Abstract
The last 15 years have witnessed a renewed and increasing interest in the visual functions of microsaccades. While considerable progress has been made on the resolution of classical debates, one specific issue that has remained controversial is the frequency of monocular microsaccades. Contrary to classical studies, several recent reports found a considerable percentage of microsaccades (up to 40%) to be monocular. These large numbers of monocular events are suspicious, particularly given that it is now clear that microsaccades serve a similar and precise gaze-centering function as larger saccades (Poletti et al, 2013). That is, while it is reasonable that precise alignment of stimuli may lead to monocular microsaccades, it is surprising to find them in experiments in which subjects observe stimuli on flat displays. Since all recent studies were based on video-based eye-trackers, systems that tend to sacrifice precision for simplicity of use, it has been hypothesized that the surge in monocular microsaccades could be the consequence of measurement artifacts. To investigate this issue, here we examined two binocular oculomotor datasets collected with two of the most precise eye-tracking methods available: Dual Purkinje Image (DPI) tracking and eye coils. DPI data were collected with the head immobilized while subjects fixated on points presented on a display. Coil data were collected during normal head-free viewing, while subjects looked at LEDs at different distances. In both datasets, we found virtually no trace of monocular microsaccades. When the parameters of the microsaccade detection algorithm were relaxed, a few monocular events were signaled; however, manual validation revealed that these events were false alarms. Our results do not exclude the existence of monocular microsaccades, but show that these are extremely rare events in standard experiments. The monocular microsaccades reported by recent studies are likely the outcome of both recording artifacts and suboptimal microsaccade detection algorithms.
Meeting abstract presented at VSS 2017