Full text
| |
| Author(s): |
Leandro Alves Neves
Total Authors: 1
|
| Document type: | Doctoral Thesis |
| Press: | São Carlos. |
| Institution: | Universidade de São Paulo (USP). Escola de Engenharia de São Carlos (EESC/SBD) |
| Defense date: | 2005-03-28 |
| Examining board members: |
Annie France Frere Slaets;
Luciano da Fontoura Costa;
Fernao Stella de Rodrigues Germano;
Paulo Mazzoncini de Azevedo Marques;
Clovis Simao Trad
|
| Advisor: | Annie France Frere Slaets |
| Abstract | |
The efficiency of systems for radiographic images acquisition, as well as for computer aided diagnosis (CAD) is verified currently by means of phantoms or images selected by the researcher, damaging the reliability of the evaluation. In this workan algorithm, which was implemented in C-language, is proposed to model the thoracic and pulmonary structures. This algorithm is flexible enough to represent anatomical variations and even some diseases, being capable of simulating their radiographic images. The structures, represented by cylinders and spheres, are built by the overlapping layers method, where a layer can be understood as a slice of the three-dimensional object. The thoracic box is modeled using control points interpolated by a B-Splines function. The dimensions of airways and arteries branches are calculated based on the flow distribution from parents to sibling branches, as well as on the patients\' anatomical characteristics. A pulmonary abscess is simulated by interrupting the growth of the branches around a region and considering a specific radiological density for this area. The inflammatory process is represented in the interstitial patterns by an increase in the thickness of the cylinders walls which simulate the bronchioles. The results obtained, evaluated by three specialists, present characteristics compatible with real anatomies and simulate adequately different pulmonary diseases. The algorithm also permitted to vary the radiological contrasts and the anatomical particularities, becoming the most complete model among those found in the specialized literature. This versatility and the degree of representation allow the verification of the influence of radiological parameters, morphometric peculiarities and stage of the diseases on the quality of the images, as well as on the performance of the CAD (AU) | |