Artificial Intelligence in Musculoskeletal Imaging: Review of Current Literature, Challenges, and Trends

Abstract

Artificial intelligence (AI) has gained major attention with a rapid increase in the number of published articles, mostly recently. This review provides a general understanding of how AI can or will be useful to the musculoskeletal radiologist. After a brief technical background on AI, machine learning, and deep learning, we illustrate, through examples from the musculoskeletal literature, potential AI applications in the various steps of the radiologist's workflow, from managing the request to communication of results. The implementation of AI solutions does not go without challenges and limitations. These are also discussed, as well as the trends and perspectives.

Keywords

machine learning - artificial intelligence - deep learning - rheumatology - musculoskeletal system

• References

• 1 Singh S, Okun A, Jackson A. Artificial intelligence: learning to play Go from scratch. Nature 2017; 550 (7676): 336-337
• 2 McCarthy JMM, Rochester N, Shannon CE. A proposal for the Dartmouth summer research project on artificial intelligence. AI Mag 2006; 27: 12-14
• 3 Hutson M. AI glossary: artificial intelligence, in so many words. Science 2017; 357 (6346): 19
• 4 Merriam Webster's Collegiate Dictionary. AI 2018 ; Available at: https://www.merriam-webster.com/dictonary/artificial intelligence . Accessed January 25, 2019
• 5 Faggella D. . What is machine learning? 2017 ; Available at: https://emerjcom/ai-glossary-terms/what-is-machine-learning/ . Accessed January 25, 2019
• 6 Krizhevsky A, Sutskever I, Hinton GE. ImageNet Classification with Deep Convolutional Neural Networks. Paper presented at: 26th Annual Conference on Neural Information Processing Systems (NIPS); December 3–8, 2012 ; Lake Tahoe, NV
• 7 Simonyan K, Zisserman A. Very deep convolutional networks for large-scale image recognition. Paper presented at: 3rd International Conference on Learning Representations (ICLR); May 7–9, 2015 ; San Diego, CA
• 8 Syed AB, Zoga AC. Artificial Intelligence in radiology: current technology and future directions. Semin Musculoskelet Radiol 2018; 22 (05) 540-545
• 9 LeCun Y, Bengio Y, Hinton G. Deep learning. Nature 2015; 521 (7553): 436-444
• 10 Litjens G, Kooi T, Bejnordi BE. , et al. A survey on deep learning in medical image analysis. Med Image Anal 2017; 42: 60-88
• 11 Kahn Jr CE, Kovatsis PG, Messersmith RN, Lehr JL. Automated entry of radiology requisition information with artificial-intelligence techniques. AJR Am J Roentgenol 1989; 153 (05) 1085-1088
• 12 Kahn Jr CE. Artificial intelligence in radiology: decision support systems. Radiographics 1994; 14 (04) 849-861
• 13 Trivedi H, Mesterhazy J, Laguna B, Vu T, Sohn JH. Automatic determination of the need for intravenous contrast in musculoskeletal MRI examinations using IBM Watson's natural language processing algorithm. J Digit Imaging 2018; 31 (02) 245-251
• 14 Lee YH. Efficiency improvement in a busy radiology practice: determination of musculoskeletal magnetic resonance imaging protocol using deep-learning convolutional neural networks. J Digit Imaging 2018; 31 (05) 604-610
• 15 American College of Radiology Appropriateness Criteria. Available at: https://www.acr.org/Clinical-Resources/ACR-Appropriateness-Criteria 2019 Accessed January 25, 2019
• 16 European Society of Radiology ESR iGuide. Available at: https://www.myesr.org/esriguide 2019 Accessed January 25, 2019
• 17 Chaudhari AS, Fang Z, Kogan F. , et al. Super-resolution musculoskeletal MRI using deep learning. Magn Reson Med 2018; 80 (05) 2139-2154
• 18 Galbusera F, Bassani T, Casaroli G. , et al. Generative models: an upcoming innovation in musculoskeletal radiology? A preliminary test in spine imaging. Eur Radiol Exp 2018; 2 (01) 29
• 19 Tang A, Tam R, Cadrin-Chênevert A. , et al; Canadian Association of Radiologists (CAR) Artificial Intelligence Working Group. Canadian Association of Radiologists white paper on artificial intelligence in radiology. Can Assoc Radiol J 2018; 69 (02) 120-135
• 20 Hosny A, Parmar C, Quackenbush J, Schwartz LH, Aerts HJWL. Artificial intelligence in radiology. Nat Rev Cancer 2018; 18 (08) 500-510
• 21 Gilles B, Pai DK. Fast musculoskeletal registration based on shape matching. Med Image Comput Comput Assist Interv 2008; 11: 822-829
• 22 Lindsey R, Daluiski A, Chopra S. , et al. Deep neural network improves fracture detection by clinicians. Proc Natl Acad Sci USA 2018; 115 (45) 11591-11596
• 23 Olczak J, Fahlberg N, Maki A. , et al. Artificial intelligence for analyzing orthopedic trauma radiographs. Acta Orthop 2017; 88 (06) 581-586
• 24 Kim DH, MacKinnon T. Artificial intelligence in fracture detection: transfer learning from deep convolutional neural networks. Clin Radiol 2018; 73 (05) 439-445
• 25 Urakawa T, Tanaka Y, Goto S, Matsuzawa H, Watanabe K, Endo N. Detecting intertrochanteric hip fractures with orthopedist-level accuracy using a deep convolutional neural network. Skeletal Radiol 2019; 48 (02) 239-244
• 26 Bien N, Rajpurkar P, Ball RL. , et al. Deep-learning-assisted diagnosis for knee magnetic resonance imaging: development and retrospective validation of MRNet. PLoS Med 2018; 15 (11) e1002699
• 27 Štajduhar I, Mamula M, Miletić D, Ünal G. Semi-automated detection of anterior cruciate ligament injury from MRI. Comput Methods Programs Biomed 2017; 140: 151-164
• 28 Lodwick GS, Wilson AJ, Farrell C, Virtama P, Smeltzer FM, Dittrich F. Estimating rate of growth in bone lesions: observer performance and error. Radiology 1980; 134 (03) 585-590
• 29 Reinus WR, Wilson AJ, Kalman B, Kwasny S. Diagnosis of focal bone lesions using neural networks. Invest Radiol 1994; 29 (06) 606-611
• 30 Kahn Jr CE, Laur JJ, Carrera GF. A Bayesian network for diagnosis of primary bone tumors. J Digit Imaging 2001; 14 (02) (Suppl. 01) 56-57
• 31 Piraino DW, Richmond BJ, Uetani M. , et al. Problems in applying expert system technology to radiographic image interpretation. J Digit Imaging 1989; 2 (01) 21-26
• 32 Fletcher C, Bridge JA, Hogendoorn P, Mertens F. WHO Classification of Tumors of Soft Tissue and Bone. 4th ed. Lyon, France: International Agency for Research on Cancer; 2013
• 33 Do BH, Langlotz C, Beaulieu CF. Bone tumor diagnosis using a naïve Bayesian model of demographic and radiographic features. J Digit Imaging 2017; 30 (05) 640-647
• 34 Kijowski R, Davis KW, Woods MA. , et al. Knee joint: comprehensive assessment with 3D isotropic resolution fast spin-echo MR imaging—diagnostic performance compared with that of conventional MR imaging at 3.0 T. Radiology 2009; 252 (02) 486-495
• 35 Shakoor D, Guermazi A, Kijowski R. , et al. Diagnostic performance of three-dimensional MRI for depicting cartilage defects in the knee: a meta-analysis. Radiology 2018; 289 (01) 71-82
• 36 Colotti R, Omoumi P, Bonanno G, Ledoux JB, van Heeswijk RB. Isotropic three-dimensional T₂ mapping of knee cartilage: development and validation. J Magn Reson Imaging 2018; 47 (02) 362-371
• 37 Colotti R, Omoumi P, van Heeswijk RB, Bastiaansen JAM. Simultaneous fat-free isotropic 3D anatomical imaging and T₂ mapping of knee cartilage with lipid-insensitive binomial off-resonant RF excitation (LIBRE) pulses. J Magn Reson Imaging 2018 ; October 14 (Epub ahead of print)
• 38 Tiulpin A, Thevenot J, Rahtu E, Lehenkari P, Saarakkala S. Automatic knee osteoarthritis diagnosis from plain radiographs: a deep learning-based approach. Sci Rep 2018; 8 (01) 1727
• 39 Norman B, Pedoia V, Noworolski A, Link TM, Majumdar S. Applying densely connected convolutional neural networks for staging osteoarthritis severity from plain radiographs. J Digit Imaging 2018 ; October 10 (Epub ahead of print)
• 40 Xue Y, Zhang R, Deng Y, Chen K, Jiang T. A preliminary examination of the diagnostic value of deep learning in hip osteoarthritis. PLoS One 2017; 12 (06) e0178992
• 41 Liu F, Zhou Z, Samsonov A. , et al. Deep learning approach for evaluating knee MR images: achieving high diagnostic performance for cartilage lesion detection. Radiology 2018; 289 (01) 160-169
• 42 Omoumi P, Michoux N, Larbi A. , et al. Multirater agreement for grading the femoral and tibial cartilage surface lesions at CT arthrography and analysis of causes of disagreement. Eur J Radiol 2017; 88: 95-101
• 43 Omoumi P, Rubini A, Dubuc JE, Vande Berg BC, Lecouvet FE. Diagnostic performance of CT-arthrography and 1.5T MR-arthrography for the assessment of glenohumeral joint cartilage: a comparative study with arthroscopic correlation. Eur Radiol 2015; 25 (04) 961-969
• 44 Jamaludin A, Lootus M, Kadir T. , et al; Genodisc Consortium. ISSLS PRIZE IN BIOENGINEERING SCIENCE 2017: Automation of reading of radiological features from magnetic resonance images (MRIs) of the lumbar spine without human intervention is comparable with an expert radiologist. Eur Spine J 2017; 26 (05) 1374-1383
• 45 Tajmir SH, Lee H, Shailam R. , et al. Artificial intelligence-assisted interpretation of bone age radiographs improves accuracy and decreases variability. Skeletal Radiol 2019; 48 (02) 275-283
• 46 Halabi SS, Prevedello LM, Kalpathy-Cramer J. , et al. The RSNA Pediatric Bone Age Machine Learning Challenge. Radiology 2019; 290 (02) 498-503
• 47 Lin H. Identification of spinal deformity classification with total curvature analysis and artificial neural network. IEEE Trans Biomed Eng 2008; 55 (01) 376-382
• 48 Balsiger F, Steindel C, Arn M. , et al. Segmentation of peripheral nerves from magnetic resonance neurography: a fully-automatic, deep learning-based approach. Front Neurol 2018; 9: 777
• 49 Castrillon JGAA, Navab N, Richiardi J. Learning with multi-site {fMRI} graph data. Presented at: 48th Asilomar Conference on Signals, Systems, and Computers; November 2–5, 2014; Pacific Grove, CA
• 50 Goldberg JE, Rosenkrantz AB. Artificial intelligence and radiology: a social media perspective. Curr Probl Diagn Radiol 2018 ; July 23 (Epub ahead of print)
• 51 Thakar S. NYU, Facebook release massive MRI dataset as part of ongoing AI project. Available at: https://www.radiologybusiness.com/topics/artificial-intelligence/nyu-facebook-release-knee-mri-dataset-ai 2018 , Accessed January 25, 2019