Eur Radiol DOI 10.1007/s00330-016-4436-x

MUSCULOSKELETAL

Intrinsic carpal ligaments on MR and multidetector CT arthrography: comparison of axial and axial oblique planes Ryan K. L. Lee 1 & James F. Griffith 1 & Alex W. H. Ng 1 & Eric K. C. Law 1 & W. L. Tse 2 & Clara W. Y. Wong 2 & P. C. Ho 2

Received: 3 March 2015 / Revised: 29 February 2016 / Accepted: 23 May 2016 # European Society of Radiology 2016

Abstract Purpose To compare axial and oblique axial planes on MR arthrography (MRA) and multidetector CT arthrography (CTA) to evaluate dorsal and volar parts of scapholunate (SLIL) and lunotriquetral interosseous (LTIL) ligaments. Methods Nine cadaveric wrists of five male subjects were studied. The visibility of dorsal and volar parts of the SLIL and LTIL was graded semi-quantitatively (good, intermediate, poor) on MRA and CTA. The presence of a ligament tear was determined on arthrosocopy and sensitivity, specificity and accuracy of tear detection were calculated. Results Oblique axial imaging was particularly useful for delineating dorsal and volar parts of the LTIL on MRA with overall ‘good’ visibility increased from 11 % to 78 %. The accuracy of MRA and CTA in revealing SLIL and LTIL tear was higher using the oblique axial plane. The overall accuracy for detecting SLIL tear on CTA improved from 94 % to 100 % and from 89 % to 94 % on MRA; the overall accuracy of detecting LTIL tear on CTA improved from 89 % to 100 % and from 72 % to 89 % on MRA Conclusion Oblique axial imaging during CT and MR arthrography improves detection of tears in the dorsal and volar parts of both SLIL and LTIL.

* James F. Griffith [email protected]

1

Department of Imaging and Interventional Radiology, Prince Of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, Hong Kong

2

Department of Orthopedics and Traumatology, Prince Of Wales Hospital, The Chinese University of Hong Kong, Hong Kong Hong Kong

Key Points • Oblique axial imaging improves SLIL and LTIL visibility and tear detection. • This improvement is greater for the LTIL than for the SLIL ligament. • Overall, CT arthrography performed better than MR arthrography. Keywords Oblique axial . Arthrogram . Scapholunate interossous ligament . Lunotriquetral interosseous ligament . MRA

Introduction The scapholunate and lunotriquetral interosseous ligaments are the most important intrinsic ligaments of the wrist and also those most prone to traumatic injury [1, 2]. Tears of these intrinsic ligaments may lead to pain and carpal instability. Each of these ligaments comprises collagenous dorsal and volar parts as well as a fibrocartilaginous or membranous central part. This central fibrocartilaginous part is thin and frequently torn [1]. Structurally, the dorsal and volar collagenous parts are the more important parts [1]. The scapholunate dorsal part is thicker (2–3 mm) and stronger than the volar part (1 mm), while the reverse is true for the lunotriquetral interosseous ligament with the volar part being thicker and stronger than the dorsal part [3]. Standard axial cross-sectional magnetic resonance (MR) imaging is frequently employed to assess the integrity of the dorsal and volar parts of the scapholunate and lunotriquetral interosseous ligaments [4–9]. Robinson et al. used oblique axial MR imaging to view the scapholunate interosseous ligament (SLIL) and lunotriquetral interosseous ligament (LTIL) in 2006 [10]. In that study of 26 patients, oblique axial imaging significantly improved

Eur Radiol

conspicuity of these intrinsic ligaments tears, particularly of the LTIL. However, this study was limited by the lack of a dedicated wrist coil and no arthroscopic correlation. Also, the use of oblique axial reformatting on CTA for visualizing the SLIL and LTIL has not been studied before. The aim of the current study was to compare axial and oblique axial imaging planes on MR arthrography (MRA) using a dedicated wrist coil and multidetector CT arthrography (CTA) to evaluate the dorsal and volar interosseous ligaments of the wrist with arthroscopic correlation.

Methods and materials This study was approved by the NTEC-PWH ethics review board. Specimens Nine wrists from five formalin-embalmed male cadavers with no history of surgery were utilized for the purpose of the study in November 2013. The average age of the cadavers was 66.2 years old (range 59–90) at the time of death. Each cadaver was disarticulated at the elbow joint and consisted of the forearm, wrist and hand. Specimens were deep frozen at −40° Celsius. Before arthrography, each specimen was thawed to room temperature for 24 h. All specimens were examined with conventional arthrography followed immediately by CT arthrography and MR arthrography. Arthroscopy of the wrists was performed 1 week after the imaging examinations. Conventional arthrography Two-compartmental arthrography of the wrist was performed by one of two musculoskeletal radiologists (A.W.H.N or R.K.L.L) under fluoroscopic guidance. A 25-gauge needle Fig. 1 Coronal localizer for oblique axial plane reformatting using reformatted coronal images. The angles are taken parallel to (a) SLIL and (b) LTIL. SLIL scapholunate interosseous ligament, LTIL lunotriquetral interosseous ligament

was placed percutaneously from a dorsal approach. The needle was first advanced into the radioscaphoid joint. Needle tip position in the radioscaphoid joint was confirmed using a small test injection of iodinated contrast agent (Omnipaque 300; GE Healthcare Ltd., Shanghai, China). The MR contrast medium mixture used was 2 ml of gadoteric acid (Gadoteric acid, Dotarem, Guerbet, Roissy, France) diluted in 200 ml saline. Three to 5 ml of the solution was injected into the radioscaphoid joint under fluoroscopic control. If communication with the midcarpal joint was present, an additional 3–5 ml of the solution was injected. If no communication was present, the midcarpal (interspace between lunate, capitate, triquetrum bones) joint was sequentially injected with 3–5 ml solute. CT arthrography All cadavers underwent CT examination of the wrist immediately after arthrography on a multidetector CT scanner (Lightspeed 64 VCT, GE Healthcare, Uppsala, Sweden) using a standard clinical CT arthrography protocol (20 × 0.625 mm acquisitions with 100 mA, 120 kV, pitch of 1:1). Scan plane covered from the distal forearm to the metacarpophalangeal joints. The images were acquired in a standard axial plane. Subsequently on a dedicated CT workstation (GE Healthcare, Björkgatan 30, Uppsala, Sweden), the images were reformatted in an oblique axial plane at right angles to the scapholunate and lunotriquetral interspaces to assess both the SLIL and the LTIL respectively (Fig. 1a and b). MR arthrography MR arthrography was performed within 30 min of intraarticular contrast agent injection. All cadaveric wrists were examined in a 3.0-T MR whole-body scanner (Achieva TXseries, Philips Medical Systems, Best, The Netherlands) using a custom-built dedicated eight-channel wrist coil arrays

Eur Radiol

(SENSE wrist coil 8, Philips Medical Systems, Best, The Netherlands). The wrist was placed in a prone position parallel to the long axis of the gantry with the centre of the bore determined by laser mark. MRI protocols were listed in Table 1. The oblique axial plane at right angles to the scapholunate and lunotriquetral interspaces was performed to assess both the SLIL and the LTIL (Table 1) (Fig. 2a and b). Mean time for MR arthrography was 25 min, excluding the time taken for conventional arthrography. The mean time for the two oblique axial MR sequences was 8 min.

to assess the inter-observer variation and one of these two radiologists (R.K.L.L) repeated the assessment 2 weeks later to assess the intra-observer variation. The radiologists did not know the CTA result while judging the MRA and vice versa. Training on how to grade ligament tears was given to one radiologist (A.W.H.N) by another radiologist (R.K.L.L.)

Arthroscopy

The visibility of the dorsal and volar parts of the SLIL and the LTIL were semi-quantitatively graded as good, intermediate and poor by the two radiologists independently and afterwards by consensus. ‘Good’ visibility was defined as the ligament clearly and unambiguously seen as a sharp outline, ‘poor’ was defined as non- or near non-visibility of the ligament, while ‘intermediate’ was defined as the ligament being visible but not sharply demarcated.

All wrist specimens underwent standard wrist arthroscopy by a hand surgeon (W.L.T) with 13 years’ experience in arthroscopy, blinded to the imaging results. The arthroscopy was performed using the technique described by Geissler et al. [11–13]. The forearm was secured to a traction tower and distraction force of 10 lbs (4.5 kg) was applied across the wrist with finger traps on the index, middle and ring fingers. The radiocarpal and midcarpal joints were assessed with a 2.7-mm arthroscope and a 30° viewing angle. The individual parts of the SLIL and LTIL were assessed by both direct and dynamic inspection by inserting a 2-mm hook probe into both the radiocarpal and mid-carpal joints to detect any tear. Image analysis Two musculoskeletal radiologists (A.W.H.N and R.K.L.L) with 18 and 8 years’ experience, respectively, evaluated all the fluoroscopic, CTA and MRA images. All images were interpreted on a dedicated picture archiving and communication system (PACS) workstation (Carestream solution working station, Carestream Health, Version 11.0, Rochester, NY, USA) and the dorsal and volar parts of the SLIL and LTIL were classified as torn or not torn. The final decision was reached by consensus. Images were also assessed individually by these two radiologists Table 1

Visibility of dorsal and volar parts of the scapholunate interosseous ligament (SLIL) and lunotriquetral interosseous ligament (LTIL)

Detection of tears of dorsal and volar parts of SLIL and LTIL The presence or absence of a complete tear in the dorsal and volar fibres of the scapholunate interosseous ligament (SLIL) and lunotriquetral ligaments (LTIL) was documented. The final diagnosis of tear on CT and MR imaging was reached by consensus. The CT or MR criterion used for diagnosis of a ligament tear was visualization of a full-thickness defect in the ligament with fluid communication from one joint space to another. We used a modified version of the five-point scale established by Scheck et al. [9] to grade overall ligament integrity. This score reflects relative diagnostic confidence and was defined as follows 1 = certainly normal; 2 = probably normal; 3 = equivocal; 4 = probably torn; and 5 = certainly torn. The CT and MRA findings were correlated with the findings on wrist arthroscopy. CTA and MRA findings were categorized as true positive (TP), false positive (FP), true

MRI sequences

Before traction Fat-saturated T1-weighted turbo spin-echo (TSE) sequence Fat-saturated T1-weighted turbo spin-echo (TSE) sequence Fat-saturated T1-weighted turbo spin-echo (TSE) sequence Fat-saturated T1-weighted turbo spine-echo (TSE) sequence Fat-saturated T1-weighted turbo spine-echo (TSE) sequence)

Plane

Slice thickness (mm) TR (ms) TE (ms) Flip angle (°) FOV (mm) Matrix

Axial

2.5

678

20

90

80x80

228x182

Coronal

1.5

679

19

90

80x80

228x182

Sagittal

1.5

516

12

90

80x80

228x182

Oblique Ax (SLIL) 2

543

20

90

80x80

228x182

Oblique Ax (LTIL) 2

543

20

90

80x80

228x182

SLIL scapholunate interosseous ligament, LTIL lunotriquetral interossoeous ligament

Eur Radiol Fig. 2 Coronal localizer for the axial oblique plane. Oblique axial images were obtained parallel to the (a) SLIL and (b) LTIL. SLIL scapholunate interosseous ligament, LTIL lunotriquetral interosseous ligament

used to analyse agreement for grading of ligament visibility and grading of ligament integrity. The following agreement criteria was applied: k > 0.8 excellent, k:0.6–0.8 good, k:0.4– 0.6 moderate, k:0.2–0.4 fair, k < 0.2 poor.[14]

negative (TN) or false negative (FN) by summarizing categories 1 (certainly normal), 2 (probably normal) and 3 (not possible for assessment but probably normal) as ‘normal’ = ‘no tear’ and categories 4 (probably pathological) and 5 (certainly pathological) as ‘pathological’ = ‘tear’. Statistical analysis

Results All computations were performed using SPSS (Chicago, IL, USA). Differences with P < 0.05 were considered statistically significant. The sensitivity, specificity and accuracy of the two different planes (axial and oblique axial) on each modality (CT and MRI) to detect dorsal and volar SLIL and LTIL tears were calculated as defined previously. Kappa statistics were Table 2 The visibility of the dorsal and volar parts of the scapholunate interosseous ligament (SLIL) and lunotriquetral interosseous ligament (LTIL) on CT (CTA) and MR (MRA) arthrography in the true axial and oblique axial planes for nine wrists

Ligament visibility SLIL Dorsal fibres Volar fibres

Overall (dorsal + volar fibres) LTIL Dorsal fibres Volar fibres

Overall (dorsal + volar fibres)

Arthroscopy Arthroscopy revealed two complete dorsal and two complete volar SLIL tears, three complete dorsal and three complete volar LTIL tears.

CTA axial

CTA oblique axial

MRA axial

MRA oblique axial

Good Fair Poor Good Fair Poor Good Fair Poor Good

56 0 44 67 0 33 61 0 39 56

% % % % % % % % % %

33 0 67 33 0 67 67 0 33 78

% % % % % % % % % %

44 % 12 % 44 % 56 % 22 % 22 % 50 % 17 % 33 % 11 %

33 0 67 33 0 67 67 0 33 78

% % % % % % % % % %

Fair Poor Good Fair Poor Good Fair Poor

22 22 56 11 33 56 28 16

% % % % % % % %

0 22 78 0 22 78 0 22

% % % % % % % %

22 % 67 % 11 % 11 % 78 % 11 % 17 % 72 %

0 22 78 0 22 78 0 22

% % % % % % % %

Eur Radiol Table 3 The number of true-positive (TP), true-negative (TN), false-positive (FP) and false-negative (FP) tears in the dorsal and volar parts of the scapholunate interosseous ligament (SLIL) and CTA axial TP/TN/FP/FN

lunotriquetral interosseous ligament (LTIL) on CT (CTA) and MR (MRA) arthrography in the true axial and oblique axial planes for nine wrists

CTA oblique axial TP/TN/FP/FN

MRA axial TP/TN/FP/FN

MRA oblique axial TP/TN/FP/FN

SLIL

Dorsal tear

2/6/1/0

3/6/0/0

2/6/0/1

2/6/0/1

LTIL

Volar tear Dorsal tear

3/6/0/0 2/7/1/0

3/6/0/0 3/6/0/0

2/6/1/0 0/6/1/2

3/6/0/0 2/7/0/0

Volar tear

1/6/0/1

3/6/0/0

0/7/0/2

1/6/1/1

Interobserver and intra-observer agreement

Detection of tears of dorsal and volar parts of SLIL and LTIL

There are good interobserver (k: 0.68, CI: 0.50–0.86) and intra-observer (k: 0.79, CI: 0.69–0.89) agreement of ligament visibility. There are also good inter-observer (k: 0.60, CI: 0.43–0.77) and intra-observer (k: 0.71, CI: 0.62–0.80) agreement of ligament integrity. Visibility of dorsal and volar parts of SLIL and LTIL For the scapholunate ligament, the oblique axial plane provided as good visibility of the dorsal and volar parts of SLIL and LTIL on both MRA and CTA compared to the axial plane (Table 2). For the LTIL, oblique axial imaging appreciably improved visibility of both the dorsal and the volar parts on MRA with ‘good’ visibility increased from overall 11 % to overall 78 % (Table 2). Similarly, good ligament visibility increased from overall 56 % for axial imaging to overall 78 % for oblique axial imaging for CTA (Table 2). Table 4 The sensitivity, specificity and accuracy of CT (CTA) and MR (MRA) arthrography in revealing scapholunate (SLIL) and lunotriquetral (LTIT) tears in the true axial and oblique axial planes for nine wrists

Ligament tear detection SLIL Dorsal fibres Volar fibres

Overall (dorsal + volar fibres) LTIL Dorsal fibres Volar fibres

Overall (dorsal + volar fibres)

The true positive (TP), false positive (FP), true negative (TN) or false negative (FN), sensitivity, specificity and accuracy for CTA and MRA for detecting dorsal and volar SLIL and LTIL tears as defined previously are shown in Table 3 and Table 4. In all instances, MRA and CTA accuracy in detecting SLIL and LTIL tear was higher using the oblique axial plane. The overall accuracy of CTA and MRA in the oblique axial plane ranged from 89 % to 100 % compared to 72 % to 94 % for the axial plane (Table 4). CTA had a higher accuracy in detecting dorsal and volar SLIL and LTIL tears than MRA. Figures 3a–d and 4a–d show the normal SLIL and LTIL in true axial and oblique axial planes on both CTA and MRA. Figure 5a–b shows a false-positive dorsal SLIL tear on MRA axial plane which was correctly recognised as intact (true negative) on oblique axial imaging. Figure 5c–d shows a false-

CTA axial

CTA oblique axial

MRA axial

MRA oblique axial

Sensitivity Specificity Accuracy Sensitivity Specificity Accuracy Sensitivity Specificity Accuracy Sensitivity

100 86 89 100 100 100 100 92 94 100

% % % % % % % % % %

100 100 100 100 100 100 100 100 100 100

% % % % % % % % % %

67 100 89 100 86 89 80 92 89 0

% % % % % % % % % %

67 100 89 100 100 100 83 100 94 100

% % % % % % % % % %

Specificity Accuracy Sensitivity Specificity Accuracy Sensitivity Specificity Accuracy

88 90 50 100 88 75 93 89

% % % % % % % %

100 100 100 100 100 100 100 100

% % % % % % % %

86 67 0 100 78 0 93 72

% % % % % % % %

100 100 50 86 78 75 93 89

% % % % % % % %

Eur Radiol Fig. 3 (a, b) CT arthrography of SLIL in (a) axial and (b) oblique axial planes. Both the dorsal (black arrow) and volar (white arrow) fibres are well delineated. The scaphoid (S) and lunate (L) are labeled. SLIL scapholunate interosseous ligament. (c, d) MR arthrography of SLIL in (c) axial and (d) oblique axial planes. Both the dorsal (black arrow) and volar (white arrow) fibres are well delineated. The scaphoid (S) and lunate (L) are labeled. SLIL scapholunate interosseous ligament

positive dorsal LTIL tear on MRA axial plane which was correctly recognised as intact (true negative) on oblique axial imaging. Figure 5e–f shows a false-negative volar LTIL tear on MRA axial plane which was correctly Fig. 4 (a, b) CT arthrography of LTIL in (a) axial and (b) oblique axial planes. Both the dorsal (black arrow) and volar (white arrow) fibres are better delineated in the oblique axial plane. The lunate (L) and triquetral (T) are labeled. (c, d) MR arthrography of LTIL in (c) axial and (d) oblique axial planes. Both the dorsal and volar fibres are better delineated in the oblique axial plane. The visibility of the dorsal and volar part was classified as ‘intermediate’ in the axial plane and ‘good’ in the oblique axial plane. The lunate (L) and triquetral (T) are labeled. LTIL lunotriquetral interosseous ligament

recognised as torn (true positive) on oblique axial imaging. Figure 6a–b shows a false-negative dorsal SLIL tear on CTA axial plane which was correctly recognised as torn (true positive) on oblique axial imaging.

Eur Radiol

Fig. 5 (a, b) MR arthrography of SLIL in (a) axial and (b) oblique axial planes. The dorsal part (white arrows) of the SLIL was categorized as ‘probably torn’ in the axial plane and as ‘certainly intact’ in the oblique axial plane. Arthroscopy confirmed intact dorsal fibres of the SLIL. This is regarded as a false positive for the axial plane and a true negative for the oblique axial plane. The visibility of the dorsal part was classified as poor in the axial plane and good in the oblique axial plane. The scaphoid (S) and lunate (L) are labeled. (c, d) MR arthrography of LTIL in (c) axial and (d) oblique axial planes. The dorsal fibres (white arrows) of the LTIL were categorized as probably torn in axial plane and as certainly intact in the oblique axial plane. Arthroscopy confirmed intact dorsal fibres of the LTIL. This was regarded as a false positive for the axial plane and a

true negative for the oblique axial plane. The visibility of dorsal fibre was classified as poor in the axial plane and good in the oblique axial plane. The lunate (L) and triquetral (T) are labeled. (e, f) MR arthrography of LTIL in (e) axial and (f) oblique axial planes. The volar fibres (white arrows) of the LTIL were categorized as probably intact in the axial plane and as certainly torn in the oblique axial plane. Arthroscopy confirmed a tear of the volar fibres of the LTIL. This was regarded as a false negative for the axial plane and a true positive for the oblique axial plane. The visibility of the volar fibre was classified as intermediate in the axial plane and poor in the oblique axial plane. The lunate (L) and triquetral (T) are labeled. SLIL scapholunate interosseous ligament, LTIL lunotriquetral interosseous ligament

Discussion

injured intrinsic ligaments of the wrist [4–9]. The structurally relevant parts of the SLIL and the LTIL are the collagenous dorsal and volar parts and it is these parts that have been assessed in this study. Isolated tear of either the dorsal or volar parts of these ligaments is considered to be a likely cause of chronic wrist pain [15]. Immobilization, functional splinting, therapeutic injection and physiotherapy or surgical repair are the treatments of choice for intrinsic ligament tears [15]. The likely reason for lessened visibility of the SLIL and the LTIL on standard MR imaging is that the scapholunate joint and, more particularly, the lunotriquetral joint lies oblique to the true axial plane which is used to examine the wrist on routine MR imaging protocols. This standard axial plane is

Structural imaging of the wrist with MRI has improved greatly since the first clinical MRI scanners were introduced in the early 1990s. The reported accuracy of MRI for detecting tears in structures such as a triangular fibrocartilaginous complex is over 90 % [4–9]. Although it is appreciated that the accuracy of standard axial MR in detecting tears of the dorsal and volar parts of SLIL and LTIL of the wrist is high, the purpose of this study was to determine if oblique axial imaging of the wrist could further increase this accuracy. The scapholunate and lunotriquetral interosseous ligaments (SLIL and LTIL) are the more structurally important and also the more commonly

Eur Radiol

Fig. 6 CT arthrography of SLIL in (a) axial and (b) oblique axial reformatted planes. The dorsal fibres (white arrows) of the SLIL were categorized as certainly torn in the axial plane and as certainly intact in the oblique axial plane. Arthroscopy confirmed a tear of the dorsal fibre of the

SLIL. This is regarded as true positive for axial plane and false negative in oblique axial plane. The visibility of dorsal fibre was classified as intermediate in axial plane and good in oblique axial plane. The scaphoid (S) and lunate (L) are labeled. SLIL scapholunate interosseous ligament

perpendicular to the long axis of the radius and ulna. However, since the proximal carpal row is C-shaped, the corresponding SLIL and LTIL are aligned obliquely to a true axial plane. This study showed that oblique axial imaging, scanning perpendicular to the articulation of the scapholunate and lunotriquetral joints, significantly improved visibility of the normal ligament as well as the detection of tears in the SLIL and LTIL on both CTA and MRA. Oblique axial imaging had a greater beneficial effect on ligament visibility and tear detection in the LTIL than the SLIL because the lunotriquetral joint aligns more obliquely to the standard axial plane and hence ligament visibility and detection of tear is therefore more difficult in this joint than the scapholunate joint with standard axial imaging. Robinson et al. [10] showed that axial oblique MR sequences helped identify tears of the intrinsic ligaments, particularly the LTIL. All LTIL tears were confidently seen with oblique axial imaging in the study of Robinson et al., whereas none could be definitely be identified with standard axial imaging. In the study of Robinson et al. the sensitivity/specificity of MR oblique axial in detecting SLIL tears were 100 %/92 % versus 83 %/88 % for true axial imaging while the sensitivity/ specificity of MR oblique axial in detecting LTIL tears were 100 %/100 % versus 0 %/100 % for true axial imaging [10]. These findings are similar to the findings in our study. The main difference between the current study and that of Robinson et al. is that 3 T MRI with a dedicated wrist coil was used in this study rather than 1.5-T MR imaging and that we also had arthroscopic correlation. This is also the first study to evaluate oblique axial reformatting on CT to detect tears of the SLIL and LTIL during CTA. We found 100 % accuracy for detecting tears in the SLIL and the LTIL on reformatted oblique axial planes. In general, the accuracy of reformatted oblique axial imaging on CTA performed better than direct oblique axial imaging with MRA due most likely to the improved spatial resolution and interactive nature of CT reformation.

The main limitations to this study were the use of cadaveric wrists and the small number of wrists included. Also, as we did not include volumetric (3D) MR imaging in the examination protocol, the diagnostic performance of reformatted oblique axial MR imaging for both the normal and torn scapholunate and lunotriquetral ligaments was not tested. We also did not assess the role of indirect MR arthrography in the current study. Conventional arthrography, CTA and MRA were performed one after another followed by arthroscopy. The time span between these investigations was variable and, as a result, the degree of joint distension for CTA or MRA may have varied due to contrast leakage. In conclusion, CT and MR arthrography in the oblique axial plane was more accurate at visualizing and detecting tears of the dorsal and volar scapholunate and lunotriquetral interosseous ligaments compared to standard axial imaging. This improvement in ligament tear detection was greatest for lunotriquetral interosseous ligament tears. Acknowledgments The scientific guarantor of this publication is Ryan Ka Lok Lee. The authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article. The work described in this paper was partially supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project No.SEG_CUHK02). One of the authors has significant statistical expertise. Institutional Review Board approval was obtained. Written informed consent was obtained from all subjects (patients) in this study. No study subjects or cohorts have been previously reported. Methodology: prospective, case-control study, performed at one institution.

References 1.

2.

Lee DH, Dickson KF, Bradley EL (2004) The incidence of wrist interosseous ligament and triangular fibrocartilage articular disc disruptions: a cadaveric study. J Hand Surg [Am] 29:676–684 Klempka A, Wagner M, Fodor S, Prommersberger KJ, Uder M, Schmitt R (2016) Injuries of the scapholunate and lunotriquetral ligaments as well as the TFCC in intra-articular distal radius

Eur Radiol

3. 4. 5.

6.

7.

8.

fractures. Prevalence assessed with MDCT arthrography. Eur Radiol 26:722–732 Berger RA (2001) The anatomy of the ligaments of the wrist and distal radioulnar joints. Clin Orthop Relat Res 383:32–40 Magee T (2009) Comparison of 3-T MRI and arthroscopy of intrinsic wrist ligament and TFCC tears. AJR Am J Roentgenol 192:80–85 Lee YH, Choi YR, Kim S, Song HT, Suh JS (2013) Intrinsic ligament and triangular fibrocartilage complex (TFCC) tears of the wrist: comparison of isovolumetric 3D-THRIVE sequence MR arthrography and conventional MR image at 3T. Magn Reson Imaging 31:221–226 Lee RK, Ng AW, Tong CS et al (2013) Intrinsic ligament and triangular fibrocartilage complex tears of the wrist: comparison of MDCT arthrography, conventional 3-T MRI, and MR arthrography. Skelet Radiol 42:1277–1285 Moser T, Dosch JC, Moussaoui A, Dietemann JL (2007) Wrist ligament tears: evaluation of MRI and combined MDCT and MR arthrography. AJR Am J Roentgenol 188:1278–1286 Schmid MR, Schertler T, Pfirrmann CW et al (2005) Interosseous ligament tears of the wrist: comparison of multi-detector row CT arthrography and MR imaging. Radiology 237:1008–1013

9.

10.

11. 12. 13.

14. 15.

Scheck RJ, Romagnolo A, Hierner R, Pfluger T, Wilhelm K, Hahn K (1999) The carpal ligaments in MR arthrography of the wrist: correlation with standard MRI and wrist arthroscopy. J Magn Reson Imaging 9:468–474 Robinson G, Chung T, Finlay K, Friedman L (2006) Axial oblique MR imaging of the intrinsic ligaments of the wrist: initial experience. Skelet Radiol 35:765–773 Slutsky DJ (2012) Current innovations in wrist arthroscopy. J Hand Surg [Am] 37:1932–1941 Ekman EF, Poehling GG (1994) Principle of arthroscopy and wrist arthroscopy equipment. Hand Clin 10:557–566 Geissler WB, Freeland AE, Savoie FH, McIntyre LW, Whipple TL (1996) Intracarpal soft-tissue lesions associated with an intraarticular fracture of the distal end of the radius. J Bone Joint Surg Am 78:357–365 Landis JR, Koch GG (1977) The measurement of observer agreement for categorical data. Biometrics 33:159–174 Lee JI, Nha KW, Lee GY, Kim BH, Kim JW, Park JW (2012) Longterm outcomes of arthroscopic debridement and thermal shrinkage for isolated partial intercarpal ligament tears. Orthopedics 35: e1204–e1209