Behavioral and pupillometric correlates of perceptual decision making in mice performing a two alternative forced choice task

Ganea, Dan Alin; Kampa, Björn M. (Thesis advisor); Spehr, Marc (Thesis advisor)

Aachen : RWTH Aachen University (2021)
Dissertation / PhD Thesis

Dissertation, RWTH Aachen University, 2021


The pupil is a complex system reflecting multiple cognitive processes in a state dependent manner. However, it remains largely unclear how various cognitive and perceptual inputs are reflected by the pupil under complex task performance. How do various internal states reflect upon task performance and how do task variables differentially drive pupil dilations. For this scope, in concomitant with pupillometry, mice were trained in the performance of a vibrotactile two-alternative forced choice task (2-AFC) which requires subjects to engage with it (active performance), in order to attain a reward following the correct discrimination of a target stimulus from two simultaneously presented stimuli. Pupillometry conducted in concomitant with task performance revealed a complex pattern of dilations in regard to the various task occurrences and related to task performance. Specifically, I show that pre-stimulus pupil size (baseline pupil size) differentiates between states of disengagement from task performance versus active engagement. In addition, when actively performing the task, post-stimulus, pupillary dilations for correct responses are larger than for error responses with this difference reflecting response confidence. In this state of active engagement our findings indicate that arousal levels, as indexed by baseline pupil size a widely used marker for arousal, do not influence performance. Importantly, by using a delayed version of the 2-AFC task, I show that even though pupillary transients mainly reflect motor output or reward anticipation following the response of the animal, they also reflect animal decision confidence prior to its response, this being an integral part of the magnitude of the dilation. Separating stimulus from response, revealed that pupillary dilations show two distinct epochs. A pre-response phase being a marker of response confidence, continuously reflecting confidence until response time and varying with task difficulty. And a post-response phase, exhibiting a marked dilation relative to the pre-response component that is locked to the response and mainly reflects the motor component of the response or possible reward anticipation. Finally, in a condition of passive performance, when the stimulus has no task relevance anymore, with reward provided automatically, pupillary dilations reflect the occurrence of stimulation and reward attainment but are reduced in magnitude relative to the active performance state. This is explained by shifts of attention from task variables which lower pupillary reactivity and the lack of decisional components which are present during active performance states. These results provide further evidence for how pupillary dilations reflect cognitive processes in a task relevant context, showing that the pupil reflects response confidence and pre-stimulus pupil size encodes attentiveness rather than general arousal. Overall, these results reveal that pupillary dilations reflect both aspects of the perceptual experience of the subject as it relates to the task and its occurrences (stimulation, reward) and also an internal representations of the confidence the subject has in its ensuing decision.Overall, the results detailed in the thesis provide new insights into the cognitive substrate of pupil dilations. Highlighting a new cognitive component reflected by the pupil, that of response confidence. In addition, this work establishes, through its various task versions, a methodical approach that will allow any future attempt when used with neuromeric methods for recording large-scale neuronal network dynamics by using multielectrode or two-photon imaging to track and disentangle loci of confidence representations in the brain.


  • Department of Biology [160000]
  • Department of Molecular and Systemic Neurophysiology [162320]