Detection of speed changes during pursuit eye movements

The human visual system is strikingly insensitive to speed changes attributed to the need to infer visual acceleration, observed during stationary fixation, indirectly by comparing velocities integrated over time. The objective of this study was to test if smooth pursuit eye movements improve the detection of speed changes. This was expected for two reasons: first, pursuit reduces the retinal image slip velocity, leading to smaller Weber fractions for velocity changes; secondly, pursuit provides acceleration-dependent retinal position cues unavailable during stationary fixation such as displacements of the target image away from the fovea due to unexpected changes in target velocity. In a first set of experiments thresholds for just noticeable speed changes were measured in ten healthy human subjects confronted with a horizontally moving target, changing its velocity unpredictably during its ramp-like movement. During stationary fixation, the Weber fraction averaged 0.13 for a starting velocity of the target being 15°/s. Smooth pursuit of the same target significantly reduced the Weber fraction down to 0.08. In a second set of experiments, the discrimination of speed changes was tested in patients (n=10) with pursuit disturbances characterized by increased retinal image slip and unidirectional retinal image displacements. These patients showed a strong perceptual bias to report speed increments and an insensitivity to speed decrements. We argue that this asymmetry is a necessary consequence of a mechanism exploiting retinal position errors for the detection of speed change, confronted with directionally biased errors due to the pursuit impairment. In summary, the detection of speed changes is facilitated by pursuit eye movements but is highly susceptible to pursuit insufficiencies.