Automatic extraction of sulcal curves on the human cortical surfaces using statistical shape model

1. Overview

The cortex of the human brain is a thin convoluted surface comprised of gyri and sulci, which are folds oriented outwards and inwards, respectively. Studying sulcal shapes has applications in neurosurgical planning, understanding normal variability of brains and developmental disorders or effects of aging. Also it is very important in spatial normalization algorithms which use sulci and gyri as features. This project focuses on statistical shape model of cortical sulci and automated method of deforming the shape model to find sulcal curves in target image.

2. Methods

In this project, we build a statistical shape model of major cortical sulci , using the unit sphere as a stereotaxic space. Sulci are projected onto the unit sphere via a near homothetic mapping procedure. Our model captures inter-subject variability of the shape of the sulci, and the mean curvature along the sulcal curves as well. The statistical shape model for the sulcal curves is intended for automatic labeling and spatial normalization of cortical surfaces extracted from magnetic resonance images. The sulcal model on the unit sphere is applied to find and label sulcal curves on the spherical maps of unseen datasets using a fully automated hierarchical deformation scheme. Sulcal curves on the outer cortex are extracted via an inverse spherical mapping procedure.

The model (in the middle), is applied to the spherical map of a new image (on the left), to extrect sulcal curves on it (lower right). Finally, the sulcal curves on the original cortex are obtained via a inverse spherical mapping process (upper right).

3. Results

The average sulcal shape (in red), and the most significant eigen variation (in blue) obtained from 108 training samples.

Model deforms to find the sulcal curves on the cortex. (Click to see the .avi movies. About 4.5MB each.)

Publications

1. Xiaodong Tao, Xiao Han, Maryam E. Rettmann, Jerry L. Prince and Christos Davatzikos, "Statistical Study on Cortical Sulci of Human Brains", Proc. of Information Processing in Medical Imaging (IPMI), p.p. 475-487, Davis, CA, 2001

2. Xiaodong Tao, Jerry L. Prince and Christos Davatzikos, "Using a Statistical Shape Model to Extract Sulcal Curves on the Outer Cortex of the Human Brain", IEEE Trans. Med. Imag., 21(5):513-524, May 2002