Abstract
Following unprecedented success on the natural language tasks, Transformershave been successfully applied to several computer vision problems, achievingstate-of-the-art results and prompting researchers to reconsider the supremacyof convolutional neural networks (CNNs) as {de facto} operators. Capitalizingon these advances in computer vision, the medical imaging field has alsowitnessed growing interest for Transformers that can capture global contextcompared to CNNs with local receptive fields. Inspired from this transition, inthis survey, we attempt to provide a comprehensive review of the applicationsof Transformers in medical imaging covering various aspects, ranging fromrecently proposed architectural designs to unsolved issues. Specifically, wesurvey the use of Transformers in medical image segmentation, detection,classification, reconstruction, synthesis, registration, clinical reportgeneration, and other tasks. In particular, for each of these applications, wedevelop taxonomy, identify application-specific challenges as well as provideinsights to solve them, and highlight recent trends. Further, we provide acritical discussion of the field's current state as a whole, including theidentification of key challenges, open problems, and outlining promising futuredirections. We hope this survey will ignite further interest in the communityand provide researchers with an up-to-date reference regarding applications ofTransformer models in medical imaging. Finally, to cope with the rapiddevelopment in this field, we intend to regularly update the relevant latestpapers and their open-source implementations at\url{https://github.com/fahadshamshad/awesome-transformers-in-medical-imaging}.