Phantom-based performance comparison of two commercial deep learning CT reconstruction algorithms with super- and normal-resolution settings

TL;DR

This study compares two deep learning CT reconstruction methods, showing that the super-resolution version improves image quality and lesion detection in abdominal scans.

Contribution

The study introduces a novel comparison of super-resolution and normal-resolution deep learning CT reconstruction algorithms using a phantom-based approach.

Findings

SR-DLR improved spatial resolution and detectability of simulated abdominal lesions compared to NR-DLR.

Image noise was reduced with SR-DLR for level-2 and level-3, but image texture was better for level-1 and level-2.

SR-DLR shows potential for reducing radiation doses in abdominal CT imaging.

Abstract

We compared a super-resolution deep learning image reconstruction (SR-DLR) algorithm with a normal-resolution (NR)-DLR algorithm according to radiation dose for abdominal computed tomography (CT).

An image-quality phantom was scanned with an energy-integrating detectors CT unit at three volume CT dose index radiation dose levels (12.7, 5.9, and 3 mGy). Images were reconstructed using a 1,0242 matrix for SR-DLR and a 5122 matrix for NR-DLR, for three DLR levels (level-1, level-2, and level-3). Noise power spectrum (NPS) and task-based transfer function (TTF) for iodine and Solid Water® inserts were computed; TTF values at 50% (f50, mm-1) were used to quantify spatial resolution. The detectability index (d’) was computed for two simulated lesions.

Noise magnitude values were lower with SR-DLR than with NR-DLR for level-2 (-27.6 ± 3.8%) and level-3 (-43.5 ± 1.4%), the opposite for level-1. Average NPS spatial frequency was higher with SR-DLR than with NR-DLR for all radiation dose levels for level-1 (55.9 ± 16.7%) and level-2 (20.1 ± 13.9%) and the opposite for level-3, except at 12.7 mGy. For both inserts, f50 was higher with SR-DLR than with NR-DLR at each radiation dose and DLR level. For simulated lesions and all DLR levels, d’ values were higher with SR-DLR than with NR-DLR (level-1, 6.0 ± 2.0%; level-2, 45.7 ± 5.0%; level-3, 75.2 ± 7.3%).

Compared to NR-DLR, SR-DLR improved spatial resolution and detectability of simulated abdominal lesions; image noise was reduced with SR-DLR only for level-2 and level-3, while image texture was better for level-1 and level-2.

Super-resolution DLR with a 1,0242 matrix size improved spatial resolution and detectability of simulated abdominal lesions compared to normal-resolution DLR. Validation in clinical settings is necessary before translation into routine practice.

The performance of a new deep learning super-resolution image reconstruction algorithm (SR-DLR) was compared to a normal-resolution (NR)-DLR algorithm using an image-quality phantom for an abdominal energy-integrating detector CT protocol.SR-DLR with a 1,0242 matrix improved spatial resolution and detectability of simulated abdominal lesions compared to NR-DLR with a 5122 matrix.Using SR-DLR, therefore, presents numerous prospects for improving abdominal CT images and a high potential for reducing the radiation doses.

The performance of a new deep learning super-resolution image reconstruction algorithm (SR-DLR) was compared to a normal-resolution (NR)-DLR algorithm using an image-quality phantom for an abdominal energy-integrating detector CT protocol.

SR-DLR with a 1,0242 matrix improved spatial resolution and detectability of simulated abdominal lesions compared to NR-DLR with a 5122 matrix.

Using SR-DLR, therefore, presents numerous prospects for improving abdominal CT images and a high potential for reducing the radiation doses.

Figures