Monte Carlo simulation of the effect of focal spot size on contrast-detail detectability

Abstract

A contrast-detail experiment was simulated using Monte Carlo methods, to test the hypothesis that quantum limitations lead to an optimum minimum focal spot size below which no further improvement in image quality may be obtained. The simulation included a variable X-ray tube focal spot size, patient equivalent water phantom, X-ray couch, automatic exposure control, anti-scatter grid and indirect digital radiography detector. A number of simplifications were necessary in order to limit the calculation time to 8 days per image. Four images were produced for each focal spot size and these were scored by eight experienced observers. The contrast-detail curves were found to improve monotonically as focal spot size was reduced, with the best images produced by a point source. This contradicts the hypothesis of quantum limitation of focal spot size. We conclude that further work is required on the optimization of focal spot size. To assist with this, a new definition of system detective quantum efficiency is suggested, that includes the focal spot modulation transfer function, but does not include scattered radiation from the patient.