An Improved Segmentation of Brain Tumor, Edema and Necrosis

An Improved Segmentation of Brain Tumor, Edema and Necrosis

Paper in proceedings (conference paper)

Image processing in biomedical applications is strongly developing issue. Many methods and approaches for image preprocessing, segmentation and visualization were described. This paper deals with image segmentation, concretely brain tumor segmentation. The main problem in medical practice is to recognize the type of brain or other tumor. There are many methods for tumor classification and one of them is perfusion imaging/analysis. Perfusion images are of very low contrast and they are devaluated by noise. The main idea is to identify the level of perfusion of contrast agent transported into the pathological tissue. The level of perfusion may decide on the type of tumor. The perfusion has to be monitored in tumor region, edema around the tumor region and in the interface between brain tumor and edema. The goal described in this paper is to propose a segmentation method to recognize brain tumor, edema and necrosis in structural magnetic resonance images (T1, T2) and create a binary mask that enables measurement in perfusion weighted images.

Image processing in biomedical applications is strongly developing issue. Many methods and approaches for image preprocessing, segmentation and visualization were described. This paper deals with image segmentation, concretely brain tumor segmentation. The main problem in medical practice is to recognize the type of brain or other tumor. There are many methods for tumor classification and one of them is perfusion imaging/analysis. Perfusion images are of very low contrast and they are devaluated by noise. The main idea is to identify the level of perfusion of contrast agent transported into the pathological tissue. The level of perfusion may decide on the type of tumor. The perfusion has to be monitored in tumor region, edema around the tumor region and in the interface between brain tumor and edema. The goal described in this paper is to propose a segmentation method to recognize brain tumor, edema and necrosis in structural magnetic resonance images (T1, T2) and create a binary mask that enables measurement in perfusion weighted images.

image segmentation, image classification, brain tumor

image segmentation, image classification, brain tumor

MIKULKA, J.; GESCHEIDTOVÁ, E.

2014

25.03.2013

978-1-934142-24-0

Proceedings of PIERS 2013 in Taipei

1559-9450

Progress In Electromagnetics

United States of America

25

28

4