Computed tomography image reconstruction: Effect of detector width to the image quality

Abstract

Using area integral model is one way of getting the projection matrix of a computed tomography system. Compared to line model, the beam path is assumed to be a strip with certain thickness. The weight of each pixel is the overlap between the strip and the corresponding pixel. The objective of this paper is to elicit the effect of the detector width to the reconstructed image quality. A first generation computed tomography simulation was performed to get the projection data on a 300 by 300 digital phantom. Multiple detector width ranging from 1 mm to 5 mm (0.2 mm interval) were used to get 21 projection data. Each projection data is reconstructed using algebraic reconstruction technique with 30 different projection matrices resulting to 30 images per projection data. These projection matrices were calculated using area integral model with different reconstruction detector width ranging from 0.2 mm to 6 mm. The reconstructed image was compared to the original phantom using structural similarity index. We observed that the structural similarity index of each projection data peaks when the reconstruction detector width matches the detector width used in getting the projection data. However, the peak decreases as the detector width increases.