Correlation of the clinical and physical image quality in chest radiography for average adults with a computed radiography imaging system

Objective:

The purpose of this study was to examine the correlation between the quality of visually graded patient (clinical) chest images and a quantitative assessment of chest phantom (physical) images acquired with a computed radiography (CR) imaging system.

Methods:

The results of a previously published study, in which four experienced image evaluators graded computer-simulated postero-anterior chest images using a visual grading analysis scoring (VGAS) scheme, were used for the clinical image quality measurement. Contrast-to-noise ratio (CNR) and effective dose efficiency (eDE) were used as physical image quality metrics measured in a uniform chest phantom. Although optimal values of these physical metrics for chest radiography were not derived in this work, their correlation with VGAS in images acquired without an antiscatter grid across the diagnostic range of X-ray tube voltages was determined using Pearson’s correlation coefficient.

Results:

Clinical and physical image quality metrics increased with decreasing tube voltage. Statistically significant correlations between VGAS and CNR (R=0.87, p<0.033) and eDE (R=0.77, p<0.008) were observed.

Conclusion:

Medical physics experts may use the physical image quality metrics described here in quality assurance programmes and optimisation studies with a degree of confidence that they reflect the clinical image quality in chest CR images acquired without an antiscatter grid.

Advances in knowledge:

A statistically significant correlation has been found between the clinical and physical image quality in CR chest imaging. The results support the value of using CNR and eDE in the evaluation of quality in clinical thorax radiography.Chest radiography is one of the most frequently performed diagnostic radiographic examinations in the UK. A Health Protection Agency report 1] in 2010 showed that chest radiographs represented 19.6% of all radiographic examinations (albeit the contribution to collective dose was small at about 0.5%), so optimisation of radiation dose and image quality in chest radiography is an important research area, especially in the era of digital imaging. It is also a legal requirement in the UK under the Ionising Radiation (Medical Exposure) Regulations 2000 2] to optimise all medical exposures.Many investigators 3–12] have shown that it is not the system (including quantum) noise that is the limiting factor in chest radiography, but rather the projected patient anatomy or “anatomical noise”. It therefore follows that any images used to optimise a digital radiographic system for chest radiography must contain clinically realistic anatomical features and noise. The assessment of image quality of digital systems is typically undertaken with physical quality metrics, such as modulation transfer function (MTF), noise power spectra (NPS), detective quantum efficiency (DQE), contrast-to-noise ratio (CNR) and threshold contrast measurements 13–19]. Although these parameters describe the inherent performance of the imaging detector extremely well, it is difficult to link them to clinical image quality (i.e. the adequacy of patient images) 20] and therefore it is difficult to use them in any optimisation exercise. Furthermore, recent work by Samei et al 21,22] has shown that these metrics are not only detector centric but also not measured under typical clinical conditions. An alternative metric, the effective DQE (eDQE), was therefore proposed by the authors of that work. This new metric was designed to provide a measure of the signal-to-noise transfer characteristics of a digital imaging system measured under clinical conditions, using a phantom designed for a specific examination type, e.g. chest radiography. More recently, the same group argued that, although the eDQE provides a more clinically realistic measure of DQE, it does not take into account the radiation risk to the patient; hence, the concept of effective dose efficiency (eDE) was proposed 23], which is the effective noise equivalent quanta (eNEQ) normalised to the effective dose. However, a link between eDE and clinical image quality had not been established.Although there is a lack of work demonstrating a link between the physical and clinical image quality, De Crop et al 24] have recently established a correlation between a contrast detail phantom and clinical chest image quality by grading radiographs of embalmed cadavers and comparing the results with those derived from the phantom. However, only three cadavers were used, limiting the statistical significance of their findings, and no pathology of interest, such as lung nodules, was available.In this study, the results of a previous optimisation study performed by our group 25] using computer-simulated postero-anterior (PA) chest images have been used to investigate their correlation with the physical image quality metrics, eDE and CNR, across the diagnostic energy range (50–125 kV). The simulated images of the previous study contained clinically realistic projected anatomy and lung nodules (simulated images were used to avoid the obvious ethical issues associated with experimenting on real patients). We have chosen CNR, as this is a simple measure of image quality that is easy to use and understand, and it is often used to obtain practical measures of object detectability. However, CNR can only really be used to assess large area contrast sensitivity, as it does not include the system MTF or any noise variations with spatial frequency. Conversely, the eDE does include system resolution and noise properties as a function of spatial frequency, so this alternative more complex metric has also been investigated. It should be noted that the physical image quality metrics described in this work are not being optimised in themselves (i.e. optimal values of CNR and eDE for chest imaging are not being investigated) but are being used to predict, using a simple phantom, how the radiographic technique affects the clinical image quality.