Picture grammars in classification and semantic interpretation of 3D coronary vessels visualisations

Abstract

The work presents the new opportunity for making semantic descriptions and analysis of medical structures, especially coronary vessels CT spatial reconstructions, with the use of AI graph-based linguistic formalisms. In the paper there will be discussed the manners of applying methods of computational intelligence to the development of a syntactic semantic description of spatial visualisations of the heart’s coronary vessels. Such descriptions may be used for both smart ordering of images while archiving them and for their semantic searches in medical multimedia databases. Presented methodology of analysis can furthermore be used for attaining other goals related performance of computer-assisted semantic interpretation of selected elements and/or the entire 3D structure of the coronary vascular tree. These goals are achieved through the use of graph-based image formalisms based on IE graphs generating grammars that allow discovering and automatic semantic interpretation of irregularities visualised on the images obtained during diagnostic examinations of the heart muscle. The basis for the construction of 3D reconstructions of biological objects used in this work are visualisations obtained from helical CT scans, yet the method itself may be applied also for other methods of medical 3D images acquisition. The obtained semantic information makes it possible to make a description of the structure focused on the semantics of various morphological forms of the visualised vessels from the point of view of the operation of coronary circulation and the blood supply of the heart muscle. Thanks to these, the analysis conducted allows fast and — to a great degree — automated interpretation of the semantics of various morphological changes in the coronary vascular tree, and especially makes it possible to detect these stenoses in the lumen of the vessels that can cause critical decrease in blood supply to extensive or especially important fragments of the heart muscle.

Semantic Modelling of Coronary Vessel Structures in Computer Aided Detection of Pathological Changes

“…The obtained results show that graph languages for describing shape features can be effectively used to describe 3D reconstructions of coronary vessels and also to formulate semantic meaning descriptions of lesions found in these reconstructions [14], [15]. Such formalisms, due to their significant descriptive power (characteristic especially for graph grammars) can create models of both examined vessels whose morphology shows no lesions and those with visible lesions bearing witness to early or more advanced stages of the ischemic heart disease.…”

CAD system for automatic analysis of CT perfusion maps

“…The additional stage of reasoning algorithm will determinate relation between lesions visible on CT/MR images and perfusion map giving more accurate progno− sis for lesion evolution, l including implementation of more image registration al− gorithm (i.e. Thirion -demon algorithm, fluid registra− tion [32]), l creating more precise brain atlas templates from avera− ged CT images acquired from large group of patience, l adaptation of the algorithm for dynamic MR perfusion maps, l enlarging the knowledge base of the system in order to state more precious diagnoses, include the bloody sup− ply territories atlas (BSTs) [33] that will provide more information for description and reasoning process [34,35]. …”

Semantic Techniques of Image Retrieval – The Example of a Structural Analysis of Coronary Arteries

“…The presented methodology of automatically creating semantic descriptions of images stored in multi-media medical databases is based on methods of semantically interpreting coronary arteries, successfully used to describe and identify lesions in coronary vascularisation images as part of previous studies by the authors [6][7][8][9]. What is important in creating sequences describing images from a database is a method of effectively transforming the image information contained in these images (which is easily perceived by a human) to a machine format which supports the intelligent, semantic selection of a specific case (easily assimilated by a computer).…”

“…One possible method consists in the proposed grammar formalisms for the structural analysis of images, in which the analysed image is treated as a hierarchical structure composed of so-called picture primitives. In their previous publications, the authors proposed using graph grammars to describe and model the spatial relations of coronary vascularisation reconstructions [6][7][8][9]. Grammars of this type generate a formal language in the form of graphs which can model the images considered here, and then the graphs obtained can be represented in the form of their characteristic descriptions.…”