Abstract
An algorithm for improved automatic segmentation of gross anatomical structures of the human brain is presented that merges the output of a tissue classification process with gross anatomical region masks, automatically defined by non-linear registration of a given data set with a probabilistic anatomical atlas. Experiments with 20 real MRI volumes demonstrate that the method is reliable, robust and accurate. Manually and automatically defined labels of specific gyri of the frontal lobe are similar, with a Kappa index of 0.657.
Cite
CITATION STYLE
Collins, D. L., Zijdenbos, A. P., Baare, W. F. C., & Evans, A. C. (1999). ANIMAL+INSECT: Improved cortical structure segmentation. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1613, pp. 210–223). Springer Verlag. https://doi.org/10.1007/3-540-48714-x_16
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
