Title :
Region extraction with standard brain atlas for analysis of MRI brain images
Author :
Nakazawa, Y. ; Saito, T.
Author_Institution :
Dept. of Electr. Eng., Kanagawa Univ., Yokohama, Japan
Abstract :
Towards the development of comprehensive analysis of data obtained by various CTs, we develop a region extraction method for recognizing brain anatomies straightforwardly from a given set of multislice MRI brain images of an individual by utilizing a standard brain atlas as the high-level topological information. The region extraction method utilizes a standard anatomical contour model of an intended brain anatomy derived from the standard brain atlas as an initial active contour model, and matches the active contour model to an input MRI brain image by means of energy minimization, where the energy functional is defined with the image intensity function, smoothness constraint of the active contour model, and the standard anatomical contour models. We apply the proposed region extraction method to recognition of the ventricle, and the results demonstrate that the active contour model is attracted to the desirable local-minimal contour and hence the method can achieve satisfactory recognition
Keywords :
biomedical NMR; brain; image recognition; medical image processing; MRI brain images; active contour model; brain anatomies; energy functional; energy minimization; high-level topological information; image intensity function; local-minimal contour; multislice MRI brain images; region extraction; smoothness constraint; standard anatomical contour model; standard anatomical contour models; standard brain atlas; ventricle recognition; Active contours; Anatomy; Brain modeling; Data analysis; Data mining; Image analysis; Image recognition; Information analysis; Magnetic resonance imaging; Standards development;
Conference_Titel :
Image Processing, 1994. Proceedings. ICIP-94., IEEE International Conference
Conference_Location :
Austin, TX
Print_ISBN :
0-8186-6952-7
DOI :
10.1109/ICIP.1994.413341
