Three-dimensional isotropic T2-weighted fast spin-echo (VISTA) knee MRI at 3.0 T in the evaluation of the anterior cruciate ligament injury with additional views: comparison with two-dimensional fast spin-echo T2-weighted sequences

Abstract

Background

There has been no comparison study to assess the diagnostic efficacy of additional anterior cruciate ligament (ACL) views in three-dimensional (3D) volume isotropic turbo spin-echo acquisition (VISTA) and two-dimensional (2D) fast spin-echo (FSE) T2-weighted (T2W) images for the diagnosis of ACL tear.

Purpose

To compare the diagnostic performances of additional ACL views on VISTA with those on the 2D FSE T2W images.

Material and Methods

This retrospective study included 78 patients who were suspected to have ACL injury and underwent both 2D TSE T2W magnetic resonance imaging (MRI) and 3D VISTA MRI of the knee between November 2012 and March 2013. The diagnostic performance of each oblique sagittal and coronal view and the combined images was evaluated for sensitivity, specificity, and accuracy for diagnosing an ACL tear. The arthroscopically and clinically confirmed diagnoses were used as the reference standard. The values were statistically analyzed using the McNemar test.

Results

The inter-observer agreement between two readers of the additional ACL views on 3D VISTA and 2D FSE T2W images were substantial on 2D FSE images and nearly concurred on the VISTA image. When considering both views of the oblique sagittal and coronal images, the inter-observer agreement between readers nearly concurred. There were no statistically significant differences in the sensitivity, specificity, and accuracy between 2D FSE images and VISTA images.

Conclusion

The performance of the additional ACL view on 3D VISTA MRI is comparable to that of 2D FSE T2W MRI in the diagnosis of ACL tear though the image quality of the 3D VISTA MRI is not equal to that of 2D FSE MRI.

Keywords

Magnetic resonance imaging (MRI)knee anterior cruciate ligament (ACL)3D isotropic VISTA

Introduction

The anterior cruciate ligament (ACL) has a complex course. The ACL originates from the posterior superior part of the medial aspect of the lateral femoral condyle, runs obliquely in an anteromedial direction and inserts in the central and medial portions of the anterior inter-condylar area of the tibia (1). Magnetic resonance imaging (MRI) diagnosis of an ACL tear is based on findings that include discontinuity and a wavy contour, focal or diffuse high signal intensity within the ligament substance on T2-weighted (T2W) image or proton density image (2). Kwon et al. (3) reported that additional MR oblique views (oblique sagittal and oblique coronal, ACL views) improve the diagnostic specificity of ACL tears. They reported that oblique images can more easily capture changes in thickness, signal intensity, and contour deformity. However, to benefit from these advantages, additional sequences have to be performed in two or three planes. The three-dimensional (3D) isotropic image can be used to make multi-planar reformatted images and can shorten the total acquisition time by obviating the need to acquire the same sequence in different planes (4 –6). Several 3D fast-field echo imaging methods have been used in the evaluation of the knee joint with multi-planar reconstruction, but insufficient soft tissue contrast leads to limitations in the detection of ligament injury (7 –11). 3D isotropic fast spin-echo T2W MR (volume isotropic turbo spin echo acquisition [VISTA]) enables thin-section data acquisition without inter-slice gap and multi-planar image reformation that may be helpful for the analysis of complex structures (12). The VISTA sequences are obtained utilizing a fast spin-echo (FSE) 3D non-selective method, which uses short, non-volume selective FSE to refocus pulses and slow shorter echo spacing to prevent chemical shift artifact (8). Gold et al. (13) reported that 3D FSE produced high-quality isotropic images with similar contrast to two-dimensional (2D) FSE. Subhas et al. (14) reported that the diagnostic performance of 3D FSE with multi-planar reformation for internal derangements of the knee is comparable to that of conventional sequences consisting of 2D FSE. However, to or knowledge, there has been no comparison study to assess the diagnostic efficacy of additional ACL views in 3D VISTA and 2D FSE T2W images for the diagnosis of ACL tear. The purpose of this study is to compare the diagnostic performances of additional ACL views on 3D VISTA image with those on 2D FSE T2W images.

Material and Methods

Case selection and clinical diagnosis

This study was approved by the Institutional Ethics Review Board of our hospital. The requirement for informed consent was waived due to the retrospective study design. We retrospectively evaluated 78 patients who were suspected to have ACL injury based on clinical history and underwent both 2D TSE T2W MRI and 3D VISTA MRI of the knee from November 2012 to March 2013. The study participants included 28 women and 50 men (mean age, 43 ± 15 years; range, 13–76 years). Patients who had previous surgical interventions such as ACL reconstruction were excluded from the study. The orthopedic surgeon had advanced fellowship training in knee repair as well as 30 years of surgical experience on the knee joint. The surgeries were performed soon after MRI, with a mean imaging to surgery interval of 16 (± 14) days. No medical treatments such as NSAIDSs or physical therapy were provided during this period. During arthroscopy, the injury to the ACL was evaluated and classified as grades 0 to 3: grade 0, intact ACL; grade 1, partial adhesion of the ligament fiber and a coarse fiber surface; grade 2, elongation of ligament fiber and laxity during probing of the ligament; and grade 3, definite disruption of the ACL (15). Once the radiologists completed their blind grading of the MR images, the surgeon and radiologists correlated the surgical and MRI findings. Patients who were categorized as grade 0 without arthroscopy were diagnosed based on physical examination findings such as negative anterior drawer test, radiologic findings of negative laximetry test, and lack of significant clinical findings such as limping and discomfort on flexion and orthogonal MR sequences (sagittal and coronal images). We excluded both oblique image series at the time of making the reference diagnosis. We only used orthogonal MR images as the ancillary diagnostic method in addition to the clinical findings. Thirty-five patients (45%) were graded 0 based on these criteria (Table 1). Out of a total of 78 patients, 27 patients (35%) underwent arthroscopy by the same orthopedic surgeon. Among these cases, eight were grade 2 and eight were grade 3 patients. Eleven patients (41%, 11/27) were graded 0 on arthroscopy. The arthroscopic diagnoses of these 11 patients were medial meniscal tear (n = 6), lateral meniscal tear (n = 2), and cartilage injuries (n = 3). In the remaining 16 patients (31%, 16/51), the ACL tear was made based on clinical manifestation, physical examination and MRI findings (other orthogonal MRI sequences and follow-up MRI) by two radiologists and clinician in consensus (16). The results of stress radiographs with a TELOS device (Telos, Laubscher, Holstein, Switzerland) and clinical findings such as response to physical therapy and symptom course were also considered in the final diagnosis. Among the 16 cases, four had complete tear, five grade 1 and seven cases had grade 2 (Table 1). Although not precise accurate diagnosis, the cases of the grade 1 and 2 had obvious symptom and physical sign. The orthogonal MR images showed elongation and loosening of the ligament in the case of grade 2 and diffuse thickening and increased signal in the ligament in the case of grade 1. Final diagnoses were made though consensus between orthopedic surgeon and both radiologists.

Table 1.

Case selection and clinical diagnosis of ACL.

	ACL tear (+)
Method to confirm ACL tear	G1	G2	G3	ACL tear (–)	Total
Arthroscopy	0	8	8	11	27
MRI and physical examination	5	7	4	35	51
Total	5	15	12	46	78

ACL, anterior cruciate ligament.

MR parameters

MR examinations were performed on a 3.0-T MR scanner (Achieva, Philips Medical Systems, Best, The Netherlands) with an 8-channel knee coil. The sequences and imaging parameters are summarized in Table 2. Additional ACL sagittal views were obtained using the following parameters: TR/TE, 2500–3000/90; field of view (FOV), 14 cm; echo train length (ETL), 16; matrix, 500 × 300; slice thickness, 2 mm. ACL coronal views were obtained using the following parameters: TR/TE, 2500–3000/90; FOV, 14 cm; ETL, 16; matrix, 500 × 320; slice thickness, 2 mm. The ACL coronal view was obtained in the plane parallel to the course of the femoral inter-condylar roof based on the conventional sagittal image and the ACL sagittal image was obtained in the plane parallel to the medial border of the lateral femoral condyle on the orthogonal coronal image (3). For 3D FSE MRI, sagittal 3D T2W FSE images with a reconstruction voxel size of 0.5 × 0.5 × 0.5 mm were obtained. The 3D isotropic sequence was performed utilizing a FSE 3D non-selective method, which uses short, non-volume selective FSE refocusing pulses and allows shorter echo spacing to prevent chemical shift artifact, a driven-equilibrium radiofrequency pulse, and an asymmetric FSE profile order. The sagittal images originally acquired from 3D FSE VISTA were subsequently reformatted into oblique sagittal and coronal images by technologists with the use of software (extended MR workspace, Philips Medical Systems). The reformation was performed with same slice thickness and the same angle of the 2D FSE additional ACL views without the inter-slice gap.

Table 2.

Imaging parameters for the MR sequences.

Imaging parameter	Coro PD FSE	Coro FS T2 FSE	Sag FS PD FSE	Sag T2 FSE	ACL view Sag T2 FSE	ACL view Coro T2 FSE	VISTA
TR (ms)	3000–3500	4500–5000	4000–4500	4500–5000	2500–3000	2500–3000	1800
TE (ms)	30–40	70–90	30–40	50–100	50–80	50–80	100
Flip angle (°)	90	90	90	90	90	90	90
Matrix size	450 × 300	450 × 300	300 × 230	430 × 300	500 × 300	500 × 320	320 × 160
Field of view (cm)	16	16	16	16	14	14	16
Slice thickness (mm)	3	3	3	3	2	2	0.5
Inter-slice gap (mm)	0.3	0.3	0.3	0.3	0	0	0
Bandwidth (kHz)	460	240	240	290	400	400	0.5
Echo train length	15	16	14	17	16	16	55
Signal average	1	1	1	1	1	1	1
Scan time (min:s)	2:22	3:23	2:47	2:00	2:04	1:55	3:20

ACL, anterior cruciate ligament; Coro, coronal; FS, fat saturation; FSE, fast spin-echo; PD, proton density; Sag, sagittal; VISTA, volume isotropic turbo spin-echo acquisition.

Image analysis

The knee MR images were interpreted independently by two fellowship-trained musculoskeletal radiologists who were unaware of the radiologic reports or clinical history. First, the readers only evaluated the 2D FSE oblique sagittal ACL views. At the next session several weeks later, they evaluated the 2D FSE oblique coronal ACL views. Finally, after another several-week period, they evaluated both the 2D FSE oblique ACL views. A similar process was used to evaluate the 3D FSE VISTA images at different times. The ACL tear group and the normal group (grade 0 group) were randomly mixed for the reading session. If the ACL was intact and depicted as a continuous linear band of normal hypo-intense signal, it was categorized as grade 0. If there was focal or diffuse high signal intensity within the substance of the ACL without loosening or wavy contour, it was given a grade 1 (Figs. 1 and 2). Grade 2 was when the ACL was elongated and showed undulating wavy contour without complete disruption. If there was evidence of definite discontinuity, the ACL was considered to have a complete tear (grade 3, Figs. 3 and 4) (3).

Fig. 1.

A 44-year-old man with knee pain after skiing. (a) Oblique sagittal T2W image (TR/TE, 2700/75) shows a focal bright signal and in the ACL (arrow). (b) Oblique coronal T2W MRI (TR/TE, 2600/60) of the ACL also shows a focal bright signal in the ACL (arrow). (c, d) Oblique sagittal and coronal images reformatted from VISTA MR show bright signal in the same portion of the ACL (arrows). The final diagnosis of was grade 1 ACL tear. Medical care without surgical intervention was provided and with relief of symptoms later.

Fig. 2.

A 24-year-old man with knee pain. (a) Oblique sagittal T2W image (TR/TE, 2700/75) shows no disruption of the ACL (long arrow). (b) Oblique sagittal image reformatted from VISTA MRI shows a focal bright signal in the mid-portion of the ACL. The arthroscopic diagnosis was synovial plica syndrome with intact ACL.

Fig. 3.

A 36-year-old man with knee pain after traffic accident. (a) Oblique sagittal T2W image (TR/TE, 2700/75) shows disruption of the ACL (long arrow) with preserved tension of the residual portion of ACL (short arrow). The diagnosis was grade 2. (b) Oblique coronal T2W MRI (TR/TE, 2600/60) of the ACL shows clear disruption of the ACL (arrow). (c, d) Oblique sagittal and coronal images reformatted from VISTA MR show complete rupture of ACL. The arthroscopic diagnosis was complete rupture of ACL (grade 3).

Fig. 4.

A 58-year-old man with knee pain after slipping down. (a) Oblique sagittal T2W image (TR/TE, 2700/75) shows disruption of the ACL (long arrow). (b) Oblique coronal T2W MRI (TR/TE, 2600/60) of the ACL also shows focal disruption of the ACL (arrow). (c, d) Oblique sagittal and coronal images reformatted from VISTA MR show a grade 2 tear in the same portion of the ACL. (e) The arthroscopic finding shows some irregularities in the surface but an intact ACL.

Statistical analysis

Inter-observer agreements of diagnostic performance were analyzed using kappa statistics. The kappa value was interpreted as follows: poor (k < 0.1); slight (0.1 ≤ k ≤ 0.2); fair (0.2 < k ≤ 0.4); moderate (0.4 < k ≤ 0.6); substantial (0.6 < k ≤ 0.8); and nearly concurred (0.8 < k ≤ 1) (17). The diagnostic performance of each oblique sagittal and coronal view and the combined images were evaluated using the sensitivity, specificity, and accuracy for diagnosing an ACL tear. The arthroscopically or clinically confirmed diagnoses were used as the reference standard. The values were statistically analyzed using the McNemar test. Statistical analyses were performed using PASW software version 18.0 (IBM Corp., Armonk, NY, USA). P values ≤ 0.05 were considered to be statistically significant.

Results

Inter-observer agreements between two readers on the each additional ACL view were substantial (kappa value, 0.768; oblique coronal, 0.790; oblique sagittal) on the 2D FSE T2W images and nearly concurred on the VISTA image (kappa value, 0.837; oblique coronal, 0.845; oblique sagittal) (Table 3). When considering views of both the oblique sagittal and coronal images, the inter-observer agreement between readers nearly concurred (kappa value, 0.821–0.847). When we classified the diagnosis as normal and abnormal, the sensitivity of the 2D FSE images for the diagnosis of ACL tears was more than 96% and the sensitivity of the VISTA images was 100% (Table 4). There was no constant trend or pattern of values for specificity and accuracy. When we subdivided the diagnoses as normal, partial tear (grades 1 and 2) and complete tear (grade 3), it resulted in similar findings in sensitivity, specificity and accuracy (Table 5). There were no statistically significant differences in the sensitivity, specificity and accuracy between 2D FSE images and VISTA images (P values, 0.063–1.000).

Table 3.

Inter-observer agreement for diagnostic performance.

Views	Kappa value	95% confidence interval	P value
OS	0.790	0.676–0.904	<0.001
OC	0.768	0.650–0.885	<0.001
OS + OC	0.847	0.746–0.947	<0.001
VOS	0.845	0.748–0.943	<0.001
VOC	0.837	0.734–0.941	<0.001
VOS + VOC	0.821	0.715–0.927	<0.001

Kappa values are presented with 95% confidence intervals.

OC, oblique corona; OS, oblique sagittal; VOC, VISTA oblique coronal; VOS, VISTA oblique sagittal.

Table 4.

Sensitivity, specificity, and accuracy of the VISTA and the spin-echo imaging in the differentiation between normal and abnormal for diagnosis of ACL tears.

Views	Sensitivity (%)		Specificity (%)		Accuracy (%)
Views	Reader 1	Reader 2	Reader 1	Reader 2	Reader 1	Reader 2
OS	96.9 (31/32)	96.9 (31/32)	91.3 (42/46)	97.8 (45/46)	93.6 (73/78)	97.4 (76/78)
OC	96.9 (31/32)	96.9 (31/32)	97.8 (45/46)	91.3 (42/46)	97.4 (76/78)	93.6 (73/78)
OS + OC	96.9 (31/32)	96.9 (31/32)	97.8 (45/46)	97.8 (45/46)	97.4 (76/78)	97.4 (76/78)
VOS	100 (32/32)	100 (32/32)	84.8 (39/46)	87.0 (40/46)	91.0 (71/78)	92.3 (72/78)
VOC	100 (32/32)	100 (32/32)	89.1 (41/46)	93.5 (43/46)	93.6 (73/78)	96.2 (75/78)
VOS + VOC	100 (32/32)	100 (32/32)	87.0 (40/46)	91.3 (42/46)	92.3 (72/78)	94.9 (74/78)

Numbers in parentheses are the numbers of patients used to calculate the percentages.

OC, oblique corona; OS, oblique sagittal; VOC, VISTA oblique coronal; VOS, VISTA oblique sagittal.

Table 5.

Sensitivity, specificity, and accuracy of the VISTA and the spin-echo imaging in the differentiation between normal, partial, and complete tear for diagnosis of ACL tears.

Views	Sensitivity (%)		Specificity (%)		Accuracy (%)
Views	Reader 1	Reader 2	Reader 1	Reader 2	Reader 1	Reader 2
OS	90.6 (29/32)	93.8 (30/32)	91.3 (42/46)	97.8 (45/46)	91.0 (71/78)	96.2 (75/78)
OC	90.6 (29/32)	93.8 (30/32)	97.8 (45/46)	91.3 (42/46)	94.9 (74/78)	92.3 (72/78)
OS + OC	90.6 (29/32)	96.9 (31/32)	97.8 (45/46)	97.8 (45/46)	94.9 (74/78)	97.4 (76/78)
VOS	93.8 (30/32)	100 (32/32)	84.8 (39/46)	87.0 (40/46)	88.5 (69/78)	92.3 (72/78)
VOC	90.6 (29/32)	93.8 (30/32)	89.1 (41/46)	93.5 (43/46)	89.7 (70/78)	93.6 (73/78)
VOS + VOC	100 (32/32)	96.9 (31/32)	87.0 (40/46)	91.3 (42/46)	92.3 (72/78)	93.6 (73/78)

Numbers in parentheses are the numbers of patients used to calculate the percentages.

OC, oblique corona; OS, oblique sagittal; VOC, VISTA oblique coronal; VOS, VISTA oblique sagittal.

Discussion

The 3D MRI techniques allow delineation of anatomical details within the imaged volume and can be useful for 3D reformation in arbitrarily chosen orientations (18). 3D sequences based on gradient-echo (GRE) imaging which produce T2* contrast by using low flip angles and long TEs result in a decrease in the signal-to-noise ratio (SNR) and the contrast-to- noise ratio (CNR) (18). In comparison, 3D VISTA is superior to 3D GRE images in tissue contrast in the knee joint (19). 3D VISTA combines high spatial resolution and T2 contrast within an acceptable scan time by using long ETL and parallel imaging (18). Kwon et al. (18) reported that 3D VISTA MRI is comparable or superior to 2D FSE for the delineation of nerve roots and the neural foramen in the cervical spine; the study authors also insisted that 3D VISTA is not feasible for the evaluation of the muscular system because the muscle signals in VISTA sequences are more heterogeneous and have more background noise. However, Jung et al. (12) reported that 3D VISTA MRI yields accuracy comparable to that of 2D FSE MRI in the diagnosis of ACL and meniscal tears. In our study, we did not find any significant differences between 3D VISTA and 2D FSE imaging in diagnostic performance such as discrimination between normal and abnormal, discrimination between normal, partial tear, and complete tear, and discrimination of the four grades of ACL injuries. The only weakness of 3D VISTA compared with 2D FSE was relatively poor image sharpness which did not influence diagnostic performance. In other 3D FSE images such as the XETA-FSE sequence of GE Health care, increased image blurring and indistinctness of the structural edges have been reported (16). Since the main usefulness of the oblique views in the knee MRI is for the proper evaluation of ACL continuity, decreased margin sharpness of the shape does not hinder the correct diagnosis of ACL tears. The useful findings of the ACL injury on MRI are disruption of the ligament fiber and signal change of the ligament itself. Nevertheless, we could find and identify whether ligament disruption or signal change exists or not sufficiently (Fig. 2). Therefore oblique views from 3D VISTA images can replace the 2D FSE oblique views in the diagnosis of ACL tears. However, the more important issue is the time needed for scanning. Our mean scan time for the VISTA source image was about 3 min 20 s, not including the time for reformation. The total scan time for oblique coronal and sagittal ACL views on 2D FSE was about 4 min. The reduced scan time was only shortened by about 20%. Jung et al. (12) reported that the diagnostic performance of the 3D VISTA image is comparable to that of 2D FSE on orthogonal views in the evaluation of cruciate ligament and meniscal tears. Thus, if we replace the orthogonal T2 FSE coronal and sagittal view with the 3D VISTA reformatted image, the time benefit exceeds 66%. While there is not sufficient time benefit to use the 3D VISTA image to replace only oblique views for ACL evaluation, there may be a satisfactory time benefit if we use the 3D VISTA image to replace some orthogonal views in addition to oblique views. However, radiologists usually do not evaluate the pathology of the cruciate ligament and meniscus solely with orthogonal coronal and sagittal images of the knee joint. They should also evaluate the cartilage, bone marrow and fatty components in the knee joint. Further investigation into the diagnostic ability of the 3D VISTA image in the evaluation of the above mentioned knee pathology should be conducted.

This study has some limitations. First, our study did not include orthogonal views in the diagnosis of ACL tear. We wanted to perform a precise comparison between the 3D VISTA image and 2D FSE image; Jung et al. (12) had already assessed the use of orthogonal views in the evaluation of the ACL. The second limitation was the small number of cases which was confirmed surgically. Only 27 (35%) of 78 patients underwent arthroscopy. However, great care was taken to diagnose the best possible standard of reference in the other cases using clinical findings as well as all other available MR sequences, consensus reading and follow-up MRI. Third, the study researchers’ awareness of whether the sequence was a 3D image or 2D image may have resulted in bias during evaluation of the ACL. Although sequence parameters were not displayed when we evaluated the image, a completely blinded study was impossible because of the less sharp margins of 3D images.

In conclusion, even though image quality of the 3D VISTA MR is not equal to 2D FSE MR, the diagnostic performance of additional ACL views on 3D VISTA MR is comparable to that of 2D FSE T2W MR in the diagnosis of ACL tears.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Samsung Biomedical Research Institute grant, #SMX1132241.

References

Duc

Zanetti

Kramer

. Magnetic resonance imaging of anterior cruciate ligament tears: evaluation of standard orthogonal and tailored paracoronal images. Acta Radiol 2005; 46: 729–733.

Katahira

Yamashita

Takahashi

. MRI imaging of the anterior cruciate ligament: value of thin slice direct oblique coronal technique. Radiat Med 2001; 19: 1–7.

Kwon

Yoon

Kim

. Which oblique plane is more helpful in diagnosing an anterior cruciate ligament tear? Clin Radiol 2009; 64: 291–297.

Jung

Jee

Park

. Supraspinatus tendon tears at 3.0 T shoulder MR arthrography: diagnosis with 3D isotropic turbo spine-echo SPACE sequence versus 2D conventional sequences. Skeletal Radiol 2012; 41: 1401–1410.

Stevens

Busse

Han

. Ankle: isotropic MR imaging with 3D-FSE-cube-initial experience in healthy volunteers. Radiology 2008; 249: 1026–1033.

Kijowski

Davis

Woods

. Knee joint: comprehensive assessment with 3D isotropic resolution fast spin-echo MR imaging-diagnostic performance compared with that of conventional MR imaging at 3.0 T. Radiology 2009; 252: 486–495.

Yao

Pitts

Thomasson

. Isotropic 3D fast spin-echo with proton-density-like contrast: a comprehensive approach to musculoskeletal MRI. Am J Roentgenol 2007; 188: W199–W201.

Gold

Chen

Koo

Hargreaves

Bangerter

. Recent advances in MRI of articular cartilage. Am J Roentgenol 2009; 193: 628–638.

Guckel

Jundt

Schnabel

. Spin-echo and 3D gradient-echo imaging of the knee joint: a clinical and histopathological comparison. Eur Radiol 1995; 21: 25–33.

10.

Heron

Calvert

. Three-dimensional gradient-echo MR imaging of the knee: comparison with arthroscopy in 100 patients. Radiology 1992; 183: 839–844.

11.

Reeder

Matz

Becker

Andelman

. MR imaging of the knee in the sagittal projection: comparison of three-dimensional gradient-echo and spin-echo sequences. Am J Roentgenol 1989; 153: 537–540.

12.

Jung

Yoon

Kwon

. Diagnosis of internal derangement of the knee at 3.0-T MR imaging: 3D isotropic intermediate-weighted versus 2D sequences. Radiology 2009; 253: 780–787.

13.

Gold

Busse

Beehler

. Isotropic MRI of the knee with 3D fast spin-echo extended echo-train acquisition (XETA): initial experience. Am J Roentgenol 2007; 188: 1287–1293.

14.

Subhas

Kao

Freire

. MRI of the knee ligaments and menisci: comparison of isotropic-resolution 3D and conventional 2D fast spin-echo sequences at 3 T. Am J Roentgenol 2011; 197: 442–450.

15.

Park

Cha

Kim

. Accuracy of MRI findings in chronic lateral ankle ligament injury: Comparison with surgical findings. Clin Radiol 2012; 67: 313–318.

16.

Ristow

Steinbach

Sabo

. Isotropic 3D fast spin-echo imaging versus standard 2D imaging at 3.0 T of the knee – image quality and diagnostic performance. Eur Radiol 2009; 19: 1263–1272.

17.

Viera

Garrett

. Understanding interobserver agreement: the kappa statistic. Fam Med 2005; 37: 360–363.

18.

Kwon

Yoon

Choi

. Three-dimensional isotropic T2-weighted cervical MRI at 3T: Comparison with two-dimensional T2-weighted sequences. Clin Radiol 2012; 67: 106–113.

19.

Seo

Yoon

Kwon

. 3D isotropic turbo spin-echo intermediate-weighted sequence with refocusing control in knee imaging: comparison study with 3D isotropic fast-field echo sequence. Acta Radiol 2011; 52: 1119–1124.