Eur Radiol DOI 10.1007/s00330-015-3638-y

ULTRASOUND

The effect of percutaneous ultrasound-guided subacromial bursography using microbubbles in the assessment of subacromial impingement syndrome: initial experience Xueqing Cheng & Man Lu & Xiao Yang & Xuanyan Guo & Fanding He & Qin Chen & Peng Gu

Received: 6 November 2014 / Revised: 14 January 2015 / Accepted: 21 January 2015 # European Society of Radiology 2015

Abstract Objective The purpose of this study was to determine the feasibility and the value of percutaneous ultrasound-guided subacromial bursography (PUSB) with contrast-enhanced ultrasound (CEUS) for assessment of rotator cuff tears. Methods Between January 2012 and September 2014, 2-D US and PUSB were prospectively performed in 135 patients with suspected rotator cuff tears from among 2,169 patients with clinically diagnosed subacromial impingement syndrome. Sixty-three patients who had undergone arthroscopy were included. The PUSB findings were compared with those from 2-D US using the McNemar test, with arthroscopy as a standard. Results Of a total of 19 full-thickness tears (FTTs), 2-D US correctly diagnosed 12 and PUSB correctly diagnosed 18 (P= 0.031). With regard to partial-thickness tears, 2-D US correctly diagnosed 31 and PUSB 35 of a total of 41 tears (P= 0.375).Accuracy in diagnosing FTTs was 81.0 % for 2-D US and 98.4 % for PUSB. 2-D US and PUSB yielded sensitivity of 63.2 % and 94.7 %, respectively, for full-thickness tears, with specificity of 88.6 % and 100.0 %, respectively.

X. Cheng : M. Lu (*) : X. Guo : F. He : Q. Chen Department of Ultrasound, Sichuan Academy of Medical Sciences & Sichuan Provincial People’s Hospital, Chengdu, China e-mail: [email protected] X. Cheng North Sichuan Medical College, Nanchong, China X. Yang Department of Orthopedics, Sichuan Academy of Medical Sciences & Sichuan Provincial People’s Hospital, Chengdu, China P. Gu Affiliated Hospital of North Sichuan Medical College, Nanchong, China

Conclusions PUSB is a safe and feasible procedure, with greater sensitivity and specificity than 2D US in diagnosing FTTs As such, PUSB improves the value of ultrasound for assessing rotator cuff pathology. Key Points • PUSB allows new application of CEUS and offers a new form of bursography. • PUSB appears to facilitate diagnosis of full -thickness rotator cuff tears. • PUSB outlines the rotator cuff tears more clearly than 2-D US. Keywords Rotator cuff tears . Subacromial impingement syndrome . Contrast-enhanced ultrasound . Bursography . Ultrasound

Introduction Subacromial impingement syndrome (SIS) is a common cause of shoulder pain and disability, especially among athletes and the elderly. The underlying causes of SIS are still not clearly understood. Many authors have ascribed both extrinsic and intrinsic factors to the development of SIS, including acromial morphology, muscle weakness, degenerative tendinopathy, and overuse of the shoulder [1, 2]. Clinically, SIS represents a spectrum of pathology ranging from subacromial bursitis to full-thickness rotator cuff tears [3, 4], which is characterized by shoulder pain and physical examination findings such as the Neer and Hawkins-Kennedy impingement signs and the painful arc sign [5]. Various imaging modalities are routinely used in the evaluation of SIS, including plain radiographs, magnetic

Eur Radiol

resonance imaging (MRI), and sonography, which are often helpful in demonstrating supportive findings for SIS, discovering alternative causes of shoulder pain, and disease staging. Radiographs are generally used for the initial evaluation, and are employed to delineate the bony anatomy of the shoulder, such as acromial morphology, and to look for evidence of os acromiale, which may lead to secondary impingement. Both MRI and ultrasound have high soft tissue resolution, and both are sensitive and specific in the evaluation of rotator cuff pathology. MRI capitalizes on areas where ultrasound could be limited, including bones marrow, articular cartilage, and labrum. However, major limiting factors for the use of MRI include long image acquisition time and contraindication in patients with non-MR-compatible hardware. In contrast, ultrasound is time-efficient, readily available, low-cost, and able to provide dynamic imaging, all of which make it suitable for image-guided therapeutic intervention. Contrast-enhanced ultrasound (CEUS) has been widely used to characterize the vascularity of lesions in organs of interest. Kurihara et al. [6] demonstrated that the perfusion characteristics seen with contrast-enhanced sonography reflect tumour blood supply, which is helpful in differentiating benign from malignant lesions. Hence, published guidelines have included the use of transabdominal CEUS for workup of liver, pancreatic, and kidney lesions [7]. In these procedures, contrast agents are often injected intravenously. More recently, many authors have investigated the intracavitary use of ultrasound contrast agents—for example, in the evaluation of obstructive bile duct diseases, tubal patency, and vesicoureteral reflux—with promising results [8, 9]. In this study, we injected the contrast agent mixture into the subacromial bursa of patients with SIS under sonographic guidance in order to observe the distribution of the mixture. To our knowledge, this is the first study to investigate the value of percutaneous ultrasound-guided subacromial bursography (PUSB) for assessing rotator cuff tears in patients with SIS by injecting ultrasonic contrast medium into the subacromial bursa.

30 days; mean, 18.9 days) of the PUSB procedure. None of the 135 patients had had previous surgery on the shoulder area or a history of uncontrolled chronic diseases such as malignant neoplasms, hypocoagulability, or infection. The group comprised 71 male and 64 female patients, ranging in age from 23 to 63 years (mean, 53.8 years), with a mean duration of 10.8 months (3–24 months) of shoulder pain. Institutional review board approval was obtained for this prospective study in correlation with arthroscopy findings. Written informed consent was obtained from all patients. Equipment The conventional 2-D and PUSB examinations were performed with a Philips iU Elite ultrasound system (Philips Medical Systems, Bothell, WA, USA) equipped with contrast-tuned imaging technology and analysis software. A 5–12 MHz transducer probe was used for conventional grayscale ultrasound, while a 3–9 MHz linear array transducer was used for the PUSB procedure. SonoVue (Bracco SpA, Milan, Italy) was used as an intrabursal contrast agent. SonoVue is a second-generation blood pool contrast-enhancing agent consisting of microbubbles containing a sulfur hexafluoride gas stabilized by a phospholipid shell [10]. Procedure 2-D ultrasound For shoulder imaging, the operator faced the patient, who was seated on a revolving stool, and the procedure was performed according to the shoulder ultrasound technical guidelines recommended by the European Society of Musculoskeletal Radiology [11].The long head of the biceps tendon, subscapularis tendon, supraspinatus tendon, subacromialsubdeltoid bursa, coracoacromial ligament, infraspinatus and teres minor tendons, and acromioclavicular joint were each scanned in turn. PUSB

Materials and methods Patient population Two hundred fifty-six patients (total of 276 shoulders) with suspected rotator cuff tears were recruited from among 2,169 consecutive patients with a clinical diagnosis of SIS by an orthopaedic surgeon, who were then referred to ultrasound at a general hospital between January 2012 and September 2014. One hundred thirty-five patients had undergone both 2-D ultrasound (US) and PUSB of their unilateral affected shoulder. Sixty-three of the 135 affected shoulders (47 %) had subsequently undergone arthroscopy within 1 month (range, 7 to

Immediately after US, percutaneous US-guided subacromial bursography (PUSB) was performed using the preset Bcontrast general^ mode. First, the operator scanned the affected shoulder to locate the thickest area of the subacromialsubdeltoid (SASD) bursa in order to determine the best injection point and approach. Contrast pulse sequencing) (CPS) was then initiated, and after sterilization of the skin on the lateral side of the affected arm, the contrast mixture was slowly injected into the subacromial bursa under gray-scale ultrasound guidance. CEUS imaging was observed simultaneously. The contrast mixture comprised 0.5 ml of SonoVue (sulfur hexafluoride-based microbubbles at 8 μl (45 μg)/ml and

Eur Radiol

2.5 μm in diameter), 0.5 ml (20 mg/ml) lidocaine hydrochloride, 1.0 ml (5+2 mg/ml) compound betamethasone, and 0.5 ml (9 mg/ml) sodium chloride. After the injection, the patient was asked to gently exercise the shoulder in order to distribute the mixture through the bursa, and the probe was moved in order to visualize the distribution of contrast mixture. The MI (mechanical index) value displayed on the screen was less than 0.2. Real-time contrast imaging was captured continuously for 5 minutes and stored on the hard drive of the system for documentation and analysis. All patients were reviewed 30 minutes after the PUSB procedure.

involvement and the size and location of the tear. Fullthickness tears appeared as non-visualization of the cuff, flattening of the bursal surface of the tendon, thinning of the cuff, irregularity of the greater tuberosity, joint effusion, SASD bursa effusion, and herniation of the deltoid muscle in the cuff [12–15]. Partial-thickness tears primarily appeared as a localized hypoechoic area only partially affecting tendon thickness, and the hypoechoic appearance did not change on short- or long-axis scans or by tilting the transducer over the tendon [15, 16]. The level of the insertion of the supraspinatus tendon at the greater tuberosity, a common site of rotator cuff tear, was carefully scanned in order to avoid anisotropic artefacts.

Image analysis PUSB imaging 2-D ultrasound imaging The conventional 2-D ultrasound was performed and evaluated independently by an experienced radiologist (X.Y.G) with more than 5 years of experience in musculoskeletal ultrasound, who was blinded to the clinical data except for the age and sex of each patient. The thickness of subacromial-subdeltoid (SASD) bursa was measured, including the hypoechoic fluid within the bursa and the two layers of echogenic peribursal fat stripes. A normal SASD bursa appeared as a hypoechoic line between two hyperechoic planes, with total thickness of less than 2 mm. In the case of bursitis, fluid effusion or/and thickening of the SASD bursa were visualized. For assessment of the rotator cuff, including the supraspinatus, infraspinatus, subscapularis, and teres minor tendons, we carefully evaluated the shape, size, echogenicity, and integrity—especially at the level of the insertion of the supraspinatus tendon at the greater tuberosity—to determine any rotator cuff abnormalities. Tendinosis, which was thought to be an early result of anterosuperior impingement [4], appeared as areas of thickening and heterogeneous hypoechoicity. We were able to classify rotator cuff tears as partial- or full-thickness tears based on the extent of tendon

Fig 1 PUSB imaging (dualframe grayscale and contrastenhanced sonograms) shows an intact supraspinatus tendon, and the hyperechoic contrast agent mixture fills subacromial bursa (arrow) well without leaking into the tendon. SUP supraspinatus tendon, HH humeral head

The percutaneous US-guided subacromial bursography was performed and interpreted independently by a second radiologist (M.L.), with 10 years of experience in musculoskeletal ultrasound and 7 years of experience in CEUS. She was also blinded to both 2-D US findings and the clinical data, except for the age and sex of each patient. The distribution of the contrast mixture was dynamically observed on PUSB imaging to detect whether the cuff was intact and to differentiate full-thickness from partialthickness tears. The size and location of the tear were also observed. The results were categorized into no tear (NT), partial-thickness tear (PTT), and full-thickness rotator cuff tear (FTT). For an intact cuff, the hyperechoic contrast mixture was dispersed only within the bursa outlining a regular superior cuff surface (Fig. 1). For bursal side partial-thickness tears, PUSB showed that the contrast mixture filled the torn portion of the bursal side of the rotator cuff, appearing as an ulcer-like crater in the supersurface of the rotator cuff on CEUS imaging. Although an incomplete tear on the articular side of the rotator cuff theoretically cannot be visualized on CEUS imaging, it could be visualized on side-by-side gray-scale ultrasonography. For full-thickness tears, PUSB showed that the

Eur Radiol

glenohumeral joint, including the sheath of the long head of the biceps tendon, was filled with contrast mixture that extravasated into the joint through the cuff defect. The presence or absence of contrast agent leaking into the cuff and/or glenohumeral joint is a key point in distinguishing an intact cuff, full-thickness tear, and partial-thickness tear. Arthroscopy One orthopaedic surgeon (X.Y.) performed all arthroscopic examinations and operative procedures. The presence or absence of a full- or partial-thickness tear was noted on arthroscopy. The arthroscopy findings were used as the gold standard in this study. Statistical analysis Statistical analyses were performed using a standard statistical software package (SPSS 13.0; SPSS Inc., Chicago, IL, USA). The 2-D US and PUSB findings were correlated with the respective arthroscopy findings. The predictive values (sensitivity, specificity, positive predictive value, negative predictive value, and Youden's index) and the accuracy of 2-D US and PUSB in detecting full-thickness cuff tears were calculated using arthroscopy as the gold standard. Data from different imaging techniques were compared using a McNemar test (α=0.05, two-sided). P values less than 0.05 were considered statistically significant.

Results Percutaneous ultrasound-guided subacromial bursography procedures were successfully performed in all 135 patients. Patients complained of no discomfort, with the exception of slight pain when the needle penetrated the subcutaneous tissue. All patient were reviewed 30 minutes after the bursography procedure, and there were no hyperechoic contrast agents in the SASD or the glenohumeral joints. No severe complications requiring further treatment, such as inflammatory arthritis, occurred either during or after the procedure. Among the 63 shoulders that had undergone arthroscopy, there were 19 full-thickness tears (FTTs), 41 partial-thickness tears (PTTs, 22 on the articular side, 19 on the bursal side), and three intact rotator cuffs confirmed by arthroscopy (Table 1). Among the patients with an intact cuff, one patient had a superior labrum anterior-to-posterior (SLAP) lesion, one had supraspinatus tendinitis, and one had degenerative fraying on the articular side of the cuff. Ultrasound correctly identified 12 (63.2 %) of the 19 FTTs, and misdiagnosed 7 FTTs as PTTs. US suggested a bursal-sided partial-thickness tear in five patients that was found to be a full-thickness tear on arthroscopy,

Table 1 Comparison of US and PUSB in detecting rotator cuff tears with arthroscopy as a standard Arthroscopy

US FTT

PUSB PTT

NT

FTT

Total PTT

NT

FTT

12

7

0

18

0

1

19

PTT NT Total

5 0 17

31 2 40

5 1 6

0 0 18

35 1 36

6 2 9

41 3 63

FTT full-thickness rotator cuff tear, PTT partial-thickness tear, NT no tear

while PUSB correctly identified all full-thickness tears but missed one. The detection rate of PUSB was statistically higher than that of US for full-thickness tears (P=0.031) (Table 2). The sensitivity and specificity of US and PUSB compared with arthroscopy in detecting full-thickness tears was 63.2 % (95 % CI, 41.5–84.9 %) and 88.6 % (95 % CI, 88.2–89.0 %), and 94.7 % (95 % CI, 84.7–104.7 %) and 100.0 %, respectively. The accuracy of US and PUSB was 81.0 % (95 % CI, 71.3– 90.6 %), 98.4 % (95 % CI, 95.3–101.5 %), respectively, in diagnosing full–thickness tears (Table 3). Ultrasound accurately diagnosed 31 of 41 partialthickness tears as seen on arthroscopy, missed four articular-side and one bursal-side partial-thickness tear as normal, and overcalled two intact cuffs as partialthickness tears. PUSB accurately detected 35 (85.4 %) of 41 partial-thickness tears (Fig. 2), missed the same four articular-side and one bursal-side partial-thickness tears as US, and missed another small bursal-side incomplete tear. There was no statistical difference, however, between US and PUSB in diagnosing partial-thickness tears (P = 0.375). On the 43 rotator cuff tears accurately detected by both 2-D US and PUSB, the tears were clearly outlined by the contrast mixture, which made for better imaging quality with PUSB than 2-D US (Figs. 3 and 4). Many of the patients had additional abnormalities (other than rotator cuff tears) on US, PUSB, and arthroscopy examination. Findings other than rotator cuff tears were not assessed in this study.

Table 2 Comparison of US and PUSB in detecting full-thickness and partial-thickness rotator cuff tear with arthroscopy as a standard Method

FTTs, n (%)

PTTs, n (%)

US PUSB P

12/19 (63.2 %) 18/19 (94.7 %) 0.031

31/41 (75.6 %) 35/41 (85.4 %) 0.375

Eur Radiol Table 3

Performance of US and PUSB in the diagnosis of full-thickness rotator cuff tears

US (CI) PUSB (CI)

Sensitivity (%)

Specificity (%)

PPV (%)

NPV (%)

Youden’s index

Accuracy (%)

63.2 (41.5,84.9) 94.7 (84.7,104.7)

88.6 (88.2,89.0) 100.0 –

70.6 (67.8,73.4) 100.0 –

84.8 (82.4,87.2) 97.8 (86.5,109.1)

0.518 (0.257,0.778) 0.947 (0.847,1.036)

81.0 (71.3,90.6) 98.4 (95.3,101.5)

Discussion With the development of ultrasonic technology, including high-frequency transducers, ultrasound has been widely used as a basic imaging modality to diagnose shoulder disorders including subacromial bursitis, calcific tendinitis, and rotator cuff tears. Ultrasound can accurately depict rotator cuff pathology, and can also be used to dynamically assess for impingement through range of motion. As such, it is best-suited for evaluation of rotator cuff tears in patients with primary impingement. It may be difficult to differentiate tendinosis, partial-thickness, and full-thickness tears, especially for less experienced musculoskeletal radiologists and for complex cases. The results of this study showed accuracy of 81.0 % (95 % CI, 71.3–90.6 %) for US, which was lower than the reported 96 % [15, 16], and 98.4 % (95 % CI, 95.3–101.5 %) for PUSB in the diagnosis of full-thickness tears compared with arthroscopy. In the present study, the difference in findings between US and PUSB in assessing full-thickness tears was statistically significant (P=0.031). Both the sensitivity and specificity of PUSB in diagnosing full-thickness tears were higher than those of 2-D US. PUSB was theoretically superior in comparison to US for diagnosis of bursal-side partial-thickness tears. Practically, in this study, US underdiagnosed four bursal-side partial-thickness tears as normal, which were detected by PUSB and confirmed by arthroscopy. There was no statistically significant difference between US and PUSB, however, in assessing partial-thickness tears (P=0.375). We attributed this to the possibility that articular-side partial-thickness tears may be not influenced by PUSB, but these cases were

Fig 2 A bursa side partial-thickness tear in a 55-year-old woman. A 2-D US showed a focal hypoechoic tendon which cannot differentiate partialthickness tears from tendinosis (arrows). B PUSB showed that CA filled

included in the Bpartial-thickness tears^ group. PUSB also accurately diagnosed six full-thickness tears that were misdiagnosed by US. This is most likely due to fibrosis and scarring at the tendon tear site simulating an intact tendon or an incomplete tear on conventional US. The fluid distension provided by PUSB allows contrast agent to enter this area and leak into the joint, highlighting the tear on CEUS imaging. PUSB missed one atypical full-thickness and two small bursal-side partial-thickness tears compared to arthroscopy. There was little contrast mixture that entered the distal part of the complete adhesion area of the bursa in these three missed cases, which may have resulted the lack of detection of the distal cuff tears. PUSB also misdiagnosed a degenerative fraying on the articular side of the rotator cuff as an articular-side incomplete tear, and missed four articular-side incomplete tears. PUSB showed no improved diagnostic capability for articular rotator cuff partial-thickness tears compared with 2-D US. The PUSB technique allows the demonstration of partialthickness bursal surface and full-thickness cuff tears, and this method appears to have a higher diagnostic yield in fullthickness rotator cuff tears. In addition, PUSB was able to show the area of the cuff and the retracted torn tendon, as outlined by contrast material, more clearly than 2-D US, which not only increases the diagnostic confidence of the radiologist but also helps the surgeon in planning the operative approach and predicting the final outcome. PUSB may also provide an excellent alternative for those patients who are unable to undergo MRI to confirm cuff tears, such as patients with cardiac pacemakers and certain metal implants or who are claustrophobic. Disadvantages of PUSB include the need

and outlined the tear area of the tendon (arrows). SUP supraspinatus tendon HH humeral head

Eur Radiol Fig 3 A full-thickness tear in a 56-year-old woman was detected by 2-D US (A) and PUSB (B). PUSB outlined the tear area more clearly than 2-D US (arrows). HH humeral head

for a needle injection into the bursa and the increased cost compared with 2-D US. Studies have reported the use of contrast-enhanced ultrasound for to assessing the vascularity of the rotator cuff tendon and the subacromial bursa tissue [17–20]. The contrast agent (microsphere) was injected intravenously, and then an ultrasound scan lasting a period of several minutes was completed immediately to observe the blood flow of the rotator cuff and the subacromial bursa tissue. Gamradt et.al [17] concluded that the majority of blood flow to the repair was derived from the peribursal soft tissue and the anchor site at three months following repair. Cadet et.al [19] showed that blood flow of the repaired rotator cuff tendon decreased with time, but that exercise significantly enhanced blood flow to the repaired rotator cuff. The present study employed the use of CEUS with a SonoVue mixture to perform the PUSB procedure in an

attempt to explore a new technique for observation of rotator cuff abnormalities. The complete PUSB procedure took approximately 10–15 minutes for each shoulder. There were no severe complications requiring further treatment, either during or after the PUSB procedure. Our study suggests that the use of SonoVue is safe for bursography, just as previous studies have proven its safety in performing hysterosalpingo-contrast sonography (HyCoSy) and cholangiography [8, 9]. Since subacromial bursography was first described by Lindblom in 1939 [21], there have been few reports on the subject until now. The accuracy of subacromial bursography for the diagnosis of rotator cuff tears has not yet reached the same level as shoulder arthrography. Studies have shown that in many procedures performed without imaging guidance or only with fluoroscopic guidance, the needle is not sited in the subacromial bursa, leading to a lower rate of procedural success. In the present study, we successfully performed all 135

Fig 4 A full-thickness tear in a 61-year-old man, which was misdiagnosed as a partial-thickness tear by 2-D US (A), was accurately detected by PUSB (B, C). A 2-D US showed the tear as a partial-thickness tear (arrows). B short-axis of the supraspinatus tendon showed the hyperechoic contrast agent (CA) filled the tear area (double arrows) from subacromial bursa (arrow), asterisks show the ruptured part of the

bursa. C short axis of long bicipital tendon (LHBT) showed the hyperechoic CA into the sheath of LHBT (arrow). D Arthroscopy confirmed the same patient as full-thickness tear (arrow). SUP supraspinatus tendon, HH humeral head, Del deltoid, LT lesser tuberosity of humerus, GT greater tuberosity of humerus

Eur Radiol

PUSB procedures under the guidance of ultrasound, with high diagnostic accuracy in assessing full-thickness tears. There were limitations to the present study. First, this was a single-centre study, with a small sample of cases of rotator cuff tears. Larger samples are needed to further study the efficacy and safety of PUSB in the assessment of rotator cuff abnormalities. In addition, for ethical and practical reasons, our study did not include a truly normal group in which to perform the PUSB procedure. In conclusion, this is a preliminary report describing a new method—percutaneous US-guided subacromial bursography (PUSB) using CEUS—for assessing rotator cuff tears in patients with subacromial impingement syndrome. PUSB appears to be a feasible and effective radiation-free imaging modality with greater specificity and sensitivity than 2-D US in diagnosing full-thickness rotator cuff tears. As such, it expands the value of ultrasound in evaluating rotator cuff pathology. Acknowledgements The scientific guarantor of this publication is Man Lu. 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. This study received funding from the Sichuan Provincial Science and Technology Fund (2013JY0183). No complex statistical methods were necessary for this paper. Institutional review board approval was obtained. Written informed consent was obtained from all subjects (patients) in this study. Methodology: prospective, diagnostic or prognostic study, performed at one institution.

References 1. Balke M et al (2013) Correlation of acromial morphology with impingement syndrome and rotator cuff tears. Acta Orthop 84:178–183 2. Nasca RJ, Salter EG, Weil CE (1998) Current concepts review. Subacromial impingement syndrome (79-A: 1854–1868, Dec. 1997). J Bone Joint Surg Am 80:1852–1853 3. Neer CN (1983) Impingement lesions. Clin Orthop Relat Res 173: 70–77 4. Harrison AK, Flatow EL (2011) Subacromial impingement syndrome. J Am Acad Orthop Surg 19:701–708

5. Park HB et al (2005) Diagnostic accuracy of clinical tests for the different degrees of subacromial impingement syndrome. J Bone Joint Surg Am 87:1446–1455 6. Kurihara N et al (2012) Vascular patterns in nodules of intraductal papillary mucinous neoplasms depicted under contrast-enhanced ultrasonography are helpful for evaluating malignant potential. Eur J Radiol 81:66–70 7. Wilson SR, Burns PN (2010) Microbubble-enhanced US in body imaging: what role? Radiology 257:24–39 8. Exacoustos C et al (2009) Contrast-tuned imaging and secondgeneration contrast agent SonoVue: a new ultrasound approach to evaluation of tubal patency. J Minim Invasive Gynecol 16:437–444 9. Luyao Z et al (2012) Percutaneous ultrasound-guided cholangiography using microbubbles to evaluate the dilated biliary tract: initial experience. Eur Radiol 22:371–378 10. Kang CH et al (2009) Supraspinatus tendon tears: comparison of 3D US and MR arthrography with surgical correlation. Skelet Radiol 38: 1063–1069 11. Daenen B et al (2007) Ultrasound of the shoulder. JBR-BTR 90:325– 337 12. Brandt TD et al (1989) Rotator cuff sonography: a reassessment. Radiology 173:323–327 13. Rafii M et al (1990) Rotator cuff lesions: signal patterns at MR imaging. Radiology 177:817–823 14. Yablon CM et al (2013) Ultrasonography of the shoulder with arthroscopic correlation. Clin Sports Med 32:391–408 15. van Holsbeeck MT et al (1995) US depiction of partial-thickness tear of the rotator cuff. Radiology 197:443–446 16. Teefey SA et al (2000) Ultrasonography of the rotator cuff. A comparison of ultrasonographic and arthroscopic findings in one hundred consecutive cases. J Bone Joint Surg Am 82:498–504 17. Gamradt SC et al (2010) Vascularity of the supraspinatus tendon three months after repair: Characterization using contrast-enhanced ultrasound. J Shoulder Elb Surg 19:73–80 18. Funakoshi T et al (2010) In vivo visualization of vascular patterns of rotator cuff tears using contrast-enhanced ultrasound. Am J Sports Med 38:2464–2471 19. Cadet ER et al (2012) Contrast-enhanced ultrasound characterization of the vascularity of the repaired rotator cuff tendon: short-term and intermediate-term follow-up. J Shoulder Elb Surg 21:597–603 20. Adler RS et al (2011) Contrast-enhanced sonographic characterization of the vascularity of the repaired rotator cuff: utility of maximum intensity projection imaging. J Ultrasound Med 30: 1103–1109 21. Hodler J et al (1988) Rotator cuff tears: correlation of sonographic and surgical findings. Radiology 169:791–794