Towards image quality, beam energy and effective dose optimisation in digital thoracic radiography

Abstract

This paper outlines how objective measurements of both image quality, in terms of signal-to-noise ratio, and effective dose may be used as tools to find the optimum kVp range for a digital chest radiography system. Measurements were made with Thoravision, an amorphous selenium-based digital chest X-ray system. The entrance surface dose and the effective dose to an anthropomorphic chest phantom were determined demonstrating how effective dose is related to beam quality. The image quality was measured using detective quantum efficiency, threshold contrast and a radiologist preference trial involving 100 patients. The results show that, despite the fact that the entrance surface dose decreases as the kVp increases, the effective dose, a better measure of the risk, reaches a minimum value between 90 and 110 kVp; however, the image quality decreases as the kVp increases. In this study the optimum kVp for chest radiography, using a selenium-based radiography system, is in the range 90-110 kVp. This is contrary to the 120- to 150-kVp range that is commonly used. Also, this study shows how objective measurements can be used to optimise radiographic technique without prolonged patient trials.

Use of a digitally reconstructed radiograph-based computer simulation for the optimisation of chest radiographic techniques for computed radiography imaging systems

“…Our group has also investigated optimising a CR imaging system for chest radiography with a phantom containing only coarse anatomical detail [18][19][20], but how our conclusions related to the diagnostic quality of the clinical image was undetermined. More recently, several groups have used a computerised voxel phantom in Monte Carlo studies [21][22][23][24][25] in an attempt to model anatomical features, but the resolution of this voxel phantom [26] is relatively coarse (approximately 4 mm long63 mm wide63 mm thick) and is therefore likely to produce images of much lower spatial resolution than a real CR image (typical pixel pitch 0.160.1 mm). Another disadvantage of this Monte Carlo phantom is that it simulates only four tissue types (soft tissue, bone, bone marrow and lung tissue) and air, thereby limiting the contribution of anatomical noise.…”

ROC analysis of storage phosphor radiography and digital selenium radiography at various tube voltages in simulated chest lesions

“…Confirmation of the accurate x-ray tube voltage is also an important quality assurance topic in diagnostic radiography [11]. The ideal tube voltage range for a digital radiography was evaluated in a patient trial for DSR [12] and in a phantom study for DSR [11], [12] and CsI/a-Si [6] but not for the storage phosphor radiography. Hence, we performed phantom study by using an anthropomorphic chest phantom and simulated lesions at various tube voltages for storage phosphor radiography compared with DSR.…”

“…Therefore, chest radiographs in adults are performed at high-voltage techniques (100-150 kV) with an exposure time less than 40 msec according to the American College of Radiology standards. A phantom study of DSR that compared different tube voltages gave first evidence for superior ability to detail detection of lesions on images at lower tube voltage [11], [12]. Furthermore, the diagnostic performance of CsI/a-Si flat-panel detector and computed radiography increased at lower peak voltage with higher Az values and visual grading analysis score, respectively [6], [13].…”

Quality parameters and assessment methods of digital radiography images

“…As a result, image quality can be improved. 25 In CR and DR, Lower kVp techniques are more likely to improve SNR and hence the contrast resolution of image. However, low kVp techniques may increase radiation dose and image blur as a result of time increasing.…”