Semin Musculoskelet Radiol 2020; 24(01): 001-002
DOI: 10.1055/s-0039-3400511
Thieme Medical Publishers 333 Seventh Avenue, New York, NY 10001, USA.

Musculoskeletal Imaging Applications of Artificial Intelligence

Soterios Gyftopoulos
1  Department of Radiology, NYU Langone Health, New York, New York
Naveen Subhas
2  Department of Radiology, Cleveland Clinic, Cleveland, Ohio
› Author Affiliations
Further Information

Publication History

Publication Date:
28 January 2020 (online)

It is our pleasure to introduce this issue of Seminars in Musculoskeletal Radiology focusing on musculoskeletal (MSK) imaging applications of artificial intelligence (AI). The articles in this issue were written by experts in the fields of MSK imaging and AI. But the information included here should be useful for all radiologists, from the AI novice to the expert. We hope this issue not only serves as an educational tool but also as an inspiration for future MSK radiologists to perform their own work with AI.

Our first article, “Artificial Intelligence Explained for Nonexperts” by Narges Razavian and Krzysztof Geras, provides a review of AI for those who are unfamiliar with this tool and its uses in imaging. Then in “Improving the Speed of MRI with Artificial Intelligence,” Drs. Johnson, Recht, and Knoll discuss the use of AI for the task of image generation, specifically focusing on using machine learning and convoluted neural networks to develop algorithms to accelerate acquisition of MRI images from undersampled data sets (fast MRI) and generate high-resolution images from low-resolution data (super-resolution MRI).

Drs. Garwood, Tai, Joshi, and Watts review the use of AI in the evaluation of knee pathology in “The Use of Artificial Intelligence in the Evaluation of Knee Pathology,” which has been the focus of many recently published studies, ranging from automated radiographic grading of knee osteoarthritis to automated segmentation of articular cartilage and menisci on knee MR examinations to the automated diagnosis of anterior cruciate ligament tears. In the article that follows, “Artificial Intelligence in the Evaluation of Body Composition,” Drs. Wang and Torriani use the setting of body composition analysis to discuss the advantages and challenges in using AI for the tasks of image segmentation, classification, and quantification.

Drs. Gorelik, Chong, and Lin provide a comprehensive overview of the current and potential future uses of AI in imaging interpretations in “Pattern Recognition in Musculoskeletal Imaging Using Artificial Intelligence.” The authors discuss the use of AI in the diagnosis of common MSK conditions including osteoarthritis, fractures, and tumors.

The current state of radiomics and segmentation in MSK imaging is discussed in “Quantification in Musculoskeletal Imaging Using Computational Analysis and Machine Learning: Segmentation and Radiomics” by Drs. Cuadra, Favre, and Omoumi. Their article reviews the most commonly used segmentation methods and also describes the radiomics pipeline, highlighting its potential impact on clinical workflow and associated challenges.

In “From Data to Value: How Artificial Intelligence Augments the Radiology Business to Create Value,” Drs. Martin-Carreras and Chen explore the use of AI for tasks outside of image generation and classification such as patient scheduling, study protocoling, and study interpretation and reporting that have the potential to improve the efficiency and value of the services we deliver to our patients and referring physicians.

Drs. Forney and McBride examine AI through the lens of radiology residents and discuss the importance of exposing and educating them to the benefits and limitations of AI in their article, “Artificial Intelligence in Radiology Residency Training.”

We would like to thank the authors for their contributions to this issue. We hope you enjoy reading these articles as much as we did.