Can ultrasound diagnose a torn rotator cuff?

Can Ultrasound Diagnose Torn Rotator Cuff?

Yes, ultrasound can effectively diagnose rotator cuff tears and is considered equivalent to MRI for this purpose when performed by experienced operators, particularly excelling at detecting full-thickness tears. 1

Diagnostic Performance

Full-Thickness Tears

Ultrasound demonstrates excellent diagnostic accuracy for full-thickness rotator cuff tears:

• Sensitivity: 90-100% (especially for supraspinatus tendon)
• Specificity: 81-100%
• Performance is statistically equivalent to MRI 2, 3, 4

The 2026 systematic review found ultrasound superior to MRI for full-thickness tears, with sensitivity of 100% for supraspinatus tears and kappa values up to 1.0 2.

Partial-Thickness Tears

Performance is more variable for partial tears:

• Sensitivity: 72-94%
• Accuracy: 76-94%
• Results depend heavily on operator expertise 2, 3
• Radiologists achieve better sensitivity (86%) compared to surgeons (57%) for partial-thickness tears 3

When to Use Ultrasound vs. MRI

Choose Ultrasound When:

• Evaluating suspected full-thickness rotator cuff tears
• MRI is contraindicated or unavailable
• Cost or accessibility is a concern
• Local expertise with musculoskeletal ultrasound is available
• Evaluating post-operative rotator cuff integrity after shoulder replacement 1
• Patient has proximal humeral hardware causing MRI artifacts 5

Choose MRI When:

• Partial-thickness tears are suspected and surgical planning is needed
• Evaluating complex intra-articular pathology (labral tears, ligament injuries)
• Patient has large body habitus or restricted range of motion
• Comprehensive pre-operative assessment is required
• Detailed evaluation of muscle atrophy and fatty infiltration is needed 2, 6

ACR Appropriateness Criteria Ratings

According to the most recent 2025 ACR guidelines 5, for suspected rotator cuff tears:

• Ultrasound: Rating 9 (usually appropriate) - when local expertise available
• MRI without contrast: Rating 9 (usually appropriate)
• MR arthrography: Rating 9 (usually appropriate) - particularly for partial tears

The guidelines explicitly state these three modalities are equivalent for evaluating rotator cuff tears, depending on local expertise 1.

Critical Caveats

Operator Dependency

Ultrasound is highly operator-dependent. Diagnostic accuracy requires:

• Experienced musculoskeletal sonographer or radiologist
• Standardized examination technique
• Adequate training in shoulder anatomy 1, 5

Limitations

• Less reliable for non-supraspinatus tendons (subscapularis, infraspinatus) 2
• Cannot adequately evaluate deep intra-articular structures (labrum, glenohumeral ligaments) 1, 7
• Limited assessment of bone marrow pathology 1
• Variable performance in female patients 2
• Poor for labral tears (ACR rating of 1 - usually not appropriate) 1

Practical Algorithm

For suspected rotator cuff tear:

1. Start with plain radiographs to exclude fracture, dislocation, or arthritis
2. If radiographs are noncontributory and rotator cuff tear is suspected:
   • Use ultrasound if experienced operator available AND primarily concerned about full-thickness tear
   • Use MRI if partial tear suspected, surgical planning needed, or ultrasound expertise unavailable
   • Both modalities have equivalent diagnostic performance when properly performed 5, 2, 3

Bottom line: Ultrasound is a valid first-line imaging modality for rotator cuff tears when expertise is available, offering comparable diagnostic accuracy to MRI for full-thickness tears at lower cost and with greater accessibility, while MRI remains preferred for comprehensive assessment and surgical planning.