A comparison of structurally and functionally defined human primary visual cortex

Early visual areas can be defined using fMRI on the basis of their retinotopic organisation. Recently, very high-resolution images of the human brain in vivo have identified areas of myelination within the grey matter, corresponding to the striate cortex (Barbier et al. 2002; Clare et al. 2002). This myelination has traditionally been understood to correspond to the human primary visual cortex (V1). To test this correspondence, we compared the location of visually identified striate in high resolution images with the location of functionally defined V1. For imaging the myeloarchitecture, a magnetisation prepared 3D FLASH sequence was used as described in Clare et al. (2002). The resulting images had a resolution of 0.3×0.3×1.5 mm. Functional MRI was performed at a lower resolution of 3×3×1.5 mm using single shot EPI at the same 16 slice locations as the structural scan. Retinotopy data were collected using expanding ring and rotating wedge stimuli. The data were transformed onto a segmented (mrGray) and flattened (mrFlatMesh) T1-weighted scan (1×1×1mm). V1 was defined by locating the upper and lower field V1/V2 borders from the rotating wedge phase map. From the high-resolution myeloarchitecture images, striate cortex was conservatively determined as those regions where a stripe was identified within the grey matter. These observer drawn maps of striate cortex were then transformed into flattened space to allow comparison with the functional data. A good level of correspondence was found between the striate cortex determined in the structural MRI and V1 determined by fMRI. While the striate cortex was not identified as a continuous band, it is hoped that more striate will be revealed by using multiple slice orientations in the same subjects. In the future these very high-resolution structural images will offer the opportunity to combine the study of myeloarchitecture with functional architecture in the living human cortex.