1 July 2003 Method to segment the brain automatically applied to a magnetic resonance imaging sequence
Author Affiliations +
Optical Engineering, 42(7), (2003). doi:10.1117/1.1580832
The purpose of this preliminary study is to segment the brain on magnetic resonance (MR) images by using a robust and fully automatic method. A 2-D T2-weighted MR sequence is used as a reference for imaging the brain. Our segmentation technique could segment the brain in each slice taken from different MR scanners using different image resolutions and MR sequences. This method is based on an integrated approach that uses image processing techniques and a priori knowledge. The originality of our work is to integrate the 3-D information of the sequence in the geometrical approach, to adapt this method for other MR imaging modalities, and to use only one MR sequence to segment the whole brain. This is a multistage process requiring the background noise to be removed, and the head mask is then obtained thanks to the combination of a recursive filling technique and geometrical properties of the brain. We describe the main features of the method, and give results for some brain studies.
Vianney Le Thuc, Abdelmalik Taleb-Ahmed, Xavier Leclerc, T. Saint Michel, "Method to segment the brain automatically applied to a magnetic resonance imaging sequence," Optical Engineering 42(7), (1 July 2003). http://dx.doi.org/10.1117/1.1580832


Image segmentation

Magnetic resonance imaging



Expectation maximization algorithms



Stereovision-based close-up dimensional inspection
Proceedings of SPIE (February 24 2005)
Pattern geomorphologic analysis for soil erosion study
Proceedings of SPIE (December 19 2008)
Road extraction by point-wise Gaussian models
Proceedings of SPIE (September 23 2003)

Back to Top