HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Ohashi, K. Articles by Berbaum, K. S. Search for Related Content PubMed PubMed Citation Articles by Ohashi, K. Articles by Berbaum, K. S.

Peroneal Tendon Subluxation and Dislocation: Detection on Volume-rendered Images—Initial Experience¹

¹ From the Department of Radiology, University of Iowa Roy J. and Lucille A. Carver College of Medicine, 200 Hawkins Dr, Iowa City, IA 52242. From the 2004 RSNA Annual Meeting. Received June 14, 2005; revision requested August 12; revision received October 26; accepted November 14; final version accepted March 20, 2006. Address correspondence to K.O. (e-mail: kenjirou-ohashi{at}uiowa.edu).

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCE IN KNOWLEDGE References

 
Institutional review board approval was received and informed consent was not required for this Health Insurance Portability and Accountability Act–compliant study. The purpose of this study was to retrospectively assess the time efficiency of three-dimensional volume-rendered images obtained from multi–detector row computed tomographic data for the diagnosis of peroneal tendon subluxation or dislocation by using the consensus interpretation of multiplanar reformatted (MPR) images as the reference standard. The reference standard was provided by two musculoskeletal radiologists, and two less experienced readers evaluated 37 images in 32 patients (24 men, eight women; mean age, 41 years; age range, 18–75 years) with acute calcaneal fractures. An analysis of variance was used to compare interpretation time, and the Wilcoxon signed rank test was used to analyze diagnostic difficulty. The average time required for diagnosis was significantly shorter with volume-rendered images than with MPR images (reader 1: 42 vs 78 seconds, P < .001; reader 2: 50 vs 69 seconds, P < .01).

© RSNA, 2007

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCE IN KNOWLEDGE References

 
Although magnetic resonance imaging and ultrasonography are the mainstays in evaluating the tendons, computed tomography (CT) is also considered reliable for the diagnosis of peroneal tendon subluxation and dislocation (1–3). Peroneal tendon subluxation and dislocation are commonly associated with calcaneal fractures (2,4), and CT is often used to assess these fractures and to plan treatment (5–7). When calcaneal fractures are associated with peroneal tendon subluxation or dislocation, this may influence treatment and prognosis (3,8,9). Therefore, detecting peroneal tendon subluxation or dislocation at CT in patients with calcaneal fractures is important.

As we reviewed multi–detector row CT scans of the ankles and feet, we noticed that peroneal tendon subluxation or dislocation seemed to be more readily detectable on three-dimensional (3D) volume-rendered images than on multiplanar reformatted (MPR) images. Typically, a radiologist generates 3D volume-rendered images and MPR images in a matter of seconds on a workstation and interactively views the images. When a radiologist evaluates calcaneal fractures with a large volume of data sets after multi–detector row CT, it is practical to review the entire study for fractures by using a bone window and then to alter the window width and window level settings to look for other abnormalities. In this setting, evaluation of peroneal tendons for dislocation is an important clinical task for which both MPR and 3D volume-rendered images are available on a 2 x 2 display. Three-dimensional volume-rendered images provide a thorough visualization of the tendons and demonstrate the relationship between tendons and bone from any arbitrary projection.

Thus, the purpose of this study was to retrospectively assess the time efficiency of 3D volume-rendered images obtained from multi–detector row CT data for the diagnosis of peroneal tendon subluxation or dislocation by using the consensus interpretation of MPR images as the reference standard.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCE IN KNOWLEDGE References

 
Patients
Our institutional review board approved this retrospective study, and patient informed consent was not required. The study was performed in accordance with the Health Insurance Portability and Accountability Act. A total of 37 multi–detector row CT studies were collected from 32 consecutive patients (24 men, eight women; age range, 18–75 years; mean age, 41 years) over a 13-month period (January 2003 to January 2004). Four additional multi–detector row CT studies were excluded because of missing CT data. Five patients underwent bilateral multi–detector row CT (24 right ankles and feet and 13 left ankles and feet). Patients were referred for evaluation of calcaneal fractures due to falls (20 patients), motor vehicle accidents (10 patients), and unknown causes (two patients).

At our institution, calcaneal fractures are routinely evaluated with CT for preoperative planning and for the detection of postoperative complications. Five patients underwent surgery before multi–detector row CT, with a mean interval of 3.4 days (range, 0–14 days) between surgery and multi–detector row CT. In the remaining 27 patients, the mean interval between multi–detector row CT and injury was 3.5 days (range, 0–21 days). A total of 23 (62%) of 37 ankles and feet had been immobilized in a cast, and three ankles and feet (8%) had metallic orthopedic hardware at the time of multi–detector row CT.

Reference Standard
The consensus reading of two musculoskeletal radiologists (K.O. and G.Y.E., with 6 and 8 years of experience, respectively, in musculoskeletal multi–detector row CT) was used as the reference standard. Diagnosis was based on MPR images that were reviewed on a computer workstation by using the criteria noted below for the two readers. Peroneal tendon subluxation or dislocation was seen on 16 (43%) of 37 studies. We did not distinguish between subluxation and dislocation. The term dislocation also includes tendons that are subluxed.

Multi–Detector Row CT
CT was performed by using either a four–detector row (Aquilion; Toshiba Medical Systems, Tustin, Calif) or a six–detector row (Emotion 6; Siemens, Malvern, Pa) helical CT scanner. For the four–detector row CT scanner, imaging parameters included 120–135 kVp, 0.5-second scanning time per gantry rotation, 75–225 mAs, 240-mm field of view, 120–180-mm reconstructed field of view, 512 x 512 matrix, 1–2-mm collimation, 3.5–7.0-mm table travel per rotation, 1–2-mm reconstruction thickness, 50%–75% overlap, and a standard soft-tissue kernel. For the six–detector row CT scanner, imaging parameters included 130 kVp, 0.5-second scanning time per gantry rotation, 75–150 mAs, 500-mm field of view, 120–180-mm reconstructed field of view, 512 x 512 matrix, 0.5-mm collimation, 3.0-mm table travel per rotation, 0.63-mm reconstruction thickness, 50% overlap, and a medium smooth (soft-tissue) kernel.

In patients with orthopedic hardware, an extended CT scale was applied for imaging with the six–detector row CT scanner. The reconstructed images were sent to a Vitrea 2 workstation (version 3.5; Vital Images, Plymouth, Minn) over an intradepartmental picture archiving and communication system (Eastman Kodak, Rochester, NY) by using the Digital Imaging and Communications in Medicine protocol.

Image Review
MPR and 3D volume-rendered images were reviewed interactively on a computer workstation. MPR images were reviewed by using a window width of 340 and a window level of 40 as the default. Three-dimensional volume-rendered images were reviewed with a soft-tissue window width of 150 and a soft-tissue window level of 85 as the default. The readers could adjust the window width and window level settings. Objects that obscured important anatomic structures, such as cast materials, were manually deleted by the readers by using the software function on the workstation.

During the evaluation of MPR images, the reported CT criteria for normal peroneal tendons were used (3). On transverse images, normal peroneal tendons were defined as being posterior to the posterolateral margin of the distal fibular cortex and medial to the superior retinaculum (Fig 1). On coronal images, the peroneal tendons were defined as being posterior to the fibular groove. Displacement of one of the peroneal tendons from its normal anatomic location was considered dislocation. No measurement criteria were used because the study sample was obtained in patients with acute calcaneal fractures for which normal bony landmarks might be violated.

View larger version (90K): [in this window] [in a new window] [Download PPT slide]   Figure 1: Transverse MPR image obtained perpendicular to the long axis of the tibia demonstrates normal peroneal tendons (long arrow) that are located posterior to the posterolateral margin of the distal fibular cortex and medial to the superior retinaculum (short arrow).

View larger version (149K): [in this window] [in a new window] [Download PPT slide]   Figure 2: Three-dimesional volume-rendered image (viewed from lateral aspect) demonstrates normal peroneal tendons (long arrow) that are located posterior to the posterolateral margin of the distal fibular cortex (short arrow). This relationship was evaluated as the reader viewed the ankle from the lateral, posterolateral, and posterior aspects by rotating the 3D volume-rendered image.

The images that were used for training were not from the patients in our study and were used with institutional review board approval, waiver of informed consent, and Health Insurance Portability and Accountability Act compliance. Half of the studies were reviewed with the MPR images presented first, followed by the volume-rendered images. The other half were reviewed with the volume-rendered images presented first, followed by the MPR images. During both review sessions, images were presented in random order. To reduce recall bias, MPR and volume-rendered images from each patient were read separately at least 4 weeks apart.

The readers were blinded to the patient's name, sex, and age, as well as to other clinical information about the patient. The readers were also blinded as to the purpose of the study. The readers were asked to give their diagnosis by using five categories, three of which indicated dislocation (definite, probable, or possible) and two of which indicated no dislocation (probably not or definitely not). Agreement between the interpretation of the volume-rendered images and the reference standard images was measured by counting the number of agreements (dislocation vs no dislocation) for each reader. Five categories were used to rate the difficulty in making a diagnosis (5, very easy; 4, easy; 3, intermediate; 2, difficult; or 1, very difficult). Reading time was recorded, as well as manipulation of the default window width and window level settings because such manipulations might increase reading time.

Statistical Analysis
For each reader, an analysis of variance of the total interpretation time of each study was performed by using BMDP2V (release 8.0; Statistical Solutions, Cork, Ireland). In these analyses, the total interpretation time was compared between studies that were negative for peroneal dislocation and those that were positive for peroneal dislocation and between display modalities (volume rendered vs MPR). Thus, the outcome was treated as a between-subject factor, and the modality was treated as a within-subject factor (a repeated measure). Subjective ratings of the discrete categorical ratings of diagnostic difficulty by using the MPR and volume-rendered images were compared by using a nonparametric Wilcoxon signed rank test (BMDP3D, release 8.0; Statistical Solutions). For each reader, separate tests were performed for negative studies, positive studies, and all studies combined.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCE IN KNOWLEDGE References

 
Time Required for Diagnosis
For both readers, the evaluation of studies with volume-rendered images required less time than the evaluation of the same studies with MPR images (reader 1: 42 vs 78 seconds, F_(1,35) = 25.05, P < .001; reader 2: 50 vs 69 seconds, F_(1,35) = 7.44, P < .01). There was no significant difference in the amount of time needed to evaluate normal images and the amount of time needed to evaluation images that showed dislocation (reader 1: 59 vs 60 seconds, F_(1,35) = 0.02, P = .897; reader 2: 55 vs 66 seconds, F_(1,35) = 2.62, P = .114). No statistically significant interaction with respect to evaluation time was found between study type (no dislocation vs dislocation) or display type (MPR images vs volume-rendered images). During the interpretation of volume-rendered images, reader 1 agreed with the reference standard in 37 of 37 studies, and reader 2 agreed with the reference standard in 36 of 37 studies.

Diagnostic Difficulty
Reader 1.—During the reading of the 21 normal studies, eight were classified as equally easy to read with MPR and volume-rendered images, 10 were classified as easier to read with volume-rendered images, and three were classified as easier to read with MPR images (P = .052). During the reading of the 16 abnormal studies, nine were classified as equally easy to read with MPR and volume-rendered images, two were classified as easier to read with volume-rendered images, and five were classified as easier to read with MPR images (P = .375). After taking the results of the normal and abnormal studies together by using a Wilcoxon signed rank test, there was no significant difference in diagnostic difficulty between MPR images and volume-rendered images for reader 1 (eight studies were easier to read with MPR images, and 12 were easier to read with volume-rendered images; P = .419). Reader 1 adjusted the default window width and window level settings for MPR images in one (3%) of 37 studies; window width and window level settings for volume-rendered were adjusted in 26 (70%) of 37 studies (Fig 3).

View larger version (99K): [in this window] [in a new window] [Download PPT slide]   Figure 3a: (a) MPR image and (b–d) 3D volume-rendered images of an ankle and foot that have been immobilized in a cast. In a, an increase in attenuation is diffusely seen in the subcutaneous fat tissues around the peroneal tendons (long arrow), with the cast materials around the ankle. The cast materials seen in b have been removed by using a software function, and the image was rotated in c to view the ankle from the lateral aspect. In d, the window width and window level settings were adjusted to better visualize the peroneal tendons (arrow).

View larger version (137K): [in this window] [in a new window] [Download PPT slide]   Figure 3b: (a) MPR image and (b–d) 3D volume-rendered images of an ankle and foot that have been immobilized in a cast. In a, an increase in attenuation is diffusely seen in the subcutaneous fat tissues around the peroneal tendons (long arrow), with the cast materials around the ankle. The cast materials seen in b have been removed by using a software function, and the image was rotated in c to view the ankle from the lateral aspect. In d, the window width and window level settings were adjusted to better visualize the peroneal tendons (arrow).

View larger version (126K): [in this window] [in a new window] [Download PPT slide]   Figure 3c: (a) MPR image and (b–d) 3D volume-rendered images of an ankle and foot that have been immobilized in a cast. In a, an increase in attenuation is diffusely seen in the subcutaneous fat tissues around the peroneal tendons (long arrow), with the cast materials around the ankle. The cast materials seen in b have been removed by using a software function, and the image was rotated in c to view the ankle from the lateral aspect. In d, the window width and window level settings were adjusted to better visualize the peroneal tendons (arrow).

View larger version (110K): [in this window] [in a new window] [Download PPT slide]   Figure 3d: (a) MPR image and (b–d) 3D volume-rendered images of an ankle and foot that have been immobilized in a cast. In a, an increase in attenuation is diffusely seen in the subcutaneous fat tissues around the peroneal tendons (long arrow), with the cast materials around the ankle. The cast materials seen in b have been removed by using a software function, and the image was rotated in c to view the ankle from the lateral aspect. In d, the window width and window level settings were adjusted to better visualize the peroneal tendons (arrow).

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCE IN KNOWLEDGE References

 
With the increasing amount of data obtained in CT studies of the foot and ankle (which frequently contain more than 400 reconstructed transverse sections), radiologists are obligated to use computer workstations for reviewing images. Advances in multi–detector row CT technology and high-speed computing have made it possible to rapidly obtain and review MPR images in any arbitrary plane and high-quality 3D images from any direction. One of the advantages of 3D images is that there is a reduction in the amount of time needed to identify abnormalities by providing an effective search pattern (10). This is important for radiologists who have to review an ever increasing number of images (11). Recently, Pelc and Beaulieu (12) applied postprocessing to CT data sets to quantify differences in attenuation values between bone and tendon and to display 3D volume-rendered images of normal tendons. Ohashi et al (13) reported on the clinical applications of this technique to characterize tendon abnormalities, including avulsions and dislocations of the superficial tendons.

To test the advantages of 3D volume-rendered images for the detection of peroneal tendon dislocation, we used a group of consecutive patients with acute calcaneal fractures. The incidence of peroneal tendon subluxation or dislocation in patients with calcaneal fractures is reported to be between 25% and 47.5% (2,4), which is similar to the incidence of peroneal tendon subluxation or dislocation in our study population (16 [43%] of 37 patients). The acute trauma setting creates difficult clinical and imaging conditions for the detection of tendon dislocation, in which diffuse soft-tissue edema, distortion of bony anatomy, and cast materials may be present. Our results suggest that volume-rendered images may have the advantage in decreasing the amount of time needed to detect peroneal tendon dislocation in nontrauma settings as well.

We used readers who had no experience in the interpretation of volume-rendered images to minimize variations between readers. We became aware of a possible reduction in viewing time with volume-rendered images after we had routinely used this display format in our clinical practice for the past 3 years, and we believed that results from inexperienced readers could be extrapolated to more experienced readers.

Eliminating the amount of time required to scroll through the two-dimensional MPR sections to inspect the relationship between the peroneal tendons and the fibula may explain the reduced interpretation time with 3D volume-rendered images. However, the subjective ratings for the difficulty of interpretation did not suggest that volume-rendered images were consistently easier to interpret than MPR images. The reason is probably because inspecting the relationship between the peroneal tendons and the lateral malleolus on MPR images is relatively easy, although reviewing MPR images is more time-consuming. Additionally, this may be related to the fact that the readers adjusted the default window width and window level settings less frequently with MPR images (one of 37 studies for readers 1 and 2) than with volume-rendered images (26 of 37 studies for reader 1 and 33 of 37 studies for reader 2).

There would be little point in using volume-rendered images if a less accurate diagnosis was the result, even if such images were faster to interpret. In human performance research, this would be referred to as a speed-accuracy trade off (14). While the focus of this investigation was not on reader detection accuracy, the sacrifice of accurate reading to improve speed was unlikely for our two readers with volume-rendered images because, except for one reader in one study, both readers classified the studies the same way as the reference standard.

Our study was limited by the relatively small sample size (37 studies and two readers). The reference standard was based on MPR images that were interpreted by consensus between two experienced musculoskeletal radiologists. We did not compare multi–detector row CT findings with physical examination findings or other imaging techniques. It is possible that transient peroneal tendon dislocation with spontaneous reduction can be missed at multi–detector row CT (15,16). This study, however, was designed to compare the time efficiency of the two methods of viewing by using the same multi–detector row CT data sets.

In summary, our initial experience provides evidence that the interpretation of 3D volume-rendered images requires less time than the interpretation of MPR images for the detection of peroneal tendon dislocation in the acute trauma setting. Time reduction in the search for certain abnormalities can be vital in managing the increasing volume of multi–detector row CT data sets.

	ADVANCE IN KNOWLEDGE

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCE IN KNOWLEDGE References

 

• The time required for the diagnosis of peroneal tendon subluxation or dislocation was significantly shorter (P < .001 for reader 1 and P < .01 for reader 2) with volume-rendered images than with multiplanar reformatted images.
  

	ACKNOWLEDGMENTS

 
The authors thank David Van Roekel, MD, and Eric Callaghan, MD, for reviewing multi–detector row CT studies of the ankles and feet and Robert T. Caldwell, MFA, for editing the manuscript.

	FOOTNOTES

 

Abbreviations: MPR = multiplanar reformatted • 3D = three-dimensional

Authors stated no financial relationship to disclose.

Author contributions: Guarantors of integrity of entire study, K.O., G.Y.E., K.S.B.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, K.O., J.M.R.; clinical studies, K.O., G.Y.E.; statistical analysis, K.O., K.S.B.; and manuscript editing, K.O., G.Y.E., K.S.B.

	References

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ADVANCE IN KNOWLEDGE References

 

1. Rosenberg ZS, Feldman F, Singson RD. Peroneal tendon injuries: CT analysis. Radiology 1986;161:743–748.[Abstract/Free Full Text]
2. Rosenberg ZS, Feldman F, Singson RD, Price GJ. Peroneal tendon injury associated with calcaneal fractures: CT findings. AJR Am J Roentgenol 1987;149:125–129.[Abstract/Free Full Text]
3. Ho RT, Smith D, Escobedo E. Peroneal tendon dislocation: CT diagnosis and clinical importance. AJR Am J Roentgenol 2001;177:1193.
4. Bradley SA, Davies AM. Computed tomographic assessment of soft tissue abnormalities following calcaneal fractures. Br J Radiol 1992;65:105–111.[Abstract]
5. Sanders R. Intra-articular fractures of the calcaneus: present state of the art. J Orthop Trauma 1992;6:252–265.[Medline]
6. Wechsler RJ, Schweitzer ME, Karasick D, Deely DM, Morrison W. Helical CT of calcaneal fractures: technique and imaging features. Skeletal Radiol 1998;27:1–6.[CrossRef][Medline]
7. Richards PJ, Bridgman S. Review of the radiology in randomized controlled trials in open reduction and internal fixation (ORIF) of displaced intraarticular calcaneal fractures. Injury 2001;32:633–636.[CrossRef][Medline]
8. Sanders R. Displaced intra-articular fractures of the calcaneus. J Bone Joint Surg Am 2000;82:225–250.[Free Full Text]
9. Rammelt S, Zwipp H. Calcaneus fractures: facts, controversies and recent developments. Injury 2004;35:443–461.[CrossRef][Medline]
10. Pickhardt PJ. Translucency rendering in 3D endoluminal CT colonography: a useful tool for increasing polyp specificity and decreasing interpretation time. AJR Am J Roentgenol 2004;183:429–436.[Free Full Text]
11. Rubin GD, Napel S, Leung AN. Volumetric analysis of volumetric data: achieving a paradigm shift. Radiology 1996;200:312–317.[Free Full Text]
12. Pelc JS, Beaulieu CF. Volume rendering of tendon-bone relationships using unenhanced CT. AJR Am J Roentgenol 2001;176:973–977.[Abstract/Free Full Text]
13. Ohashi K, El-Khoury GY, Bennett DL. MDCT of tendon abnormalities using volume-rendered images. AJR Am J Roentgenol 2004;182:161–165.[Abstract/Free Full Text]
14. Luce RD. Response times: their role in inferring elementary mental organization. New York, NY: Oxford Science, 1986; 56–57, 81–90.
15. Shellock FG, Feske W, Frey C, Terk M. Peroneal tendons: use of kinematic MR imaging of the ankle to determine subluxation. J Magn Reson Imaging 1997;7:451–454.[Medline]
16. Neustadter J, Raikin SM, Nazarian LN. Dynamic sonographic evaluation of peroneal tendon subluxation. AJR Am J Roentgenol 2004;183:985–988.[Abstract/Free Full Text]

This Article Abstract Figures Only Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Ohashi, K. Articles by Berbaum, K. S. Search for Related Content PubMed PubMed Citation Articles by Ohashi, K. Articles by Berbaum, K. S.