Abstract

In this research, three-dimensional images of brain tumors have been constructed from a series of Magnetic Resonance Images (MRIs) for successive T1-weighted slices, using a simple automatic technology. The MRIs are pre-processed through bilateral filtering to reduce noise and retaining the edges between brain tissues. Spatial enhancement technique is adopted to transform image gray levels by highlighting the band density of the region of interest (tumor) and expanding them into full range. Sobel edge detection is used by morphological operations to extract the tumor area of the image of each slice. The contour of the tumor area in each segmented slice is used to create 3D of the tumor. The relative size of the tumor then measured depending on brain tissue after stripping the skull. The result showed that the adopted technique can efficiently detect and extract the brain tumor from the consecutive slices which can be used to produce 3-D image of the tumor and to identify the tumor's location inside the skull.

Keywords

The article was added to IASJ on 2014-09-01

377 Total full text downloads since the date of addition