Comparative analysis of radiation dose and low contrast detail detectability using routine paediatric chest radiography protocols

Highlights

• 1-Considerable difference in IQ and dose exist for paediatric CXRs between hospitals.

• 2-Differences exist in paediatric CXR protocols different between hospitals.

• 3-Routine paediatric CXRs protocols require further optimisation.

Abstract

Objectives

To compare low contrast detail (LCD) detectability and radiation dose for routine paediatric chest X-ray (CXR) imaging protocols among various hospitals.

Methods

CDRAD 2.0 phantom and medical grade polymethyl methacrylate (PMMA) slabs were used to simulate the chest region of four different paediatric age groups. Radiographic acquisitions were undertaken on 17 X-ray machines located in eight hospitals using their existing CXR protocols. LCD detectability represented by image quality figure inverse (IQF_inv) was measured physically using the CDRAD analyser software. Incident air kerma (IAK) measurements were obtained using a solid-state dosimeter.

Results

The range of IQF_inv, between and within the hospitals, was 1.40–4.44 and 1.52–2.18, respectively for neonates; 0.96–4.73 and 2.33–4.73 for a 1-year old; 0.87–1.81 and 0.98–1.46 for a 5-year old and 0.90–2.39 and 1.27–2.39 for a 10-year old.

The range of IAK, between and within the hospitals, was 8.56–52.62 μGy and 21.79–52.62 μGy, respectively for neonates; 5.44–82.82 μGy and 36.78–82.82 μGy for a 1-year old; 10.97–59.22 μGy and 11.75–52.94 μGy for a 5-year old and 13.97–100.77 μGy and 35.72–100.77 μGy for a 10-year old.

Conclusions

Results show considerable variation, between and within hospitals, in the LCD detectability and IAK. Further radiation dose optimisation for the four paediatric age groups, especially in hospitals /X-ray rooms with low LCD detectability and high IAK, are required.

Introduction

In radiography, the priority is to produce images with a sufficient level of quality to achieve diagnosis. However, the radiation dose to the patient should also be taken into consideration in order to avoid unnecessary radiation exposure [1]. The importance of this is even more apparent for paediatric radiology since children are up to a factor of 10 more radiosensitive than adults [2,3]. By keeping the radiation exposure as low as reasonably practicable (ALARP), both professionals and legislators strive to minimise this risk while maintaining image quality (IQ). The balance between radiation dose and IQ is often referred to as optimisation and is one of the fundamental principles of radiographic practice [4]. Optimisation is not easy to achieve because of differences in imaging systems performance, patient size variations and differences in clinical imaging protocols that are in routine use. Paediatric dose optimisation is even more challenging than that for adults due to the wide variation in paediatric sizes which makes the determination of optimum exposure parameters more difficult [4,5]. The above factors could lead to IQ and radiation dose differences, between and within hospitals, for the same clinical investigations. IQ differences may affect the diagnostic accuracy. Furthermore, differences in radiation dose affect the risk to the patient.

Among the different X-ray examinations, chest radiography (CXR) is the most common examination in children as a result of common respiratory conditions e.g. pneumonia and is also invaluable for resolving a broad range of clinical problems [[6], [7], [8]]. However, the optimisation of IQ and radiation dose for CXR is considered to be difficult because of the wide range of tissue types and medical indications within the chest region. These tissue types vary between high (e.g. mediastinum) and low (e.g. lung) X-ray attenuation characteristics [9,10]. Often there are a number of clinical protocols that are available and suitable for undertaking CXRs [10]. Consequently, questions have arisen as to what extent do standard clinical protocols for undertaking paediatric X-ray examinations vary between imaging systems and departments and what is the resultant impact on IQ and radiation dose? In the United Kingdom (UK), there is a lack of data about the likely differences, between and within hospitals, in terms of radiation dose and IQ for paediatric CXRs. The aim of this study is to compare low contrast detail (LCD) detectability and radiation dose for paediatric CXR protocols across a selection of hospitals using the CDRAD 2.0 phantom.