A quality control assessment for digital x-ray radiography through pixel values based signal and noise characteristics

Abstract

Abstract

Quality of x-ray image is a major factor for the correct clinical diagnoses in radiology. A quality control (QC) programme for medical x-ray imaging is essential to preserve quality of the images. Implementation of an effective quality assurance programme to ensure the high quality of images is a responsibility of a medical physicist in the hospital. However lack of relevant, expensive QC tools has become a challenge to maintain the quality of medical x-ray equipment in developing countries such as Sri Lanka. This article presents a simple, low cost QC procedure to measure the quality of a digital x-ray detector and image reader in occasionally through the analysis of signal and noise of digital images. The mean of the pixel value within 5 mm x 5 mm region of interest (ROI) of a uniform digital image is used to estimate signal and standard deviation (SD) of pixel value within the ROI used to estimate noise. This study suggests good baseline tests for QC through the linear relationship of pixel value based signals versus ln [mAs] and SNR versus ln [mAs]. Stable signal, noise and SNR characteristics of an image will indicate stability of the quality of the imaging device. Then present experiment creates base line for signal and noise assessment for future analysis. Likewise signal to noise ratio (SNR) characteristics can be used to compare image performance between two equipment or two set up.