New image quality and dose reduction technique for pediatric digital radiography - 03/02/21

Doi : 10.1016/j.diii.2021.01.009

Joelle A. Feghali ^a,⁎ , Greg Chambers ^a, Julie Delépierre ^a, Sophie Chapeliere ^a, Inès Mannes ^a,^b, Catherine Adamsbaum ^a,^b
^a AP–HP, Bicêtre Hospital, Pediatric Radiology Department, 94270 Le Kremlin-Bicêtre, France
^b Medical School, Paris-Saclay University, 94270 Le Kremlin-Bicêtre, France

^⁎Corresponding author.

Sous presse. Épreuves corrigées par l'auteur. Disponible en ligne depuis le Wednesday 03 February 2021
Cet article a été publié dans un numéro de la revue, cliquez ici pour y accéder

Highlights

•	A novel approach to optimise pediatric digital radiography (DR) was introduced using a simulation tool and a specific image quality grid.
•	S-Vue™ image denoising was applied to thorax, pelvis, abdomen and spine DR images and its effect was assessed through simulation tool and image quality grid.
•	S-Vue™ can yield digital radiography images of pediatric patients with high diagnostic quality at half the standard radiation dose.

Le texte complet de cet article est disponible en PDF.

Abstract

Purpose

The purpose of this study was to test a new post-processing and denoising engine for patient dose reduction while maintaining diagnostic image quality (IQ) in pediatric digital radiography (DR).

Materials and methods

Pediatric DR images of the thorax, pelvis, abdomen and spine obtained in 174 patients (102 males, 72 females; mean age, 2±1.8 [SD] years; age range: 6 months–9 years) were retrieved. Artificial noise was added to the images to simulate acquisitions at 50%, 32% and 12.5% of the routine dose levels. A total of 696 images corresponding to four dose levels were post-processed using S-Vue™ and further blindly scored by three pediatric radiologists using a scoring grid of 4–6 criteria specifically defined per anatomical area. The mean score was assessed for each area and weight class (5–15 and 15–30kg) and compared across the simulated low dose images. Paired Wilcoxon test was used with a threshold difference of 0.5 (half a criterion) between scores to highlight a significant reduction in image quality. Inter-rater reliability was assessed using intraclass correlation coefficient (ICC).

Results

Only the 50% reduced dose images showed non-inferiority when compared to routine images for all of areas and weight classes (P<0.01). Very good inter-rater reliability of the overall scores was observed for the pelvis in the 5–15kg weight class (ICC=0.85) for images at full dose, 50% and 32% reduced dose. For the remaining areas (thorax, abdomen and spine) and weight classes, inter-rater reliability was moderate (ICC: 0.3–0.6).