Figure 3.

(a) Digital phantom used in simulation studies emulating an extremity imaging scan with i) fat, ii) muscle, iii) bone, iv) line pairs, and v) a uniform disc. For noise evaluations, sample variance was calculated in the disc interior indicated with a circle. Spatial resolution was estimated using the edge response between the disc and fat background. (b) Systems with varying degrees of source and detector blur were simulated to investigate reconstruction performance over a range of scenarios. These scenarios are lettered a-g and permit two experiments where 1) the total blur is constant and the proportion of source and detector blur is varied; and 2) the proportion of source and detector blur is constant and the total blur is varied.