Segmentação do ventrículo esquerdo utilizando técnicas de processamento de imagens

AUTOR(ES)
DATA DE PUBLICAÇÃO

2008

RESUMO

Image processing techniques are largely used with various areas of science, including medicine. These techniques are used mainly as a way to improve medical exams, making then faster and reducing the observers influence on their execution. This work presents a method that combines various image processing techniques and has as its main objective to extract the left ventricle internal cavity border on 2-D echocardiography images, using both short-axis and long-axis images. First the images were cut. Then the program applies the high boost filtering, thresholding, elevation filtering and LoG. Then watershed is applied, followed by radial search, interpolation on the candidate points and border attenuation. The program also uses an extra step to obtain the borders on image sequences. This extra step consists on using information from a previous obtained segmentation. The results obtained from the presented method are compared with the manual segmentation made by a specialist medic. Three different error measures are used, and results from these tests shows that the semi-automatic method presented achieved good results.

ASSUNTO(S)

engenharia eletrica busca radial segmentação processamento de imagens médicas filtros morfológicos ecocardiografia

Documentos Relacionados