AN ANALYSIS OF DOSE AND IMAGE QUALITY IN DIGITAL PAEDIATRIC IMAGING

Richard Gould, Phil Doyle, Sonyia McFadden, Ciara Hughes

Research output: Contribution to conferenceAbstract

Abstract

Since the advent of digital imaging increasing numbers of radiological examinations are being performed on infants and children. Evidence suggests there is much scope for dose reduction in paediatrics through use of additional copper filtration, optimum tube kilo-voltages and an appropriate choice of scatter removal technique. The current study attempts to propose optimum settings for imaging neonate to adolescence patient sizes.
Anthropomorphic phantoms (CIRS, Virgina, USA) were used to examine which radiographic factors provide maximum image quality for the lowest dose penalty. Anthropomorphic phantoms are clinically realistic in terms of patient size and irradiation geometry, X-ray attenuation and scatter producing properties. The phantoms were imaged with a caesium iodide flat panel detector (Carestream DRX-1). Images were acquired using 50 to 90 kVp with inherent filtration only and with 0.2mm of additional copper filtration. Measurements were repeated using an anti-scatter (AS) grid, a 15cm air gap and with no scatter removal technique.
Image data was analysed using signal-to-noise ratio (SNR) as an objective measurement of image quality in the chest, abdomen, pelvis and lumbar spine. The SNR was computed as a ratio of the linearised mean pixel value (signal) and the linearised standard deviation (noise) of regions of interest drawn using Image J software (NIH, USA). A figure of merit (FOM), calculated using effective dose was used to assist in evaluating the optimum balance between good diagnostic image quality (high SNR) and the associated radiation dose.
For the neonate, 5 year old and 10 year old phantom sizes, better image quality was observed using no scatter removal technique, followed by use of a 15cm air gap. Use of an AS grid provided the poorest image quality (lowest SNR values). SNR values peaked in the range 70 to 80 kVp for the examination types simulated, indicating these tube potentials are optimum. The addition of 0.2 mm Cu filtration provided no significant difference to measured SNR values. However when the effective dose is taken into account using a FOM, significant dose savings were realised.
The results indicate that image quality and patient dose is optimum for general paediatric radiography using tube potentials of 70 to 80 kVp with 0.2 mm Cu additional filtration and no scatter reduction method. This optimised technique would reduce patient doses by more than half, relative to conventional adult settings.
LanguageEnglish
Publication statusPublished - Sep 2013
EventEuropean Radiation Research conference - Dublin
Duration: 2 Sep 2013 → …

Conference

ConferenceEuropean Radiation Research conference
CityDublin
Period2/09/13 → …

Fingerprint

dosage
signal to noise ratios
tubes
figure of merit
examination
grids
cesium iodides
copper
abdomen
pelvis
spine
chest
air
radiography
penalties
standard deviation
attenuation
pixels
computer programs
irradiation

Cite this

Gould, R., Doyle, P., McFadden, S., & Hughes, C. (2013). AN ANALYSIS OF DOSE AND IMAGE QUALITY IN DIGITAL PAEDIATRIC IMAGING. Abstract from European Radiation Research conference, Dublin, .
Gould, Richard ; Doyle, Phil ; McFadden, Sonyia ; Hughes, Ciara. / AN ANALYSIS OF DOSE AND IMAGE QUALITY IN DIGITAL PAEDIATRIC IMAGING. Abstract from European Radiation Research conference, Dublin, .
@conference{3bd29d2a8dc64fc3b0e860fe89e575ab,
title = "AN ANALYSIS OF DOSE AND IMAGE QUALITY IN DIGITAL PAEDIATRIC IMAGING",
abstract = "Since the advent of digital imaging increasing numbers of radiological examinations are being performed on infants and children. Evidence suggests there is much scope for dose reduction in paediatrics through use of additional copper filtration, optimum tube kilo-voltages and an appropriate choice of scatter removal technique. The current study attempts to propose optimum settings for imaging neonate to adolescence patient sizes.Anthropomorphic phantoms (CIRS, Virgina, USA) were used to examine which radiographic factors provide maximum image quality for the lowest dose penalty. Anthropomorphic phantoms are clinically realistic in terms of patient size and irradiation geometry, X-ray attenuation and scatter producing properties. The phantoms were imaged with a caesium iodide flat panel detector (Carestream DRX-1). Images were acquired using 50 to 90 kVp with inherent filtration only and with 0.2mm of additional copper filtration. Measurements were repeated using an anti-scatter (AS) grid, a 15cm air gap and with no scatter removal technique.Image data was analysed using signal-to-noise ratio (SNR) as an objective measurement of image quality in the chest, abdomen, pelvis and lumbar spine. The SNR was computed as a ratio of the linearised mean pixel value (signal) and the linearised standard deviation (noise) of regions of interest drawn using Image J software (NIH, USA). A figure of merit (FOM), calculated using effective dose was used to assist in evaluating the optimum balance between good diagnostic image quality (high SNR) and the associated radiation dose.For the neonate, 5 year old and 10 year old phantom sizes, better image quality was observed using no scatter removal technique, followed by use of a 15cm air gap. Use of an AS grid provided the poorest image quality (lowest SNR values). SNR values peaked in the range 70 to 80 kVp for the examination types simulated, indicating these tube potentials are optimum. The addition of 0.2 mm Cu filtration provided no significant difference to measured SNR values. However when the effective dose is taken into account using a FOM, significant dose savings were realised.The results indicate that image quality and patient dose is optimum for general paediatric radiography using tube potentials of 70 to 80 kVp with 0.2 mm Cu additional filtration and no scatter reduction method. This optimised technique would reduce patient doses by more than half, relative to conventional adult settings.",
author = "Richard Gould and Phil Doyle and Sonyia McFadden and Ciara Hughes",
year = "2013",
month = "9",
language = "English",
note = "European Radiation Research conference ; Conference date: 02-09-2013",

}

Gould, R, Doyle, P, McFadden, S & Hughes, C 2013, 'AN ANALYSIS OF DOSE AND IMAGE QUALITY IN DIGITAL PAEDIATRIC IMAGING' European Radiation Research conference, Dublin, 2/09/13, .

AN ANALYSIS OF DOSE AND IMAGE QUALITY IN DIGITAL PAEDIATRIC IMAGING. / Gould, Richard; Doyle, Phil; McFadden, Sonyia; Hughes, Ciara.

2013. Abstract from European Radiation Research conference, Dublin, .

Research output: Contribution to conferenceAbstract

TY - CONF

T1 - AN ANALYSIS OF DOSE AND IMAGE QUALITY IN DIGITAL PAEDIATRIC IMAGING

AU - Gould, Richard

AU - Doyle, Phil

AU - McFadden, Sonyia

AU - Hughes, Ciara

PY - 2013/9

Y1 - 2013/9

N2 - Since the advent of digital imaging increasing numbers of radiological examinations are being performed on infants and children. Evidence suggests there is much scope for dose reduction in paediatrics through use of additional copper filtration, optimum tube kilo-voltages and an appropriate choice of scatter removal technique. The current study attempts to propose optimum settings for imaging neonate to adolescence patient sizes.Anthropomorphic phantoms (CIRS, Virgina, USA) were used to examine which radiographic factors provide maximum image quality for the lowest dose penalty. Anthropomorphic phantoms are clinically realistic in terms of patient size and irradiation geometry, X-ray attenuation and scatter producing properties. The phantoms were imaged with a caesium iodide flat panel detector (Carestream DRX-1). Images were acquired using 50 to 90 kVp with inherent filtration only and with 0.2mm of additional copper filtration. Measurements were repeated using an anti-scatter (AS) grid, a 15cm air gap and with no scatter removal technique.Image data was analysed using signal-to-noise ratio (SNR) as an objective measurement of image quality in the chest, abdomen, pelvis and lumbar spine. The SNR was computed as a ratio of the linearised mean pixel value (signal) and the linearised standard deviation (noise) of regions of interest drawn using Image J software (NIH, USA). A figure of merit (FOM), calculated using effective dose was used to assist in evaluating the optimum balance between good diagnostic image quality (high SNR) and the associated radiation dose.For the neonate, 5 year old and 10 year old phantom sizes, better image quality was observed using no scatter removal technique, followed by use of a 15cm air gap. Use of an AS grid provided the poorest image quality (lowest SNR values). SNR values peaked in the range 70 to 80 kVp for the examination types simulated, indicating these tube potentials are optimum. The addition of 0.2 mm Cu filtration provided no significant difference to measured SNR values. However when the effective dose is taken into account using a FOM, significant dose savings were realised.The results indicate that image quality and patient dose is optimum for general paediatric radiography using tube potentials of 70 to 80 kVp with 0.2 mm Cu additional filtration and no scatter reduction method. This optimised technique would reduce patient doses by more than half, relative to conventional adult settings.

AB - Since the advent of digital imaging increasing numbers of radiological examinations are being performed on infants and children. Evidence suggests there is much scope for dose reduction in paediatrics through use of additional copper filtration, optimum tube kilo-voltages and an appropriate choice of scatter removal technique. The current study attempts to propose optimum settings for imaging neonate to adolescence patient sizes.Anthropomorphic phantoms (CIRS, Virgina, USA) were used to examine which radiographic factors provide maximum image quality for the lowest dose penalty. Anthropomorphic phantoms are clinically realistic in terms of patient size and irradiation geometry, X-ray attenuation and scatter producing properties. The phantoms were imaged with a caesium iodide flat panel detector (Carestream DRX-1). Images were acquired using 50 to 90 kVp with inherent filtration only and with 0.2mm of additional copper filtration. Measurements were repeated using an anti-scatter (AS) grid, a 15cm air gap and with no scatter removal technique.Image data was analysed using signal-to-noise ratio (SNR) as an objective measurement of image quality in the chest, abdomen, pelvis and lumbar spine. The SNR was computed as a ratio of the linearised mean pixel value (signal) and the linearised standard deviation (noise) of regions of interest drawn using Image J software (NIH, USA). A figure of merit (FOM), calculated using effective dose was used to assist in evaluating the optimum balance between good diagnostic image quality (high SNR) and the associated radiation dose.For the neonate, 5 year old and 10 year old phantom sizes, better image quality was observed using no scatter removal technique, followed by use of a 15cm air gap. Use of an AS grid provided the poorest image quality (lowest SNR values). SNR values peaked in the range 70 to 80 kVp for the examination types simulated, indicating these tube potentials are optimum. The addition of 0.2 mm Cu filtration provided no significant difference to measured SNR values. However when the effective dose is taken into account using a FOM, significant dose savings were realised.The results indicate that image quality and patient dose is optimum for general paediatric radiography using tube potentials of 70 to 80 kVp with 0.2 mm Cu additional filtration and no scatter reduction method. This optimised technique would reduce patient doses by more than half, relative to conventional adult settings.

M3 - Abstract

ER -

Gould R, Doyle P, McFadden S, Hughes C. AN ANALYSIS OF DOSE AND IMAGE QUALITY IN DIGITAL PAEDIATRIC IMAGING. 2013. Abstract from European Radiation Research conference, Dublin, .