A Tensor Factorization Method for 3D Super-Resolution with Application to Dental CT

TL;DR

This paper introduces a tensor-factorization-based method for 3D super-resolution in dental CT that is faster and slightly more accurate than existing iterative techniques, with fewer hyperparameters and easier parameter tuning.

Contribution

The paper presents a novel tensor factorization approach for 3D image super-resolution that significantly reduces computation time and maintains or improves image quality compared to state-of-the-art methods.

Findings

2-order-of-magnitude faster reconstruction time

Slightly improved PSNR and segmentation quality

Low sensitivity to hyperparameter variations

Abstract

Available super-resolution techniques for 3D images are either computationally inefficient prior-knowledge-based iterative techniques or deep learning methods which require a large database of known low- and high-resolution image pairs. A recently introduced tensor-factorization-based approach offers a fast solution without the use of known image pairs or strict prior assumptions. In this article this factorization framework is investigated for single image resolution enhancement with an off-line estimate of the system point spread function. The technique is applied to 3D cone beam computed tomography for dental image resolution enhancement. To demonstrate the efficiency of our method, it is compared to a recent state-of-the-art iterative technique using low-rank and total variation regularizations. In contrast to this comparative technique, the proposed reconstruction technique gives a 2-order-of-magnitude improvement in running time -- 2 minutes compared to 2 hours for a dental volume of 282$\times$266$\times$392 voxels. Furthermore, it also offers slightly improved quantitative results (peak signal-to-noise ratio, segmentation quality). Another advantage of the presented technique is the low number of hyperparameters. As demonstrated in this paper, the framework is not sensitive to small changes of its parameters, proposing an ease of use.