This work is about the development of a computational system to allow an automatic and objective evaluation of mammographic equipment performance considering its image quality. The information should be obtained from computational analysis of radiographic images of some phantoms, which should be developed for evaluation of each specific parameter. Thus, the final proposal is to present a "package" tool that will include phantoms for acquirement of the radiographic images and softwares for evaluation of these images, resulting in the automatic determination of quality parameters related to the equipment under examination.Parameters that should be measured include the Modulation Transfer Function (MTF), the Sensitometric Curve and the Noise Power Spectrum (NPS). Thus, this work will allow the evaluation of those three basic physical parameters, recommended for the ICRU (International Commission on Radiation Units and Measurements), for evaluation of a radiological image system: space resolution (by MTF), contrast resolution (by sensitometric curve) e noise (by NPS). In addition, another important parameter, that comes from the combination of these others three previous parameters, also should be considered: the Detective Quantum Efficiency (DQE). The DQE gives an evaluation of image system considering the efficiency in radiation photon detection, and its measurement will provide, in addition to an objective evaluation image quality, also an indication of the relationship between mammographic image quality and the radiation dose absorbed by the patient. Thus, this work will contribute, not only in to mammography quality control but also to the dose reduction during a breast image examination, which represents an important contribution to this research area.
News published in Agência FAPESP Newsletter about the scholarship: