Miriam Spering

Adaptive behaviour in a highly dynamic and complex environment necessitates the ability to transform sensory input into motor output at different timescales. The goal of this dissertation was to probe the sensorimotor processes that underlie precise movement control in rich sensory environments. In four experimental studies, I studied different types of tracking eye movements in response to stimuli that mirror some of the complexities found in the natural environment (complex pattern motion, target acceleration, and audiovisual signals). In the first study, I investigated the dynamic motion integration processes that give rise to the coherent perception of 2D pattern motion. Comparing eye movements to observers’ perceptual experience revealed shared motion integration mechanisms that drive continuous tracking and motion perception (Chapter 2). Accurate sensing of complex motion signals is a fundamental ability that allows humans to interact with objects moving along complex trajectories. In Chapter 3, I investigated how observers track and manually intercept accelerating targets. Results showed that observers continuously monitor the changing target speed but fail to account for acceleration to predict and intercept targets. Natural interaction with moving visual objects typically involves multiple senses. In two studies, I demonstrated profound and rapid effects of audiovisual integration on goal-directed actions. I found that auditory and visual signals were weighted according to their uncertainty to guide interceptive movements (Chapter 4). Finally, presenting audiovisual distractors during continuous visual tracking revealed multisensory response enhancement of reflexive movement inhibition at short latencies of ~100 milliseconds (Chapter 5). These findings were obtained using eye movements as a continuous indicator of human sensorimotor integration, highlighting the use of eye movements as a model system to study adaptive sensorimotor behaviour in dynamic environments. I propose a model that describes how eye movements at different timescales are affected by the dynamic processing of complex visual and audiovisual signals. The present findings can be used to inform training protocols for interceptive sports athletes and to assess sensorimotor deficits in clinical populations.

View record

We live in a dynamic visual environment, which requires perceiving moving objects around us and acting accordingly. However, we do not yet fully understand how visual information informs perception and motor actions. This dissertation examines the perception-action link by testing how motion and expectation signals are processed for perception and for eye movements as an example of human motor action. I focus on two types of human eye movements triggered by distinct brain mechanisms: ocular torsion, the eyes’ rotation about the line of sight triggered by rotational motion, and smooth pursuit, the eyes’ continuous tracking of translational motion. Torsion is mostly controlled by subcortical brain areas, but might share early-stage cortical processing of sensory signals with perception. In contrast, smooth pursuit is controlled by subcortical and cortical areas and might therefore be more closely linked to perception, sharing both motion and expectation signal processing with perception. To test the torsion-perception link, I utilized a perceptual illusion induced by visual rotational motion. Results show that torsional velocity correlates with the perceptual illusion, potentially suggesting shared motion processing (Chapter 2.1). However, anticipatory torsion can only be elicited by trial repetition, but not by cognitive cues that induce expectation (Chapter 2.2). These results show that similar visual motion signals might drive reflexive torsion and perception. Expectation signals appear to be less effective in driving torsion. Probing the pursuit-perception link, I found dissociations between how each system processes motion and expectation signals. When integrating diverse motion signals across space, pursuit was biased to the average motion direction, whereas perception showed no consistent bias (Chapter 3). When investigating the role of expectation, I found that anticipatory pursuit followed the expected direction, whereas perception was biased in the opposite direction (Chapter 4). Overall, this dissertation reveals that perception and eye movements likely share early-stage motion processing, even for reflexive eye movements such as torsion. But perception and eye movements differ in how they utilize higher-level motion or expectation signals. The dissociations might indicate how each system optimally meets different functional demands: Perception relies on object segregation, whereas eye movements rely on signal integration.

View record

Natural tasks, such as catching a ball, involve the decision whether, when, and where to act. This dissertation examines the relationship between eye and hand movements during goal-directed manual interceptions that require rapid sensorimotor decisions. Human observers viewed and predicted the motion path of a briefly presented moving target and intercepted it at its assumed end position. Observers naturally tracked the moving target to guide interceptive hand movements. To probe the tight eye-hand link, I investigated the effect of perceptual-motor training on eye and hand movement quality (Chapter 2). Results indicate a mutual benefit of training eye and hand movements concurrently. Eye movement training alone was not sufficientto improve hand movement accuracy. However, training that required an active sensorimotor decision (eye or hand interception) enhanced eye movement quality.Next, I tested the role of eye movements during go/no-go decisions. Observers predicted whether targets passed through (go required) or missed (no-go required) a strike box. Observers' eye movements differentiated between decision outcome (go vs. no-go) on a trial-by-trial basis with an overall accuracy of 76% (Chapter 3). Moreover, I found that different eye movement phases were linked to a two-stage decision process. Whereas eye velocity during pursuit initiation corresponded to go/no-go decision accuracy, pursuit maintenance was related to successful interception timing (Chapter 4).Finally, I investigated the role of movement constraints on decision accuracy by manipulating response modality (button press vs. interceptive hand movement) and eye movements (free viewing vs. fixation; Chapter 5). Decision formation occurred earlier but less accurately when an interceptive hand movement had to be planned and executed. Eye movements (compared to fixation) enhanced decision accuracy regardless of response modality. These results indicate that perceptual decision formation occurs dynamically, relying on the continuous updating of sensory information until an action is required.In sum, this dissertation provides evidence that eye movements are directly related to neural signatures of perceptual decision making. Furthermore, eye and hand movements show interdependencies during visual predictions and manual interception. This work highlights the potential of studying eye movements as continuous readouts of ongoing sensorimotor and cognitive processes during natural tasks.

View record