Determination of the modulation transfer function using the edge method: Influence of scattered radiation

The edge method for measuring the modulation transfer function (MTF) has recently gained popularity due to its simplicity and appropriateness particularly for digital imaging systems. Often edge test devices made of rather thin metal sheets are used, which are semitransparent to x rays and may generate scattered radiation. The effect of this scattered radiation on the determined MTF was investigated both theoretically (assuming an ideal detector) and experimentally using a CsI-based digital detector. It was found that the MTF increases due to the scattered radiation for all spatial frequencies larger than 0 mm(-1). The theoretical model developed in this study predicts that the maximum error compared to the true detector MTF is given by S/A, where A is the attenuated fraction and S is the scattered fraction reaching the detector, relative to the incident radiation. Theoretical and experimental results are in good agreement for radiation qualities corresponding to general radiography (RQA3, RQA5, and RQA7), whereas for chest beam quality (RQA9) the experimentally observed MTF error is larger than predicted by the simple model, possibly because the energy response of the CsI-based detector differs from that of an ideal one. The theoretical MTF error reaches a value of 18% for a 0.25 mm thick lead edge of RQA9. Since the MTF enters squared into the determination of the detective quantum efficiency (DQE), an error of at least 36% in DQE may result when using this edge test device. In conclusion, the use of fully absorbing edge material is advised for MTF determination with the edge method.