Our lab celebrates its 10th anniversary this month! On this occasion, we looked back at 10 of the most significant papers that we wrote, and how they may pave the way for our future research directions in the lab.
1. Where it all started; the microsaccade as a “genuine” motor act: Hafed, 2011
We study microsaccades. The lab started shortly after our description of how the superior colliculus contributes to microsaccade generation (Hafed et al., 2009). In Hafed, 2011, we explained how this discovery, as well as our earlier discovery of the links between microsaccades and covert visual attention (Hafed and Clark, 2002), set an interesting foundation for the next decade of research on these eye movements. We were now in a position to consider each individual microsaccade as a “genuine” motor output of the brain. This meant that there should be both sensory and motor processes associated with each individual microsaccade, which modulate brain state in predictable manners. A significant fraction of our later work in the decade was directly motivated by this article.
2. A one-to-one link between individual microsaccades and performance changes in covert visual attention tasks: Hafed, 2013
Motivated by Hafed, 2011 and related studies, we directly tested the hypothesis that perception is modulated strongly, but briefly, in association with each individual microsaccade. This idea is well accepted for large saccades, and the intuition that a microsaccade is a “genuine” saccade (Hafed, 2011) supports this hypothesis. In this study, we validated the hypothesis. Then, we argued that in covert attention studies, in which a cue is presented during fixation and perceptual performance is altered (presumably by attention shifts), the performance alteration is in reality time-locked to microsaccade generation. In other words, attentional effects in cueing tasks are time-locked to when individual microsaccades occur. This was the first explicit extension of our findings from a decade earlier, in Hafed and Clark, 2002, that attentional effects were observed only when microsaccades occurred.
3. Neuronal correlates of covert visual attention in superior colliculus and frontal eye field neurons time locked to individual microsaccade occurrence: Chen et al., 2015
Given Hafed, 2013, we were now in a position to explore the neural mechanisms of these surprising observations. We found that even without any attentional tasks at all, all the most classic neuronal correlates of covert visual attention can be observed in two “attention-related” brain areas just when a microsaccade occurs. These results were significant because they demonstrated the first neural evidence for a tight, almost deterministic link between individual microsaccades and perceptual changes previously attributed to covert visual attention. These results also demonstrated remarkable congruence between neurophysiological mechanisms and our earlier perceptual observations.
4. Vision, perception, and attention through the lens of microsaccades: Hafed et al., 2015
This article synthesized our findings on the links between microsaccades and covert visual attention, since our very first demonstration of them (Hafed and Clark, 2002), and also since our renewed insights mentioned in the papers above. These insights were only possible through neurophysiological experiments on microsaccades in primates. In this article, we made a very strong prediction that performance changes in simple, yet classic, attentional cueing tasks may be fully accounted for by the peri-microsaccadic changes in perception that we demonstrated in our lab in its first decade of existence. This has the potential to significantly recast interpretations of a large fraction of cognitive neuroscience experiments with enforced gaze fixation.
5. Microsaccadic rhythmicity and peri-microsaccadic effects can account for all the effects previously attributed to covert visual attention in Posner cueing tasks: Tian et al., 2016
The overall synthesis of all of the ideas above resulted in a highly daring model that we described in this paper: microsaccadic rhythmicity (repetitive microsaccade occurrence; Hafed and Ignashchenkova, 2013) coupled with peri-microsaccadic changes (the papers above) are entirely sufficient to explain the most classic attentional effects in Posner cueing tasks (attentional capture and inhibition of return). This large paper, combining computational modeling and a large amount of experiments (again with remarkable congruence), provided a highly surprising culmination of the ideas laid out at the beginning of the decade in Hafed, 2011. It also laid important foundations for additional insights on why microsaccades occurred at all in our tasks in the first place, which we have later validated with additional experiments.
6. Form and function in the organization of the primate superior colliculus: Hafed and Chen, 2016
The primate superior colliculus has been heavily studied for more than 50 years, but investigations of its topographic organization were surprisingly sparse. Indeed, all existing models of superior colliculus topographic representation are based on only one seminal study from 1972, but the data in this study were very sparse. In the set of experiments described in this paper, we discovered a highly surprising property of superior colliculus that we, and others, had not appreciated before. We found that the superior colliculus represents the upper visual field (above the line of sight) very differently from the lower visual field (below the line of sight). This multi-faceted finding represented remarkable optimization of this brain structure to the statistical properties of natural images to which we may want to make eye movements. Thus, the superior colliculus is optimized to the ecological constraints in which our brain operates. This idea fits so well with renewed efforts in the field to explore structural and functional organization of brain circuits from an ecological perspective.
7. Visual functions of the primate superior colliculus: Chen et al., 2018
Along the same lines as in Hafed and Chen, 2016 above, we further explored the impacts of known statistical regularities in natural images on the properties of superior colliculus visual responses. We found that superior colliculus neurons detect low spatial frequencies (most present in natural images) much better (and earlier) than high spatial frequencies. Moreover, eye movement reaction times directly reflect such dependencies of superior colliculus neurons. This study marked an important continuation of our efforts to understand the full gamut of visual functions of the superior colliculus. What was most amazing to us is that the superior colliculus is actually very sensitive to patterned stimuli (more so than to simple spots of light). The great majority of past work on the primate superior colliculus had focused on using just simple spots of light as targets for saccadic eye movements – the spots were a useful experimental tool to direct eye movements and facilitate the study of the motor functions of the superior colliculus. However, our results showed that the superior colliculus in primates, like in rodents, may be viewed as an early visual area as much as it can be viewed as a late motor structure.
8. Foveal vision by the primate superior colliculus: Chen et al., 2019
What was really missing from all of the above was whether and how foveal visual input is represented in the superior colliculus. For almost fifty years, models of collicular topography emphasized that the superior colliculus serves the purpose of orienting gaze away from the current object (that is, representing the periphery). However, this left many questions unanswered, particularly on how tiny microsaccades can be visually controlled. In this study, we found that the superior colliculus is as foveal a visual structure as the primary visual cortex. This is important because the superior colliculus provides an important alternative visual pathway in the brain, which needs to be further investigated.
9. Memory-guided microsaccades: Willeke et al., 2019
The end of the decade allowed us to integrate our earlier work on microsaccades (described above) with our later work on foveal vision (and other visual functions) in the primate superior colliculus (also described above). We showed that microsaccades can be very easily memory-guided – that is, generated to a remembered blank foveal location without any target. Besides explicitly busting a surprisingly prevalent myth in the field (and public discourse) that microsaccades are “involuntary” eye movements, this observation allowed us to explore neural responses in the superior colliculus for microsaccade generation with and without visual targets. We surprisingly found that some collicular neurons only emit a motor burst if a microsaccade is towards a visual target; the very same neurons are completely silent if the very same motor output is generated without any visual target. In other words, if the microsaccade was directed to a “blank” (as in memory-guided microsaccades), the very same neurons were not contributing to the motor control of the eye movement. This work was the first follow up on our 2009 work (Hafed et al., 2009) describing microsaccade-related motor discharge, and it demonstrated an interesting functional diversity of movement-related neurons, including so-called visually-dependent movement neurons. Until now, we had only known that there are microsaccade-related responses in the superior colliculus. What was missing, and what needs further investigation, is full understanding of the functional diversity of cell types contributing to microsaccade generation.
10. Retinal origins of perceptual saccadic suppression: Idrees et al., 2019
The superior colliculus receives inputs from other brain areas, as well as directly from the retina itself. In this study, we explored the retinal contributions to perceptual effects associated with eye movements. To our utmost surprise, we found that visual processing mechanisms in the retina can account for a surprisingly large part of perceptual properties of peri-saccadic perception. This study motivates a great deal of future research in the next decade!!