Skip to main content
SpringerLink
Log in

Ruling out rotator cuff tear in shoulder radiograph series using deep learning: redefining the role of conventional radiograph

  • Musculoskeletal
  • Published:
European Radiology Aims and scope Submit manuscript

Abstract

Objective

To develop a deep learning algorithm that can rule out significant rotator cuff tear based on conventional shoulder radiographs in patients suspected of rotator cuff tear.

Methods

The algorithm was developed using 6793 shoulder radiograph series performed between January 2015 and June 2018, which were labeled based on ultrasound or MRI conducted within 90 days, and clinical information (age, sex, dominant side, history of trauma, degree of pain). The output was the probability of significant rotator cuff tear (supraspinatus/infraspinatus complex tear with > 50% of tendon thickness). An operating point corresponding to sensitivity of 98% was set to achieve high negative predictive value (NPV) and low negative likelihood ratio (LR−). The performance of the algorithm was tested with 1095 radiograph series performed between July and December 2018. Subgroup analysis using Fisher’s exact test was performed to identify factors (clinical information, radiography vendor, advanced imaging modality) associated with negative test results and NPV.

Results

Sensitivity, NPV, and LR− were 97.3%, 96.6%, and 0.06, respectively. The deep learning algorithm could rule out significant rotator cuff tear in about 30% of patients suspected of rotator cuff tear. The subgroup analysis showed that age < 60 years (p < 0.001), non-dominant side (p < 0.001), absence of trauma history (p = 0.001), and ultrasound examination (p < 0.001) were associated with negative test results. NPVs were higher in patients with age < 60 years (p = 0.024) and examined with ultrasound (p < 0.001).

Conclusion

The deep learning algorithm could accurately rule out significant rotator cuff tear based on shoulder radiographs.

Key Points

• The deep learning algorithm can rule out significant rotator cuff tear with a negative likelihood ratio of 0.06 and a negative predictive value of 96.6%.

• The deep learning algorithm can guide patients with significant rotator cuff tear to additional shoulder ultrasound or MRI with a sensitivity of 97.3%.

• The deep learning algorithm could rule out significant rotator cuff tear in about 30% of patients with clinically suspected rotator cuff tear.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2

Similar content being viewed by others

Abbreviations

AUC:

Area under the receiver operating characteristic curve

CNN:

Convolutional neural network

Cutoff98% :

Cutoff point for an expected sensitivity of 98%

Cutoffoptimal :

Optimal cutoff point determined by Youden’s J statistic

DICOM:

Digital Imaging and Communications in Medicine

FCN:

Fully connected network

NPV:

Negative predictive value

VAS:

Visual analog scale

References

  1. de Jesus JO, Parker L, Frangos AJ, Nazarian LN (2009) Accuracy of MRI, MR arthrography, and ultrasound in the diagnosis of rotator cuff tears: a meta-analysis. AJR Am J Roentgenol 192:1701–1707

    Article  Google Scholar 

  2. Roy JS, Braen C, Leblond J et al (2015) Diagnostic accuracy of ultrasonography, MRI and MR arthrography in the characterisation of rotator cuff disorders: a systematic review and meta-analysis. Br J Sports Med 49:1316–1328

    Article  Google Scholar 

  3. Amini B, Beckmann NM, Beaman FD et al (2018) ACR Appropriateness Criteria® shoulder pain-traumatic. J Am Coll Radiol 15:S171–S188

  4. Small KM, Adler RS, Shah SH et al (2018) ACR Appropriateness Criteria® shoulder pain—atraumatic. J Am Coll Radiol 15:S388–S402

  5. Pearsall AW 4th, Bonsell S, Heitman RJ, Helms CA, Osbahr D, Speer KP (2003) Radiographic findings associated with symptomatic rotator cuff tears. J Shoulder Elbow Surg 12:122–127

    Article  Google Scholar 

  6. Hardy DC, Vogler JB 3rd, White RH (1986) The shoulder impingement syndrome: prevalence of radiographic findings and correlation with response to therapy. AJR Am J Roentgenol 147:557–561

    Article  CAS  Google Scholar 

  7. Cone RO 3rd, Resnick D, Danzig L (1984) Shoulder impingement syndrome: radiographic evaluation. Radiology 150:29–33

    Article  Google Scholar 

  8. Peh WC, Farmer TH, Totty WG (1995) Acromial arch shape: assessment with MR imaging. Radiology 195:501–505

    Article  CAS  Google Scholar 

  9. Huang LF, Rubin DA, Britton CA (1999) Greater tuberosity changes as revealed by radiography: lack of clinical usefulness in patients with rotator cuff disease. AJR Am J Roentgenol 172:1381–1388

    Article  CAS  Google Scholar 

  10. Liotard JP, Cochard P, Walch G (1998) Critical analysis of the supraspinatus outlet view: rationale for a standard scapular Y-view. J Shoulder Elbow Surg 7:134–139

    Article  CAS  Google Scholar 

  11. Hyvonen P, Paivansalo M, Lehtiniemi H, Leppilahti J, Jalovaara P (2001) Supraspinatus outlet view in the diagnosis of stages II and III impingement syndrome. Acta Radiol 42:441–446

    Article  CAS  Google Scholar 

  12. Hussain A, Muzzammil M, Butt F, Valsamis EM, Dwyer AJ (2018) Effectiveness of plain shoulder radiograph in detecting degenerate rotator cuff tears. J Ayub Med Coll Abbottabad 30:8–11

    PubMed  Google Scholar 

  13. Chin K, Chowdhury A, Leivadiotou D, Marmery H, Ahrens PM (2017) The accuracy of plain radiographs in diagnosing degenerate rotator cuff disease. Shoulder Elbow. https://doi.org/10.1177/1758573217743942

  14. Simel DL, Samsa GP, Matchar DB (1991) Likelihood ratios with confidence: sample size estimation for diagnostic test studies. J Clin Epidemiol 44:763–770

    Article  CAS  Google Scholar 

  15. Jaeschke R, Guyatt GH, Sackett DL (1994) Users’ guides to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients? The Evidence-Based Medicine Working Group. JAMA 271:703–707

    Article  CAS  Google Scholar 

  16. Narasimhan R, Shamse K, Nash C, Dhingra D, Kennedy S (2016) Prevalence of subscapularis tears and accuracy of shoulder ultrasound in pre-operative diagnosis. Int Orthop 40:975–979

    Article  CAS  Google Scholar 

  17. Jacobson JA (2011) Shoulder US: anatomy, technique, and scanning pitfalls. Radiology 260:6–16

    Article  Google Scholar 

  18. Okoroha KR, Mehran N, Duncan J et al (2017) Characterization of rotator cuff tears: ultrasound versus magnetic resonance imaging. Orthopedics 40:e124–e130

    Article  Google Scholar 

  19. Ide J, Maeda S, Takagi K (2005) Arthroscopic transtendon repair of partial-thickness articular-side tears of the rotator cuff: anatomical and clinical study. Am J Sports Med 33:1672–1679

    Article  Google Scholar 

  20. Moosmayer S, Lund G, Seljom US et al (2019) At a 10-year follow-up, tendon repair is superior to physiotherapy in the treatment of small and medium-sized rotator cuff tears. J Bone Joint Surg Am 101:1050–1060

    Article  Google Scholar 

  21. McCallister WV, Parsons IM, Titelman RM, Matsen FA 3rd (2005) Open rotator cuff repair without acromioplasty. J Bone Joint Surg Am 87:1278–1283

    PubMed  Google Scholar 

  22. Harryman DT 2nd, Mack LA, Wang KY, Jackins SE, Richardson ML, Matsen FA 3rd (1991) Repairs of the rotator cuff. Correlation of functional results with integrity of the cuff. J Bone Joint Surg Am 73:982–989

    Article  Google Scholar 

  23. Yamamoto A, Takagishi K, Osawa T et al (2010) Prevalence and risk factors of a rotator cuff tear in the general population. J Shoulder Elbow Surg 19:116–120

    Article  Google Scholar 

  24. Sayampanathan AA, Andrew TH (2017) Systematic review on risk factors of rotator cuff tears. J Orthop Surg (Hong Kong) 25:2309499016684318

    Google Scholar 

  25. Koh KH, Han KY, Yoon YC, Lee SW, Yoo JC (2013) True anteroposterior (Grashey) view as a screening radiograph for further imaging study in rotator cuff tear. J Shoulder Elbow Surg 22:901–907

    Article  Google Scholar 

  26. Ono K, Yamamuro T, Rockwood CA Jr (1992) Use of a thirty-degree caudal tilt radiograph in the shoulder impingement syndrome. J Shoulder Elbow Surg 1:246–252

    Article  CAS  Google Scholar 

  27. Mayerhoefer ME, Breitenseher MJ, Roposch A, Treitl C, Wurnig C (2005) Comparison of MRI and conventional radiography for assessment of acromial shape. AJR Am J Roentgenol 184:671–675

    Article  Google Scholar 

  28. Bigliani LU, Ticker JB, Flatow EL, Soslowsky LJ, Mow VC (1991) The relationship of acromial architecture to rotator cuff disease. Clin Sports Med 10:823–838

    CAS  PubMed  Google Scholar 

  29. Hu J, Shen L, Albanie S, Sun G, Wu E (2018) Squeeze-and-excitation networks. Proceedings of the IEEE conference on computer vision and pattern recognition, pp 7132-7141

  30. Glorot X, Bengio Y (2010) Understanding the difficulty of training deep feedforward neural networks. Proceedings of the thirteenth international conference on artificial intelligence and statistics, pp 249-256

  31. Hinton G, Srivastava N, Swersky K (2012) Neural networks for machine learning lecture 6a overview of mini-batch gradient descent. Available via https://www.cs.toronto.edu/~hinton/coursera/lecture6/lec6.pdf

  32. Simonyan K, Vedaldi A, Zisserman A (2014) Deep inside convolutional networks: visualising image classification models and saliency maps. Available via https://arxiv.org/pdf/1312.6034.pdf. Accessed May 30 2019

  33. Philbrick KA, Yoshida K, Inoue D et al (2018) What does deep learning see? Insights from a classifier trained to predict contrast enhancement phase from CT images. AJR Am J Roentgenol 211:1184–1193

    Article  Google Scholar 

  34. Yao L, Prosky J, Poblenz E, Covington B, Lyman K (2018) Weakly supervised medical diagnosis and localization from multiple resolutions. arXiv preprint arXiv:180307703

  35. DeLong ER, DeLong DM, Clarke-Pearson DL (1988) Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 44:837–845

    Article  CAS  Google Scholar 

  36. Clopper CJ, Pearson ES (1934) The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika 26:404–413

    Article  Google Scholar 

  37. Boonstra AM, Schiphorst Preuper HR, Balk GA, Stewart RE (2014) Cut-off points for mild, moderate, and severe pain on the visual analogue scale for pain in patients with chronic musculoskeletal pain. Pain 155:2545–2550

    Article  Google Scholar 

  38. Bachmann LM, Kolb E, Koller MT, Steurer J, ter Riet G (2003) Accuracy of Ottawa ankle rules to exclude fractures of the ankle and mid-foot: systematic review. BMJ 326:417

    Article  Google Scholar 

  39. Foad A, Wijdicks CA (2012) The accuracy of magnetic resonance imaging and magnetic resonance arthrogram versus arthroscopy in the diagnosis of subscapularis tendon injury. Arthroscopy 28:636–641

    Article  Google Scholar 

  40. Garavaglia G, Ufenast H, Taverna E (2011) The frequency of subscapularis tears in arthroscopic rotator cuff repairs: a retrospective study comparing magnetic resonance imaging and arthroscopic findings. Int J Shoulder Surg 5:90–94

    Article  Google Scholar 

  41. Ono Y, Sakai T, Carroll MJ, Lo IK (2017) Tears of the subscapularis tendon: a critical analysis review. JBJS Rev 5:1–12

    Article  Google Scholar 

  42. Whiting P, Rutjes AW, Reitsma JB, Glas AS, Bossuyt PM, Kleijnen J (2004) Sources of variation and bias in studies of diagnostic accuracy: a systematic review. Ann Intern Med 140:189–202

    Article  Google Scholar 

  43. Lenza M, Buchbinder R, Takwoingi Y, Johnston RV, Hanchard NC, Faloppa F (2013) Magnetic resonance imaging, magnetic resonance arthrography and ultrasonography for assessing rotator cuff tears in people with shoulder pain for whom surgery is being considered. Cochrane Database Syst Rev. https://doi.org/10.1002/14651858.CD009020.pub2:CD009020

Download references

Acknowledgments

The authors sincerely thank Jeongmin Choi for her contribution in the data collection.

Funding

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. NRF-2019R1F1A1060126), National Research Foundation of Korea (NRF) grant funded by the Ministry of Education (No. 2017R1D1A1B03033610), and grant from the SNUBH Research Fund (No. 13-2019-006).

Author information

Authors and Affiliations

  1. Department of Radiology, Seoul National University Bundang Hospital, 82 Gumi-ro, 173 Beon-gil, Bundang-gu, Seongnam-si, Gyeonggi-do, 13620, South Korea

    Youngjune Kim, Dongjun Choi, Kyong Joon Lee, Yusuhn Kang, Joong Mo Ahn, Eugene Lee, Joon Woo Lee & Heung Sik Kang

Authors
  1. Youngjune Kim
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Dongjun Choi
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Kyong Joon Lee
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Yusuhn Kang
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Joong Mo Ahn
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Eugene Lee
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Joon Woo Lee
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Heung Sik Kang
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding authors

Correspondence to Kyong Joon Lee or Yusuhn Kang.

Ethics declarations

Guarantor

The scientific guarantor of this publication is Yusuhn Kang.

Conflict of interest

The authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.

Statistics and biometry

One of the authors (Dongjun Choi) has significant statistical expertise.

Informed consent

Written informed consent was waived by the institutional review board.

Ethical approval

Institutional review board approval was obtained.

Methodology

• retrospective

• experimental

• performed at one institution

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

ESM 1

(DOCX 196 kb)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Kim, Y., Choi, D., Lee, K.J. et al. Ruling out rotator cuff tear in shoulder radiograph series using deep learning: redefining the role of conventional radiograph. Eur Radiol 30, 2843–2852 (2020). https://doi.org/10.1007/s00330-019-06639-1

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00330-019-06639-1

Keywords

Search

Navigation