A Review of Deep Learning CT Reconstruction: Concepts, Limitations, and Promise in Clinical Practice

Abstract Purpose of Review Deep Learning reconstruction (DLR) is the current state-of-the-art method for CT image formation. Comparisons to existing filter back-projection, iterative, and model-based reconstructions are now available in literature. This review summarizes prior methods, introduces DLR, then reviews recent findings from DLR a physics clinical perspective. Recent Findings has been shown allow noise magnitude reductions relative filtered back-projection without suffering “plastic” or “blotchy” texture that was found objectionable with most iterative solutions. Clinically, early reader studies have reported increases subjective quality scores successfully implemented DLR-enabled dose reductions. Summary The future bright; deep learning methods only started tackle problems this space via addressing reduction. Artifact mitigation spectral applications likely be candidates applications.