The use of deep learning in medical imaging to improve spine care: A scoping review of current literature and clinical applications

• Published in: North American Spine Society journal (NASSJ) Vol. 15; p. 100236
• Main Authors: Constant, Caroline, Aubin, Carl-Eric, Kremers, Hilal Maradit, Garcia, Diana V. Vera, Wyles, Cody C., Rouzrokh, Pouria, Larson, Annalise Noelle
• Format: Journal Article
• Language: English
• Published: United States Elsevier Inc 01.09.2023; Elsevier
• Subjects: Artificial intelligence; Clinical care; Deep learning; Imaging; Machine learning; Review; Spine; Systematic Reviews/Meta-Analyses; Deep learning; Spine; Imaging; Machine learning; Clinical care; Review; Artificial intelligence
• ISSN: 2666-5484; 2666-5484
• DOI: 10.1016/j.xnsj.2023.100236

•Most deep learning (DL) in imaging studies focused on spinal conditions' detection and diagnosis.•A total of 92% of DL in imaging studies developed a new model while 8% validated a pre-existing one.•DL in medical imaging showed promising performance in improving clinical spine care.•Implementation or demonstration of DL in real-world situations was rare. Artificial intelligence is a revolutionary technology that promises to assist clinicians in improving patient care. In radiology, deep learning (DL) is widely used in clinical decision aids due to its ability to analyze complex patterns and images. It allows for rapid, enhanced data, and imaging analysis, from diagnosis to outcome prediction. The purpose of this study was to evaluate the current literature and clinical utilization of DL in spine imaging. This study is a scoping review and utilized the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology to review the scientific literature from 2012 to 2021. A search in PubMed, Web of Science, Embased, and IEEE Xplore databases with syntax specific for DL and medical imaging in spine care applications was conducted to collect all original publications on the subject. Specific data was extracted from the available literature, including algorithm application, algorithms tested, database type and size, algorithm training method, and outcome of interest. A total of 365 studies (total sample of 232,394 patients) were included and grouped into 4 general applications: diagnostic tools, clinical decision support tools, automated clinical/instrumentation assessment, and clinical outcome prediction. Notable disparities exist in the selected algorithms and the training across multiple disparate databases. The most frequently used algorithms were U-Net and ResNet. A DL model was developed and validated in 92% of included studies, while a pre-existing DL model was investigated in 8%. Of all developed models, only 15% of them have been externally validated. Based on this scoping review, DL in spine imaging is used in a broad range of clinical applications, particularly for diagnosing spinal conditions. There is a wide variety of DL algorithms, database characteristics, and training methods. Future studies should focus on external validation of existing models before bringing them into clinical use.