A Cartesian coordinate system for human cerebral cortex

The most commonly used method for specifying the locations of functional areas in the human cerebral cortex is the coordinate system of Talairach and Tournoux (Co-planar Stereotaxic Altas of The Human Brain (1988) Georg Thieme Verlag, Stuttgart). It was designed to locate subcortical nuclei by reference to an axis joining the anterior and posterior commissures. The coordinate system has difficulties, however, when applied to cortical locations: (1) it can be difficult to locate the posterior commissure (PC); (2) the fundamental axis is short, and errors in specifying the axis lead to large errors at the cortical surface; (3) there is no normalisation for brain size. We sought to rectify these problems with a new coordinate system, the Sydney system, in which the fundamental axis runs in the medial sagittal plane from the anterior edge of the corpus callosum to the posterior end of the parieto-occipital sulcus. Normalisation is achieved by dividing all distances by the length of the fundamental axis. Using functionally important points and anatomical landmarks on cadaveric specimens and magnetic resonance images (MRI), three-dimensional coordinates were measured in both the Talairach and Sydney systems. The Sydney system has the following advantages over the Talairach system: (1) the fundamental axis is more than four times longer and is easier to identify; (2) the Sydney system is more precise, in that it reduces the spread of points across the sample; (3) the normalised coordinates allow locations to be compared across individuals, regardless of brain size. We conclude that for the mapping of cortical areas, the Sydney system is potentially an improvement on Talairach's.