For the decision the problem of early identification and differential diagnostics of diseases of lungs is offered to use the principles of construction the multiagent of intellectual systems. Work of the first agent is based on segmentation of the raster image of fluorogram. The task solved by the second agent - the textural analysis of the image. The third agent of multiagentsystem has to aggregate the decisions made by agents of the bottom level. The models of decisive modules intended for classification of raster images of the fluorogram, are based on neuronetwork technologies and fuzzy logic of decision-making. Decisive modules are constructed on hybrid technology and allow to unite advantages of trained systems of recognition of images (neural networks are used) and the expert estimates received on the basis of the analysis of experience of the radiologist by formation of decisive rules on the basis of fuzzy logic of decision-making.