A bifield stimulation method for rapidly obtaining retinotopic maps in human occipital cortex using functional MRI was compared to conventional unifield stimulation. While maintaining central fixation, each participant viewed the conventional display, consisting of a single rotating checkerboard wedge and, in a separate run, the bifield display, consisting of two symmetrically placed rotating checkerboard wedges (a "propeller" configuration). Both stimulus configurations used wedges with 30 degree polar angle width, 6.8 degrees visual angle extension from fixation, and 8.3 Hz contrast polarity reversal rate. Retinotopic maps in each condition were projected onto a distortion corrected computationally flattened cortical surface representation obtained from a high-resolution structural MRI. An automated procedure to localize borders between early visual areas revealed, as expected, that map precision increased with duration of data acquisition for both conditions. Bifield stimulation required 40% less time to yield maps with similar precision to those obtained using conventional unifield stimulation.
Copyright 2002 Wiley-Liss, Inc.