MBIR Training for a 2.5D DL network in X-ray CT

TL;DR

This paper presents a deep learning approach that emulates model-based iterative reconstruction in X-ray CT, achieving similar image quality faster and with less computational cost.

Contribution

The authors develop a modified Unet-based 2.5D deep learning network trained to replicate MBIR image quality, reducing computation time while maintaining high image fidelity.

Findings

DL-MBIR images match MBIR in texture and noise characteristics

The method reduces computational cost compared to traditional MBIR

Results show successful emulation of MBIR operator by the DL network

Abstract

In computed tomographic imaging, model based iterative reconstruction methods have generally shown better image quality than the more traditional, faster filtered backprojection technique. The cost we have to pay is that MBIR is computationally expensive. In this work we train a 2.5D deep learning (DL) network to mimic MBIR quality image. The network is realized by a modified Unet, and trained using clinical FBP and MBIR image pairs. We achieve the quality of MBIR images faster and with a much smaller computation cost. Visually and in terms of noise power spectrum (NPS), DL-MBIR images have texture similar to that of MBIR, with reduced noise power. Image profile plots, NPS plots, standard deviation, etc. suggest that the DL-MBIR images result from a successful emulation of an MBIR operator.