Segmentation and modeling of scattered RTG irradiation on quality of skiagraphy images in clinical conditions

During the clinical RTG examination, the secondary (scattered) irradiation is produced. This unwanted irradiation is spread into surrounding in the examination room. Unfortunately, is such clinical workspace other skiagraphy images are stored as well. When skiagraphy image is irradiated, its surface and structure may be irreversibly deteriorated. This affect has a severe impact on quality and reproducibility of the clinical information: imagined structures shed contrast and spatial features. Therefore, modelling and estimation of the scattered irradiation is, from clinical point of view, substantially important. Moreover, there are not any methods objectively evaluating a noise level caused by the scattered irradiation. We have used the multiregional segmentation model which is able to extract areas which are affected by the scattered irradiation. On a base of this approach we specify a finite histogram band specifying area of the scattered irradiation. In this regard, we have specified four image features which are used for specification of the irradiation level. In our analysis, we experimentally analysed estimated level of the scattered irradiation in a dependence of a distance skiagraphy image from the RTG device.