A fast deformable region model for brain tumor boundary extraction

We present a modified deformable region model for extraction of a brain tumor boundary in 2D MR images. The deformable region model tolerates a rough initial plan when compared with the active contour model. However, it is time consuming to compute and compare the gray level distribution of the object and all its boundary points. Using a point sampling technique, the number of boundary point processed is greatly reduced. Performance of our modified deformable region model is evaluated on a MR image. The modified model is fast while similar results are obtained.