Feasibility testing of a new paradigm to compare Decision-making in cognitive and motor domains
Poster No:
2063
Submission Type:
Abstract Submission
Authors:
Ksenia Germanova1, Ksenia Panidi2, Pavel Novikov2, Vadim Nikulin3, Maria Nazarova4
Institutions:
1HSE Univresity, Moscow, Moscow, 2HSE University, Moscow, Moscow, 3Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Saxony, 4Aalto University, Espoo, Espoo
First Author:
Co-Author(s):
Introduction:
Every movement is a choice made under uncertainty. Recent studies suggest that motor control can be viewed through the lens of neuroeconomics. Motor Decision-Making (MDM) is extensively studied in healthy subjects. We suggested extending these principles to motor rehabilitation. The goal of the present study is to develop and test the feasibility of a new experimental paradigm for motor decision-making under risk. We aim to compare risk preferences in motor and cognitive domains in healthy participants to transfer this paradigm to the clinical population.
Methods:
20 young healthy participants (10 females, age: 24 ± 2.57) were enrolled in the study. The experiment comprised two parts conducted on the same day.
Cognitive part. In the cognitive part, participants performed a series of 80 binary lottery choice questions. Each question presented a choice between a sure amount and a lottery where a bigger amount can be won with a specific probability (Fig. 1).
Motor part. In the motor part, we used custom-made 'Risk&Reach' software (https://risk-n-reach.azurewebsites.net/) on the touch-screen interface when a participant has to make reaching movements with antigravitational support (Fig. 2). This part consisted of three phases: (1) individualization of the trial duration and the most distant goal position; (2) Three types of motor baselines; (3) individualized motor lotteries. All tasks in the motor part involved performing a movement with a dominant hand. After the individualization phase, participants performed three baseline tasks (160 trials each): (a) reaching towards a single goal, (b) reaching towards the same goal in the presence of an additional goal, (c) reaching towards the same goal associated with reward in the presence of an additional goal.
For the motor lotteries, participants had to perform a series of 80 motor choices with probabilities and rewards corresponding to the choices in the cognitive part. To create such lotteries in the motor domain we employed a model-based approach, mapping the probability of successfully hitting a goal using the baseline (a) data. Similar to the cognitive part, in the motor domain participants were presented with a sequence of binary choices between a sure and a risky option (Fig. 2). A sure option implied performing a reaching movement towards a proximal motor goal with a 100% probability of hit, while a risky option implied making a reaching towards a distant goal with a less than 100% probability of hit. The position of the risky (distant) goal corresponded to the hit probability. Both risky and sure options were rewarded with points converted into valuable rewards.
Analysis. We compared probabilities from three datasets (baselines a, b, and c) averaged within 8 target points. Model-predicted probabilities from baseline(a) were compared to observed probabilities in baselines (b) and (c). We used two-sided t-test with Bonferroni correction to account for multiple comparisons, and GLM model to investigate the effect of task type on motor performance characteristics.
Cognitive part. In the cognitive part, participants performed a series of 80 binary lottery choice questions. Each question presented a choice between a sure amount and a lottery where a bigger amount can be won with a specific probability (Fig. 1).
Motor part. In the motor part, we used custom-made 'Risk&Reach' software (https://risk-n-reach.azurewebsites.net/) on the touch-screen interface when a participant has to make reaching movements with antigravitational support (Fig. 2). This part consisted of three phases: (1) individualization of the trial duration and the most distant goal position; (2) Three types of motor baselines; (3) individualized motor lotteries. All tasks in the motor part involved performing a movement with a dominant hand. After the individualization phase, participants performed three baseline tasks (160 trials each): (a) reaching towards a single goal, (b) reaching towards the same goal in the presence of an additional goal, (c) reaching towards the same goal associated with reward in the presence of an additional goal.
For the motor lotteries, participants had to perform a series of 80 motor choices with probabilities and rewards corresponding to the choices in the cognitive part. To create such lotteries in the motor domain we employed a model-based approach, mapping the probability of successfully hitting a goal using the baseline (a) data. Similar to the cognitive part, in the motor domain participants were presented with a sequence of binary choices between a sure and a risky option (Fig. 2). A sure option implied performing a reaching movement towards a proximal motor goal with a 100% probability of hit, while a risky option implied making a reaching towards a distant goal with a less than 100% probability of hit. The position of the risky (distant) goal corresponded to the hit probability. Both risky and sure options were rewarded with points converted into valuable rewards.
Analysis. We compared probabilities from three datasets (baselines a, b, and c) averaged within 8 target points. Model-predicted probabilities from baseline(a) were compared to observed probabilities in baselines (b) and (c). We used two-sided t-test with Bonferroni correction to account for multiple comparisons, and GLM model to investigate the effect of task type on motor performance characteristics.
·Lotteries in the cognitive domain
·Lotteries in the motor domain
Results:
The presence of a reward significantly increased the probability of the distant goal being hit compared to the trials without a reward (p = 0.002) and compared to the predicted probabilities (p = 0.032). Our analysis indicated that participants chose risky options significantly more often in motor-based compared to cognitive lotteries (p = 0.001).
Conclusions:
The presence of choice did not reveal any effect on the hit probability, however, the presence of reward significantly increased the probability of reaching the distant goal. Second, we observed that participants displayed more risk-taking behavior in the motor domain compared to the cognitive domain. Concluding, we showed the feasibility of the developed MDM paradigm for reaching movements to transfer it to patients with motor impairment.
Emotion, Motivation and Social Neuroscience:
Reward and Punishment
Emotion and Motivation Other
Higher Cognitive Functions:
Decision Making 2
Executive Function, Cognitive Control and Decision Making
Motor Behavior:
Motor Behavior Other 1
Keywords:
Design and Analysis
Motor
Other - motor decision-making
1|2Indicates the priority used for review
Provide references using author date format
Gallivan, J. P., Chapman, C. S., Wolpert, D. M., & Flanagan, J. R. (2018). Decision-making in sensorimotor control. Nature Reviews Neuroscience, 19(9), 519-534.