The goal of VISAFIX was to investigate in humans and monkeys how miniature eye movements are controlled when we stare at a complex visuel scene and what could be the consquences of them on the visual processing. <br />
Visual exploration of a complex visual scene is a dynamical process involving rapid eye movements of various amplitudes. Even during the fixation of a limited portion of the scene, microsaccades are produced that share many of the properties of larger amplitude saccades. However, these miniature movements have been considered for a long time as the product of some stochastic activity within the end-path of the oculomotor system. More recently, this explanation has been re-evaluated since it was found that both saccades and microsaccades are movements whose amplitude and direction are controlled by common processes of adjustment.. However, this view did not take into account the visual processes that feed the oculomotor system. The goal of VISAFIX was to better determine those visual signals that guide microsaccades, their neural substrate in the brainstem and their impact on the cortical coding of object location and low-level visual perception. This research has important consequences on our understanding of the nature of visual exploration (active vision) as well as the respective roles of noise and information in controlling behavior.
VISAFIX takes advantage of the strong interdisciplinary staff of a single CNRS team, who is able to tackle these questions in psychophysics (humans and monkeys), electrophysiology of brainstem structures (monkeys) and optical imaging of cortical dynamics (monkeys). A key strategy is to compare fixation eye movements in both humans and monkeys using behavioral and psychophysical tasks. To investigate the role of brainstem nuclei in the generation of microsaccades, we used electrophysiology mapping and reversible inactivation techniques. We also recorded cortical activities in awake, fixating monkeys using voltage-sensitive dye imaging, a technique uniquely mastered in Europe by our research team.
We have shown that the statistical properties of natural scenes do change the pattern of fixation eye movements, through a saliency map. Features locations in such saliency map interfere with the nearly stochastic activity of fixation eye movements to bias direction and amplitude of microsaccades. This result suggests that, rather than the mere image statistics, it is the saliency maps that affect/control microsaccades, similar to large exploratory saccades. These goal-oriented microsaccades are generated by the same brainstem nuclei as the large, exploratory saccades, in particular the Omnipause and Inhibitory burst neurons. We have also showed that the cortical representation of feature location in areas V1 is affected by microsaccades although in a damped way. Such dampening of eye movements induced shits in cortical activities may explain in part perceptual stability during eye movements. Finally, we have investigated how the pattern of eye movements changes with different aspects of static and moving stimuli and re-evaluated their impact of low-level perceptual decisions such as grating detection/discrimination or texture recognition.
The VISAFIX project was conducted within a single research team led by Dr Guillaume Masson. This single partner project was based on the complementary expertise existing within the team and the absence of any other French partners working on this topic. The project started in April 2011 and terminated in October 2014. He was funded by ANR with 456K€ on a total cost of 1.1M€.
VISAFIX has led to 5 research articles in the best international scientific journals of the field (J Neurosci, Vision Research, Nature Neuroscience, J Neurophysiol). 4 other articles are still in preparation or have been submitted within the last couple o
While observing natural scenes, we are making sequences of visual fixations during which our eyes are instable and constantly making small movements: the so-called microsaccades. In the 60’s, an intense scientific debate questioned the nature and the role of such microsaccades. A renewed interest in this question arose only recently because some experimental work casted doubt on the classical conception of vision and oculomotricity coupling. First, fixation instability is the byproduct of a dynamical motor system at the origin of a movements from microsaccades to large saccades. Our recent work published in Science (2009) shows that the superior colliculus is playing a central role in this motor continuum. Second, there is a growing body of evidence showing that fixation instability is influenced by visual and cognitive factors. Noisy and ambiguous stimuli, in particular, are affecting their statistical distribution. This observation underlies the probable link between motor and perceptive fluctuations. Other work suggests that microsaccades may play an important role in perception, but the psychophysical or behavioral framework needs to be established. Lastly, it has been shown that cortical activity dynamics is modulated by these microsaccades. Our previous work (Nature Neuroscience 2000; PNAS, 2002) indeed showed the existence of a manifest visual input during saccades. Physiological work showed that cortical dynamic states are changing as a function of fixation instability. However, these relationships have not been studied at the level of neuronal population and on different visual cortical areas. Moreover, how such motor fluctuations affect cortical representation of visual information, such as position, is still an unresolved question.
VISAFIX is an interdisciplinary project conducted by a single CNRS team: DyVA (INCM). The competences of 5 permanent researchers will be gathered (behavior and psychophysics, neurophysiology and optical imaging on the awake behaving monkey) in order to develop a research program embracing all these questions. What are the neural bases of microsaccades? How do they depend on sensorial processes? Do they interact with perception? How do these microsaccades influence the population activity dynamics and information representation as measured with optical imaging methods? VISAFIX offers a unique way to approach this phenomenon at different levels along the nervous system of human and non-human primates. We propose to take benefit of these small eye movements to inspect how dynamics at cortical-sensorial and subcortical-motor levels interact to continuously control eye, and therefore retinal, position. The intrication between the four different tasks will be forming the interdisciplinary approach that we promote. For instance, perturbing sub-cortical networks in monkeys will elicit more fixation eye movements and we will investigate their consequences at perceptual (motion selection) and visual cortical dynamics (population dynamics in V1, V2 and V4) levels. Conversely, both cortical dynamics and fixation behavior will driven by complex motion stimuli such as motion clouds.
VISAFIX is a remastered version of a research project submitted last year at the “AAP Blanc” that was selected on a complementary list, but could not be financed finally. We took into account the helpful critics of last year reviewers in order to propose an ambitious project though more compact and efficient.
Monsieur Guillaume Masson (CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE - DELEGATION REGIONALE PROVENCE CORSE) – email@example.com
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
INCM - CNRS CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE - DELEGATION REGIONALE PROVENCE CORSE
Help of the ANR 456,144 euros
Beginning and duration of the scientific project: - 36 Months