Objective: Eye movements provide important insights into the pathophysiology of neurological disorders. Here two paradigms were devised to quantify effects of motivation on motor control and decision making.

Background: The speed-accuracy trade-off can be overcome by reward, invigorating movements and improving response precision. In PD, characterised by a dopaminergic deficit, reduced reward sensitivity suggests a higher cost for controlling intrinsic neuronal noise(1). We asked whether motivational control costs extend to two tasks, limited by decision- and memory noise. Task1: Choosing between a larger number of options, responses become slower (Hick’s law). If there is a cost for improving the effective signal-to-noise-ratio (SNR), motivation should result in reaction time (RT) scaling(2). Task2: Sequences held in working memory; Latencies increase and spatial precision falls with the length of the sequence keeping with lower SNR(3). If motivation reduces noise in spatial memory, an improvement should be observed with longer sequences of saccades, or longer memory retention intervals.

Methods: 22 healthy subjects performed both saccadic tasks recorded by an infrared eye tracker. Both involved three incentive levels (-50p, 0p, 50p). Reward/loss was calculated depending on the performance compared to their running average. Task1: Participants fixated centrally while either 2, 4 or 8 placeholders indicated possible target locations. A central arrow indicated the target location for the eye movement. Task2: While participants fixated centrally, either a single target or a sequence of 4 was shown. Participants were required to make the saccade or sequence of saccades to the remembered locations.

Results: Task1: RTs were faster with reward and loss. Amplitudes decreased with larger numbers of possible targets, and peak velocity was correspondingly slower. Reward increased velocity, but this was strongest when just two potential targets were shown. Task2:RTs were faster with reward. Latencies were longer, and error was larger with longer sequences. For single memory-guided saccades, accuracy was improved by both incentives. For the sequence of 4 saccades, both incentives increased speed but at the cost of diminished accuracy.

Conclusions: We demonstrate that motivation has characteristic effects on action selection and memory. Asymmetrical effects are observed where reward but not penalty invigorated saccade velocity, when choices were simple. RTs for choice and memory were improved by both incentives. Energisation of motor commands may depend on neural reward signals, such as dopamine, whereas noise reduction in decision-making may be valence-independent.

References: 1 Manohar et al 2015. 2 Holmes et al 2006. 3 Zingale et al 1987.

« Back to 2018 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/motivational-effects-on-action-selection-and-memory/