Sonographic Detection of Baker s Cysts: Comparison with MR Imaging

Transcription

1 Eric E. Ward 1 Jon. Jacobson 2 David P. Fessell 2 Curtis W. Hayes 2 Marnix van Holsbeeck 1 Received June 20, 2000; accepted after revision ugust 14, Presented at the annual meeting of the merican Roentgen Ray Society, San Francisco, pril May Department of Radiology, Henry Ford Hospital, 2799 W. Grand lvd., Detroit, MI Department of Radiology, University of Michigan Medical Center, 1500 E. Medical Center Dr., TC-2910G, nn rbor, MI ddress correspondence to J.. Jacobson. JR 2001;176: X/01/ merican Roentgen Ray Society Sonographic Detection of aker s Cysts: Comparison with MR Imaging OJECTIVE. The objective of this study was to assess the ability of sonography to reveal aker s cysts using MR imaging as a gold standard. MTERILS ND METHODS. The study group consisted of 36 consecutive knees in 36 patients evaluated with both MR imaging and sonography. Inclusion criteria included axial proton density weighted or T2-weighted MR images, a sonography report that documented the evaluation of the popliteal region of the knee, and sonographic and MR images that were available for review. The MR images were retrospectively reviewed for the presence of aker s cyst (fluid signal between the semimembranosus and medial gastrocnemius tendons). Sonography reports were compared with the MR imaging results. The sonographic images were also retrospectively reviewed to determine whether any characteristic findings on sonography were significantly associated with the presence of aker s cyst on MR imaging. RESULTS. Retrospective review of MR images revealed 21 aker s cysts, one myxoid liposarcoma, one meniscal cyst, and 13 examinations with normal findings. The sonography reports revealed that the 21 aker s cysts were correctly diagnosed, whereas the meniscal cyst and myxoid liposarcoma were misdiagnosed as aker s cysts. Retrospective review of sonographic images showed a 100% sensitivity, specificity, positive predictive value, negative predictive value, and accuracy in the diagnosis of aker s cyst when hypoechoic or anechoic fluid was present between the semimembranosus and medial gastrocnemius tendons. No other sonographic characteristics were significant. CONCLUSION. Identification of fluid between the semimembranosus and medial gastrocnemius tendons in communication with a posterior knee cyst indicates aker s cyst with 100% accuracy. P opliteal or aker s cyst represents fluid distention of a bursa between the gastrocnemius and semimembranosus tendons via a communication with the knee joint [1]. lthough called the gastrocnemiosemimembranosus bursa, it represents a composite of two bursae: a bursa anterior to the medial gastrocnemius tendon (the subgastrocnemius bursa) and a bursa between the tendons of the gastrocnemius and semimembranosus tendons [2]. Frequency of aker s cysts identified on MR imaging ranges from 10% to 41% [3]. aker s cysts may be seen with many joint abnormalities, such as internal derangement, osteoarthrosis, or inflammatory arthritis; the most common associations include joint effusion, meniscal tear, and degenerative joint disease [1]. The diagnosis of aker s cyst is effectively made with MR imaging because fluid distention of the gastrocnemiosemimembranosus bursa is well depicted on T2-weighted axial MR images [1]. The appearance of aker s cysts on sonography has been described in the literature [4 8]. Fluid distention of the gastrocnemiosemimembranosus bursa is characteristic on sonography, similar to that on MR imaging [8]. However, studies determining the effectiveness of sonography in the diagnosis of aker s cysts are limited by small sample sizes, lack of gold standards, and undefined sonographic criteria. In this study, we evaluate the ability of sonography to reveal aker s cyst using MR imaging as a gold standard. In addition, we determine the sonographic characteristics critical to the diagnosis of aker s cyst enabling differentiation from other posterior soft-tissue masses or cysts. Materials and Methods Seven hundred eighty-nine consecutive sonographic examinations of the knee performed between September 1994 and November 1998 were identified from a computer search of radiology reports. Of these 789 cases, 83 patients were initially included as also having undergone MR imaging of JR:176, February

2 Ward et al. the ipsilateral knee. dditional inclusion criteria included axial T2- or proton density weighted MR images, sonographic evaluation of the popliteal region of the knee documented in the sonography report, and sonographic and MR images available for review. This resulted in 36 knee examinations of 36 patients. Retrospective review of the axial T2- or proton density weighted MR images obtained on a 1.5-T scanner (Signa; General Electric Medical Systems, Milwaukee, WI) was then performed by one of the investigators who was unaware of the sonographic imaging and report results. MR images were evaluated for the presence of a popliteal or aker s cyst. The criterion for a popliteal cyst was the presence of high-signal-intensity fluid between the medial gastrocnemius and semimembranosus tendons. ny other posterior soft-tissue mass or cyst not representing a aker s cyst was noted, and patient records were reviewed for any surgical or pathologic follow-up in these cases. The MR imaging results were used as the gold standard for the sonographic data. ecause the MR images were analyzed retrospectively, the parameters varied. However, the standard sequences used in this investigation included axial proton density or T2- weighted fast spin-echo MR images with fat saturation (TR range/te range, /56 102; 5- mm slice thickness; 0- to 2-mm slice gap; matrix; 1 2 excitations; cm field of view). Sonography Report nalysis The first part of the study compared the sonography reports with the retrospective MR imaging results to determine the effectiveness of sonography for diagnosis of aker s cyst. Each sonographic examination of the knee (models HDI 3000 and 5000, dvanced Technology Laboratories, othell, W; or model 5200, coustic Imaging, Phoenix, Z) was completed by one of five musculoskeletal radiologists with experience in musculoskeletal sonography using 5-MHz curvilinear and 7- to 10-MHz linear transducers. It is not known what information or correlative imaging studies were available at the time of the sonographic examinations. The sonography report stated either that a aker s cyst was present or absent or, alternatively, that the sonographic evaluation of the posterior knee showed unremarkable or normal findings. The sonography reports were then classified on the basis of this information as positive or negative for aker s cysts. These data were compared with the retrospective MR imaging data for calculation of sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. Retrospective Sonographic Image nalysis Fig year-old asymptomatic man. xial sonogram shows semimembranosus tendon (solid arrow) and medial gastrocnemius tendon (open arrow) with no intervening fluid. Semimembranosus tendon appears artifactually hypoechoic because of anisotropy (normal hyperechoic tendon appears hypoechoic if tendon is not perpendicular to the ultrasound beam). Note that top of image is posterior; left side of image is medial. M = medial gastrocnemius muscle, F = femur. For the second part of the study, the aker s cysts and other posterior soft-tissue masses or cysts were retrospectively characterized on sonography to determine whether any differentiating features existed. Establishing a specific diagnosis was not included in this retrospective portion of the study because this task would be limited without the benefit of real-time imaging and by relying on static images from the sonographic study. Inclusion criteria for this aspect of the study included a aker s cyst or other posterior softtissue mass or cyst present on MR imaging and axial sonographic images of the medial popliteal region. If the medial gastrocnemius and semimembranosus tendons were not visible on the axial sonographic images (Fig. 1), the case was deemed inadequate and discarded for the purpose of this retrospective part of the study. This resulted in 19 cases. retrospective review of the sonographic images was then completed by a fellowship-trained musculoskeletal radiologist with experience in musculoskeletal sonography. This individual was aware of the project design but was unaware of the MR imaging findings and the prospective sonography report results. ll available sonographic images were assessed for anechoic or hypoechoic fluid extending posteriorly between the semimembranosus and medial gastrocnemius tendons (Figs. 2 and 3) and for the presence of any soft-tissue mass or cyst of the posterior aspect of the knee. If an abnormal soft-tissue mass or cyst in the popliteal region was identified on the sonographic images, size (in centimeters), echogenicity (anechoic, hypoechoic, hyperechoic, or mixed echogenicity relative to muscle echogenicity), and internal septations (Fig. 3) were recorded. Evidence of rupture of a cystic abnormality (anechoic or hypoechoic area surrounding the mass or cyst) (Fig. 4) and intracystic intraarticular body (focal intrinsic echogenic object) (Fig. 5) were also recorded. The volume of the cyst was estimated by multiplying the three dimensions (anteroposterior, transverse, sagittal) if visible on the retrospective images. It was then determined whether there was a statistically significant association between any of the retrospective sonographic data and the MR imaging results using the Fisher s exact test. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were also calculated for each of the sonographic characteristics. Sonographic Technique lthough the technique for aker s cyst evaluation on sonography varied among each of the five prospective examiners, a general description of sonographic technique can be made. Examination began at the level of the mid calf in the transverse plane for visualization of the medial gastrocnemius, lateral gastrocnemius, and soleus muscles in the midline. fter the most medial aspect of the medial gastrocnemius muscle was identified, imaging in the transverse plane was continued superiorly to the level of the knee joint. Visualization of the semimembranosus tendon medial to the gastrocnemius tendon indicates the location where aker s cysts are found (Figs. 1 3). ecause the medial gastrocnemius and semimembranosus tendons are not completely parallel to each other, one of the normally hyperechoic tendons may appear artifactually hypoechoic because of anisotropy (Fig. 1). This occurs when a tendon is not imaged perpendicular to the ultrasound beam [9]. It is important not to mistake this artifactual hypoechogenicity for hypoechoic fluid. ngling the transducer in the cephalad and caudal directions will eliminate anisotropy and allow the correct identification of the tendon. Imaging of a aker s cyst in the sagittal plane is then completed to determine its cephalocaudal extent and to evaluate its inferior margin for rupture (Fig. 4). Results The 36 patients in the study group had a mean age of 46 years (age range, 7 84 years); 53% (n = 19) were male, and 47% (n = 17) were female. Of the patients with aker s cysts, the average age was 48 years (range, 7 84 years), and of those without aker s cysts, it was 47 years (range, years), yielding no significant difference. The mean time interval between the sonography and MR imaging 374 JR:176, February 2001

3 Sonography and MR Imaging of aker s Cysts Fig year-old woman with aker s cyst., xial sonogram of posterior knee shows aker s cyst (arrowheads) with fluid (solid straight arrow) between semimembranosus tendon (curved arrow) and medial gastrocnemius tendon (open arrow). Note subgastrocnemius component (asterisk) of aker s cyst. Note that top of image is posterior; right side of image is medial. M = medial gastrocnemius muscle., xial proton density weighted MR image with fat saturation reveals aker s cyst (arrowheads) with fluid (black arrow) between semimembranosus tendon (curved white arrow) and medial gastrocnemius tendon (open arrow). Note subgastrocnemius component (asterisk) of aker s cyst. M = medial gastrocnemius muscle. Fig year-old boy with aker s cyst., Sagittal sonogram of posterior knee shows aker s cyst (arrowheads). Note septation (solid arrow). Note that top of image is posterior; left side of image is superior. Open arrows = medial gastrocnemius tendon, T = tibia., Sagittal proton density weighted MR image reveals aker s cyst (arrowheads). Note septation (solid arrow). Open arrows = medial gastrocnemius tendon. JR:176, February

4 Ward et al. Fig year-old man with rupture of aker s cyst., Sagittal sonogram of posterior knee shows irregular hypoechogenicity (arrows) surrounding the inferior border of aker s cyst (arrowheads). Note that top of image is posterior; left side of image is superior. M = medial gastrocnemius muscle., Coronal proton density weighted MR image with fat saturation reveals high-signal-intensity fluid (arrows) extending inferiorly from aker s cyst (arrowheads). Fig year-old woman with intraarticular body in aker s cyst., xial sonogram shows echogenic intraarticular body (arrow) in aker s cyst (arrowheads). Note that top of image is posterior., Sagittal proton density weighted MR image reveals intermediate signal intraarticular body (arrow) in aker s cyst (arrowheads). examinations was 78 days. Of the patients with aker s cysts, the average time interval was 84 days (range, days), and of those without aker s cysts, it was 71 days (range, days), yielding no significant difference between the two groups. Sonographic examination occurred before the MR imaging examination in 31 patients, after the MR imaging examination in four, and on the same day as the MR imaging examination in one. Given the retrospective nature of this study, it is not known whether symptoms were referable to the posteromedial aspect of the knee. Retrospective analysis of the 36 MR imaging examinations of the knee revealed 21 pa- 376 JR:176, February 2001

5 Sonography and MR Imaging of aker s Cysts TLE 2 Sonographic Characteristics of Posterior Knee Soft-Tissue Masses or Cysts TLE 1 Sonography Report Sonography Reports Versus Retrospective MR Imaging Results MR Imaging Results aker s Cyst Other Cyst or Mass No Cyst or Mass Total aker s cyst Other masses No cyst or mass Total tients with aker s cysts (58%) and 15 patients without aker s cysts (42%). Of these 15 patients, 13 had no soft-tissue mass or cyst in the popliteal fossa. The two remaining patients had posterior cystic abnormalities proved by surgery to represent a myxoid liposarcoma and a meniscal cyst with coexisting meniscal tear. Sonography Report nalysis Evaluation of the sonography reports (Table 1) revealed that all 23 of the posterior soft-tissue masses or cysts shown on MR imaging were identified on sonography. In the diagnosis of a posterior soft-tissue mass or cyst (including a aker s cyst) versus a normal popliteal region, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy equaled 100%. ll 21 aker s cysts revealed on MR imaging were correctly diagnosed on sonography. However, the two other posterior soft-tissue masses or cysts (one myxoid liposarcoma and one meniscal cyst) were misdiagnosed as aker s cysts. In regard to the myxoid liposarcoma, sonography was performed before the MR imaging examination (47-day interval). In regard to the meniscal cyst, sonography was performed after the MR imaging examination (27- day interval). The sonographic diagnosis was correct in the 13 cases without posterior soft-tissue abnormality or aker s cyst. In the diagnosis of aker s cyst versus other findings (normal or other posterior soft-tissue mass or cyst), this resulted in an overall sensitivity of 100%, specificity of 87%, positive predictive value of 91%, negative predictive value of 100%, and accuracy of 94% when comparing the prospective sonography reports with the MR images. MR Imaging Results Retrospective Sonographic Interpretation aker s Cyst (n = 17) Other Cyst or Mass (n = 2) p % No. % No. Posterior soft-tissue mass or cyst Fluid between semimembranosus and medial gastrocnemius tendons nechoic Hypoechoic Mixed echogenicity Size (average) 10 cm cm 3 Septations Intraarticular body Rupture Note. Dash ( ) indicates p value not calculated. Retrospective Sonographic Image nalysis Regarding the retrospective review of the sonographic images, four of the 23 cases representing aker s cysts or other posterior soft-tissue mass or cyst on MR imaging were discarded because the semimembranosus and medial gastrocnemius tendons could not be identified on the sonographic images. Of the remaining 19 cases, retrospective review of the sonographic images revealed that all 19 of the posterior soft-tissue masses or cysts visible on MR imaging were identified on the sonographic images. The 19 posterior soft-tissue masses or cysts corresponded to 17 aker s cysts, one meniscal cyst, and one myxoid liposarcoma. Of the 17 aker s cysts, anechoic or hypoechoic fluid was identified between the semimembranosus and medial gastrocnemius tendons in all 17 (100%) (Table 2). This fluid communicated with the posterior cystic abnormality, creating a crescent shape that was concave toward midline as it wrapped around the medial gastrocnemius muscle and tendon (Figs. 2 and 3). Of the two other posterior soft-tissue masses or cysts representing myxoid liposarcoma (Fig. 6) and meniscal cyst (Fig. 7), no fluid was identified between the semimembranosus and medial gastrocnemius tendons. There was a significant association (p = 0.006) between the presence of fluid between the semimembranosus and medial gastrocnemius tendons on sonography and the presence of aker s cyst on MR imaging. This finding of fluid between the semimembranosus and medial gastrocnemius tendons on sonography resulted in a sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of 100% in differentiating aker s cyst from other posterior soft-tissue masses or cysts. In regard to other sonographic characteristics of the 19 soft-tissue masses (Table 2), 10 (59%) of the 17 aker s cysts were anechoic, four (23%) were hypoechoic, three (18%) were of mixed echogenicity, and none (0%) were hyperechoic. The one meniscal cyst appeared to be of mixed echogenicity, whereas the one myxoid liposarcoma appeared hypoechoic. The average size of the aker s cysts was 10 cm 3 (range, cm 3 ), whereas the myxoid liposarcoma measured 22.9 cm 3. Septations were identified in eight (47%) of the 17 aker s cysts (Fig. 3) and in both the meniscal cyst and myxoid liposarcoma. n intraarticular body was present in one of the 17 aker s cysts (Fig. 5). Evidence for rupture or leak (Fig. 4) was seen in three (18%) of the 17 aker s cysts. Overall, besides fluid between the semimembranosus and medial gastrocnemius tendons, there were no sonographic characteristics significantly associated with the finding of aker s cyst on MR imaging using the Fisher s exact test. The sonographic features of aker s cyst were then analyzed to determine their ability to indicate a diagnosis of aker s cyst versus other soft-tissue masses or cysts. When the soft-tissue mass or cyst appeared anechoic, sensitivity was 59%, specificity was 100%, positive predictive value was 100%, negative predictive value was 22%, and accuracy was 63% in the diagnosis of aker s cyst. If the soft-tissue mass or cyst appeared hypoechoic, sensitivity was 24%, specificity was 50%, positive predictive value was 80%, negative predictive value was 7%, and accuracy was 26%. If the soft-tissue mass or cyst JR:176, February

6 Ward et al. appeared to be of mixed echogenicity, sensitivity was 18%, specificity was 50%, positive predictive value was 75%, negative predictive value was 7%, and accuracy was 21% in the diagnosis of aker s cyst. The presence of septations resulted in a sensitivity of 47%, specificity of 0%, positive predictive value of 80%, negative predictive value of 0%, and accuracy of 42%. The presence of an intraarticular body resulted in sensitivity of 6%, specificity of 100%, positive predictive value of 100%, negative predictive value of 11%, and accuracy of 16%. If the soft-tissue mass or cyst appeared as a rupture or leak, sensitivity was 21%, specificity was 100%, positive predictive value was 100%, negative predictive value was 13%, and accuracy was 26% in the diagnosis of aker s cyst. Discussion aker s cysts develop when there is distention of the gastrocnemiosemimembranosus bursa [1]. The fluid distending the bursa typically originates from the knee joint. Communication between the posterior knee joint and the gastrocnemiosemimembranosus bursa has been shown to increase with age, possibly because of degenerative thinning of the joint capsule [10] compounded by joint effusion and increased intraarticular pressure associated with internal derangement [1]. Fifty percent of patients who are Fig year-old man with myxoid liposarcoma., xial sonogram shows relatively hypoechoic and septate cystic mass (arrowheads). Note that top of image is posterior; left side of image is lateral. F = femur, TV = transverse., xial T1-weighted MR image reveals heterogeneous, predominately intermediate-signal-intensity myxoid liposarcoma (arrowheads). C, xial proton density weighted MR image with fat saturation reveals predominately high-signal-intensity mass (arrowheads). Note lack of extension between semimembranosus tendon (solid arrow) and medial gastrocnemius tendon (open arrow). 50 years old or older will show findings of this communication [11]. The 15- to 20-mm transverse slitlike opening allows a joint effusion to enter the bursa [10], typically in knee flexion when the quadriceps tendon compresses the suprapatellar recess and intraarticular pressure is increased [12]. Fluid within a aker s cyst may communicate freely with the knee joint [13]. However, the presence of a valvelike mechanism may allow joint fluid to communicate in only one direction [13]. The gastrocnemiosemimembranosus bursa is a composite of two bursae [2]. The subgastrocnemius bursa between the medial gastrocnemius tendon and medial femoral condyle is the point C 378 JR:176, February 2001

7 Sonography and MR Imaging of aker s Cysts Fig year-old girl with meniscal cyst., Sagittal sonogram shows hypoechoic meniscal cyst (curved arrow) in contact with hyperechoic meniscus (open arrows) and hypoechoic meniscal tear (solid straight arrows). Note that top of image is posterior; left side of image is superior. F = femur, c = hyaline cartilage., Sagittal proton density weighted MR image reveals meniscal cyst (curved arrow) in continuity with meniscal tear (straight arrow). C, xial proton density weighted MR image with fat saturation reveals meniscal cyst (curved arrow) with signal intensity of fluid without extension between semimembranosus tendon (undulating arrow) and medial gastrocnemius tendon (arrowhead). of communication with the posterior joint capsule [2]. The posterior extension of a aker s cyst represents the second bursa between the medial gastrocnemius tendon and semimembranosus tendon that commonly communicates with the subgastrocnemius bursa [2]. oth sonography and MR imaging depict these characteristic features of a aker s cyst (Figs. 2 and 3). Retrospective review of the sonographic images showed that aker s cysts extend posterior between the medial gastrocnemius and semimembranosus tendons in a crescent-shaped configuration (Figs. 2 and 3). nechoic or hypoechoic fluid can be identified wrapping around the medial gastrocnemius tendon in the axial plane concave toward midline. The presence of fluid between the semimembranosus and medial gastrocnemius tendons was the most helpful sonographic feature in the diagnosis of aker s cyst. In the presence of a posterior soft-tissue knee mass or cyst, identification of fluid between the semimembranosus and medial gastrocnemius tendons communicating with the soft-tissue mass or cyst indicated aker s cyst as the cause, excluding other soft-tissue masses or cysts (100% sensitivity, specificity, positive predictive value, negative predictive value, and accuracy). Regarding the other sonographic features of aker s cysts, 59% were anechoic, 23% were C JR:176, February

8 Ward et al. hypoechoic, and 18% showed mixed echogenicity relative to muscle. lthough the cause of increased echogenicity within a aker s cyst was not proven in this study, possibilities include hemorrhage and thickened synovium. Overall, we found no significant difference between aker s cyst and other posterior soft-tissue knee mass or cyst on the basis of echogenicity or size or the presence of septations, intraarticular body, or rupture. However, a purely anechoic cystic structure was associated with 100% specificity and 100% positive predictive value in the diagnosis of aker s cyst versus mass. In addition, the presence of an intraarticular body (Fig. 5) or cyst rupture (Fig. 4) revealed a similar 100% specificity and 100% positive predictive value. Given the low sensitivities of these sonographic characteristics, the finding of fluid between the semimembranosus and medial gastrocnemius tendons remains the most useful sonographic feature in the diagnosis of aker s cysts with a sensitivity of 100%. Of note, the myxoid liposarcoma (Fig. 6) was hypoechoic with septations, and the meniscal cyst (Fig. 7) was of mixed echogenicity with septations. Neither showed communication with fluid between the semimembranosus and medial gastrocnemius tendons. Comparison of the sonography reports and the MR images revealed that 23 of the 23 posterior soft-tissue masses or cysts were identified on sonography. lthough 21 of the aker s cysts were correctly diagnosed, two of the other soft-tissue masses or cysts, representing a myxoid liposarcoma and a meniscal cyst, were misdiagnosed as aker s cysts on sonography. Retrospective review of the sonographic images in these two cases revealed that anechoic or hypoechoic fluid was not shown between the semimembranosus and medial gastrocnemius tendons. It is unclear if the examiner misinterpreted that fluid was present between the semimembranosus and medial gastrocnemius tendons or if the examiner did not even assess the fluid in this region. The results of this study show that noting the absence of fluid between the semimembranosus and medial gastrocnemius tendons would have indicated that the cystic masses were not aker s cysts. The reported incidence of aker s cysts in the general population ranges from 10% to 41% [3]. The incidence of aker s cysts in our study group is 58%. This finding may relate to a possible higher incidence of internal derangement in our patient population. This finding could also be associated with a higher threshold of the referring physicians in ordering MR imaging examinations, thus increasing the pretest probability of internal derangement. Limitations of this study include inherent operator dependence in the acquisition of the sonographic data and images. Kappa statistics for interobserver variability were not calculated. Limitations also exist in the retrospective interpretation of sonographic images. The low number of soft-tissue masses or cysts not representing aker s cysts may limit generalization of these results to larger populations. In addition, MR imaging was used as a gold standard and pathologic proof was not present in the diagnosis of aker s cysts. Lastly, neither color nor power Doppler imaging was used in this study, so the ability of Doppler imaging to differentiate aker s cyst from other abnormalities could not be assessed. The results of this study show the effectiveness of sonography in the diagnosis of aker s cysts. nalysis of the sonography reports revealed that all the posterior soft-tissue masses or cysts were depicted on sonography. Retrospective review of the sonographic images showed that when a cystic posterior soft-tissue mass is identified on sonography, visualization of communicating anechoic or hypoechoic fluid between the semimembranosus and medial gastrocnemius tendons indicates aker s cyst and excludes other cystic soft-tissue masses or cysts such as tumor or meniscal cyst. This distinction is especially important when a complicated aker s cyst appears heterogeneous and simulates a soft-tissue neoplasm. References 1. Miller TT, Staron R, Koenigsberg T, Levin TL, Feldman F. MR imaging of aker cysts: association with internal derangement, effusion, and degenerative arthropathy. Radiology 1996;201: Rauschning W. Popliteal cysts and their relation to the gastrocnemio-semimembranous bursa: studies on the surgical and functional anatomy. cta Orthop Scand 1979;179: Janzen DL, Peterfy CG, Forbes JR, Tirman PFJ, Genant HK. Cystic lesions around the knee joint: MR imaging findings. JR 1994;163: Laine HR, Harjula, Peltokallio P. Ultrasound in the evaluation of the knee and patellar regions. J Ultrasound Med 1987;6: Toolanen G, Lorentzon R, Friberg S, Dahlström H, Öberg L. Sonography of popliteal masses. cta Orthop Scand 1988;59: Østergaard M, Court-Payen M, Gideon P, et al. Ultrasonography in arthritis of the knee: a comparison with MR imaging. cta Radiol 1995;36: ndonopoulos P, Yarmenitis S, Sfountouris H, Siamplis D, Zervas C, ounas. aker s cyst in rheumatoid arthritis: an ultrasonographic study with high resolution technique. Clin Exp Rheumatol 1995;13: Helbich TH, reitenseher M, Trattnig S, Nehrer S, Erlacher L, Kainberger F. Sonomorphologic variants of popliteal cysts. J Clin Ultrasound 1998;26: Crass JR, van de Vegte GL, Harkavy L. Tendon echogenicity: ex vivo study. Radiology 1988; 167: Lindgren PG, Willén R. Gastrocnemio-semimembranosus bursa and its relation to the knee joint. I. natomy and histology. cta Radiol 1977;18: Lindgren PG, Willén R. Gastrocnemio-semimembranosus bursa and its relation to the knee joint. II. Post mortem radiography. cta Radiol 1977;18: Doppman JL. aker s cyst and the normal gastrocnemio-semimembranosus bursa. JR 1965;94: Lindgren PG, Rauschning W. Clinical and arthrographic studies on the valve mechanism in communicating popliteal cysts. rch Orthop Trauma Surg 1979;95: JR:176, February 2001