Polygonal and Analytical Surfaces in Medical Image Visualization Tasks
2011
Katrina Boločko, Aleksandrs Glazs

The task of medical image processing and analysis is very topical, because the analysis of such images allows physicians to obtain accurate and timely information about the patient and possible diagnosis. One of the most important tasks of medical image processing is the three-dimensional visualization of medical objects. The three-dimensional visualization helps physicians see the exact position of the pathology zone and plan the necessary surgical intervention. The three-dimensional models in medical image processing should meet the following requirements: high quality model fast model construction speed and ability to review the model in real time. The goal of this work is to propose a method of three-dimensional model construction that would meet all the necessary requirements. Two types of three-dimensional models were considered in this work – analytical models and polygonal models. Classical polygonal model construction methods meet two out of three requirements: fast speed and real time mode, but have low quality model (with possible distortions and defects. Classical analytical model construction methods meet one out of three requirements: high quality model, but are not able to work in real time at all. Therefore, since analytical models are not able to work with real time tasks, the possibility to improve the quality of the polygonal model was considered in this paper In this paper, a new polygonal model construction method, which has better model quality, is proposed. The proposed method was compared with an existing analytical model creation method [3]. The analytical model's positive aspects were the following: the high quality of the model and the possibility of a thorough model analysis, as, e.g., exact volume estimation of the analytical model using integrals [6]. However for further research, the proposed polygonal method has been chosen, because it matched all three criteria (high quality, fast speed, real time). The proposed polygonal method was compared with an existing medical image processing software 3D-Doctor [1]. Model creation speed and model quality were used as criteria for the comparison. The experimental results show that the proposed method has a higher model quality, as well as the speed of model construction using the proposed method is at least 10 times faster than applying 3D-Doctor. During the course of work, the following conclusions were made: the analytical model should be used for medical object visualization only as an additional tool, because of the failure to work in real time. However, the analytical model does provide the possibility to calculate the volume of the model precisely using integrals. The proposed polygonal model construction method may be used in real time, and, as indicated by the experimental results, the quality of the model and model construction speed are better, than in the existing 3D-Doctor software.


Keywords
Medical images, image analysis, visualization, pologonal models, analytic models

Boločko, K., Glazs, A. Polygonal and Analytical Surfaces in Medical Image Visualization Tasks. Technologies of Computer Control. Vol.48, 2011, pp.7-13. ISSN 1407-7493.

Publication language
Latvian (lv)
The Scientific Library of the Riga Technical University.
E-mail: uzzinas@rtu.lv; Phone: +371 28399196