing MR images. The main advantage of this software is that it runs
on any commercially available PC station under Windows 95 or 98 operating
systems (Microsoft, Redmond, WA).
For volume measurements, on the transverse T1-weighted scans the rectus muscles were out-lined manually with a mouse by the same observer, who was unaware of the endocrine status or any other data of the subjects except their names and ages. The muscle area was computed by the software on the basis of the enclosed pixel count and pixel size. Summing the areas and multiplying by slice thickness gave the single muscle volumes. Because of the close vicinity of the rectus superior and the levator palpebrae superioris muscles, they were often difficult to differentiate in any plane, so they were evaluated together as the superior muscle group.
On the third coronal slice behind the posterior pole of the globe, the cross-sectional area (Ameas), the long diameter (Dlong), and the perpendicular short diameter (Dshort) of the muscles were measured (Fig. 1A). Diameter perpendicular to the long axis of the muscle at the largest extent of the muscle belly (maximal diameter, Dmax) was also measured in the transverse plane for the medial and lateral recti (Fig. 1B) and in the sagittal plane for the inferior rectus muscle and the superior muscle group (Fig. 1C). Each area was measured two times and each diameter, three times, and the averages of these values were used for further calculations.
|
|
|
Calculations and Statistics
To assess whether two diameters estimate volume more precisely than one, we
hypothesized the coronal muscle projections to be regular ellipsoids and
calculated an approximate cross-sectional area for each muscle as follows:
 |
Calculated and measured areas, as well as long, short, and maximal diameters, were compared with each other and with muscle volumes using paired samples t tests. Data acquired from the left and right orbits were compared with independent samples t tests. Pearson's correlation coefficient and two-by-two tables were used for correlation calculations. The level of statistical significance was set at a p value of less than 0.05.
To analyze the intraobserver variation, the same observer measured images of 40 randomly selected orbits a second time after a 6- to 9-month interval. For assessment of interobserver variation, 20 orbits were selected randomly and were reevaluated by a second examiner. The data of the reevaluated measurements were compared with those of the original measurements, and correlations were calculated.
Statistical analyses were done with the SAS for Windows 8.1 software (SAS Institute, Cary, NC).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
The overall intraobserver and interobserver variabilities were very good (r = 0.961 and 0.897, respectively).
Except for the lateral rectus muscle, the muscle volumes were significantly larger in the patient group than in the control group. Cross-sectional areas of the inferior rectus muscle and the superior muscle group, but not the medial and lateral recti, were also larger in the patients' group.
The calculated and measured cross-sectional areas correlated well in every muscle in the Graves' ophthalmopathy group (Tables 2,3,4,5). The long and short diameters in the coronal plane correlated strongly with the coronal areas. For the same parameters, the correlation was somewhat weaker in the control subjects.
|
|
|
|
In the inferior rectus muscle and the superior muscle group, muscle volumes were best estimated by the measured coronal cross-sectional areas, followed by the calculated coronal areas and the short or long diameter. Although maximal diameters were significantly larger in patients than in control subjects, maximal diameters showed the weakest correlations with volumes in these muscles. In contrast, maximal diameters exhibited excellent correlation with muscle volumes in the medial and lateral recti. The only other parameter showing a comparably high correlation with muscle volume in the medial rectus was the coronal area.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
MR imaging, with its own limitations, such as cost and patient claustrophobia, obviates the use of ionizing radiation and can acquire cross-sectional scans in any plane. The ability of MR imaging to detect edema has been used for almost two decades for monitoring the effectiveness of therapy in Graves' ophthalmopathy. Elevated T2 relaxation time in extraocular muscles is considered a sign of disease activity [13, 14]. Even so, that sign does not reliably indicate the grade of muscle dysfunction [9]. The potential of MR imaging for volume measurement has recently been introduced in Graves' ophthalmopathy [3, 4]. Volume estimation from two-dimensional and three-dimensional databases has special software and hardware requirements. However, despite minor limitations, this method seems to be a hopeful method for the future.
Firbank and Coulthard [3] developed a semiautomated technique for volume estimation on coronal MR images, but the method could be used in only a quarter of the 19 patients and seven control subjects studied; the rest had to be evaluated manually. Muscle volumes obtained by this group differ substantially from our results. Their data showed the medial rectus muscle in patients and the lateral rectus in healthy subjects to have the largest volume. We found the inferior rectus muscle to be the largest in both the Graves' ophthalmopathy group and the control group, followed by the superior muscle group and the medial and lateral rectus muscles. One possible cause for the discrepancies might be that the measurement technique of Firbank and Coulthard exhibits the largest inaccuracy (>10%) just for the medial and lateral recti. Another explanation may be the different planes used for volume measurement in the two studies, which means different regions of the orbit have limited visualization. Also, ethnic differences between the two study groups may have accounted for the discrepancies. However, we think that in coronal slices the muscles are relatively difficult to delineate in the region of the orbital apex and close to the globe. Measurement in the transverse plane avoids this drawback, so that method seems to be more suitable for volume estimation.
If we use muscle volume as an outcome indicator of therapy, image postprocessing can be fairly time-consuming for each follow-up examination. With repeated assessment of an easy-to-measure parameter that correlates well with muscle volume changes, this time could be significantly reduced. The coronal is the only plane in which all rectus muscles can be visualized on the same slice. Extraocular muscles are cylindric objects, with a narrowing at their origin at Zinn's ligament in the orbital apex; similarly, their front ends become thinner as they approach their insertion at the equator of the globe [15, 16]. Therefore, these muscles' coronal cross sections, which are nearly ellipsoids, could be ideal candidates for evaluation in the mid orbital region. For this reason we chose consistently the same slice approximately 1 cm posterior to the globe in every study subject. The oblique course of the muscles in the orbit explains why their cross sections in the coronal plane are inevitably distorted, which means that the measured cross-sectional areas and diameters do not equal those that would be measured in a slice perpendicular to the long axis of the muscle. Using coronal slices, both transverse and sagittal diameters of the superior and inferior recti, the transverse diameter of the lateral rectus, and the cross-sectional areas of all these muscles differ from their true values. The medial rectus muscle, coursing almost parallel with the medial orbital wall, is theoretically not affected in this respect. The lateral rectus is the muscle whose cross section is distorted to the greatest extent, which may explain the limited value of its cross section for volume estimation. Of course, these considerations are true only if the imaged coronal plane is perpendicular to the midsagittal plane, as in our case. Some examiners use surface coils and image the orbits separately, applying two distinct coronal planes perpendicular to each optic nerve. In that case, more or less all rectus muscles become implicitly distorted.
Single muscles can be best visualized in the plane parallel to their course—that is, the superior and inferior recti in the sagittal plane and the medial and the lateral recti in the transverse plane. Ozgen and Ariyurek [8] suggested measuring the greatest diameter of the muscles in the corresponding plane and using it as an indicator for volume changes. These "real" diameters are not subject to anatomic distortion. Unfortunately, their CT study about establishing a normal range for extraocular muscle thickness did not include any diameter measurements in the sagittal plane.
We found that, although the preselected coronal scan used in our study may or may not intersect the muscles at their greatest extension, long and short diameters and coronal areas correlate better with muscle volume than maximal diameters do, at least in the inferior recti and superior muscle group in the patients with Graves' ophthalmopathy. The volume of the medial rectus muscle can also be well estimated using the cross-sectional area measured on the same slice.
For the lateral rectus muscle, we think it best to estimate muscle volume changes on the basis of the greatest diameter measured in the transverse plane. The volume of this muscle shows only fair correlation with either coronal area or any other diameter.
The excellent correlation between the measured and the calculated areas for every muscle in our patients indicates that the cross section of the enlarged muscle is almost an ellipsoid, the area of which can be well estimated by measuring its long and short diameters. In contrast, normal-sized muscles of the control individuals, especially the medial rectus muscle, often exhibit an almost crescent-shaped or irregular coronal projection that can be responsible for the weaker correlations. In general, one coronal diameter signals volume changes less reliably than both diameters, which is also a consequence of the ellipsoid coronal projection of enlarged muscles.
Our results show that in patients with Graves' ophthalmopathy, the volume of the rectus muscles, except the lateral rectus, can be accurately estimated by simple measurement of their cross-sectional area in a properly chosen coronal MR slice. The largest transverse diameter in the lateral rectus can be used for the same purpose. If one wishes to save time during image analysis, acquisition of the long and short diameters in the same coronal slice can be an alternative for the coronal area. Deduction of muscle volume from a single coronal diameter is generally not advisable.
It is not clear whether changes occurring in muscle shape and length after therapy have a significant influence on the use of the previously referred to correlations. To answer this question, further prospective follow-up studies among patients with Graves' ophthalmopathy are required. Long-term clinical studies are necessary to ascertain correlation between volume and function of extraocular muscles. However, MR imaging is a uniquely useful diagnostic tool in Graves' ophthalmopathy because it can concurrently provide a rapid assessment of muscle size and edema.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
] / 4).
