Poster | 6th Internet World Congress for Biomedical Sciences |
Juan Gutierrez Aguado(1), Isaac Llorens i Eixea(2), Ricardo Ferris Castell(3)
(1)(2)Facultat de Fisica. Universitat de Valencia - Burjassot. Spain
(3)Universitat de Valencia - Burjasot. Spain
Contact address: |
Juan Gutierrez Aguado Facultat de Fisica. Universitat de Valencia C/ Dr. Moliner, 50 Burjassot Valencia 46100 Spain Juan.Gutierrez@uv.es |
[New Technology] |
[Medical Electronics & Engineering] |
[Radiology & Nuclear Medicine] |
In this paper it is presented a greedy strategy to find the final contour of an object in a medical image, starting from an initial roughly outline. It is based on a B-snake, a set of points approximated by a cubic polynomial. This curve so defined moves towards the desired features guided by a functional energy term. Energy has two terms, the first one controls smoothness, and another one related to image gradient. The object contour is where energy has a minimum.
B-snake application on medical images shows a well behavior, leading to robust methods, and being able to obtain precisely contours regardless image noise or bad outline initializations.
This is a well-established approach in features extraction that leads to accurate and complete object detection. Therefore, it allows an isolation of important objects in medical images like radiographies or MRI. An extension of this method would allow a 3D reconstruction of whole organs in a body from a set of computer axial tomographies.
[New Technology] |
[Medical Electronics & Engineering] |
[Radiology & Nuclear Medicine] |