Moving objects viewed through stationary eyes appear to move, but during smooth-pursuit eye movements, stationary objects in the field are not perceived as moving. Moreover a moving object tracked with the eyes seems to move, although its retinal image is virtually stationary (Leigh & Zee, 1991). This implies that the brain compares retinal motion and eye position to compute the perceived movement of objects (Lisberger & Movshon, 1999). We measured Blood Oxygenation Level-Dependent signals by means of functional MRI (Siemens-Varian 3T scanner) to investigate brain regions underlying this computation.
Seven subjects, who gave informed consent, viewed a display consisting of high-contrast vertical white stripes, with a central cross that they fixated constantly, on a black background. To maintain attention constant, they had to look for and report occasional small changes in the shape of the cross. Stimulation conditions, each exposed 8 times for 27 s, included: (1) fixation cross stationary, stripes oscillating horizontally back and forth with sinusoidal time-course (2.4 s period; 10 deg amplitude); (2) stripes stationary, cross oscillating, producing the same retinal motion as in (1) (Leigh & Zee, 1991); (3) stripes and cross both stationary; and (4) stripes and cross moving together, producing perceived motion of the entire display with only small residual slip of the retinal image.
Our results suggest that the temporo-occipital cortical area MT+, thought to be homologous to monkey MT & MST (Zeki et al. 1991; Watson et al. 1993), is involved in computing external object motion. MT+ was strongly activated by motion of the stripes when the eye was stationary but showed no obvious activation when identical retinal motion was generated by smooth pursuit eye movements. MT+ was also activated when the eyes moved with the stripes, i.e. without significant retinal motion.
These results, which imply that MT+ receives an extraretinal signal about eye position as well as information about retinal motion, are compatible with a recent report that neurons in monkey MT respond differently to retinal motion caused by stimulus movement and eye movement (Thiele et al. 2002).
This work was supported by the University of Pisa, the Medical Research Council and the Oxford McDonnell Centre for Cognitive Neuroscience.