Purpose: To test the sensitivity of automatic methods for evaluating CDMAM test images with respect to noise.
Materials and methods: CDMAM test images were analyzed with two computer programs. The images were made with different tube loads [mAs]. The other exposure conditions remained constant. They were analyzed with the CDCOM program, which is offered by the EUREF as a free download, and with the CDMAM Image Checker (CDIC), which was developed by the authors.
Results: The determination of the sensitivity in one image always delivers the same result when the same type of computer program is used. This means that the precision of both programs is sufficient. The dose sensitivity of CDIC is two times higher than the sensitivity of CDCOM. However, the required entrance dose (ESAK) for a faultless evaluation with the CDIC program is in the range of 10 mGy. The nominal sensitivity values for the CDCOM program attain a higher level. Differences in dose of more than 5 % should be detectable by both programs.
Conclusion: Methods that dispense with visual inspections to determine the performance of X-ray units for mammography can be applied in the acceptance test or the yearly constancy tests according to the German X-ray directive ( section sign 16). The CDCOM program cannot be characterized fully because the data is not complete. Finally the detection methods are not clear. Therefore, the CDCOM program can be called a black box method, while the CDIC has to be called an open source method (general public license).