How to Dictate a Shoulder MRI: The Cuff Tendon by Tendon, in the Surgeon's Words

The shoulder MRI report has one primary reader, and that reader is deciding between a scope, an open repair, and physical therapy. Dictate for the orthopedist and everyone else is served automatically. The organizing principle is the same one that fixes most MSK dictation: structures, not sequences. Nobody wants a tour of your coronal obliques; they want the supraspinatus.

The cuff: four tendons, four sentences

Every report addresses each tendon by name, even when normal:

• Supraspinatus — where most of the pathology lives, anterior fibers first.
• Infraspinatus — tears here usually extend from the supraspinatus; say whether the tear crosses the interval between them.
• Subscapularis — the most under-called tendon on routine shoulder MRI, best seen on axials. A subscapularis tear changes the surgical approach, and it drags the biceps into the report (below).
• Teres minor — almost always normal; name it anyway, because "the rest of the cuff" is not a structure.

The vocabulary ladder is fixed: normal → tendinosis (intermediate intrasubstance signal, no fiber discontinuity — "tendinitis" is not an MRI diagnosis) → partial-thickness tear → full-thickness tear. Each rung has its own required descriptors, and this is where the report earns its keep.

Partial thickness, in surgical terms

A partial-thickness tear gets three descriptors, because all three change management:

1. Surface: articular-sided, bursal-sided, or intrasubstance. Articular-sided tears of the supraspinatus footprint are the common pattern; bursal-sided tears behave differently and are the ones the impingement-decompression conversation cares about.
2. Depth, as a fraction of tendon thickness: the widely taught surgical threshold is 50% — a tear involving more than half the tendon's thickness is discussed for repair or completion-and-repair, under half is usually debrided or rehabbed. Give millimeters of depth and the fraction ("involving approximately 60% of tendon thickness").
3. Location along the footprint: anterior, mid, posterior fibers, and the anteroposterior length of involvement.

"Partial tear of the supraspinatus" with none of the three is a finding the surgeon must re-derive from the images, which means the report added a billing code and nothing else.

Full thickness: the repair-planning numbers

A full-thickness tear means fluid signal crosses the tendon from articular to bursal surface. The sentence set that follows is the repair plan's raw material:

• Dimensions: anteroposterior width and mediolateral extent, in centimeters. Full-thickness is not the same as full-width; a 4 mm pinhole perforation and a complete tendon detachment are both "full thickness," and only the numbers distinguish them.
• Retraction: how far the torn tendon stump has pulled back, in centimeters, with a landmark — to the lateral humeral head, to the level of the humeral head apex, to the glenoid. Retraction to the glenoid is a different operation, and sometimes no operation.
• Which tendons: a tear extending into the infraspinatus or anteriorly into the rotator interval and subscapularis is a different repair than an isolated supraspinatus tear of the same area.

The Goutallier comment: the sentence everyone forgets

Repairability does not live in the tendon; it lives in the muscle. A perfectly describable tear attached to a fat-replaced muscle will fail repair, and the surgeon needs that information from you, on the sagittal sequence lateral to where the trainee's eye stops. The standard vocabulary is the Goutallier classification of fatty infiltration, graded 0–4 per muscle: 0, normal; 1, fatty streaks; 2, fat less than muscle; 3, fat equal to muscle; 4, fat greater than muscle. The widely taught decision point is grade 2 — significant fatty infiltration (grade 2 and above) predicts worse repair outcomes, and grades 3–4 with retraction push the conversation toward tendon transfer or reverse arthroplasty rather than repair.

The report does not need the numeral to be useful, but it needs the content: one sentence per chronically torn tendon's muscle — "moderate fatty infiltration and volume loss of the supraspinatus muscle (Goutallier 3)" — plus atrophy when present. A chronic full-thickness tear dictated without any muscle comment is the single most common structural omission in resident shoulder reports.

The labrum, honestly, without arthrography

On a routine non-arthrographic shoulder MRI, the labrum deserves modesty. Sensitivity for labral tears — SLAP lesions especially — is meaningfully lower without intra-articular contrast distending the joint, and the anterosuperior quadrant is a minefield of normal variants: the sublabral foramen and the Buford complex (absent anterosuperior labrum with a cord-like middle glenohumeral ligament) are anatomy, not pathology, and calling either a tear sends a healthy shoulder to arthroscopy.

So the honest sentence structure is calibrated to what the study can support. When the labrum is clearly intact: say so. When there is a definite displaced tear or a paralabral cyst (which is a labral tear's footprint until proven otherwise — and worth a sentence about the spinoglenoid notch and denervation changes when present): say so with confidence. In between: "no displaced labral tear; the sensitivity of non-contrast MRI for non-displaced labral pathology is limited, and MR arthrography is recommended if clinical suspicion is high." That sentence is not hedging — it is routing, the same move as sending a complex renal cyst to a dedicated protocol.

The rest of the checklist

Fixed slots, one line each: long head of the biceps — located in the groove, subluxed, or dislocated (a medially dislocated biceps is a subscapularis tear until proven otherwise), plus its signal; acromioclavicular joint and acromial morphology, including an os acromiale when present, because it changes decompression planning; glenohumeral cartilage — a cuff report that ignores established arthropathy sets the surgeon up for an intraoperative surprise; bones — marrow edema, cysts at the greater tuberosity footprint, a Hill-Sachs impaction deformity after instability events; and the axillary recess and quadrilateral space on the way out.

The phrases attendings correct

• "Rotator cuff tear" without a tendon name, surface, depth, or size. The cuff is four structures, not one.
• "High-grade partial tear" with no percentage or millimeters. Grade it in the units the repair decision uses.
• "Supraspinatus tear" on a chronic study with no muscle comment. Tendon describes the injury; muscle predicts the outcome.
• "Intact labrum" on a non-distended joint read by a resident at 2 a.m. Claim what the study supports.
• "Correlate clinically." The orthopedist examined the shoulder an hour ago; correlate radiologically instead — with the priors.

A model report

EXAMINATION: MRI right shoulder without contrast
HISTORY: 58-year-old male, chronic right shoulder pain
and weakness. No acute trauma.
TECHNIQUE: Multiplanar multisequence MRI of the right
shoulder without intravenous or intra-articular
contrast.
COMPARISON: None.
FINDINGS:
Supraspinatus: Full-thickness tear of the anterior
fibers measuring 1.8 cm AP x 1.5 cm mediolateral, with
tendon retraction to the level of the humeral head
apex. Moderate fatty infiltration of the supraspinatus
muscle (Goutallier 3) with volume loss.
Infraspinatus: Mild tendinosis; no tear. Muscle normal.
Subscapularis: Intact. Teres minor: Normal.
Biceps: Long head located within the groove; mild
tendinosis.
Labrum: No displaced labral tear. Non-contrast MRI is
limited for non-displaced labral pathology.
AC joint: Moderate osteoarthritis with inferior
osteophyte. No os acromiale.
Glenohumeral joint: Mild cartilage thinning; no
high-grade chondral defect. Small joint effusion.
Bones: Cystic change at the greater tuberosity
footprint. No fracture.
IMPRESSION:
1. Full-thickness anterior supraspinatus tear
   (1.8 x 1.5 cm) retracted to the humeral head apex,
   with Goutallier grade 3 fatty infiltration — relevant
   to repairability.
2. Intact subscapularis; biceps located, mildly
   tendinotic.
3. Moderate AC joint osteoarthritis.

Every sentence in that model maps to a decision: repair size, repair feasibility, approach, decompression. That is the test for everything you dictate on this study. For the normal scaffold, start from the normal shoulder MRI template, and browse the template library for the matching knee and lumbar structures. And if you want to watch a thirty-second cuff dictation become exactly this report — tendons slotted, Goutallier line prompted when the tear is chronic — run a shoulder case in the live demo, no signup needed.