JBJS | Instructional Course Lectures, The American Academy of Orthopaedic Surgeons - Magnetic Resonance Imaging of the Shoulder*†

The Journal of Bone & Joint Surgery, Volume 79, Issue 6

Instructional Course Lecture | June 01, 1997

Instructional Course Lectures, The American Academy of Orthopaedic Surgeons - Magnetic Resonance Imaging of the Shoulder*†

RICHARD J. HERZOG, M.D.‡, PHILADELPHIA, PENNSYLVANIA

View Disclosures and Other Information

An Instructional Course Lecture, The American Academy of Orthopaedic Surgeons

The Journal of Bone & Joint Surgery. 1997; 79:934-53

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

text A A A

Introduction

The foundation of successful patient care is an accurate clinical diagnosis. The accomplishment of this goal begins with a complete history and physical examination. If the initial evaluation does not provide the information needed to explain the symptoms, additional diagnostic testing is usually needed. The additional data provided by these tests become useful clinical information only when they are integrated with the patient's history and the results of the physical examination and other ancillary tests. Imaging studies may also be part of the preoperative workup in order to confirm the presence of a pathological condition and to determine its extent. In the past, radiographic imaging studies, such as plain radiography and computed tomography, and radionuclide studies, have played a major role in the diagnostic evaluation of patients who have a musculoskeletal disorder. These studies focused mainly on the detection of osseous abnormalities. With the recent development and implementation of magnetic resonance imaging, it is now possible to evaluate non-invasively the soft-tissue structures of the body, such as muscles, tendons, and ligaments, which are the structures frequently responsible for most of the symptoms in patients who have musculoskeletal dysfunction.

The goal of any imaging study is to define accurately the pathomorphological changes in a specific tissue, organ, or part of the body. Whenever possible, objective diagnostic criteria should be employed when the findings of an imaging study are reported. Objective categorization of pathological changes facilitates the interpretation and communication of abnormalities detected on a test, and these same criteria can then be used when follow-up tests are performed to assess the effects of different forms of therapy, such as operative intervention or non-operative rehabilitation. The reproducibility and reliability of all objective diagnostic criteria must be evaluated rigorously in prospective, blinded studies before they are implemented⁶⁴.

*Printed with permission of The American Academy of Orthopaedic Surgeons. This article will appear in Instructional Course Lectures, Volume 47, The American Academy of Orthopaedic Surgeons, Rosemont, Illinois, March 1998.

†No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.

‡210 West Rittenhouse Square, Apartment 1805, Philadelphia, Pennsylvania 19103.

View Larger

TABLE I ROUTINE PROTOCOL FOR MAGNETIC RESONANCE IMAGING OF THE SHOULDER

*1 = made with the upper extremity in neutral or slight external rotation, and 2 = a field of view of sixteen centimeters for all sequences.

Plane*	Relaxation Time (msec)	Echo Time (msec)	No. of Excitations	Image Matrix	Thickness (mm)	Skip (mm)	Comments
Oblique coronal							Parallel to long axis of body of scapula, including soft tissues anterior and posterior to humeral head
1	2500	Dual-echo minimum/70	1	256 by 192	4	0.5	Classic multi-echo
Oblique sagittal							Perpendicular to long axis of scapula from base of coracoid through humeral head
1	1500	Minimum	1	256 by 192	4	0.5
2	4000	Approximately 90	2	256 by 192	4, same location as 1	0.5	Fast spin echo-echo train length-8; fat-saturated band width, 32kHz
Axial							From superior to acromiclavicular joint through glenoid and humeral neck
1	1500	Minimum	1	256 by 192	3	0.5
2	4000	Approximately 90	2	256 by 192	3, same location as 1	0.5	Fast spin echo-echo train length-8; fat-saturated band width, 32 kHz

Add to My JBJS

TABLE I ROUTINE PROTOCOL FOR MAGNETIC RESONANCE IMAGING OF THE SHOULDER

*1 = made with the upper extremity in neutral or slight external rotation, and 2 = a field of view of sixteen centimeters for all sequences.

Plane*	Relaxation Time (msec)	Echo Time (msec)	No. of Excitations	Image Matrix	Thickness (mm)	Skip (mm)	Comments
Oblique coronal							Parallel to long axis of body of scapula, including soft tissues anterior and posterior to humeral head
1	2500	Dual-echo minimum/70	1	256 by 192	4	0.5	Classic multi-echo
Oblique sagittal							Perpendicular to long axis of scapula from base of coracoid through humeral head
1	1500	Minimum	1	256 by 192	4	0.5
2	4000	Approximately 90	2	256 by 192	4, same location as 1	0.5	Fast spin echo-echo train length-8; fat-saturated band width, 32kHz
Axial							From superior to acromiclavicular joint through glenoid and humeral neck
1	1500	Minimum	1	256 by 192	3	0.5
2	4000	Approximately 90	2	256 by 192	3, same location as 1	0.5	Fast spin echo-echo train length-8; fat-saturated band width, 32 kHz

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1 Proton-density-weighted oblique sagittal image showing an osseous proliferation projecting off the inferior margin of the acromioclavicular joint (arrow) and impinging on the anterosuperior margin of the supraspinatus myotendinous junction.

View Large | Download Slide(.ppt)
Add to My JBJS

Figs. 2-A through 2-D: Proton-density-weighted oblique sagittal images. Fig. 2-A: A type-I acromion (arrow) and no impingement on the cuff.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-B A type-II acromion (arrowheads) and a mildly thickened coracoacromial ligament (arrow) associated with moderate narrowing of the supraspinatus outlet.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-C A type-III acromion with a hook of the anteroinferior margin (arrow).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-D A large osseous ridge (arrow) projecting off the anteroinferior margin of a type-I acromion. This represents an acquired deformity of the acromion that results in narrowing of the supraspinatus outlet similar to that seen with a type-III acromion.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3-A: A proton-density-weighted axial image made through the acromioclavicular joint, showing an os acromiale (arrow).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3-B: The os acromiale (curved arrow) is also delineated on the proton-density-weighted oblique sagittal image, where a mildly thickened coracoacromial ligament (straight arrow) inserts into it.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4-A: A proton-density-weighted oblique coronal image showing increased signal intensity within a thickened supraspinatus segment of the cuff (arrow).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4-B: A T2-weighted oblique coronal image showing persistently increased signal intensity within the substance of the supraspinatus tendon (arrow) but no tear extending to the bursal or articular surface of the cuff.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5-A: A T2-weighted oblique coronal image showing a small, deep partial-thickness tear (arrow) involving the articular surface of the cuff.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5-B: A T2-weighted oblique coronal image of a different shoulder, showing an irregular oblique tear (arrow) involving the bursal side of the cuff.

View Large | Download Slide(.ppt)
Add to My JBJS

Figs. 6-A, 6-B, and 6-C: A small full-thickness tear (arrow) at the insertion site of the rotator cuff, posterior to the rotator interval. Fig. 6-A: T2-weighted oblique coronal image.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 6-B: T2-weighted oblique sagittal image.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 6-C T2-weighted axial image.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 7-A: A proton-density-weighted oblique coronal image showing a tapered, completely torn supraspinatus tendon (curved arrow) and tissue demonstrating intermediate signal intensity (straight arrow) in the region of the gap in the tendon.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 7-B: A T2-weighted oblique coronal image showing high-signal-intensity fluid within the glenohumeral joint but no fluid extending through the defect in the cuff. The reactive tissue within the defect (arrow) demonstrates intermediate signal intensity.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 8 A proton-density-weighted oblique sagittal image of a large, chronic full-thickness tear involving the supraspinatus and infraspinatus segments of the cuff, showing moderate-to-severe atrophy of the supraspinatus (curved arrow) and infraspinatus (straight arrow) muscles.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 9 A proton-density-weighted axial image showing a completely detached subscapularis tendon (curved arrow). The biceps tendon (straight arrow) is still located within the bicipital groove, which is covered by an intact transverse ligament.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 10 A T2-weighted axial image showing the biceps tendon (straight arrow) to be medially subluxated and resting on the medial wall of the bicipital groove. The deep fibers of the subscapularis tendon (curved arrow) are detached from the lesser tuberosity, but the superficial fibers are still continuous with the fibrous fascia covering the bicipital groove.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 11 A fat-saturated T2-weighted axial image showing edematous changes in the lateral segment of the teres minor muscle (arrow) due to an acute stretch-induced injury. There is no tear of the posterior aspect of the labrum or posterior capsular stripping.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 12 A T2-weighted oblique sagittal image showing a ganglion cyst (curved arrow) located within the supraspinous fossa and extending through the spinoglenoid notch. There is increased signal intensity within the infraspinatus muscle (straight arrow) due to subacute denervation.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 13-A: A proton-density-weighted axial image made after a recent anterior dislocation of the shoulder. The anterior aspect of the glenoid labrum is torn and detached (arrow).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 13-B: The extent of the labral tear (arrows) is defined on the proton-density-weighted oblique sagittal image.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 14 A proton-density-weighted oblique coronal image showing an avulsion of the inferior aspect of the capsule from the neck of the humerus (arrow).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 15 A T2-weighted axial image showing a tear through the base of the posterior aspect of the glenoid labrum (arrow) along with posterior capsular stripping.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 16-A: A proton-density-weighted oblique coronal image showing a type-III tear of the superior aspect of the labrum extending anteriorly and posteriorly (SLAP lesion) (arrow).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 16-B: A gradient-echo axial image also showing a type-III tear of the superior aspect of the labrum extending anteriorly and posteriorly (arrows).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 17 A T2-weighted oblique coronal image showing a minimally displaced fracture of the greater tuberosity (arrow), which was the result of a skiing injury. The rotator cuff is intact.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 18 A proton-density-weighted oblique coronal image showing a large osteonecrotic focus in the superomedial segment of the humeral head, without collapse of the subchondral bone plate, associated with chronic use of steroids.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 19 A proton-density-weighted oblique sagittal image showing subchondral cysts (arrow) within the posteroinferior margin of the glenoid.

Magnetic Resonance Imaging

Complete assessment of all of the soft tissues and osseous structures of the shoulder girdle is mandatory to achieve a comprehensive evaluation of dysfunction of the shoulder. Magnetic resonance imaging studies should be performed in three orthogonal planes for all patients. The images are made with the upper extremity comfortably positioned at the side in slight abduction and neutral or slight external rotation⁹. The axis of the different scan planes is determined by the orientation of the scapulohumeral axis, which is defined by an initial axial localizing sequence. The coronal sequence is oriented parallel to the long axis of the body of the scapula and the supraspinatus tendon and perpendicular to the glenoid fossa. The coronal sequence extends from the coracoid process through the entire rotator cuff mechanism—that is, the tendinous cuff and the muscles. The sagittal sequence is oriented perpendicular to the coronal sequence and extends from the suprascapular notch through the rotator cuff and into the adjacent deltoid muscle. Both the coronal and the sagittal sequences are oriented obliquely to the true coronal and sagittal planes of the body because of the normal rotation of the scapula on the chest wall. Therefore, these sequences are correctly designated oblique coronal and oblique sagittal. The axial sequence is oriented perpendicular to the glenoid fossa and extends from the superior margin of the acromioclavicular joint through the quadrilateral space.

The magnetic resonance imaging sequences typically used to evaluate the musculoskeletal system are the spin-echo, gradient-echo, and short tau inversion recovery (STIR) sequences. They provide excellent contrast and spatial resolution for the evaluation of normal anatomical structures and pathological changes within the body. Spin-echo T1-weighted sequences (a short relaxation time and a short echo time) are optimum to display tissues containing fat or blood and provide excellent anatomical detail. Spin-echo proton-density-weighted sequences (a long relaxation time and a short echo time) are also used to assess soft-tissue and osseous anatomy. Spin-echo T2-weighted sequences (a long relaxation time and a long echo time) are used to delineate tissue containing fluid or edema. Short tau inversion recovery and spin-echo fat-saturated T2-weighted sequences are particularly sensitive to fluid or edema in soft tissues and osseous structures.

The high signal intensity in cancellous bone on T1-weighted sequences is due to the large amount of fat within the cancellous bone. Any process that replaces this fat results in decreased signal intensity on the spin-echo T1-weighted sequence. If the abnormal process contains increased free water (as with an inflammatory focus or malignant cells), there is increased signal intensity in the cancellous bone on spin-echo T2-weighted and short tau inversion recovery sequences. If the process replacing the marrow fat contains fibrous tissue or additional mineralization, there is low signal intensity on all of the magnetic resonance images.

With magnetic resonance imaging, it is now possible to follow non-invasively the evolution of injury, repair, and remodeling of tissue as well as the changes of tissue-aging. Acute injury to vascularized tissue incites an inflammatory response, with an increase in tissue hydration that is detected on spin-echo T2-weighted and short tau inversion recovery sequences as a focus of high signal intensity. With tissue-healing, remodeling, and fibrosis, the signal intensity on the T2-weighted and short tau inversion recovery sequences decreases because of the increased amount of collagen and the decreased amount of fluid in the tissue. Aging involves changes in both the cellular and the extracellular components of tissue². Tissue-aging that is accompanied by desiccation of the tissue demonstrates decreased signal intensity on spin-echo T2-weighted sequences. Tissue-aging that is accompanied by fatty infiltration demonstrates increased signal intensity on spin-echo T1-weighted sequences, and that accompanied by myxoid degeneration of collagen demonstrates increased signal intensity on spin-echo T2-weighted or short tau inversion recovery sequences because of the increased hydration of the degenerated tissue.

Recent advances in magnetic resonance imaging, including faster imaging sequences and improved surface coil design, have markedly increased the amount of information provided by these images. Our routine imaging examination of the shoulder, performed with a 1.5-tesla magnetic resonance imaging system, is different for each plane (Table I). The type of equipment employed markedly affects the quality of the images and their potential usefulness. While patients tolerate some low-field-strength open magnetic resonance imaging systems better than they do high-field-strength systems because they are not enclosed closely by the low-field-strength unit, the quality of the images generated by these open systems is typically poorer than that achieved with high-field-strength systems.

The magnetic resonance imaging sequences used for most musculoskeletal evaluation, including that of the shoulder, are similar. A complete discussion of the physics of magnetic resonance imaging and the different sequences that are currently used are outside the scope of this paper^8,37,38. Absolute contraindications to magnetic resonance imaging include the presence of a cardiac pacemaker, a metallic foreign body in the eye or spine, a cerebral aneurysm clip, some types of infusion pumps, bone or nerve stimulators, and an ocular or cochlear implant⁶⁰. It is the responsibility of both the clinician who orders the magnetic resonance imaging study and the facility that performs it to screen patients adequately before the examination.

Rotator Cuff and Impingement

The rotator cuff is one of the largest tendinous structures in the body and, because of the functional demands placed on it, it is prone to overload and failure. Rotator cuff disease represents a spectrum of pathological changes within the cuff, including microscopic or macroscopic failure of fibers associated with tissue edema, hemorrhage, and fibrosis. These conditions are superimposed on the normal changes in the cuff due to aging. The primary mechanisms of injury to the cuff appear to be extrinsic impingement and intrinsic overload. A patient who has impingement syndrome has pain on attempted use of the shoulder, particularly with overhead activities. The pain is precipitated by entrapment or abrasion of the rotator cuff and the peritendinous bursa and soft tissue. This may occur under a degenerated acromioclavicular joint or subjacent to the coracoacromial arch (the arch formed by the coracoid process, the coracoacromial ligament, and the acromion). In the supraspinatus outlet (the space between the humeral head and the coracoacromial arch), the rotator cuff mechanism may impinge against a thickened coracoacromial ligament; an osseous ridge projecting off the anteroinferior margin of the acromion at the insertion of the coracoacromial ligament; or a curved, tilted, or hooked acromion. With the arm abducted and externally rotated, the undersurface of the cuff may also impinge against the posterosuperior margin of the glenoid labrum and the rim of the glenoid. Repetitive abrasion of the rotator cuff mechanism may precipitate bursal or peritendinous inflammation or failure of the cuff. Such failure represents a continuum of pathological changes within the cuff, beginning with hemorrhage and edema, followed by inflammation and fibrosis, and possibly ending with a partial or full-thickness tear⁴⁴.

Disruption of the fibers secondary to abrasion is associated with edema or hemorrhage, or both, in the cuff and in the peritendinous tissues. If an inflammatory reaction is incited within the cuff by failure of the collagen fibers, then the term tendinitis is appropriate to describe the pathological condition. The response of the cuff to the initial injury depends on the functional demands placed on the shoulder as well as on the reparative capacity of the cuff. There is still controversy concerning the ability of a cuff to heal and how the aging process affects this ability. Recent investigation suggests that the vascularity of a cuff may be maintained with aging⁷ but the altered structure of the collagen and proteoglycans in the aging cuff may affect its response to chronic microtrauma or acute macrotrauma.

As failure of the cuff evolves from acute or subacute tendinitis into a more chronic condition, possibly associated with a partial or full-thickness tear, biopsy of the cuff may demonstrate hyaline or myxoid degeneration of the collagen fibers, fibrovascular proliferation, and few or no inflammatory cells. The term tendinosis and not tendinitis is probably appropriate to describe such a cuff²⁸. Tendinosis may represent an abortive healing response of a tendon from chronic extrinsic impingement or intrinsic overload. Tendinitis may be superimposed on tendinosis if there is new failure of the fibers in a weakened, degenerated cuff.

While this spectrum of abnormalities may cause similar problems about the shoulder, including pain, decreased strength, and a limited range of motion, the prognoses and therapeutic options vary widely depending on the degree of the pathological changes within the cuff. The optimum role of an imaging study is to define the nature and extent of these pathological changes and to provide a comprehensive assessment of all of the structures of a shoulder that may precipitate or perpetuate failure of the cuff.

With magnetic resonance imaging, it is possible to define precisely the morphology of the acromioclavicular joint, the coracoacromial arch, and the rotator cuff. Magnetic resonance imaging demonstrates the location at which a cuff may be impinging on an area of osseous proliferation¹⁰ (Fig. 1). It is also possible to detect bursal inflammation, which is typically associated with tears of the rotator cuff, including partial-thickness tears of the articular surface and intratendinous tears¹⁴. Impingement—that is, pushing against—is a physical phenomenon and can be detected on magnetic resonance images. However, the diagnosis of impingement syndrome, which is a painful symptom complex secondary to the repetitive abrasion and inflammation of the rotator cuff or the peritendinous tissue, or both, as a result of impingement can only be made clinically. Magnetic resonance images are typically made with the upper extremity slightly abducted and in neutral or slight external rotation. This position does not precipitate symptoms of impingement in most patients; therefore, until dynamic magnetic resonance imaging studies can be performed with the upper extremity abducted or in forward elevation and internal rotation it is not possible to assess completely the dynamic relationships of the cuff to the acromioclavicular joint and the coracoacromial arch.

Changes in the width and configuration of the subacromial space with scapular protraction and retraction were studied with magnetic resonance imaging by Solem-Bertoft et al.⁶⁷. In a limited number of normal subjects, the anterior opening of the subacromial space became smaller with scapular protraction. In young patients, degenerative changes of the acromioclavicular joint or the acromion are rare. A torn cuff is more likely to be due to impingement of the cuff against the glenoid labrum or rim or to be related to chronic tensile overload and secondary fatigue failure of the cuff fibers. With decreased tensile strength of the cuff, additional stress may precipitate a partial or full-thickness tear. Direct-impaction injuries of the rotator cuff may also occur in athletes who engage in contact sports¹. If biomechanical imbalance results from a torn rotator cuff or is secondary to primary instability of the shoulder, secondary impingement on the cuff may occur and elicit symptoms. Impingement may also result from a decrease in the functional area of the supraspinatus outlet due to post-traumatic changes of the humeral head or the acromion or to hypertrophy of the myotendinous unit of the rotator cuff secondary to physical conditioning.

Before ultrasound and magnetic resonance imaging were developed, plain radiography was the primary diagnostic imaging study used to evaluate pain in the shoulder. While plain radiographs are helpful in the evaluation of osseous anatomy and abnormalities, they provide only indirect evidence of pathological changes of the rotator cuff. The best indicator of a torn rotator cuff on an anteroposterior radiograph of the shoulder made with the upper extremity in neutral rotation is a distance of less than six millimeters between the humeral head and the acromion⁸². Unfortunately, this is a very late finding in the natural history of degeneration of the cuff and, when present, usually indicates a very large or massive tear. The supraspinatus outlet radiograph has recently been used to assess the shape of the acromion, but because a plain radiograph represents a two-dimensional projection image of a three-dimensional structure it is frequently difficult to determine the true shape of the acromion. Interobserver variability in the assessment of the shape also makes it difficult to interpret this radiograph²⁴.

With the direct multiplanar capabilities of magnetic resonance imaging, it is possible to depict all of the osseous and soft-tissue structures that make up the coracoacromial arch and that may encroach on the supraspinatus outlet. It is important to assess the morphology of the acromial process. The precise shape of the acromion can be determined on the oblique sagittal image immediately lateral to the acromioclavicular joint and is defined as straight (type I; Figs. 2-A and 2-D), curved (type II; Fig. 2-B), or hooked (type III; Fig. 2-C). In addition to assessing the shape of the acromion, it is important to determine whether there is an abnormal tilt of the acromion inferolaterally, anteriorly, or posteriorly, which may decrease the volume of the supraspinatus outlet. The detection of an os acromiale is also important, as this has been associated with a tear of the rotator cuff. It is easy to detect an os acromiale if the axial magnetic resonance images include the entire acromion and acromioclavicular joint. Unfortunately, many magnetic resonance imaging studies of the shoulder do not include a complete set of axial images; however, as long as the study contains an oblique sagittal sequence, it is still possible to diagnose an os acromiale by observing the location of the insertion of the coracoacromial ligament. If the ligament inserts on an ossicle that is separate from the remainder of the acromion then an os acromiale is present⁸⁰ (Figs. 3-A and 3-B).

The appearance of a normal rotator cuff on a magnetic resonance image is similar to that of other tendons in the body. With its high collagen content, a normal cuff demonstrates little or no signal intensity on spin-echo sequences⁸³. A normal cuff may demonstrate increased signal intensity, typically medial to its insertion site, because of the orientation of the collagen fibers in the cuff relative to the main magnetic field in the magnetic resonance scanner. When collagen fibers are oriented approximately 55 degrees to the main magnetic field, increased signal intensity within the substance of the cuff may be detected on T1-weighted and protondensity-weighted sequences. This area of increased signal intensity is not seen on a T2-weighted sequence. This phenomenon is referred to as the magic-angle effect and should not be mistaken for evidence of a tear or degeneration of the cuff⁷⁰. In addition, a normal cuff does not have any morphological changes such as attenuation, hypertrophy, or fraying.

Abnormalities of the rotator cuff are seen as altered morphology along with abnormal signal intensity. Inflammation, degeneration, or intrasubstance tears, or all three, whether elicited by extrinsic impingement or intrinsic overload, are detected on a T1-weighted or proton-density-weighted sequence as a focus of increased signal intensity because of increased hydration of the tissue. In one study, eight of ten patients who had clinically suspected impingement syndrome had abnormal signal intensity within the rotator cuff on magnetic resonance images²⁷. Degeneration and inflammation were demonstrated in all biopsy specimens from the cuff of five of these eight patients.

The signal intensity on a T2-weighted sequence depends on the degree of hydration because of the pathological changes in the collagen fascicles and the type of reactive tissue in the tendon. Generally, the greater the hydration of the tissue, the higher the signal intensity on the T2-weighted sequence. The morphology of the cuff may be altered (attenuated or hypertrophied), and the margins of the cuff may be ill defined but not focally torn. If there is fluid or edema in the tissues surrounding a degenerated or inflamed cuff, it will be detected on a spin-echo T2-weighted or short tau inversion recovery sequence as a focus of high signal intensity. Fluid within the subacromial or subdeltoid bursa is detected easily on fat-saturated T2-weighted sequences. Kjellin et al.²⁸, in a study of cadavera, compared the abnormalities detected in the rotator cuff on magnetic resonance images with the histological findings. Areas of the cuff that demonstrated abnormally increased signal intensity on spin-echo proton-density-weighted sequences but no increase in intensity on T2-weighted sequences corresponded to areas of eosinophilic, fibrillar, or mxyoid degeneration as well as areas of fibrosis. Areas demonstrating increased signal intensity on a T2-weighted sequence corresponded to areas of severe degeneration and disruption of the fibers (Figs. 4-A and 4-B). Degenerated supraspinatus tendons demonstrating increased signal intensity on T2-weighted and gradient-echo sequences have also been shown to contain interstitial tears, fibrocartilaginous metaplasia, and fatty infiltration⁴⁰.

A partial-thickness tear of the rotator cuff is detected on spin-echo T2-weighted sequences as a focal area of disrupted fibers on the bursal or articular surface of the cuff (Figs. 5-A and 5-B) or within the substance of the cuff. It is frequently possible to determine whether the partial tear involves more or less than 50 per cent of the thickness of the cuff, which may help the surgeon to determine whether the cuff should be debrided or repaired. Discontinuity or detachment of the tendon along with fluid extending from the articular to the bursal surface appears to be the most specific finding on magnetic resonance images for the diagnosis of a full-thickness tear¹¹ (Figs. 6-A, 6-B, and 6-C). By imaging the tear in at least two planes, it is possible to measure the size and determine the location of the tear accurately. Unfortunately, fluid is not always detected extending through a full-thickness tear, particularly when the tear is chronic and has generated a fibrous reaction in the peritendinous tissue⁵⁶ (Figs. 7-A and 7-B). In these cases, assessment of the morphology of the cuff or detection of retraction of the cuff may provide the information needed to make an accurate diagnosis. With magnetic resonance imaging, it is also possible to determine the degree of retraction of the cuff, the quality of the torn cuff fibers, and the degree of associated muscle atrophy (Fig. 8). The residual function of the muscles of a torn cuff may be determined by assessment of the muscle morphology^41,42.

Isolated full-thickness tears of the subscapularis segment of the cuff may be difficult to detect clinically¹⁷; however, magnetic resonance imaging provides an excellent means to detect these tears⁵² (Fig. 9). On magnetic resonance images, it is also possible to detect partial detachment of the deep fibers of the superior segment of the subscapularis tendon that may be associated with medial subluxation of the biceps tendon (Fig. 10). If the deep fibers are detached completely, medial intra-articular dislocation of the biceps tendon is possible. In dissection studies, Clark and Harryman⁷ detected a tendinous sleeve surrounding the biceps tendon where it enters the bicipital groove, with the subscapularis fibers located under and the supraspinatus fibers located over the biceps tendon. It also appears that the superficial subscapularis fibers are continuous with a fibrous fascia that extends over the biceps tendon⁸¹.

Iannotti et al.²³ reported on the efficacy of magnetic resonance imaging in the evaluation of ninety-one patients who had an operative procedure for dysfunction of the shoulder and in the assessment of fifteen asymptomatic volunteers. With regard to the detection of a complete tear, magnetic resonance imaging was 100 per cent sensitive and 95 per cent specific. Tendinitis was defined arthroscopically as an area of hyperemia on the undersurface of the cuff or as thickening of the subacromial bursa. Degeneration or a partial tear was defined arthroscopically as fraying or fibrillation of the cuff. Magnetic resonance imaging had a sensitivity of 82 per cent and a specificity of 85 per cent with regard to the differentiation between tendinitis and degeneration of the cuff and a sensitivity of 93 per cent and a specificity of 87 per cent with regard to differentiation of a normal tendon from one showing signs of impingement. Those authors concluded that high-resolution magnetic resonance imaging is an excellent non-invasive tool for the diagnosis of disorders of the rotator cuff mechanism. The magnetic resonance images in that study were made and interpreted by musculoskeletal radiologists with extensive experience in magnetic resonance imaging.

In addition to the study by Iannotti et al.²³, there have been reports recently on the high accuracy of magnetic resonance images made with the new fast-imaging techniques to detect full-thickness tears of the cuff^54,65,68. Robertson et al.⁵⁸ recently demonstrated that full-thickness tears can be diagnosed accurately with magnetic resonance imaging with little interobserver variation with regard to the correct diagnosis. However, interobserver variation was much greater for the diagnosis of partial-thickness tears, tendinitis, and normal cuffs.

The sensitivity and specificity of magnetic resonance imaging in the diagnosis of a partial-thickness tear of the rotator cuff is lower than those for detection of a full-thickness tear. Traughber and Goodwin⁷⁵, in a comparison of the results of magnetic resonance imaging with the findings at arthroscopy in twenty-eight patients, detected five of five full-thickness tears but only four of nine partial-thickness tears. Hodler et al.²² compared standard magnetic resonance imaging with magnetic resonance arthrography for the diagnosis of partial-thickness tears. Whereas only one of thirteen partial tears was detected on standard magnetic resonance images, six of the thirteen tears were detected on magnetic resonance arthrograms. Both Palmer et al.⁵⁰ and Karzel and Snyder²⁶ reported on the improved rate of detection with magnetic resonance arthrography compared with standard magnetic resonance imaging for partial and full-thickness tears. Palmer et al. performed both fat-suppressed and non-fat-suppressed oblique coronal T1-weighted sequences for thirty-seven patients along with magnetic resonance arthrography. They detected sixteen of sixteen full-thickness tears with both techniques but only three of eight partial-thickness tears without fat suppression; all eight were detected with fat suppression. Since magnetic resonance arthrography is more invasive, costly, and time-consuming than standard magnetic resonance imaging, its efficacy must be proved with well designed prospective clinical studies before it can be recommended on a routine basis. It is possible that in selected patients, such as athletes who perform repetitive overhead activities, magnetic resonance arthrography with special positioning of the shoulder may be efficacious⁷³.

Occasionally, a patient is seen because of recurrent symptoms after repair of the rotator cuff. The same magnetic resonance imaging studies performed preoperatively can be performed to evaluate the cuff postoperatively⁴⁹. Gaenslen et al.¹⁵ evaluated twenty-nine patients (thirty shoulders) postoperatively with magnetic resonance imaging, and they correctly diagnosed sixteen of nineteen full-thickness and five of six partial-thickness tears detected at the time of a reoperation on the cuff. If there is any type of metallic hardware in the shoulder, fat-saturated spin-echo sequences should be avoided because of amplification of metallic artefacts. Karzel and Snyder²⁶ reported a higher rate of detection of postoperative full-thickness tears with magnetic resonance arthrography than with conventional magnetic resonance imaging. Whereas the size of a recurrent tear appears to be related to the degree of dysfunction of the shoulder²¹, it is important to remember that a full-thickness tear after an operation may be asymptomatic. Calvert et al.³ performed arthrography on twenty patients after repair of the rotator cuff and demonstrated leakage of contrast material, indicating a full-thickness tear, in eighteen of them. Seventeen of the eighteen patients were asymptomatic at the time of the arthrography.

Although magnetic resonance imaging is very sensitive in the detection of tendon disorders, it cannot determine their clinical importance. Knowledge of the appearance of the rotator cuff on magnetic resonance images of asymptomatic subjects is necessary before the clinical relevance of changes detected in the cuffs of symptomatic patients can be determined^34,46,63. Sher et al.⁶³ imaged the shoulders of ninety-six asymptomatic individuals, the youngest of whom was nineteen years old. The over-all prevalence of full-thickness tears was 15 per cent (fourteen), and the over-all prevalence of partial-thickness tears was 20 per cent (nineteen). However, when the different age-groups were considered, no full-thickness tear and only one partial-thickness tear (4 per cent) was diagnosed in the twenty-five subjects who were nineteen to thirty-nine years old, and one full-thickness tear (4 per cent) and six partial-thickness tears (24 per cent) were diagnosed in the twenty-five subjects who were forty to sixty years old. Of the forty-six individuals who were more than sixty years old, thirteen (28 per cent) were diagnosed with a full-thickness tear and twelve (26 per cent), with a partial-thickness tear. Miniaci et al.³⁴ evaluated thirty shoulders in twenty asymptomatic subjects who were seventeen to forty-nine years old. Although they detected abnormal signal intensity within the cuffs of these subjects, none had evidence of a full-thickness tear. Needell et al.⁴³ evaluated the prevalence of peritendinous and bone abnormalities in the same asymptomatic cohort of ninety-six patients studied by Sher et al.⁶³. In 75 per cent of the subjects, they detected osteoarthrosis of the acromioclavicular joint, which correlated more closely with age than with abnormalities of the rotator cuff. Unfortunately, they did not categorize the type of osteoarthrosis; specifically they did not state whether there were osteophytes projecting off the inferior margin of the joint line. The prevalence of subacromial spurs was 12 per cent (three of twenty-six) in the subjects who were nineteen to thirty-nine years old and 52 per cent (twenty-five of forty-eight) in the subjects who were sixty-one to eighty-eight years old, and the prevalence of cysts of the humeral head was 8 per cent (two of twenty-six) and 40 per cent (nineteen of forty-eight), respectively. Both of these changes correlated closely with the severity of the rotator cuff abnormalities. The presence of fluid in the subacromial or subdeltoid bursa also correlated with abnormalities of the cuff. While cysts of the humeral head, subacromial spurs, and fluid within the subacromial or subdeltoid bursa may indicate an abnormal rotator cuff, it cannot be inferred that their presence reflects a symptomatic condition.

Muscle Injury

Muscle injury is one of the most frequent causes of musculoskeletal dysfunction related to sports activities. Muscle contusions occur most frequently in the lower extremities but may also occur in the shoulder girdle. Acute disruption of muscle fibers may result in intramuscular hemorrhage or a hematoma as well as incite an inflammatory response associated with interstitial edema. Because of the acute pain associated with a muscle injury, it may be difficult to determine the precise location and severity of the injury with a physical examination. Before the availability of magnetic resonance imaging, radiographic studies were of little value in the assessment of acute muscle injuries. On plain radiographs, there may be obliteration of the fat planes surrounding an injured muscle secondary to the perimuscular edema. With computerized tomography, the size or contour of a muscle may be seen to be altered but the detection of intramuscular hemorrhage, edema, or hematoma is limited. Because of the excellent soft-tissue contrast resolution provided by magnetic resonance imaging, it is now possible to determine the nature and extent of a muscle injury.

A muscle contusion is indicated by abnormal signal intensity and morphology of a muscle on magnetic resonance images. On spin-echo sequences, a normal muscle demonstrates intermediate signal intensity on T1-weighted and proton-density-weighted sequences and low signal intensity on T2-weighted sequences. In a contused muscle, the interstitial edema or hemorrhage, or both, is seen as high signal intensity on T2-weighted sequences. Since hemorrhage infiltrates through a muscle and mixes with the interstitial edema, it is not possible to separate it from the edematous muscle tissue. With a grade-1 contusion (failure of microstructural fiber), the size of the muscle may be slightly increased and the margins may have a feathery appearance because of the extension of interstitial edema into the perimuscular tissue³⁵. Edematous changes in the adjacent subcutaneous fat are also frequently detected. With a grade-2 muscle contusion (a partial macroscopic tear), there is a focus of disrupted muscle fibers in addition to the altered signal intensity from the interstitial edema and hemorrhage. A grade-3 muscle contusion appears similar to a grade-2 contusion, except that there is complete disruption of the muscle fibers. A muscle hematoma is depicted as a focus of intermediate or high signal intensity on a T1-weighted sequence, depending on the age and chemical composition of the hematoma, and as a focus of high signal intensity on a T2-weighted sequence. It is also possible to detect the sequelae of a muscle contusion, which include muscle atrophy, fibrosis, calcification, and ossification¹⁸.

Muscle strain is probably the most common type of injury of the myotendinous unit. A muscle strain is an acute, stretch-induced injury secondary to excessive indirect force generated by eccentric muscular contraction. Such strains may involve the muscles stabilizing the shoulder joint. The pain elicited by an acute muscle strain typically occurs during an athletic activity or immediately at its termination. The pathological changes in an acutely strained muscle include microtraumatic disruption of the muscle fibers near its myotendinous junction associated with edema and hemorrhage. Because the tendon of a multipennate muscle extends into the muscle belly, the symptoms of a strain may be located anywhere within the muscle and not merely at its ends. A grade-1 muscle strain appears similar to a grade-1 contusion on magnetic resonance images. The muscle may be enlarged secondary to interstitial edema or hemorrhage, or both, and there is increased signal intensity within the muscle on a spin-echo fat-saturated T2-weighted sequence (Fig. 11). Clinically, a grade-2 muscle strain presents as muscle pain associated with a loss of strength. Pathologically, there is a macroscopic partial tear of the myotendinous unit. On the magnetic resonance image, there is a partial tear of the muscle fibers associated with edema or hemorrhage, or both. With a grade-3 strain, there is complete disruption of the myotendinous unit. In addition to the evaluation of muscle strains, it is also possible to detect the changes of delayed-onset muscle soreness^48,61 and to study muscle physiology with magnetic resonance imaging^53,62.

Magnetic resonance imaging can also detect changes in a muscle after injury to its innervation. Fleckenstein et al.¹³ and Uetani et al.⁷⁸ reported on the appearance of acute, subacute, and chronic muscle denervation on magnetic resonance images. With acute denervation, there is no detectable difference in the appearance of normally and abnormally innervated muscles. With subacute denervation, there is an increase in the T1 and T2 relaxation values of the denervated muscles. Short tau inversion recovery sequences or fat-suppressed spin-echo T2-weighted sequences are most sensitive for the detection of the increased signal intensity related to the prolonged T2 relaxation value of a muscle. Findings of subacute denervation due to paralabral ganglion cysts compressing the suprascapular nerve at the level of the spinoglenoid notch have been detected in the infraspinatus muscle⁷⁹ (Fig. 12). In chronically denervated muscle, atrophy and fatty infiltration is best identified on T1-weighted sequences. The determination of exactly which muscle groups are affected by denervation may make it possible to explain functional loss and to determine treatment. Uetani et al. demonstrated that the abnormal signal intensity initially identified in denervated muscle reverted to normal with the return of neural function after therapeutic intervention. To my knowledge, it is not yet known whether the abnormal signal intensity detected in the infraspinatus muscle in patients who have a ganglion cyst reverts to normal after the cyst is removed.

Whenever a patient is seen for muscular dysfunction, it is necessary to rule out a primary neuropathy or myopathy as the cause. The diagnosis of a neuropathy or a myopathy that alters the morphology or signal intensity of a muscle is possible on magnetic resonance images. Direct evaluation of the brachial plexus is also possible, along with assessment of the major nerves innervating the shoulder girdle. Masses in the suprascapular and spinoglenoid notch or in the quadrilateral space that may be compressing the suprascapular or axillary nerves can be easily diagnosed with magnetic resonance imaging as long as the study includes the appropriate anatomical regions. Direct communication between the referring physician and the radiologist is frequently needed if the maximum amount of information is to be generated for the assessment of a patient who has a suspected neuropathy or myopathy.

Although magnetic resonance imaging is extremely sensitive to pathological changes within a muscle, it lacks specificity. Any pathological process that alters muscle morphology, incites an inflammatory response, or increases hydration of the muscle can be detected with magnetic resonance imaging. The results of even a benign procedure, such as an intramuscular injection, can be detected on a magnetic resonance image as a focus of abnormal signal intensity in the muscle and perifascial tissue. Precise correlation of the abnormal findings on the magnetic resonance images with the history and the findings of the physical examination is necessary to determine the clinical relevance of abnormal findings and to help to narrow down the differential diagnosis.

Instability

Instability of the shoulder can present in a variety of ways, and the radiographic approach should be tailored to the clinical situation. For the diagnosis of acute dislocation, the role of imaging is to confirm the presence and direction of the dislocation, to identify associated fractures or other injuries, and to evaluate the adequacy of reduction. Plain radiographs are usually adequate for the evaluation of an acute dislocation. Computerized tomography helps to clarify the nature of a complex fracture-dislocation in cases of a failed reduction with intra-articular bone fragments and in cases of chronic dislocation. Magnetic resonance imaging may help the physician to evaluate the etiology of an irreducible dislocation caused by a dislocated biceps tendon or an avulsed rotator cuff or labrum.

Magnetic resonance imaging in the evaluation of instability is typically applicable to patients who have symptoms or signs suggestive of recurrent subluxation or dislocation of the shoulder and for whom the history and physical examination are inconclusive. Conventional magnetic resonance imaging provides objective evidence of instability by detecting abnormalities of the capsulolabral complex, humeral head, and glenoid. Magnetic resonance imaging also provides information concerning the remainder of the shoulder girdle that may be causing the dysfunction.

In order to diagnose a torn labrum or torn glenohumeral ligaments accurately, it is necessary to know the normal range of appearance of these structures^30,45. A normal anterior aspect of the labrum may be triangular, crescentic, cleaved, notched, or round and may be a different shape on each side of the body. Even though the labrum is composed of fibrocartilage, it may demonstrate a globular or linear focus of increased signal intensity on T1-weighted, gradient-echo, and proton-density-weighted images. As with the rotator cuff, the orientation of the collagen fibers in the labrum with respect to the main magnetic field may cause increased signal intensity that is detected in the labrum on T1-weighted and proton-density-weighted images. The posterior aspect of the labrum is normally smaller than the anterior aspect and is typically triangular. Investigators describing the appearance of a normal glenoid labrum have typically used magnetic resonance images of asymptomatic patients^30,45. Unfortunately, this does not guarantee that the labra were normal.

An acute traumatic dislocation may damage the labrum, capsulolabral complex, supporting myotendinous structures, rim of the glenoid, or humeral head. The role of magnetic resonance imaging is to define precisely the location and extent of the injury of these structures. In the acute situation, the presence of joint fluid helps in the evaluation of the labrum. The labrum may be torn, detached, or displaced after dislocation of the shoulder³³. An anterior tear frequently is detected on magnetic resonance images as a mass of tissue contiguous with deformed or stripped glenohumeral ligaments (Figs. 13-A and 13-B). A torn or deformed middle or inferior glenohumeral ligament may be detected with magnetic resonance arthrography^5,51, but it may be difficult to assess with conventional magnetic resonance imaging unless there is a large joint effusion. A strain or tear of the subscapularis muscle or tendon may be associated with an anterior dislocation. Rarely, there is stripping of the inferior aspect of the capsule from the humeral neck with an acute dislocation. This condition is difficult to diagnose clinically, but it can be diagnosed on a magnetic resonance image (Fig. 14).

Additional findings indicative of previous anterior traumatic dislocation include a Hill-Sachs deformity, a Bankart lesion, and capsular stripping. A Hill-Sachs deformity is an osteochondral defect in the superior aspect of the posterolateral margin of the humeral head that results from impaction of the humeral head on the rim of the glenoid at the time of anterior dislocation of the shoulder. This deformity is detected on a magnetic resonance image as an indentation in the humeral head. If there has been a recent episode of anterior instability, there may be high signal intensity in the cancellous bone surrounding the deformity on a T2-weighted image. A cyst sometimes forms subjacent to a Hill-Sachs deformity in patients who have a history of multiple dislocations. A Bankart lesion is a tear of the anterior capsulolabral complex that occurs with anterior dislocation of the shoulder. If there is also a fracture of the anterior part of the glenoid rim, the abnormality is referred to as an osseous Bankart lesion. The detection of a chronic osseous Bankart lesion may be subtle on a magnetic resonance image, but acute or subacute lesions may be associated with osseous edema in the rim of the glenoid or with anterior capsular stripping. While such stripping with associated pericapsular edema indicates an episode of anterior instability, a redundant anterior aspect of the capsule can be seen in asymptomatic subjects and is of uncertain clinical relevance⁴⁵.

Signs of posterior instability include tears of the posterior aspect of the labrum and posterior capsular stripping (Fig. 15). For patients in whom instability is clinically suspected but the labrum appears normal on the initial standard imaging sequences, additional sequences made with the upper extremity in internal and external rotation may help to demonstrate a torn labrum. Special positioning of the upper extremity has also been used to assess the posterosuperior segment of the labrum⁷².

Several normal findings or anatomical variants depicted on a magnetic resonance image may be mistaken for a torn labrum. There is a normal zone of hyaline cartilage interposed between the labrum and the rim of the glenoid that manifests increased signal intensity on T1-weighted, proton-density-weighted, and gradient-echo sequences and that should not be misdiagnosed as a tear. Normal anatomical variants, such as a sublabral foramen at the level of the subscapularis recess or a Buford complex, may also be mistaken for a labral tear⁷⁷. If there is no fluid in the shoulder joint, abutment of the glenohumeral ligaments against the anterior aspect of the labrum may simulate a deformed or torn labrum³⁰.

With use of arthroscopy as the so-called gold standard, the accuracy of magnetic resonance imaging in the detection of labral tears exceeds 90 per cent. Legan et al.²⁹, with use of magnetic resonance imaging, accurately diagnosed fifty-four of fifty-seven anterior labral tears, and Gusmer et al.²⁰ diagnosed thirty-seven of thirty-seven anterior labral tears detected at operation. Legan et al. diagnosed four anterior labral tears that were not present at operation and Gusmer et al., three. Anterior labral tears seem to be easier to detect than posterior tears²⁹. The clinical value of magnetic resonance images is directly affected by factors such as the type of imaging equipment used, the completeness of the imaging sequences, the type of cohort population, and the specialization of the radiologists interpreting the studies. Garneau et al.¹⁶ reported poor interobserver and intraobserver variability in the detection of labral abnormalities with magnetic resonance imaging.

Other findings of instability on magnetic resonance images include osseous changes in the anterior or posterior margin of the glenoid; glenoid dysplasia⁷⁶ or abnormal tilt; subluxation of the humeral head; and abnormalities of the rotator cuff, especially the subscapularis muscle and tendon. Glenoid labral cysts are typically associated with labral tears and may indicate the direction of instability. Tirman et al.⁷¹ reported evidence of a labral cyst on the magnetic resonance images of twenty patients; all of the cysts were associated with a labral tear. Eleven of the twenty patients had clinical evidence of instability in the direction of the labral tear and cyst. The labral tear and cyst were located superiorly in nine patients, posteriorly in nine, and anteriorly in two. The cyst extended into the spinoglenoid notch in six patients, the suprascapular notch in three, and both in four. In eight patients, the labral tear as well as a communication between the joint and the cyst was confirmed at an operation.

The shoulder joint normally contains only a small amount of fluid, making evaluation of the joint capsule and the glenohumeral ligaments difficult. Magnetic resonance arthrography has been used to overcome this limitation by the injection of saline solution or a dilute solution of a magnetic resonance contrast agent into the joint^12,74. Distention of the joint capsule facilitates the evaluation of the capsulolabral complex similar to the way that it does with computerized tomographic arthrography. Several recent studies have demonstrated the superiority of magnetic resonance arthrography compared with computerized tomographic arthrography for the evaluation of instability of the shoulder^6,25,59. In addition, magnetic resonance arthrography is reportedly more accurate than standard magnetic resonance imaging for the detection of capsulolabral abnormalities^6,26. The major disadvantage of magnetic resonance arthrography is that it transforms a standard non-invasive magnetic resonance imaging study into a limited invasive study, which is more time-consuming and more expensive. In addition, magnetic resonance contrast agents have not been approved by the Food and Drug Administration for intra-articular use. Although magnetic resonance arthrography may be useful in a select subset of patients, such as those who have posterosuperior impingement⁷², it must be clarified whether its increased sensitivity comes at the expense of decreased specificity and, in addition, how the information affects the management of the patient and the outcome. Considering the recent report by Gusmer et al.²⁰, which showed conventional magnetic resonance imaging to be highly accurate for the detection of labral abnormalities, a conventional study should probably remain the initial screening examination if magnetic resonance imaging is needed to evaluate a patient who has symptoms of instability.

Although labral tears are typically associated with clinical or occult instability, they also may be found in patients who have a stable shoulder and pain, clicking, or locking. This has been reported in athletes who participate in overhead activities⁵⁵ as well as in non-athletes⁴⁷. A torn labrum may appear frayed, blunted, or attenuated and may contain abnormal signal intensity extending to its articular surface. Superior labral tears involving the biceps anchor are detected both in athletes who participate in overhead sports and in non-athletes. Snyder et al.⁶⁶ reported tears of the superior aspect of the labrum that may extend anteriorly or posteriorly (SLAP lesions), and they classified these tears according to the type of labral deformity and the integrity of the biceps anchor. A type-I lesion represents fraying of the superior aspect of the labrum, which is attached to the rim of the glenoid; a type-II lesion, avulsion of the superior aspect of the labrum and the biceps anchor from the superior aspect of the rim; a type-III lesion, a tear of the superior aspect of the labrum, which is displaced into the shoulder joint (similar to a bucket-handle tear of the meniscus); and a type-IV lesion, a tear of the superior aspect of the labrum associated with a partially detached biceps anchor (part of the biceps inserts into the torn labrum and part, into the supraglenoid tubercle). The reported accuracy of magnetic resonance imaging in detecting these lesions has varied greatly in the literature. Some studies demonstrated the need for intra-articular contrast medium to make the diagnosis²⁶, whereas in other studies routine magnetic resonance imaging sequences successfully demonstrated the tear of the superior aspect of the labrum that extended anteriorly or posteriorly^4,39. Both oblique coronal and axial sequences are needed for optimum detection of the presence and extent of the tear (Figs. 16-A and 16-B).

The application of magnetic resonance imaging for the assessment of instability of the shoulder is highly dependent on the physician managing the patient. If an orthopaedic surgeon uses arthroscopy to repair Bankart-type lesions and arthrotomy to repair all other types of instability, magnetic resonance imaging may be beneficial in the preoperative planning³⁶. In difficult cases, magnetic resonance imaging may be needed to provide objective evidence of instability. While magnetic resonance imaging appears to have contributed important information to the care of patients^36,72, it is still necessary for each clinician to determine its appropriate role, in his or her particular medical community, in the assessment of patients who have symptoms of instability of the shoulder.

Suder et al.⁶⁹ evaluated patients after traumatic primary anterior dislocation of the shoulder and found that conventional magnetic resonance imaging was only moderately reliable in the preoperative evaluation of labral tears and Hill-Sachs deformities and that it provided little useful information concerning capsulolabral lesions. Green and Christensen¹⁹, in a study of the magnetic resonance images of thirty-three patients who had possible anterior instability of the shoulder, detected only twenty-one of twenty-eight labral tears that were present at operation. They concluded that "magnetic resonance imaging is not useful in the surgical planning for most patients with obvious anterior shoulder instability." In a study of forty-one labral tears that were detected at operation, Liu et al.³¹ accurately diagnosed thirty-seven tears on physical examination but only twenty-four on magnetic resonance images. Two tears that were diagnosed on physical examination and with magnetic resonance imaging were not detected at operation. It is hoped that with the continued improvement in magnetic resonance imaging technology, such as three-dimensional imaging of the labrum³², its value in the assessment of instability of the shoulder will continue to improve.

Additional Applications

Trauma

After a direct-impaction injury to an extremity, it is fairly common for an individual to have normal findings on plain radiography despite pain involving an osseous structure. Before the availability of magnetic resonance imaging, the precise etiology of this pain was unclear; it was uncertain whether it was related to injury of soft-tissue or osseous structures, or both. With the exquisite sensitivity of magnetic resonance imaging for the detection of bone marrow edema, it became readily apparent that many of these patients had areas of edema in the cancellous bone at the site of an osseous injury. These areas were designated bone contusions or bruises, and it was hypothesized that the osseous edema was due to trabecular microfractures. Recently, Rangger et al.⁵⁷ reported that biopsy of two foci of osseous edema in a knee demonstrated a reparative process of fractured cancellous bone. Bone contusions may occur secondary to an extrinsic impaction injury or to bones impacting against one another as a result of acute instability or malalignment. It is important to detect a bone contusion since it may explain the etiology of pain and eliminate the need for additional diagnostic tests.

In the shoulder, contusions in the posterior superolateral margin of the humeral head and the anterior margin of the glenoid may be detected after an episode of anterior instability. Contusions of the lesser tuberosity and the posterior margin of the glenoid are seen after a posterior dislocation. Contusions of the greater tuberosity are frequently detected after a fall while skiing. In addition, magnetic resonance imaging can detect non-displaced or minimally displaced fractures of the greater tuberosity in these patients (Fig. 17). These fractures are frequently difficult to detect on plain radiographs.

Osteonecrosis

Atraumatic osteonecrosis of the shoulder may be detected in patients who are taking steroids (Fig. 18) or those who have sickle-cell or marrow-storage disease. The appearance of the osteonecrosis depends on its stage of evolution. The diagnosis is based on detection of a well marginated focus of abnormal subchondral cancellous bone that may be surrounded by a rim of reactive bone. Collapse of the subchondral bone plate as well as secondary degenerative changes may be detected. This appears to be less common in the humeral head than in the femoral head, most likely because of the non-weight-bearing function of the humeral head. Osteonecrosis of the humeral head may also be detected after a comminuted fracture that disrupts its blood supply.

Tumor

Patients who have a malignant process involving the scapula, the humeral head, or the humeral neck may be seen for pain in the shoulder. Because of osteopenia of the shoulder girdle, which is common in older patients, it is often difficult to detect early osseous destruction on plain radiographs. As in other areas of the body, magnetic resonance imaging provides an excellent means with which to evaluate the presence and extent of either primary or metastatic tumors of the shoulder girdle. T1-weighted sequences are used to define the interface between the tumor and the fat in the cancellous bone or at the cortical surface, and fat-suppressed T2-weighted or short tau inversion recovery sequences are used to define extension of the tumor into soft tissue. With magnetic resonance imaging, it is possible to detect benign tumors, such as a scapular exostosis, that are precipitating symptoms of a snapping scapula.

Magnetic resonance imaging also provides the most comprehensive evaluation of soft-tissue tumors. Tumors that originate in synovial tissue, such as diffuse or focal pigmented villonodular synovitis and synovial chondromatosis, or tumors that are located within the muscles or perimuscular tissue of the shoulder girdle may be assessed completely with magnetic resonance imaging. As in other joints of the body, pigmented villonodular synovitis typically presents with foci of nodular tissue demonstrating low signal intensity on spin-echo T1-weighted and T2-weighted sequences because of its hemosiderin content.

Arthritis and Osteoarthrosis

Patients who have a systemic arthritis, such as rheumatoid arthritis or ankylosing spondylitis, may be seen for symptoms about the shoulder. Plain radiographs are usually sufficient to evaluate these patients, unless secondary complications such as an insufficiency fracture or infection develop.

Osteoarthrosis of the glenohumeral joint is relatively infrequent, and it may be difficult to detect on plain radiographs. Because the articular cartilage of the humeral head and glenoid fossa is normally two to three millimeters thick, it is extremely difficult to detect isolated lesions of the articular cartilage. Once subchondral edema or cystic changes are present as part of a degenerative condition, they are easily detected with the appropriate proton-density or T2-weighted sequences (Fig. 19). Small loose bodies are difficult to detect unless there is at least a little fluid within the joint. Small marginal osteophytes of the rim of the glenoid or the junction of the humeral head and neck are also difficult to detect with magnetic resonance imaging because of the soft tissues inserting into these regions. Magnetic resonance imaging provides an excellent assessment of arthropathy of the rotator cuff by depicting the osseous alterations of the glenohumeral joint and the supporting soft-tissue structures.

Inflammation and Infection

Soft-tissue infection, septic arthritis, and osteomyelitis of the shoulder girdle are unusual causes of pain about the shoulder but they must be considered, particularly in a patient who is immunosuppressed or is receiving steroids. Any inflammatory process involving the fascia or the muscles can be detected with magnetic resonance imaging because of the increased tissue hydration resulting from the influx of cellular elements and the associated interstitial edema. One of the main applications of magnetic resonance imaging in patients who have a soft-tissue infection is the determination of whether there is a focal abscess or evidence of muscle necrosis. Magnetic resonance imaging is extremely sensitive for detecting a joint effusion, osseous edema, and destruction of bone due to infection.

A common cause of pain about the shoulder is calcific bursitis. While bursitis is typically precipitated by the deposition of hydroxyapatite crystals, and is therefore readily diagnosed on plain radiographs, it can also be caused by an infectious agent such as tuberculosis. Fluid within the subacromial and subdeltoid bursa is easily detected on magnetic resonance images. In addition to detection of calcification within a fluid-filled bursa, it is possible to assess the condition of the synovial tissue and the subjacent rotator cuff on magnetic resonance images. It is also possible to detect so-called rice bodies within the bursa, which may be present with rheumatoid bursitis or, rarely, some infectious agents.

Overview

An imaging study should be ordered only when its results will directly affect the management of a patient. As a result of the static nature of most imaging studies, including magnetic resonance imaging, their interpretation is predominantly focused on the detection of isolated structural abnormalities. But static images are a snapshot of the evolution of a pathological condition, and their value is enhanced if they can elucidate the natural history of a pathological process. Magnetic resonance imaging has greatly increased our understanding of dysfunction of the shoulder. Magnetic resonance imaging can detect not only pathological conditions that are precipitating symptoms but also alterations in tissue resulting from chronic microtrauma or aging that do not precipitate symptoms. These subclinical abnormalities are not false-positive findings, as they represent pathological changes in the tissue. Although they may have no clinical relevance in the explanation of a patient's current symptoms, they may have prognostic importance if they reflect altered biomechanical properties of tissue. In the future, this information may be used to guide athletic training or to optimize patient rehabilitation.

While magnetic resonance imaging has improved the diagnosis and treatment of musculoskeletal disorders, its efficacy can be enhanced further by a closer working relationship among physicians who care for patients who have musculoskeletal disorders. Prospective, controlled studies are still needed to compare the cost-effectiveness of the various imaging and diagnostic modalities employed to evaluate the shoulder. Redundant and unnecessary studies must be eliminated if we hope to control the spiraling costs of health care.

References

Blevins, F. T.; Hayes, W. M.; and Warren, R. F.: Rotator cuff injury in contact athletes. Am. J. Sports Med.,24: 263-267, 1996.24263 1996 [PubMed]

Buckwalter, J. A.; Woo, S. L-Y.; Goldberg, V. M.; Hadley, E. C.; Booth, F.; Oegema, T. R.; and Eyre, D. R.: Current concepts review. Soft-tissue aging and musculoskeletal function. J. Bone and Joint Surg.,75-A: 1533-1548, Oct. 1993.75-A1533 1993

Calvert, P. T.; Packer, N. P.; Stoker, D. J.; Bayley, J. I. L.; and Kessel, L.: Arthrography of the shoulder after operative repair of the torn rotator cuff. J. Bone and Joint Surg.,68-B(1): 147-150, 1986.68-B(1)147 1986

Cartland, J. P.; Crues, J. V., III; Stauffer, A.; Nottage, W.; and Ryu, R. K.: MR imaging in the evaluation of SLAP injuries of the shoulder: findings in 10 patients. AJR: Am. J. Roentgenol.,159: 787-792, 1992.159787 1992 [PubMed]

Chandnani, V. P.; Gagliardi, J. A.; Murnane, T. G.; Bradley, Y. C.; DeBerardino, T. A.; Spaeth, J.; and Hansen, M. F.: Glenohumeral ligaments and shoulder capsular mechanism: evaluation with MR arthrography. Radiology,196: 27-32, 1995.19627 1995 [PubMed]

Chandnani, V. P.; Yeager, T. D.; DeBerardino, T.; Christensen, K.; Gagliardi, J. A.; Heitz, J. R.; Baird, D. E.; and Hansen, M. F.: Glenoid labral tears: prospective evaluation with MR imaging, MR arthrography, and CT arthrography. AJR: Am. J. Roentgenol.,161: 1229-1235, 1993.1611229 1993 [PubMed]

Clark, J. M., and Harryman, D. T., II: Tendons, ligaments, and capsule of the rotator cuff. Gross and microscopic anatomy. J. Bone and Joint Surg.,74-A: 713-725, June 1992.74-A713 1992

Crues, J. V., III: Technical considerations. In MRI of the Knee, edited by J. H. Mink, M. A. Reicher, J. V. Crues, III, and A. L. Deutsch. Ed. 2, pp. 1-23. New York, Raven Press, 1993.

Davis, S. J.; Teresi, L. M.; Bradley, W. G.; Ressler, J. A.; and Eto, R. T.: Effect of arm rotation on MR imaging of the rotator cuff. Radiology,181: 265-268, 1991.181265 1991 [PubMed]

Farley, T. E.; Neumann, C. H.; Steinbach, L. S.; and Petersen, S. A.: The coracoacromial arch: MR evaluation and correlation with rotator cuff pathology. Skel. Radiol.,23: 641-645, 1994.23641 1994

Farley, T. E.; Neumann, C. H.; Steinbach, L. S.; Jahnke, A. J.; and Petersen, S. S.: Full-thickness tears of the rotator cuff of the shoulder: diagnosis with MR imaging. AJR: Am. J. Roentgenol.,158: 347-351, 1992.158347 1992 [PubMed]

Flannigan, B.; Kursunoglu-Brahme, S.; Snyder, S.; Karzel, R.; Del Pizzo, W.; and Resnick, D.: MR arthrography of the shoulder: comparison with conventional MR imaging. AJR: Am. J. Roentgenol.,155: 829-832, 1990.155829 1990 [PubMed]

Fleckenstein, J. L.; Watumull, D.; Conner, K. E.; Ezaki, M.; Greenlee, R. G., Jr.; Bryan, W. W.; Chason, D. P.; Parkey, R. W.; Peslock, R. M.; and Purdy, P. D.: Denervated human skeletal muscle: MR imaging evaluation. Radiology,187: 213-218, 1993.187213 1993 [PubMed]

Fukuda, H.; Hamada, K.; Nakajima, T.; and Tomonaga, A.: Pathology and pathogenesis of the intratendinous tearing of the rotator cuff viewed from en bloc histologic sections. Clin. Orthop.,304: 60-67, 1994.30460 1994 [PubMed]

Gaenslen, E. S.; Satterlee, C. C.; and Hinson, G. W.: Magnetic resonance imaging for evaluation of failed repairs of the rotator cuff. Relationship to operative findings. J. Bone and Joint Surg.,78-A: 1391-1396, Sept. 1996.78-A1391 1996

Garneau, R. A.; Renfrew, D. L.; Moore, T. E.; el-Khoury, G. Y.; Nepola, J. V.; and Lemke, J. H.: Glenoid labrum: evaluation with MR imaging. Radiology,179: 519-522, 1991.179519 1991 [PubMed]

Gerber, C., and Krushell, R. J.: Isolated rupture of the tendon of the subscapularis muscle. Clinical features in 16 cases. J. Bone and Joint Surg.,73-B(3): 389-394, 1991.73-B(3)389 1991

Greco, A.; McNamara, M. T.; Escher, R. M.; Trifilio, G.; and Parienti, J.: Spin-echo and STIR MR imaging of sports-related muscle injuries at 1.5 T. J. Comput. Assist. Tomog.,15: 994-999, 1991.15994 1991

Green, M. R., and Christensen, K. P.: Magnetic resonance imaging of the glenoid labrum in anterior shoulder instability. Am. J. Sports Med.,22: 493-498, 1994.22493 1994 [PubMed]

Gusmer, P. B.; Potter, H. G.; Schatz, J. A.; Wickiewicz, T. L.; Altchek, D. W.; O'Brien, S. J.; and Warren, R. F.: Labral injuries: accuracy of detection with unenhanced MR imaging of the shoulder. Radiology,200: 519-524, 1996.200519 1996 [PubMed]

Harryman, D. T., II; Mack, L. A.; Wang, K. Y.; Jackins, S. E.; Richardson, M. L.; and Matsen, F. A.: Repairs of the rotator cuff. Correlation of functional results with integrity of the cuff. J. Bone and Joint Surg.,73-A: 982-989, Aug. 1991.73-A982 1991

Hodler, J.; Kursunoglu-Brahme, S.; Snyder, S. J.; Cervilla, V.; and Karzel, R. P.: Rotator cuff disease: assessment with MR arthrography versus standard MR imaging in 36 patients with arthroscopic confirmation. Radiology,182: 431-436, 1992.182431 1992 [PubMed]

Iannotti, J. P.; Zlatkin, M. B.; Esterhai, J. L.; Kressel, H. Y.; Dalinka, M. K.; and Spindler, K. P.: Magnetic resonance imaging of the shoulder. Sensitivity, specificity, and predictive value. J. Bone and Joint Surg.,73-A: 17-29, Jan. 1991.73-A17 1991

Jacobson, S. R.; Speer, K. P.; Moor, J. T.; Janda, D. H.; Saddemi, S. R.; MacDonald, P. B.; and Mallon, W. J.: Reliability of radiographic assessment of acromial morphology. J. Shoulder and Elbow Surg.,4: 449-453, 1995.4449 1995

Jahnke, A. H., Jr.; Petersen, S. A.; Neumann, C.; Steinbach, L.; and Morgan, F.: A prospective comparison of computerized arthrotomography and magnetic resonance imaging of the glenohumeral joint. Am. J. Sports Med.,20: 695-701, 1992.20695 1992 [PubMed]

Karzel, R. P., and Snyder, S. J.: Magnetic resonance arthrography of the shoulder. A new technique of shoulder imaging. Clin. Sports Med.,12: 123-136, 1993.12123 1993 [PubMed]

Kieft, G. J.; Bloem, J. L.; Rozing, P. M.; and Obermann, W. R.: Rotator cuff impingement syndrome: MR imaging. Radiology,166: 211-214, 1988.166211 1988 [PubMed]

Kjellin, I.; Ho, C. P.; Cervilla, V.; Haghighi, P.; Kerr, R.; Vangness, C. T.; Friedman, R. J.; Trudell, D.; and Resnick, D.: Alterations in the supraspinatus tendon at MR imaging: correlation with histopathologic findings in cadavers. Radiology,181: 837-841, 1991.181837 1991 [PubMed]

Legan, J. M.; Burkhard, T. K.; Goff, W. B., II; Balsara, Z. N.; Martinez, A. J.; Burks, D. D.; Kallman, D. A.; O'Brien, T. J.; and LaPoint, J. M.: Tears of the glenoid labrum: MR imaging of 88 arthroscopically confirmed cases. Radiology,179: 241-246, 1991.179241 1991 [PubMed]

Liou, J. T.; Wilson, A. J.; Totty, W. G.; and Brown, J. J.: The normal shoulder: common variations that simulate pathologic conditions at MR imaging. Radiology,186: 435-441, 1993.186435 1993 [PubMed]

Liu, S. H.; Henry, M. H.; Nuccion, S.; Shapiro, M. S.; and Dorey, F.: Diagnosis of glenoid labral tears. A comparison between magnetic resonance imaging and clinical examinations. Am. J. Sports Med.,24: 149-154, 1996.24149 1996 [PubMed]

Loehr, S. P.; Pope, T. L., Jr.; Martin, D. F.; Link, K. M.; Monu, J. U.; Hunter, M.; and Reboussin, D.: Three-dimensional MRI of the glenoid labrum. Skel. Radiol.,24: 117-121, 1995.24117 1995

McCauley, T. R.; Pope, C. F.; and Jokl, P.: Normal and abnormal glenoid labrum: assessment with multiplanar gradient-echo MR imaging. Radiology,183: 35-37, 1992.18335 1992 [PubMed]

Miniaci, A.; Dowdy, P. A.; Willits, K. R.; and Vellet, A. D.: Magnetic resonance imaging evaluation of the rotator cuff tendons in the asymptomatic shoulder. Am. J. Sports Med.,23: 142-145, 1995.23142 1995 [PubMed]

Mink, J. H.: Muscle injuries. In MRI of the Knee, edited by J. H. Mink, M. A. Reicher, J. V. Crues, III, and A. L. Deutsch. Ed. 2, pp. 406-408. New York, Raven Press, 1993.

Minkoff, J., and Cavaliere, G.: Glenohumeral instabilities and the role of magnetic resonance imaging techniques. The orthopedic surgeon's perspective. Magnet. Reson. Imag. Clin. North America,1: 105-123, 1993.1105 1993

Mirowitz, S. A.: Fast scanning and fat-suppression MR imaging of musculoskeletal disorders. AJR: Am. J. Roentgenol.,161: 1147-1157, 1993.1611147 1993 [PubMed]

Mirowitz, S. A.; Apicella, P.; Reinus, W. R.; and Hammerman, A. M.: MR imaging of bone marrow lesions: relative conspicuousness on T1-weighted, fat-suppressed T2-weighted, and STIR images. AJR: Am. J. Roentgenol.,162: 215-221, 1994.162215 1994 [PubMed]

Monu, J. U.; Pope, T. L., Jr.; Chabon, S. J.; and Vanarthos, W. J.: MR diagnosis of superior labral anterior posterior (SLAP) injuries of the glenoid labrum: value of routine imaging without intraarticular injection of contrast material. AJR: Am. J. Roentgenol.,163: 1425-1429, 1994.1631425 1994 [PubMed]

Nakagaki, K.; Ozaki, J.; Tomita, Y.; and Tamai, S.: Magnetic resonance imaging of rotator cuff tearing and degenerative tendon changes: correlation with histologic pathology. J. Shoulder and Elbow Surg.,2: 156-164, 1993.2156 1993

Nakagaki, K.; Ozaki, J.; Tomita, Y.; and Tamai, S.: Alterations in the supraspinatus muscle belly with rotator cuff tearing: evaluation with magnetic resonance imaging. J. Shoulder and Elbow Surg.,3: 88-93, 1994.388 1994

Nakagaki, K.; Ozaki, J.; Tomita, Y.; and Tamai, S.: Function of supraspinatus muscle with torn cuff evaluated by magnetic resonance imaging. Clin. Orthop.,318: 144-151, 1995.318144 1995 [PubMed]

Needell, S. D.; Zlatkin, M. B.; Sher, J. S.; Murphy, B. J.; and Uribe, J. W.: MR imaging of the rotator cuff: peritendinous and bone abnormalities in an asymptomatic population. AJR: Am. J. Roentgenol.,166: 863-867, 1996.166863 1996 [PubMed]

Neer, C. S., II: Impingement lesions. Clin. Orthop.,173: 70-77, 1983.17370 1983 [PubMed]

Neumann, C. H.; Petersen, S. A.; and Jahnke, A. H.: MR imaging of the labral-capsular complex: normal variations. AJR: Am. J. Roentgenol.,157: 1015-1021, 1991.1571015 1991 [PubMed]

Neumann, C. H.; Holt, R. G.; Steinbach, L. S.; Jahnke, A. H., Jr.; and Petersen, S. A.: MR imaging of the shoulder: appearance of the supraspinatus tendon in asymptomatic volunteers. AJR: Am. J. Roentgenol,158: 1281-1287, 1992.1581281 1992 [PubMed]

Neviaser, T. J.: The GLAD lesion: another cause of anterior shoulder pain. Arthroscopy,9: 22-23, 1993.922 1993 [PubMed]

Nurenberg, P.; Giddings, C. J.; Stray-Gundersen, J.; Fleckenstein, J. L.; Gonyea, W. J.; and Peshock, R. M.: MR imaging-guided muscle biopsy for correlation of increased signal intensity with ultrastructural change and delayed-onset muscle soreness after exercise. Radiology,184: 865-869, 1992.184865 1992 [PubMed]

Owen, R. S.; Iannotti, J. P.; Kneeland, J. B.; Dalinka, M. K.; Deren, J. A.; and Oleaga, L.: Shoulder after surgery: MR imaging with surgical validation. Radiology,186: 443-447, 1993.186443 1993 [PubMed]

Palmer, W. E.; Brown, J. H.; and Rosenthal, D. I.: Rotator cuff: evaluation with fat-suppressed MR arthrography. Radiology,188: 683-687, 1993.188683 1993 [PubMed]

Palmer, W. E.; Brown, J. H.; and Rosenthal, D. I.: Labral-ligamentous complex of the shoulder: evaluation with MR arthrography. Radiology,190: 645-651, 1994.190645 1994 [PubMed]

Patten, R. M.: Tears of the anterior portion of the rotator cuff (the subscapularis tendon): MR imaging findings. AJR: Am. J. Roentgenol.,162: 351-354, 1994.162351 1994 [PubMed]

Polak, J. F.; Jolesz, F. A.; and Adams, D. F.: NMR of skeletal muscle. Differences in relaxation parameters related to extracellular/intracellular fluid spaces. Invest. Radiol.,23: 107-112, 1988.23107 1988 [PubMed]

Quinn, S. F.; Sheley, R. C.; Demlow, T. A.; and Szumowski, J.: Rotator cuff tendon tears: evaluation with fat-suppressed MR imaging with arthroscopic correlation in 100 patients. Radiology,195: 497-500, 1995.195497 1995 [PubMed]

Rafii, M.; Firooznia, H.; Bonamo, J. J.; Minkoff, J.; and Golimbu, C.: Athlete shoulder injuries: CT arthrographic findings. Radiology,162: 559-564, 1987.162559 1987 [PubMed]

Rafii, M.; Firooznia, H.; Sherman, O.; Minkoff, J.; Weinreb, J.; Golimbu, C.; Gidunal, R.; Schinella, R.; and Zaslav, K.: Rotator cuff lesions: signal patterns at MR imaging. Radiology,177: 817-823, 1990.177817 1990 [PubMed]

Rangger, C.; Klestil, T.; Kathrein, A.; Inderster, A.; and Hamid, L.: Influence of magnetic resonance imaging on indications for arthroscopy of the knee. Clin. Orthop.,330: 133-142, 1996.330133 1996 [PubMed]

Robertson, P. L.; Schweitzer, M. E.; Mitchell, D. G.; Schlesinger, F.; Epstein, R. E.; Friedman, B. G.; and Fenlin, J. M.: Rotator cuff disorders: interobserver and intraobserver variation in diagnosis with MR imaging. Radiology,194: 831-835, 1995.194831 1995 [PubMed]

Sano, H.; Kato, Y.; Haga, K.; Iroi, E.; and Tabata, S.: Magnetic resonance arthrography in the assessment of anterior instability of the shoulder: comparison with double-contrast computed tomography arthrography. J. Shoulder and Elbow Surg.,5: 280-285, 1996.5280 1996

Shellock, F. G.: MRI biological effects, safety, and patient management. In MRI of the Knee, edited by J. H. Mink, M. A. Reicher, J. V. Crues, III, and A. L. Deutsch. Ed. 2, pp. 31-46. New York, Raven Press, 1993.

Shellock, F. G.; Fukunaga, T.; Mink, J. H.; and Edgerton, V. R.: Exertional muscle injury: evaluation of concentric versus eccentric actions with serial MR imaging. Radiology,179: 659-664, 1991.179659 1991 [PubMed]

Shellock, F. G.; Fukunaga, T.; Mink, S. H.; and Edgerton, V. R.: Acute effects of exercise on MR imaging of skeletal muscle: concentric vs eccentric actions. AJR: Am. J. Roentgenol.,156: 765-768, 1991.156765 1991 [PubMed]

Sher, J. S.; Uribe, J. W.; Posada, A.; Murphy, B. J.; and Zlatkin, M. B.: Abnormal findings on magnetic resonance images of asymptomatic shoulders. J. Bone and Joint Surg.,77-A: 10-15, Jan. 1995.77-A10 1995

Sidor, M. L.; Zuckerman, J. D.; Lyon, T.; Koval, K.; Cuomo, F.; and Schoenberg, N.: The Neer classification system for proximal humeral fractures. An assessment of interobserver reliability and intraobserver reproducibility. J. Bone and Joint Surg.,75-A: 1745-1750, Dec. 1993.75-A1745 1993

Singson, R. D.; Hoang, T.; Dan, S.; and Friedman, M.: MR evaluation of rotator cuff pathology using T2-weighted fast spin-echo technique with and without fat suppression. AJR: Am. J. Roentgenol.,166: 1061-1065, 1996.1661061 1996 [PubMed]

Snyder, S. J.; Banas, M. P.; and Karzel, R. P.: An analysis of 140 injuries to the superior glenoid labrum. J. Shoulder and Elbow Surg.,4: 243-248, 1995.4243 1995

Solem-Bertoft, E.; Thuomus, K.-Å.; and Westerberg, C.-E.: The influence of scapular retraction and protraction on the width of the subacromial space. An MRI study. Clin. Orthop.,296: 99-103, 1993.29699 1993 [PubMed]

Sonin, A. H.; Peduto, A. J.; Fitzgerald, S. W.; Callahan, C. M.; and Bresler, M. E.: MR imaging of the rotator cuff mechanism: comparison of spin-echo and turbo spin-echo sequences. AJR: Am. J. Roentgenol.,167: 333-338, 1996.167333 1996 [PubMed]

Suder, P. A.; Frich, L. H.; Hougaard, K.; Lundorf, E.; and Wulff Jakobsen, B.: Magnetic resonance imaging evaluation of capsulolabral tears after traumatic primary anterior shoulder dislocation. A prospective comparison with arthroscopy of 25 cases. J. Shoulder and Elbow Surg.,4: 419-428, 1995.4419 1995

Timins, M. E.; Erickson, S. J.; Estkowski, L. D.; Carrera, G. F.; and Komorowski, R. A.: Increased signal in the normal supraspinatus tendon on MR imaging: diagnostic pitfall caused by the magic-angle effect. AJR: Am. J. Roentgenol.,165: 109-114, 1995.165109 1995 [PubMed]

Tirman, P. F.; Feller, J. F.; Janzen, D. L.; Peterfy, C. G.; and Bergman, A. G.: Association of glenoid labral cysts with labral tears and glenohumeral instability: radiologic findings and clinical significance. Radiology,190: 653-658, 1994.190653 1994 [PubMed]

Tirman, P. F.; Bost, F. W.; Garvin, G. J.; Peterfy, C. G.; Mall, J. C.; Steinbach, L. S.; Feller, J. F.; and Crues, J. V., III: Posterosuperior glenoid impingement of the shoulder: findings at MR imaging and MR arthrography with arthroscopic correlation. Radiology,193: 431-436, 1994.193431 1994 [PubMed]

Tirman, P. F.; Bost, F. W.; Steinbach, L. S.; Mall, J. C.; Peterfy, C. G.; Sampson, T. G.; Sheehan, W. E.; Forbes, J. R.; and Genant, H. K.: MR arthrographic depiction of tears of the rotator cuff: benefit of abduction and external rotation of the arm. Radiology,192: 851-856, 1994.192851 1994 [PubMed]

Tirman, P. F.; Stauffer, A. E.; Crues, J. V., III; Turner, R. M.; Nottage, W. M.; Schobert, W. E.; Rubin, B. D.; Janzen, D. L.; and Linares, R. C.: Saline magnetic resonance arthrography in the evaluation of glenohumeral instability. Arthroscopy,9: 550-559, 1993.9550 1993 [PubMed]

Traughber, P. D., and Goodwin, T. E.: Shoulder MRI: arthroscopic correlation with emphasis on partial tears. J. Comput. Assist. Tomog.,16: 129-133, 1992.16129 1992

Trout, T. E., and Resnick, D.: Glenoid hypoplasia and its relationship to instability. Skel. Radiol.,25: 37-40, 1996.2537 1996

Tuite, M. J., and Orwin, J. F.: Anterosuperior labral variants of the shoulder: appearance on gradient-recalled-echo and fast spin-echo MR images. Radiology,199: 537-540, 1996.199537 1996 [PubMed]

Uetani, M.; Hayashi, K.; Matsunaga, N.; Imamura, K.; and Ito, N.: Denervated skeletal muscle: MR imaging. Work in progress. Radiology,189: 511-515, 1993.189511 1993 [PubMed]

Uppal, G. S.; Uppal, J. A.; and Dwyer, A. P.: Glenoid cysts mimicking cervical radiculopathy. Spine,20: 2257-2260, 1995.202257 1995 [PubMed]

Uri, D. S.; Kneeland, J. B.; and Herzog, R.: Os acromiale: evaluation of markers for identification on sagittal and coronal oblique MR images. Skel. Radiol.,26: 31-34, 1997.2631 1997

Walch, G.; Nove-Josserand, L.; Levigne, C.; and Renaud, E.: Tears of the supraspinatus tendon associated with "hidden" lesions of the rotator interval. J. Shoulder and Elbow Surg.,3: 353-360, 1994.3353 1994

Weiner, D. S., and Macnab, I.: Superior migration of the humeral head. A radiological aid in the diagnosis of tears of the rotator cuff. J. Bone and Joint Surg.,52-B(3): 524-527, 1970.52-B(3)524 1970

Zlatkin, M. B.; Iannotti, J. P.; Roberts, M. C.; Esterhai, J. C.; Dalinka, M. K.; Kessel, H. Y.; Schwartz, J. S.; and Lenkinski, R. E.: Rotator cuff tears: diagnostic performance of MR imaging. Radiology,172: 223-229, 1989.172223 1989 [PubMed]

Topics

View Large | Download Slide(.ppt)
Add to My JBJS

Figs. 2-A through 2-D: Proton-density-weighted oblique sagittal images. Fig. 2-A: A type-I acromion (arrow) and no impingement on the cuff.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-B A type-II acromion (arrowheads) and a mildly thickened coracoacromial ligament (arrow) associated with moderate narrowing of the supraspinatus outlet.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-C A type-III acromion with a hook of the anteroinferior margin (arrow).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3-A: A proton-density-weighted axial image made through the acromioclavicular joint, showing an os acromiale (arrow).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3-B: The os acromiale (curved arrow) is also delineated on the proton-density-weighted oblique sagittal image, where a mildly thickened coracoacromial ligament (straight arrow) inserts into it.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4-A: A proton-density-weighted oblique coronal image showing increased signal intensity within a thickened supraspinatus segment of the cuff (arrow).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5-A: A T2-weighted oblique coronal image showing a small, deep partial-thickness tear (arrow) involving the articular surface of the cuff.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 5-B: A T2-weighted oblique coronal image of a different shoulder, showing an irregular oblique tear (arrow) involving the bursal side of the cuff.

View Large | Download Slide(.ppt)
Add to My JBJS

Figs. 6-A, 6-B, and 6-C: A small full-thickness tear (arrow) at the insertion site of the rotator cuff, posterior to the rotator interval. Fig. 6-A: T2-weighted oblique coronal image.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 6-B: T2-weighted oblique sagittal image.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 6-C T2-weighted axial image.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 13-A: A proton-density-weighted axial image made after a recent anterior dislocation of the shoulder. The anterior aspect of the glenoid labrum is torn and detached (arrow).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 13-B: The extent of the labral tear (arrows) is defined on the proton-density-weighted oblique sagittal image.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 14 A proton-density-weighted oblique coronal image showing an avulsion of the inferior aspect of the capsule from the neck of the humerus (arrow).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 15 A T2-weighted axial image showing a tear through the base of the posterior aspect of the glenoid labrum (arrow) along with posterior capsular stripping.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 16-A: A proton-density-weighted oblique coronal image showing a type-III tear of the superior aspect of the labrum extending anteriorly and posteriorly (SLAP lesion) (arrow).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 16-B: A gradient-echo axial image also showing a type-III tear of the superior aspect of the labrum extending anteriorly and posteriorly (arrows).

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 17 A T2-weighted oblique coronal image showing a minimally displaced fracture of the greater tuberosity (arrow), which was the result of a skiing injury. The rotator cuff is intact.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 19 A proton-density-weighted oblique sagittal image showing subchondral cysts (arrow) within the posteroinferior margin of the glenoid.

View Larger

TABLE I ROUTINE PROTOCOL FOR MAGNETIC RESONANCE IMAGING OF THE SHOULDER

*1 = made with the upper extremity in neutral or slight external rotation, and 2 = a field of view of sixteen centimeters for all sequences.

Plane*	Relaxation Time (msec)	Echo Time (msec)	No. of Excitations	Image Matrix	Thickness (mm)	Skip (mm)	Comments
Oblique coronal							Parallel to long axis of body of scapula, including soft tissues anterior and posterior to humeral head
1	2500	Dual-echo minimum/70	1	256 by 192	4	0.5	Classic multi-echo
Oblique sagittal							Perpendicular to long axis of scapula from base of coracoid through humeral head
1	1500	Minimum	1	256 by 192	4	0.5
2	4000	Approximately 90	2	256 by 192	4, same location as 1	0.5	Fast spin echo-echo train length-8; fat-saturated band width, 32kHz
Axial							From superior to acromiclavicular joint through glenoid and humeral neck
1	1500	Minimum	1	256 by 192	3	0.5
2	4000	Approximately 90	2	256 by 192	3, same location as 1	0.5	Fast spin echo-echo train length-8; fat-saturated band width, 32 kHz

Add to My JBJS

TABLE I ROUTINE PROTOCOL FOR MAGNETIC RESONANCE IMAGING OF THE SHOULDER

*1 = made with the upper extremity in neutral or slight external rotation, and 2 = a field of view of sixteen centimeters for all sequences.

Plane*	Relaxation Time (msec)	Echo Time (msec)	No. of Excitations	Image Matrix	Thickness (mm)	Skip (mm)	Comments
Oblique coronal							Parallel to long axis of body of scapula, including soft tissues anterior and posterior to humeral head
1	2500	Dual-echo minimum/70	1	256 by 192	4	0.5	Classic multi-echo
Oblique sagittal							Perpendicular to long axis of scapula from base of coracoid through humeral head
1	1500	Minimum	1	256 by 192	4	0.5
2	4000	Approximately 90	2	256 by 192	4, same location as 1	0.5	Fast spin echo-echo train length-8; fat-saturated band width, 32kHz
Axial							From superior to acromiclavicular joint through glenoid and humeral neck
1	1500	Minimum	1	256 by 192	3	0.5
2	4000	Approximately 90	2	256 by 192	3, same location as 1	0.5	Fast spin echo-echo train length-8; fat-saturated band width, 32 kHz