An MRI of shoulder is one of the most frequently ordered musculoskeletal examinations in modern radiology, offering unmatched soft-tissue contrast for evaluating the rotator cuff, labrum, capsule, biceps tendon, and surrounding bursae. Unlike radiographs, which primarily show bone, or ultrasound, which is operator dependent, magnetic resonance imaging delivers multiplanar views of every anatomic structure in the glenohumeral joint. For radiologic technologists, orthopedic surgeons, and primary care clinicians, mastering shoulder MRI protocols is essential to producing diagnostic-quality images and interpreting them correctly.
The shoulder is a remarkably mobile joint, sacrificing stability for range of motion, which makes it uniquely vulnerable to injury. Throwing athletes, overhead laborers, and aging patients alike present with pain, weakness, and decreased function that often require advanced imaging. An MRI scan can distinguish between a partial-thickness tear, a full-thickness retracted tear, tendinosis, calcific deposits, and adhesive capsulitis, each of which carries a very different treatment plan and prognosis.
This guide walks through every component of a modern shoulder MRI study, from patient positioning and coil selection to the specific pulse sequences used in each plane. We will cover when to add intra-articular gadolinium contrast for an MR arthrogram, how to recognize common pathology on T1 and T2 weighted images, and the indications for non-contrast versus contrast-enhanced protocols. Whether you are preparing for a board exam or scanning patients tomorrow, the information here will sharpen your clinical decision-making.
Shoulder imaging accounts for roughly 15 percent of all musculoskeletal MRI volume in the United States, and that share continues to grow as the population ages and weekend-warrior injuries climb. Insurance carriers increasingly require documented conservative therapy failure before approving an MRI, so understanding the appropriate use criteria matters for ordering providers. A well-justified, well-protocoled scan reduces repeat imaging, lowers costs, and accelerates the path to definitive treatment, whether that is physical therapy, an injection, or arthroscopic surgery.
For radiologic technologists, the shoulder presents technical challenges that the knee or lumbar spine do not. Off-isocenter positioning, dedicated surface coils, oblique scan planes aligned to the supraspinatus tendon, and motion control during a 30 to 45 minute exam all demand careful attention. Producing images that allow radiologists to confidently identify a 2-millimeter labral tear requires understanding both the physics of MRI and the three-dimensional anatomy of the joint. You can sharpen those skills with our free MRI knowledge questions and answers resource.
Patients reading this guide will find practical information about what to expect, how long the scan takes, why the machine is so loud, whether contrast is necessary, and how to prepare. Demystifying the experience reduces anxiety and improves cooperation, which in turn improves image quality. We will also address common questions about claustrophobia, metal implants, and the difference between conventional MRI and MR arthrography. By the end, you will understand the why behind every step of the examination.
Finally, this article serves as both a clinical reference and a study aid. Each section ties the imaging findings back to anatomy and pathology, with examples drawn from common board exam scenarios. The accompanying quizzes reinforce key concepts, and the FAQ addresses the questions that come up most often in clinical practice. Bookmark this page, share it with colleagues, and use it as a launching point for deeper study of shoulder MRI.
Patient lies supine with the affected arm at the side, palm facing up or in slight external rotation. The shoulder is positioned as close to isocenter as possible. Sandbags or straps minimize motion during the 30 to 45 minute acquisition.
A dedicated multi-channel shoulder coil wraps around the joint to maximize signal-to-noise ratio. Flexible coils accommodate varying body habitus while phased-array designs improve resolution for small structures like the labrum and biceps anchor.
Axial images cover from the AC joint to below the glenoid. Oblique coronal images parallel the supraspinatus tendon. Oblique sagittal images run perpendicular to the glenoid face, optimal for evaluating the rotator cuff in cross section.
Standard protocol includes T1, T2, proton density, and fat-suppressed sequences. Fat saturation highlights edema, fluid, and pathology by suppressing bright marrow and subcutaneous signal that would otherwise mask subtle findings.
The technologist reviews images before the patient leaves to verify coverage, motion-free acquisition, and adequate fat suppression. Repeat sequences are performed immediately if artifacts compromise diagnostic interpretation by the radiologist.
Understanding shoulder anatomy is the foundation for interpreting any MRI of the shoulder. The glenohumeral joint is a ball-and-socket articulation between the humeral head and the shallow glenoid fossa of the scapula. Stability is provided not by bony congruity but by a complex of soft-tissue restraints, including the glenoid labrum, joint capsule, glenohumeral ligaments, and the four rotator cuff tendons. Each of these structures has a characteristic appearance on MRI that the interpreting radiologist must recognize.
The rotator cuff consists of the supraspinatus, infraspinatus, teres minor, and subscapularis muscles. Their tendons converge to form a cuff around the humeral head, controlling rotation and providing dynamic stability. The supraspinatus runs through the subacromial space and is the most commonly torn tendon, particularly at its critical zone approximately one centimeter from the greater tuberosity insertion. On oblique coronal images, the supraspinatus should appear uniformly dark on all sequences.
The glenoid labrum is a fibrocartilaginous ring that deepens the socket and serves as the attachment point for the long head of the biceps tendon superiorly and the inferior glenohumeral ligament inferiorly. Labral tears are classified by location using clock-face positions, with SLAP tears occurring at the superior labrum from anterior to posterior. The labrum normally appears as a uniformly black triangular structure on axial and coronal images.
The long head of the biceps tendon originates from the supraglenoid tubercle and superior labrum, then courses through the bicipital groove of the humerus. Subluxation, dislocation, or rupture of this tendon is a frequent finding in older patients and overhead athletes. On axial images, the biceps tendon should sit centrally within the groove, held in place by the transverse humeral ligament and the medial sling formed by the subscapularis tendon and superior glenohumeral ligament.
The acromioclavicular joint sits superior to the rotator cuff and is a common site of osteoarthritis, particularly in patients over 50. Hypertrophic changes, inferior osteophytes, and capsular distension can impinge on the underlying supraspinatus tendon. The shape of the acromion itself, classified as type 1 flat, type 2 curved, or type 3 hooked, influences impingement risk. Sagittal images best demonstrate acromial morphology and its relationship to the rotator cuff.
The subacromial-subdeltoid bursa is a potential space between the rotator cuff and the deltoid muscle. Normally it contains only a thin film of lubricating fluid, but bursitis from impingement, inflammation, or rotator cuff disease causes distension that appears bright on fluid-sensitive sequences. Recognizing the difference between bursal fluid, joint effusion, and pathologic fluid collections is a fundamental skill. Review additional educational content through our mri medical abbreviation for terminology used in shoulder reports.
Finally, the deltoid, trapezius, and periscapular muscles surround the joint and provide power for movement. Muscle quality, including fatty infiltration and atrophy, is assessed using the Goutallier classification on sagittal T1 images. Advanced fatty replacement of the rotator cuff musculature indicates chronic tear and predicts poor surgical outcomes, often pushing treatment toward conservative management or reverse total shoulder arthroplasty rather than primary repair.
T1-weighted images provide excellent anatomic detail with bright fat signal and dark fluid. They are essential for assessing bone marrow, fatty infiltration of rotator cuff muscles using the Goutallier classification, and post-contrast enhancement when intravenous gadolinium is administered. The high contrast between subcutaneous fat and underlying structures makes T1 ideal for surveying overall anatomy at the start of interpretation.
In shoulder MRI, T1 sequences are obtained in all three planes. Sagittal T1 specifically evaluates the rotator cuff cross-section and muscle bulk, while coronal T1 shows the supraspinatus tendon and acromiohumeral interval. Pathologic findings on T1 include fatty infiltration appearing as bright streaks within muscle bellies, dark marrow signal from infiltration, and loss of normal fat planes around tendons indicating fibrosis.
T2-weighted and proton density sequences with fat saturation are the workhorses of shoulder imaging. By suppressing bright fat signal, these sequences make fluid, edema, and pathology jump off the screen as bright signal. Rotator cuff tears appear as bright fluid signal traversing the tendon, bone bruises show as ill-defined marrow edema, and labral tears appear as linear bright signal disrupting the normally dark labrum.
Fat suppression can be achieved through spectral fat saturation, STIR, or Dixon techniques. STIR is more uniform across the field of view but has lower signal-to-noise ratio. Dixon methods provide simultaneous in-phase, out-of-phase, water-only, and fat-only images in a single acquisition, increasingly favored on modern scanners for shoulder protocols where field inhomogeneity near the lung apex degrades spectral fat saturation.
MR arthrography adds intra-articular gadolinium contrast diluted in saline and lidocaine, injected under fluoroscopic or ultrasound guidance before scanning. The bright contrast distends the joint capsule and outlines the labrum, articular cartilage, and undersurface of the rotator cuff, dramatically improving sensitivity for subtle tears that conventional MRI may miss, particularly partial articular-sided cuff tears and labral pathology.
Arthrogram protocols emphasize T1 fat-saturated sequences in all three planes to maximize contrast between bright gadolinium and surrounding tissues. An ABER position view, with the arm abducted and externally rotated, is often added to tension the anterior inferior labrum and detect Bankart lesions. Conventional T2 sequences are also included to evaluate paralabral cysts, bone edema, and extra-articular pathology not affected by intra-articular contrast.
The oblique coronal plane must be prescribed parallel to the long axis of the supraspinatus muscle and tendon as visualized on the axial localizer. Off-axis prescriptions create partial volume averaging that mimics or hides rotator cuff tears, leading to false-positive or false-negative interpretations. A two-degree adjustment can change a confident diagnosis into a diagnostic dilemma.
Common pathology identified on MRI of shoulder spans the spectrum from acute trauma to chronic degeneration. Rotator cuff tears are by far the most frequent finding, affecting an estimated 22 percent of the general population and rising sharply with age. Tears are classified by thickness as partial or full, by location as articular-sided, bursal-sided, or intratendinous, and by size in millimeters or by Cofield grading. Full-thickness tears appear as bright fluid signal traversing the entire tendon on fat-saturated T2 images, often with retraction of the torn end.
Tendinosis, also called tendinopathy, represents degenerative change without frank tearing. It appears as intermediate signal intensity within the tendon on fluid-sensitive sequences, brighter than the normal dark tendon but not as bright as free fluid. Tendinosis often coexists with partial tears and is a common cause of impingement-related shoulder pain. Distinguishing tendinosis from a low-grade partial tear can be challenging and may require arthrography for confirmation.
Labral tears are the second major category. Bankart lesions involve the anterior-inferior labrum and result from anterior shoulder dislocations, often with an associated Hill-Sachs impaction fracture on the posterolateral humeral head. Reverse Bankart and reverse Hill-Sachs lesions occur in posterior dislocations. SLAP tears affect the superior labrum where the biceps anchor attaches and are graded I through IV based on biceps tendon involvement and displacement. Paralabral cysts are an indirect sign of labral tear.
Adhesive capsulitis, or frozen shoulder, presents with thickening of the rotator interval and inferior axillary recess. Thickening of the coracohumeral ligament beyond four millimeters and obliteration of the fat triangle between the coracoid and the coracohumeral ligament are characteristic findings. Edema and enhancement of the capsule are best appreciated on fat-saturated T2 and post-contrast T1 images, respectively. This diagnosis is often clinical, but MRI helps exclude other causes of restricted motion.
Calcific tendinitis appears as low-signal foci within the rotator cuff tendons on all sequences, corresponding to hydroxyapatite crystal deposition. Surrounding edema during the resorptive phase causes severe pain and bright signal on T2. Plain radiographs remain the primary imaging modality for confirming calcific tendinitis, but MRI is often the first study ordered for shoulder pain. Recognizing the pattern prevents misdiagnosis as a tear or mass lesion in older adults.
Acromioclavicular joint arthrosis and impingement findings are nearly universal in older patients. Subacromial spurs, inferior osteophytes from the AC joint, and acromial morphology all contribute to mechanical impingement of the rotator cuff. Greater tuberosity cortical irregularity, cystic change, and reactive bone marrow edema are secondary signs of chronic impingement. The combination of these findings with rotator cuff tendinosis or partial tear supports the diagnosis of impingement syndrome.
Less common but important findings include glenohumeral arthritis, osteonecrosis of the humeral head, suprascapular nerve entrapment by paralabral cysts, biceps tendon dislocation, and neoplasms. The radiologist must systematically evaluate every structure on every sequence to avoid missing these uncommon but clinically significant diagnoses. A search pattern that proceeds from bones to tendons to labrum to soft tissues ensures comprehensive coverage of every shoulder MRI examination.
MR arthrography deserves special discussion because it remains the gold standard for evaluating subtle labral pathology and post-surgical shoulders. The procedure begins with fluoroscopic or ultrasound-guided puncture of the glenohumeral joint, typically through a rotator interval or posterior approach. A mixture of dilute gadolinium, saline, iodinated contrast, and sometimes lidocaine is injected to distend the capsule. The patient is then transferred to the MRI scanner for immediate imaging.
The injected contrast outlines intra-articular structures with bright signal on T1 fat-saturated sequences. Labral tears appear as contrast extending into the substance of the labrum or undercutting its attachment to the glenoid. Partial articular-sided rotator cuff tears, which are easy to miss on conventional MRI, become obvious as contrast tracking into the tendon from the joint side. SLAP tears, Bankart variants, and rotator interval pathology are dramatically better visualized with arthrography.
Indications for MR arthrogram include young patients with suspected labral injury, throwing athletes with internal impingement, post-operative evaluation of rotator cuff repair, and assessment of recurrent shoulder instability. In older patients with suspected full-thickness rotator cuff tears, conventional MRI is usually sufficient and arthrography adds little. Cost, additional time, and procedural risk drive selective use of arthrography rather than universal application.
Direct MR arthrography uses intra-articular gadolinium, while indirect arthrography involves intravenous gadolinium with delayed imaging after exercise to allow contrast to diffuse into the joint. Indirect arthrography is less invasive but produces lower-quality joint distension and is largely abandoned in modern practice. Direct arthrography remains the standard despite the need for fluoroscopy or ultrasound guidance for the injection step.
The ABER view, with the arm abducted to 90 degrees and externally rotated, is a hallmark of MR arthrography. This position places tension on the anterior inferior labrum and the inferior glenohumeral ligament, opening up tears that may appear normal in the neutral position. The ABER sequence is typically a single oblique axial T1 fat-saturated acquisition added at the end of the arthrogram protocol. Patient comfort in this position is critical for motion-free images.
Complications of MR arthrography are rare but include vasovagal reaction during injection, transient post-injection pain lasting one to two days, and the extremely rare septic arthritis. Strict sterile technique, appropriate needle size, and slow contrast injection minimize these risks. Patients should be counseled that the immediate post-procedure soreness is expected and typically resolves quickly with rest and ice. For patients comparing imaging approaches, our mri alternatives discusses when other modalities may be appropriate.
Image interpretation of an arthrogram requires familiarity with normal variants that mimic pathology. The sublabral foramen, Buford complex, and sublabral recess are all normal anatomic variations that can be mistaken for tears by inexperienced readers. Knowing the typical locations and appearances of these variants, particularly the absent anterior superior labrum of the Buford complex, prevents over-diagnosis and unnecessary surgical intervention. Cross-correlation with arthroscopy is the ultimate truth standard.
Practical tips for both technologists and patients ensure that an MRI of shoulder produces diagnostic-quality images on the first attempt. Start with patient communication. Explain the duration of the exam, the loud knocking and buzzing sounds from gradient coils, the importance of holding still, and the call button or squeeze ball available throughout. A patient who understands what is happening tolerates the scan better and produces fewer motion artifacts that would require repeat sequences.
Positioning the shoulder at isocenter requires creative use of pads and straps. Patients with large body habitus may not be able to center the shoulder under the main magnet bore, leading to off-center artifacts and field inhomogeneity. In these cases, lateral decubitus positioning or use of a wide-bore scanner may be necessary. Some facilities use the contralateral shoulder hanging off the table to allow the affected side to drop closer to isocenter.
Hand positioning influences the appearance of the biceps tendon and rotator cuff. Neutral position with the thumb pointing up is standard, but slight external rotation places the supraspinatus tendon perpendicular to the magnetic field and reduces magic angle artifact at its insertion. Internal rotation, while sometimes used to evaluate posterior structures, increases magic angle artifact and is generally avoided unless specifically requested by the radiologist for a particular clinical question.
Sequence parameters should be optimized for the field strength of the scanner. At 3 Tesla, higher signal-to-noise ratio allows thinner slices and higher matrices, improving spatial resolution. However, 3T also amplifies susceptibility artifacts near metal implants and increases specific absorption rate concerns. At 1.5T, more averages and thicker slices may be needed to achieve diagnostic quality, but artifacts from prior surgical hardware are more manageable.
Motion artifact is the most common reason for non-diagnostic shoulder MRI. Even small movements blur tendon edges and mimic or obscure pathology. Strategies to minimize motion include shortening acquisition time with parallel imaging, using comfortable positioning aids, ensuring adequate analgesia for painful shoulders, and choosing sequences with built-in motion compensation such as radial or PROPELLER acquisitions. The technologist should monitor scout images during the exam and abort sequences with obvious motion early.
Post-processing techniques can salvage suboptimal acquisitions. Multiplanar reformatting of isotropic 3D sequences allows the radiologist to create new planes after the patient leaves. Image filtering and contrast adjustment may improve visualization of subtle findings. However, these techniques cannot create information that was never acquired. The best approach remains careful protocol design and meticulous attention to image quality during the initial acquisition.
Finally, communicate with the ordering provider and the radiologist when something unusual appears during the scan. A subtle finding flagged in real time can prompt additional sequences or views while the patient is still on the table. Whether this is an unexpected mass, a contrast extravasation, or a pathologic fracture, immediate communication shortens diagnostic timelines and improves patient care. This collaborative approach distinguishes excellent imaging practices from average ones and contributes to better outcomes for every shoulder patient.