Scatter and veiling glare estimation based on sampled primary intensity.

Abstract

Scatter and veiling glare are predominant sources of error in videodensitometric iodine quantification. Standard beam stop techniques such as lead strips or an array of lead discs, placed before the patients, have previously been used to measure scatter and veiling glare in digital radiographic images. However, these techniques significantly increase patient x-ray exposure. In order to overcome this limitation, a scatter measurement technique based on sampled primary intensity has been investigated. This technique uses an array of apertures in a lead sheet to sample the primary x-ray intensity. The scatter-glare intensity in these locations is calculated by subtracting the sampled primary intensity from an open field image which contains both primary and scatter-glare. The calculated scatter-glare values can be interpolated or combined with digital filtration to estimate the scatter-glare intensity on a pixel by pixel basis. The technique was evaluated using a Lucite step phantom and an anthropomorphic chest phantom. The average rms percentage errors of scatter and veiling glare estimation using bi-cubic interpolation and digital filtration techniques were 8.02% and 7.53%, respectively. The average rms percentage errors of primary intensity estimation using bi-cubic interpolation and digital filtration techniques were 10.01% and 8.91%, respectively. The x-ray exposure-area product (EAP) from the aperture array was only 4.38% of the EAP from the open field. These results indicate that the scatter-glare intensity can be accurately estimated with minimal x-ray exposure using sampled primary intensity.