Quantitative MRI: a reliable protocol for measurement of cerebral gyrification using stereology

Abstract

Stereology is a sampling technique used to generate mathematically unbiased estimates of geometric properties of three-dimensional structures based on two-dimensional slices of the object. Using this method, volume and surface area estimates may be generated in a robust, reliable and time-efficient manner based on magnetic resonance (MR) image data. We present a method of determining the isoperimetric ratio, which may be used as an index of gyrification of the human cerebrum, using stereological volume and surface area estimates applied to in vivo MR image data. This parameter may be of use in identifying subtle cortical developmental malformations in diseases such as epilepsy and schizophrenia. We also conduct a series of inter- and intrarater repeatability tests based on 20 normal brains, which demonstrated the reliability and robustness of stereological techniques.