Printer Friendly

The magic-angle effect of the supraspinatus tendon.

Shoulder pain is a common complaint in patients older than 40 and is most often attributed to tears within the rotator cuff musculature resulting from anatomical impingement and trauma. (1) The rotator cuff muscles within the shoulder are frequently damaged, even in asymptomatic patients. In a 1995 study by Sher et al, (2) more than one third of patients found to be nor-real with a physical examination were later found to have a rotator cuff injury.

The rotator cuff comprises 4 muscles (subscapularis, supraspinatus, infraspinatus and teres minor) that act along with the deltoid to elevate and rotate the arm. Impingement and pain from rotator cuff injuries are routinely evaluated with magnetic resonance (MR) imaging and are most often found within the distal supraspinatus tendon near the point of insertion on the humeral head. (3,4)

The supraspinatus originates from the supraspinous fossa on the scapula and inserts on the greater tubercle of the humerus. (See Fig. 1.) On the distal end, the supraspinatus extends below the acromion process of the scapula, which may cause it to be injured or pinched in overhead arm movements, and is affected in 95% of all cuff tears. (5)


The most common injuries found within the distal supraspinatus tendon are generally described as a full-thickness tear, a partial-thickness tear or tendinopathy. (6) In a full-thickness tear, the supraspinatus tendon is torn through the entire anterior-to-posterior width of the tendon; in chronic cases following complete separation of the tendon, the medial end retracts as a result of atrophy of the supraspinatus muscle. By comparison, a partialthickness tear is an incomplete tear most commonly found on the articular or inferior surface causing a defect within part of the tendon. Unlike the tears, the pathologic changes described as tendinopathy are found within the tendon and result from histological changes that may cause focal swelling and weakening of the tendon. (6)

Although rotator cuff injury can occur in any individual, the rate of incidence increases with age. In a study of 180 patients treated with arthroscopy, the average age for a partial-thickness tear was 42, while the average patient with a full-thickness tear was 56 years old. (5)

The normal supraspinatus tendon appears uniformly low in signal intensity with all MR imaging pulse sequences; its signal intensity has been compared to that found in cortical bone. (6) Perceptible signal intensity within the normally black supraspinatus tendon is considered evidence of a rotator cuff disorder. Moderate-to-high, full-thickness tears demonstrate a fluid-filled gap, and retraction of tendinous margins is often evident, making these tears readily identified. Studies have shown that the greater the injury, the higher the accuracy of MR imaging evaluation. (7-10) Although MR evaluation of full-thickness tears has a high degree of accuracy, the sensitivity for partial-thickness tears and tendinopathy is relatively poor. (11-14) Research studies correlating MR imaging with arthroscopic findings have shown that the low detection accuracy has resulted in some partial tears being missed while others are diagnosed as false positives. (15-18) Degenerative changes or overuse injury may also result in tendinopathy, which adds to the normal variability of the tendon's signal, making it difficult to discern tears within the tendon. (19) Increased signal has also been found following tendon repair (20) and histopathologic studies evaluating scarring. (21)

Although increased MR signal intensity within the distal portion of the supraspinatus tendon typically represents injury, (22) this is also a frequent observation in healthy subjects due to a high signal resulting from a magic-angle effect. (23) The highly organized collagen fibers within the supraspinatus tendon produce a high signal on images with a short echo time, proton density (PD) and T1. This artifact occurs when the long axes of the collagen fibers are oriented at 55[degrees] to the main magnetic field. In most high-field MR scanners, the main magnetic field runs through the patient on the head-to-toe axis. Due to the shape of the shoulder, the well-organized collagen fibers within the supraspinatus tendon may be oriented at 55[degrees] from the main magnetic field, resulting in the magic-angle or high-signal artifact. (5,23) In the study conducted by Timins et al, (23) the magic-angle effect was found in all subjects (5 healthy volunteers and 3 cadaveric specimens) studied who did not have any injury to the supraspinatus tendon. Unfortunately, this artifact occurs in the same region where cuff tears and degenerative tendinopathy are prevalent, making it difficult to evaluate the supraspinatus tendon for injury. (24)

MR Weighting

The signal used to create MR images is primarily generated by hydrogen within the body, which is found most abundantly in the form of water and fat. Based on differences in their magnetization, specific MR sequences that vary the TR (repetition time) and the TE (echo time) can greatly alter the contrast of both fat and water. (See Fig. 2.) In general, T1-weighted images are characterized by bright fat and dark water (short TR and short TE) and are most often used to demonstrate normal anatomy. In contrast, T2-weighted images are characterized by bright water and dark fat (long TR and long TE) and are most often used to demonstrate pathology. For structures not seen well with T1 or T2 weighting, PD-weighted images are characterized by bright neural tissue and dark cortical bone (long TR and short TE). Together, these 3 types of MR imaging can greatly alter the contrast of anatomical structures and are routinely used in examinations to provide the information necessary for an accurate diagnosis. (25)


Study Materials and Methods

A descriptive study was conducted to evaluate the incidence of a magic-angle effect on PD-weighted MR images of the supraspinatus tendon as seen in the coronal oblique plane. Imaging was performed on a 1.5T system (Signa Excite, GE Medical Systems, Milwaukee, Wis) with a dedicated shoulder surface coil (Medical Advances, Milwaukee, Wis) that was positioned over the shoulder and secured with a binding strap. The coil was 6.5 inches (16.5 cm) in diameter and was contoured to the shape of the shoulder. Unlike previous studies in which patients were imaged with the shoulder in a relaxed position, the patients in this study were positioned with as much external rotation as possible. The subjects were centered in the gantry, and images were obtained by laterally off-centering the field of view to the position of the shoulder following external rotation.

Coronal oblique images were obtained parallel to the supraspinatus tendon, as demonstrated on an initial transaxial gradient-echo series. A slice thickness of 3.5-mm with a 0.5-mm intersection gap was used. The image matrix and field of view are included on selected images. Fat suppression technique was applied and has been shown to improve sensitivity of MR evaluation of rotator cuff injuries. (26)

To demonstrate the appearance of the supraspinatus tendon on a coronal oblique MR image, a sample image along with a corresponding labeled drawing is shown in Figure 3. The supraspinatus muscle originates from the body of the scapula and extends between the head of the humerus and the acromion process of the scapula to insert on the greater tubercle of the humerus. Based on this anatomical arrangement, the supraspinatus tendon can be injured by getting pinched between the head of the humerus and the acromion process of the scapula due to overhead movements of the arm or trauma. (5)


For this study, 300 consecutive patients were selected from those symptomatic patients referred for MR evaluation of the shoulder with both PD and T2-weighted sequences. To comply with policies established in the Code of Federal Regulations (Title 45, CFR, Part 46) for the protection and safety of human subjects used in any type of research, including biomedical and behavioral research, the Institutional Review Board for the Protection of Human Research Subjects reviewed and granted approval for this research activity. The regulations, policies, procedures and ethical commitments of Fort Hays State University, Hays, Kan, required that individuals not be involved in activities that would place them at risk; before the examination, patients gave written consent for their results to be used in this research study.

In cases of tendon repair consistent with a healing response or surgical correction, increased signal intensity has been shown on PD images. (20) This increased signal intensity on PD sequences has been attributed to the presence of fat, (27,28) subclinical degeneration (21,29,30) intermingling of connective tissue, (31) and alterations in vascularity, (29,30) Previous studies have also shown that rotator cuff muscles are frequently damaged in asymptomatic patients (2) so the injury evaluated during the MR examination could have developed over the patient's lifetime. To avoid these variables, all patients with a reported injury were removed from the sample group.

To eliminate patients with an actual injury, the written MR reports generated by a board-certified radiologist were reviewed by an American Registry of Radiologic Technologists (ARRT)-registered radiographer, and those reporting any injury to the supraspinatus tendon were eliminated from the sample group. The MR images of the patients without an injury to the supraspinatus tendon were independently evaluated by 2 ARRT-registered radiographers who had successfully completed baccalaureate course work in sectional anatomy and pathology, and were in the process of completing the clinical competency requirements to sit for the ARRT examination in MR. To prepare for this study, the radiographers read the radiologist's written report and were provided examples of the magic-angle effect by previous investigators. (6,23,32) Within a 6-month period following the MR exam, 2 radiographers independently examined the oblique coronal images nearest the center of the supraspinatus tendon and compared the signal within the distal tendon on PD and T2-weighted images. The image evaluation was limited to a comparison of signal on PD and T2-weighted images and no intensity scales were used in this study. Patients were classified as either having higher signal with PD or T2-weighted images. Images found to have a higher signal with the PD sequence were classified as having the magic-angle effect since an injury would most likely have a stronger signal on T2-weighted images. (6,32)


Of the 300 consecutive patients evaluated in this study, 136 patients were reported to have an injury to the supraspinatus as described by a board-certified radiologist and were eliminated from the sample group. The remaining 164 patients--comprised of 72 women and 92 men ranging in age from 26 to 72 years old, with an average age of 54--were found not to have an injury as determined by multiple MR sequences.

Of the 164 patients found without an injury within the supraspinatus tendon, both reviewers found the same 156 patients to have low signal within the tendon on both PD and T2 images. A representative sample is shown in Figure 4A and 4B. On the PD image (see Fig. 4A), the distal end of the suparaspinatus tendon (marked with the arrowhead) appears as a region of low signal extending from the origin of the muscle on the supraspinous fossa of the scapula. By comparison, the T2-weighted image (see Fig. 4B) appears much the same with a low density in the distal end of the supraspinatus tendon. However, since T2 images generate a much stronger signal from water, the fluid within the shoulder joint space (arrow) has a much stronger signal compared to the corresponding PD-weighted image depicting the same slice of anatomy.


Of the 164 patients found not to have any injury to the supraspinatus, both reviewers found the same 8 patients (5%) to have the high signal artifact on the PD sequence compared to the T2-weighted images. As shown in Figure 5A and 5B, the magic-angle effect is readily apparent in the distal end of the supraspinatus in the PD-weighted image. (See Fig. 5A.) In the corresponding T2-weighted image depicting the same slice of anatomy (see Fig. 5B), the signal within the tendon is greatly diminished (arrow) compared to the PD-weighted image. Because only normal examinations were included in the sample group, the patient images showing a stronger signal within the distal tendon (see Figs. 5A and 5B) were easily distinguished from the patient images considered negative for the magic-angle effect. (See Fig. 4.)



As shown by the results of this study, the high-signal effect found on PD-weighted images appears much like tendinopathy or a tear within the distal supraspinatus tendon. Unfortunately, this artifact may be misinterpreted as a false positive for a rotator cuff injury if not compared with T2-weighted images. On T2-weighted images, a real injury also has a strong signal due to the fluid concentrated at the site of the injury. By comparison, a magic-angle effect has a weaker signal on T2-weighted images due to the longer TE. (6)

Even though patients with injury to the supraspinatus tendon were excluded from the sample group examined in this study, images of a patient with tendinopathy within the supraspinatus tendon are included as Figure 6. As shown in this case, even a relatively minor injury to the rotator cuff generates intratendinous signal on T2-weighted images. As shown previously by Tuite, (5) the larger the injury to the supraspinatus tendon, the greater the signal on T2-weighted images.


Very much like Figure 5A, which demonstrated the intratendinous signal generated by the magic-angle effect on PD sequences, a partial-thickness tear also has a high signal compared to the surrounding tendon with PD weighting as shown in Figure 7.


In previous studies, the magic-angle effect has been described in MR evaluation of tendons in the ankle, wrist and rotator cuff of the shoulder. (24) Increased signal within the distal portion of the supraspinatus tendon during MR imaging is a frequent observation in normal shoulders. (29,31,33,34) In a study of 15 asymptomatic subjects, Mirowitz (29) found that all subjects (100%) had a relative increase in signal within the supraspinatus tendon despite the absence of any injury shown in the other sequences, including T2-weighted images. In a similar study, Liou et al (34) evaluated 60 asymptomatic shoulders determined to be normal by MR evaluation and focal signal intensity was present in 95% (57 of the 60) of the shoulders within the distal tendon on the PD-weighted images. Likewise, Neumann et al (31) evaluated PD and T2-weighted coronal images of 55 asymptomatic shoulders and found signal intensity in 89% (49 of 55) of the subjects that decreased in intensity on T2-weighted images.

In this study, the magic-angle effect was found in 5% (8 of 164) of the cases evaluated. As described by previous investigators, (5,24) the magic-angle effect is found when the highly organized collagen fibers within the supraspinatus tendon form a 55[degrees] angle, or the magic-angle, to the main magnetic field. As shown by Timins et al, (23) the angle of the humeral shaft relative to the main magnetic field affects the magic-angle signal apparent within the supraspinatus tendon. Because previous studies demonstrating a much greater rate of incidence were performed with either cadaveric specimens or healthy subjects with the arm placed in a relaxed position, our findings suggest the external rotation of the arm used in this study greatly reduced the incidence of the magic-angle effect. The low incidence rate found in this study suggests that external rotation of the hand alters the supraspinatus tendon orientation to the main magnetic field, thereby reducing the incidence of the 55[degrees] angle to the main magnetic field, or the magic-angle.

Previous studies evaluating patients without injury have reported a much higher or nearly universal incidence rate of magic-angle effect. (29,31,33,34) Although this disparity might be attributed to technical differences between the studies, this conclusion is not supported by the research focused on comparing technical differences in the MR evaluation of rotator cuff injuries. Previous investigators found no changes in sensitivity for MR evaluation of the shoulder for extremity and high-field MR (35,36) for fast spin echo compared to conventional spin echo sequences (37) and gradient echo compared to fast spin sequences. (38) The only notable difference was that the intensity of magic-angle artifact increased by using specific technical factors (ie, a high flip angle at TE values less than 30 ms). (39) However, since the intensity of the artifact was not measured, the technical factors used in our study were comparable to those studies finding a much higher incidence of the magic-angle effect.

Limitations of the Study

Even though patients with an injury evaluated by MR were removed from the sample group, the films were reviewed by only 1 radiologist, so there may have been instances in which an injury was not reported. In cases of an acute injury, the signal would have been stronger on T2-weighted images so the patient would not have been included in the magic-angle effect cases. However, in cases where the injury had been repaired, the highly variable signal may have inadvertently been included for the 5% of the patients where the artifact was reported.

The individuals independently reviewing the films in this study had very similar knowledge of sectional anatomy and pathology that may have contributed to their very high agreement levels. Further, the patients involved in this study were taken from only 1 clinical facility, and the results may be slightly different at other facilities due to differences in technology and imaging protocols.


(1.) McConville OR, Iannotti JP. Partial-thickness tears of the rotator cuff: evaluation and management. J Am Acad Orthop Surg. 1999;7(1):32-43.

(2.) Sher JS, Uribe JW, Posada A, Murphy BJ, Zlatkin MB. Abnormal findings on magnetic resonance images of asyptomatic shoulders. J Bone Joint Surgery. 1995;77(1):10-15.

(3.) Fritz RC, Stoller DW. MR imaging of the rotator cuff. Magn Reson Imaging Clin N Am. 1997;5 (4):435-454.

(4.) Tuite MJ. MR imaging of sports injuries to the rotator cuff. Magn Resort Imaging Clin N Am. 2003;11 (2):207-219, v.

(5.) Tuite MJ. Shoulder, rotator cuff injury (MRI). Available at: Accessed July 20, 2004.

(6.) Zlatikin MB. MRI of the shoulder: imaging techniques and diagnostic pitfalls. MRI of the Shoulder. 2003;3:29-45.

(7.) Fritz RC. Magnetic resonance imaging of sports-related injuries to the shoulder: impingement and rotator cuff. Radiol Clin North Am. 2002;40(2):217-34, vi.

(8.) Evancho AM, Stiles RG, Fajman WA, Flower SP, et al. MR imaging diagnosis of rotator cuff tears. A JR Am J Roentgenol. 1998; 151 (4):751-754.

(9.) Morrison DS, Ofstein R. The use of magnetic resonance imaging in the diagnosis of rotator cuff tears. Orthopedics. 1990;13(6):633-637.

(10.) Zlatikin MB, Iannotti JP, Roberts MC, Esterhai JL, et al. Rotator cuff tears: diagnostic performance of MR imaging. Radiology. 1989; 172 (1): 223-239.

(11.) Balich SM, Sheley RC, Brown TR, Sauser DD, Quinn SF. MR imaging of the rotator cuff tendion: interobserver agreement and analysis of interpretive errors. Radiology. 1997;204(1):191-194.

(12.) Ertl JP, Kovacs G, Burger RS. Magnetic resonance imaging of the shoulder in the primary care setting. Med Sci Sports Exerc. 1998;30(4 Suppl):S7-11.

(13.) Nakatani T, Fujita K, Iwasaki Y, Sakai H, Kurosaka M. MRI-negative rotator cuff tears. Magn Reson Imaging. 2003;21(1):41-45.

(14.) Wnorowski DC, Levinshohn EM, Chamberlain BC, McAndrew DL. Magnetic resonance imaging assessment of the rotator cuff: is it really accurate? Arthroscopy. 1997;13(6):710-719.

(15.) Quinn SF, Sheley RC, Demlow TA, Szumowski J. Rotator cuff tendon tears: evaluation with fat-suppressed MR imaging with arthroscopic correlation in 100 patients. Radiology. 1995;195 (5):497-500.

(16.) Balich SM, Sheley RC, Brown TR, Sauser DD, Quinn SF. MR imaging of the rotator cuff tendon: interobserver agreement and analysis of interpretive errors. Radiology. 1997;204:1991-1994.

(17.) Teefey SA, Robin DA, Middletown WD, Hildebolt CF, Leibold RA, Yamaguchi K. Detection and quantification of rotator cuff tears. Comprasion of utrasonographic, magnetic resonance imaging, and artoropscopic findings in seventy-one consecutive cases. J Bone Joint Surg Am. 2004; 86-A(4):708-716.

(18.) Yamakawa S, Hashizume H, Ichikawa N, Itadera E, Inoue H. Comparative studies of MRI and operative findings in rotator cuff tear. Acta Med Okayama. 2001;55 (5):261-268.

(19.) Wright T, Yoon C, Schmit BE Shoulder MRI refinements: differentiation of rotator cuff tear from artifacts and tendonosis, and reassessment of normal findings. Semin Ultrasound CT MR. 2001;22 (4):383-395.

(20.) Gusmer PB, Hollis PG, William DD, O'Brein SJ. MR imaging of the shoulder after rotator cuff repair. A JR Am J Roentgenol. 1997;168:559-563.

(21.) Kjellin I, Ho CP, Cervilla V, Haghighi P, et al. Alternation in the supraspinatus tendon at MR imaging: correlation with histopathologic findings in cadavers. Radiology. 1991;18(3):837-841.

(22.) Tuite MJ, Asinger D, Orwin JF. Angled oblique sagittal MR imaging of rotator cuff tears: comparison with standard oblique sagittal images. Skeletal Radiol. 2001;30 (5):262-269.

(23.) Timins ME, Erickson SJ, Estokowski LD, Carrera GE Komorowski RA. Increased signal in the normal supraspinatus tendon on MR imaging: diagnostic pitfall caused by the magic-angle effect. AJR Am J Roentgenol. 1995;165(1):109-114.

(24.) Hayes CW, Parellada JA. The magic-angle effect in musculoskeletal MR imaging. Top Magn Reson Imaging. 1996;8(1):51-56.

(25.) Westbrook C, Kaut C. MRI in Practice. 2nd ed. Boston, Mass: Blackwell Publishing; 1998.

(26.) Funke M, Kopka L, Vosshenrich R, Oestmann JW, Grabbe E. MR arthrography in the diagnosis of rotator cuff tears. Standard spin-echo alone or with fat suppression? Acta Radiol. 1996;37(5):627-632.

(27.) Heuck A, Appel M, Kaiser E, Luttke, Lukas P. MR imaging of the shoulder: potential overinterpretation of normal findings in the rotator cuff. Radiology. 1990;177(P):103.

(28.) Rafii M, Firooznia H, Sherman O. Rotator cuff lesions: signal patterns at MR imaging. Radiology. 1990;177:817-823.

(29.) Mirowitz SA. Normal rotator cuff: MR imaging with conventional and fat-suppression techniques. Radiology. 1991;180:735-740.

(30.) Vahlensieck M, Pollack M, Lang P, Grampp S, Genant H. Two segments of the supraspinous muscle; cause of high signal intensity at MR imaging? Radiology. 1993;186:449-454.

(31.) Neumann CH, Hold RG, Steinbach LS, Jahnke AH Jr., Peterson SA. MR imaging of the shoulder: appearance of the supraspinatus tendon in asymptomatic volunteers. AJR Am J Roentgenol. 1992;158:1281-1287.

(32.) Jones AO. Magnetic resonance imaging of the supraspinatus tendon: the significance of signal intensity alterations at the 'critical zone'. Australas Radiol. 1998;42(2):106-113.

(33.) Kaplan PA, Bryans KC, Davick JP, Otte M, et al. MR imaging of the normal shoulder: variants and pitfalls. Radiology. 1992;184:519-524.

(34.) Liou JTS, Wilson AJ, Totty WG, Brown JJ. The normal shoulder: common variations that simulate pathologic conditions at MR imaging. Radiology. 1993;186:435-441.

(35.) Sakurai G, Ozaki J, Tomita Y, Kondo T, Tamai S. Incomplete tears of the subscapularis tendon associated with tears of the snpraspinatus tendon: cadveric and clinical studies. J Shoulder Elbow Surg. 1998;7(5):510-515.

(36.) Zlatkin MB, Hoffman C, Shellock FG. Assessment of the rotator cuff and glenoid labrum using an extremity MR system: MR results compared to surgical findings from a multi-center study. J Magn Reson Imaging. 2004;195:623-631.

(37.) Carrino JA, McCauley TR, Katz LD, Smith RC, Lange RC. Rotator cuff: evaluation with fast spin-echo versus conventional spin-echo MR imaging. Radiology. 1997;202 (2): 533-539.

(38.) Sahin-Akyar C, Miller TT, Staron RB, McCarthy DM, 0Feldman E Gradient-echo versus fat-suppressed fast spin-echo MR imaging of rotator cuff tears. A JR Am J Roentgenol. 1998;171(1):223-227.

(39.) Zurlo JV, Blacksin MF, Karimi S. The influence of flip angle on the magic-angle effect. Skeletal Radiol. 2000;29(10):593-596.


This study could not have been completed without the substantial efforts of Craig Boro, Ben Gibbs, Terri Krier, Brandon Reicks, Jessica Ridgway and Kaely Steinert. This publication was made possible by National Institutes of Health grant number P20 RR016475 from the IDeA Networks of Biomedical Research Excellence Program of the National Center for Research Resources.

Michael E. Madden, Ph.D., R.T.(R)(CT)(MR), is director of medical diagnostic imaging and radiologic technologies at Fort Hays State University, Hays, Kan.

Reprint requests may be sent to the American Society of Radiologic Technologists, Communications Department, 15000 Central Ave. SE, Albuquerque, NM 87123-3917.
COPYRIGHT 2006 American Society of Radiologic Technologists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2006 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Madden, Michael E.
Publication:Radiologic Technology
Geographic Code:1USA
Date:May 1, 2006
Previous Article:Advanced CT technologist needed.
Next Article:Admission requirements for radiography programs.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |