Abstract
Purpose: fMRI-based retinotopic mapping is a fundamental technique for the investigation of the human visual cortex. We acquired retinotopic mapping data at a magnetic field strength of 7 Tesla and tested whether high resolution retinotopic maps of the early visual areas can be obtained. Methods: Two subjects with normal vision underwent T2* MRI scanning of the occipital lobe during visual stimulation (7 Tesla, Magnetom, Siemens). We stimulated the retina with a portion of a phase reversing (6 reversals per second) circular checkerboard stimulus (8 deg radius; 62 cd/m2 mean luminance; 95% contrast) that stepped through polar angles or eccentricities of the visual field in accordance with established retinotopic mapping techniques (Engel et al., 1997). Seven 36 s cycles of the stimulus were presented for each condition. Every 2 seconds, 25 slices parallel to the calcarine sulcus were acquired. Three voxel sizes were tested: 2.5, 1.4, and 1.1 mm isotropic voxels. After motion and distortion correction, fMRI signals were projected to the flattened representation of T1 weighted images acquired at 3 Tesla (Trio, Siemens), Fourier analysed, and correlated with the stimulus fundamental frequency. Results: Retinotopic mapping yielded phase maps that allowed for the identification of V1 and other early visual areas for each voxel size tested. Mean correlation coefficients were determined for the representation of the horizontal meridian in V1 and found to increase from 0.57±0.09 to 0.92±0.03 with increasing voxel size. Conclusions: Retinotopic mapping benefits from the use of a field strength of 7 Tesla: The visual cortex can be sampled at a high resolution aiding a detailed description of the visual field representations. Alternatively, with 2.5 mm isotropic voxels, mapping beyond the early visual areas is facilitated, as a greater expanse of occipital and parietal cortex responds significantly than for the same measurement duration at lower field strengths.