TL;DR
This paper demonstrates that microsaccades, tiny involuntary-like eye movements, can be voluntarily controlled by humans and monkeys based on memory and abstract instructions, not just visual cues. This finding challenges the long-held view of microsaccades as purely reflexive and reveals a more nuanced role in cognitive processes and neural circuitry.
What Problem Does This Paper Address?
For decades, microsaccades have been largely characterized as involuntary, spontaneous eye movements that occur during fixation. While some research has hinted at their potential for voluntary control, a definitive demonstration has been lacking. This paper aims to bridge this gap by providing robust behavioral and neural evidence that microsaccades can be deliberately generated, even in the absence of visual guidance, by leveraging memory and abstract task instructions. This is crucial because microsaccades are increasingly understood to play significant roles in visual perception, attention, and eye movement control, making their voluntary nature a key piece of the puzzle.
Methods
The researchers adapted a "memory-guided saccade" task for microsaccades. In this task, human and monkey participants were shown a brief visual stimulus (a flash) and then had to remember its location during a "memory interval." After this interval, a "go signal" (disappearance of a fixation spot) instructed them to make an eye movement towards the remembered location. Crucially, the visual stimulus was no longer present, meaning the movement had to be guided by memory alone.
Key methodological aspects include:
- Behavioral Tasks:
- Memory-guided microsaccade task: Participants (humans and monkeys) generated eye movements towards a remembered stimulus location after a delay, without visual guidance.
- Corrective, visually-guided microsaccades: These were measured after the remembered target reappeared, serving as a baseline for uninstructed, visually guided movements.
- Control tasks: Human participants also performed tasks involving button presses or mouse pointing to indicate remembered target locations, allowing researchers to disentangle eye movement control from spatial working memory itself.
- Neural Recordings: In monkeys, electrophysiological recordings were made from neurons in the superior colliculus (SC), a midbrain area known to be involved in eye movement control. This allowed researchers to examine neural activity associated with memory-guided versus visually-guided microsaccades.
- Data Analysis: Rigorous statistical analyses were employed to compare reaction times, success rates, directional accuracy, and amplitude scaling between memory-guided and visually-guided movements. Specialized tools for circular statistics were used for directional data.
Datasets
The study utilized data collected from:
- Three adult male rhesus macaque monkeys: Data included both behavioral eye movement recordings and neural recordings from the superior colliculus.
- Seven human subjects: Behavioral data was collected from these participants.
No pre-existing datasets were used for training or evaluation; all data was collected specifically for this study.
Key Insights and Ideas
- Microsaccades are not just involuntary: The most striking insight is that microsaccades can be voluntarily triggered based on memory and abstract instructions, directly challenging the prevailing view of them as purely reflexive.
- Memory-guided microsaccades are accurate: These memory-guided movements were not only timely but also spatially and temporally accurate, mirroring the precision of visually-guided movements.
- Superior Colliculus involvement: Neural recordings revealed specific neurons in the superior colliculus that fired exclusively for memory-guided microsaccades, but not for similar-sized visually-guided ones, pointing to distinct neural circuitry for voluntary, memory-driven eye movements.
- Foveal spatial working memory expansion: An interesting behavioral finding was an "overshoot" in movement amplitude for very small target eccentricities in memory-guided tasks, which was also observed in non-eye-movement tasks. This suggests that spatial working memory itself might expand or distort the representation of foveal space.
Key Results
- Behavioral Evidence:
- Monkeys and humans successfully generated memory-guided microsaccades with high success rates (>80%) and reaction times comparable to larger saccades.
- These movements were directionally accurate, with errors significantly smaller than those of uninstructed, corrective, visually-guided microsaccades.
- Microsaccade amplitudes scaled with remembered target eccentricity, though with a slight overshoot for very small eccentricities.
- Control tasks confirmed that the overshoot was related to spatial working memory representation, not solely to eye movement execution.
- Neural Evidence:
- Superior colliculus neurons showed movement-related activity bursts for memory-guided microsaccades.
- A significant subset of these neurons (22%) exhibited activity exclusively for memory-guided microsaccades, not for visually-guided ones of similar size. This "ER-Neuron" category provides strong neural evidence for distinct voluntary control mechanisms.
- Other SC neuron types (visual, visual-motor, motor, and visually-dependent) were also identified, showing a spectrum of responses related to visual input and movement execution.
Limitations and Future Work
- Overshoot in Memory-Guided Movements: While the authors attribute the overshoot to spatial working memory representation, it still represents a deviation from perfect accuracy for very small movements. Future work could further investigate the precise mechanisms of this foveal expansion.
- ER-Neuron Mechanisms: The specific neural mechanisms underlying the exclusive firing of ER-neurons for memory-guided microsaccades warrant further investigation. Understanding how these neurons are recruited and modulated could shed light on voluntary control.
- Cortical vs. Subcortical Roles: The study highlights subcortical (SC) involvement in working memory and voluntary control. Future research could explore the interplay between these subcortical circuits and higher cortical areas in generating these movements.
- Generalizability: While the study demonstrates voluntary control in a specific memory-guided task, further research could explore the extent to which this voluntary control applies to other types of microsaccades and cognitive tasks.
- Awareness: The authors acknowledge that a lack of awareness might contribute to the perception of microsaccades as involuntary. Future studies could directly investigate the relationship between awareness and voluntary microsaccade generation.
Summary updated