Deep learning reconstruction with single-energy metal artifact reduction in pelvic computed tomography for patients with metal hip prostheses

Abstract

Purpose The aim of this study was to assess the impact of the deep learning reconstruction (DLR) with single-energy metal artifact reduction (SEMAR) (DLR-S) technique in pelvic helical computed tomography (CT) images for patients with metal hip prostheses and compare it with DLR and hybrid iterative reconstruction (IR) with SEMAR (IR-S). Materials and methods This retrospective study included 26 patients (mean age 68.6 ± 16.6 years, with 9 males and 17 females) with metal hip prostheses who underwent a CT examination including the pelvis. Axial pelvic CT images were reconstructed using DLR-S, DLR, and IR-S. In one-by-one qualitative analyses, two radiologists evaluated the degree of metal artifacts, noise, and pelvic structure depiction. In side-by-side qualitative analyses (DLR-S vs. IR-S), the two radiologists evaluated metal artifacts and overall quality. By placing regions of interest on the bladder and psoas muscle, the standard deviations of their CT attenuation were recorded, and the artifact index was calculated based on them. Results were compared between DLR-S vs. DLR and DLR vs. IR-S using the Wilcoxon signed-rank test. Results In one-by-one qualitative analyses, metal artifacts and structure depiction in DLR-S were significantly better than those in DLR; however, between DLR-S and IR-S, significant differences were noted only for reader 1. Image noise in DLR-S was rated as significantly reduced compared with that in IR-S by both readers. In side-by-side analyses, both readers rated that the DLR-S images are significantly better than IR-S images regarding overall image quality and metal artifacts. The median (interquartile range) of the artifact index for DLR-S was 10.1 (4.4–16.0) and was significantly better than those for DLR (23.1, 6.5–36.1) and IR-S (11.4, 7.8–17.9). Conclusion DLR-S provided better pelvic CT images in patients with metal hip prostheses than IR-S and DLR.

The impact of large language models on radiology: a guide for radiologists on the latest innovations in AI

“…The inception of Deep Learning (DL) has catalyzed a significant progression in artificial intelligence (AI) [1], unlocking numerous possibilities, especially in diagnostic radiology-an arena pivotal for accurate imaging data interpretation. This progression is attributed mainly to the emergence of Convolutional Neural Networks (CNNs) [2,3], which have markedly enhanced image recognition, segmentation, analysis, and improvement of image quality [1,[4][5][6][7][8][9][10][11][12][13][14][15]. This represents a foundational shift in automated feature extraction from imaging data, consequently reducing the time and expertise required for interpreting medical images.…”

Enhancing image quality in computed tomography angiography follow-ups after endovascular aneurysm repair: a comparative study of reconstruction techniques

“…In a phantom study, AiCE plus SEMAR performed the best in reducing artifacts compared to other algorithms [ 25 ]. Moreover, studies have shown that combining DLR with SEMAR further reduces image noise and artifacts, thus improving image quality, in CT images with metal implants [ 26 , 27 ]. However, the usefulness of AiCE with SEMAR (AiCE + SEMAR) in EVAR patients has not been reported.…”

Deep Learning Reconstruction Plus Single-Energy Metal Artifact Reduction for Supra Hyoid Neck CT in Patients With Dental Metals

“…Because of the improved image quality, DLR is expected to spread widely into daily clinical practice. However, although DLR can also be used in combination with SEMAR, 14 no published reports have evaluated the quality of combined DLR-SEMAR images compared with DLR alone or with Hybrid IR-SEMAR.…”