Abstract
Background: Retinotopic mapping using fMRI objectively determines the functional topography of visual cortex. Behavioral visual field mapping (perimetry) is a subjective method because it depends on the patient's judgment. Comparison between retinotopic and perimetric maps is difficult. Our goal was to develop an fMRI-based visual field test that is objective, provides high spatial resolution, and allows comparison of behavioral and brain activation data as well as longitudinal observation of visual performance.
Methods: In 10 normally-sighted subjects, retinotopic mapping (checkerboard stimuli, 8Hz flicker) and fMRI-based perimetry were performed using a 3T Philips Intera MR Scanner (FEEPI, TR=2000ms, 28 slices, 2mm, no gap, resolution 128 × 128). For fMRI-based perimetry, small flickering stimuli (∼2.5 to 12.5 degrees eccentricity, M-scaled) were presented at 48 positions, with two randomly positioned stimuli appearing simultaneously in each trial (2000ms), one in the left and one in the right hemifield. BOLD response, stimulus detection, and reaction times were acquired in five runs (30 repetitions/ position). Data were analyzed with Brain Voyager QX software (GLM, linear correlation maps, ROI analysis) separately for the left and right hemisphere. Retinotopic maps were used to determine the consistency and validity of fMRI-based perimetry results.
Results: Retinotopic maps in healthy subjects were comparable to results reported in the literature. FMRI-based perimetry yielded locally specific activation maps at the locations expected based on retinotopic maps. We found interindividual differences in BOLD amplitudes of activation maps.
Conclusions: fMRI-based perimetry is an objective tool for testing visual field function with a higher spatial resolution than conventional retinotopic mapping and permits point-by-point relation of behavioral parameters (stimulus detection, reaction times) to locally specific activation of early visual cortical areas. Thus, it is also suitable for the assessment of visual cortex function in visually impaired patients.
Funded by VA Grant C-2726 to JFR.