Display options
Cite
Nissinen T, Suoranta S, Saavalainen T, et al. Detecting pathological features and predicting fracture risk from dual-energy X-ray absorptiometry images using deep learning. Bone Rep. 2021;14:101070doi: 10.1016/j.bonr.2021.101070.
Nissinen, T., Suoranta, S., Saavalainen, T., Sund, R., Hurskainen, O., Rikkonen, T., Kröger, H., Lähivaara, T., & Väänänen, S. P. (2021). Detecting pathological features and predicting fracture risk from dual-energy X-ray absorptiometry images using deep learning. Bone reports, 14101070. https://doi.org/10.1016/j.bonr.2021.101070
Nissinen, Tomi, et al. "Detecting pathological features and predicting fracture risk from dual-energy X-ray absorptiometry images using deep learning." Bone reports vol. 14 (2021): 101070. doi: https://doi.org/10.1016/j.bonr.2021.101070
Nissinen T, Suoranta S, Saavalainen T, Sund R, Hurskainen O, Rikkonen T, Kröger H, Lähivaara T, Väänänen SP. Detecting pathological features and predicting fracture risk from dual-energy X-ray absorptiometry images using deep learning. Bone Rep. 2021 Apr 24;14:101070. doi: 10.1016/j.bonr.2021.101070. eCollection 2021 Jun. PMID: 33997147; PMCID: PMC8102403.
Copy Download .nbib Format: NLM
Share it on

Bone Rep. 2021 Apr 24;14:101070. doi: 10.1016/j.bonr.2021.101070. eCollection 2021 Jun.

Detecting pathological features and predicting fracture risk from dual-energy X-ray absorptiometry images using deep learning.

Bone reports

Tomi Nissinen, Sanna Suoranta, Taavi Saavalainen, Reijo Sund, Ossi Hurskainen, Toni Rikkonen, Heikki Kröger, Timo Lähivaara, Sami P Väänänen

Affiliations

  1. Department of Applied Physics, University of Eastern Finland, POB1627, 70211 Kuopio, Finland.
  2. Department of Orthopaedics, Traumatology and Hand Surgery, Kuopio University Hospital, POB1777, 70211 Kuopio, Finland.
  3. Department of Clinical Radiology, Kuopio University Hospital, POB1777, 70211 Kuopio, Finland.
  4. Institute of Clinical Medicine, University of Eastern Finland, POB1627, 70211 Kuopio, Finland.

PMID: 33997147 PMCID: PMC8102403 DOI: 10.1016/j.bonr.2021.101070

Abstract

Dual-energy X-ray absorptiometry (DXA) is the gold standard imaging method for diagnosing osteoporosis in clinical practice. The DXA images are commonly used to estimate a numerical value for bone mineral density (BMD), which decreases in osteoporosis. Low BMD is a known risk factor for osteoporotic fractures. In this study, we used deep learning to identify lumbar scoliosis and structural abnormalities that potentially affect BMD but are often neglected in lumbar spine DXA analysis. In addition, we tested the approach's ability to predict fractures using only DXA images. A dataset of 2949 images gathered by Kuopio Osteoporosis Risk Factor and Prevention Study was used to train a convolutional neural network (CNN) for classification. The model was able to classify scoliosis with an AUC of 0.96 and structural abnormalities causing unreliable BMD measurement with an AUC of 0.91. It predicted fractures occurring within 5 years from the lumbar spine DXA scan with an AUC of 0.63, meeting the predictive performance of combined BMD measurements from the lumbar spine and hip. In an independent test set of 574 clinical patients, AUC for lumbar scoliosis was 0.93 and AUC for unreliable BMD measurements was 0.94. In each classification task, neural network visualizations indicated the model's predictive strategy. We conclude that deep learning could complement the well established DXA method for osteoporosis diagnostics by analyzing incidental findings and image reliability, and improve its predictive ability in the future.

© 2021 The Author(s).

Keywords: Deep learning; Degeneration; Dual-energy X-ray absorptiometry; Fracture risk; Lumbar spine; Scoliosis; Trabecular bone score

Conflict of interest statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References

  1. Commun Biol. 2019 Oct 25;2:390 - PubMed
  2. Nature. 2015 May 28;521(7553):436-44 - PubMed
  3. Radiology. 2020 Apr;295(1):136-145 - PubMed
  4. J Bone Miner Res. 2010 Nov;25(11):2332-40 - PubMed
  5. Osteoporos Int. 2004 Nov;15(11):847-54 - PubMed
  6. Am J Epidemiol. 1999 Sep 1;150(5):511-6 - PubMed
  7. Am J Epidemiol. 2018 Jan 1;187(1):94-102 - PubMed
  8. Paediatr Child Health. 2007 Nov;12(9):771-6 - PubMed
  9. Osteoporos Int. 1992 May;2(3):135-40 - PubMed
  10. Sci Rep. 2018 Jan 29;8(1):1727 - PubMed
  11. Osteoporos Int. 1999;9 Suppl 2:S2-8 - PubMed
  12. Osteoporos Int. 2007 Aug;18(8):1033-46 - PubMed
  13. J Magn Reson Imaging. 2019 Apr;49(4):1029-1038 - PubMed
  14. J Bone Joint Surg Am. 2010 Mar;92(3):743-53 - PubMed
  15. Radiol Clin North Am. 2012 Jul;50(4):613-28 - PubMed
  16. Maturitas. 2014 Dec;79(4):449-55 - PubMed
  17. Ann Intern Med. 1991 Dec 1;115(11):837-42 - PubMed
  18. Osteoporos Int. 2008 Apr;19(4):385-97 - PubMed
  19. Front Bioeng Biotechnol. 2018 Jun 27;6:75 - PubMed
  20. J Bone Miner Res. 2008 Dec;23(12):1892-904 - PubMed
  21. Bone. 2008 Apr;42(4):775-87 - PubMed
  22. J Bone Miner Res. 2010 Jul;25(7):1487-95 - PubMed
  23. Radiology. 2019 Nov;293(2):405-411 - PubMed
  24. Nat Med. 2020 Jan;26(1):77-82 - PubMed
  25. J Bone Miner Res. 2020 Jan;35(1):28-35 - PubMed
  26. Comput Math Methods Med. 2019 Feb 19;2019:6357171 - PubMed
  27. J Bone Miner Res. 2016 Feb;31(2):341-6 - PubMed
  28. IEEE Trans Image Process. 2005 Sep;14(9):1360-71 - PubMed
  29. Calcif Tissue Int. 2020 Apr;106(4):378-385 - PubMed
  30. Semin Musculoskelet Radiol. 2016 Jul;20(3):246-253 - PubMed
  31. Osteoporos Int. 2004 Apr;15(4):274-80 - PubMed
  32. Osteoporos Int. 2014 Jan;25(1):1-21 - PubMed
  33. J Bone Miner Res. 2005 Mar;20(3):501-8 - PubMed
  34. Invest Radiol. 1990 Jan;25(1):6-18 - PubMed
  35. Osteoporos Int. 2010 Jun;21(6):919-29 - PubMed

Publication Types