Validation of MTF measurement for digital mammography quality control.

The modulation transfer function (MTF) describes the spatial resolution properties of imaging systems. In this work, the accuracy of our implementation of the edge method for calculating the presampled MTF was examined. Synthetic edge images with known MTF were used as gold standards for determining the robustness of the edge method. These images simulated realistic data from clinical digital mammography systems, and contained intrinsic system factors that could affect the MTF accuracy, such as noise, scatter, and flat-field nonuniformities. Our algorithm is not influenced by detector dose variations for MTF accuracy up to 1∕2 the sampling frequency. We investigated several methods for noise reduction, including truncating the supersampled line spread function (LSF), windowing the LSF, applying a local exponential fit to the LSF, and applying a monotonic constraint to the supersampled edge spread function. Only the monotonic constraint did not introduce a systematic error; the other methods could result in MTF underestimation. Overall, our edge method consistently computed MTFs which were in good agreement with the true MTF. The edge method was then applied to images from a commercial storage-phosphor based digital mammography system. The calculated MTF was affected by the size (sides of 2.5, 5, or 10cm) and the composition (lead or tungsten) of the edge device. However, the effects on the MTF were observed only with regard to the low frequency drop (LFD). Scatter nonuniformity was dependent on edge size, and could lead to slight underestimation of LFD. Nevertheless, this negative effect could be minimized by using an edge of 5cm or larger. An edge composed of lead is susceptible to L-fluorescence, which causes overestimation of the LFD. The results of this work are intended to underline the need for clear guidelines if the MTF is to be given a more crucial role in acceptance tests and routine assessment of digital mammography systems: the MTF algorithm and edge object test tool need to be publicly validated.

[1]  Hilde Bosmans,et al.  Practical method for detected quantum efficiency (DQE) assessment of digital mammography systems in the radiological environment , 2002, SPIE Medical Imaging.

[2]  F R Verdun,et al.  Estimation of the noisy component of anatomical backgrounds. , 1999, Medical physics.

[3]  E. Samei,et al.  A method for measuring the presampled MTF of digital radiographic systems using an edge test device. , 1998, Medical physics.

[4]  P. Greer,et al.  Evaluation of an algorithm for the assessment of the MTF using an edge method. , 2000, Medical physics.

[5]  P. Judy,et al.  The line spread function and modulation transfer function of a computed tomographic scanner. , 1976, Medical physics.

[6]  I A Cunningham,et al.  Signal and noise in modulation transfer function determinations using the slit, wire, and edge techniques. , 1992, Medical physics.

[7]  Kenneth A Fetterly,et al.  Performance evaluation of a "dual-side read" dedicated mammography computed radiography system. , 2003, Medical physics.

[8]  Ulrich Neitzel,et al.  Accuracy of a simple method for deriving the presampled modulation transfer function of a digital radiographic system from an edge image. , 2003, Medical physics.

[9]  Andrew D. A. Maidment,et al.  Conditioning data for calculation of the modulation transfer function. , 2003, Medical physics.

[10]  R. Waggener,et al.  Half-value layer and intensity variations as a function of position in the radiation field for film-screen mammography. , 1999, Medical physics.

[11]  Hilde Bosmans,et al.  Use of MTF calculation in global and local resolution assessment in digital mammography , 2003, SPIE Medical Imaging.

[12]  C E Metz,et al.  Transfer function analysis of radiographic imaging systems. , 1979, Physics in medicine and biology.

[13]  Hilde Bosmans,et al.  Practical MTF calculation in digital mammography: a multicenter study , 2004, SPIE Medical Imaging.

[14]  Kunio Doi,et al.  A simple method for determining the modulation transfer function in digital radiography , 1992, IEEE Trans. Medical Imaging.