Intelligent System for Multi-Layer Computed Tomography Shots Parameterization

摘要

Accessibility of computer tomography scan devices led to the development of computer-aided diagnosis systems. Their task is to increase the accuracy and speed of diagnosis process. Such systems provide for medics an additional data extracted from computer tomography scans. The set of computer tomography layers of multi-layer shot provides comprehensive information about inner structure of the examining object. Currently medics are just beginning to use the computer-aided diagnosis systems in their daily medical practice. But improvement of computer-aided diagnosis systems is actively developing field of research. Nowadays computer tomography is widespread in medical clinics. Complex of diagnostic procedures is frequently performed with use of contrasting dye. Although, tomography study with contrast dye use has contraindications and is much more expensive for patients. Along with development of tomography shots computer reconstruction methods, processing and analysis of images are improving. Ways of visualization improvement and images reconstruction were well investigated owing to their relative universality, while methods of images segmentation and objects of interest classification are developed to lesser extent as the applied approaches substantially depend on a solving task. Parametrization of the computer tomography shot objects allows extending diagnostic data available to medics and is a necessary step on the way to completely automatic classification of the studying objects on the basis of the received parameters values. In some cases the evaluation of numerical parameters of small anatomical structures is required. For example, if their pathological growth exists or their identification, measurement of volume and carrying out other examining are necessary. This article proposes data processing algorithms and the software as the basis of the intellectual system for parameterization of the small objects occupying a small part of the initial tomographic picture without use of contrast dye. The software, developed according to proposed algorithms, evaluates parameters of small objects and sinuses, which occupy only small part of initial computed tomography layers, with high accuracy without use of dye.