Related Experiment Video
Updated: May 24, 2026
06:12
Virtual Prism Adaptation Therapy: Protocol for Validation in Healthy Adults
Published on: February 12, 2020
Probabilistic models of state estimation predict visuomotor transformations during prism adaptation.
Masaki Yamamoto1, Hiroshi Ando
1Division of Human Biology, Department of Rehabilitation Science, Graduate School of Health Sciences, Kobe University, Kobe, Japan. yamamotoma@kawasakigakuen.ac.jp
Visual Neuroscience
|March 7, 2012
Summary
This study developed a Kalman filter model to predict state-space estimation during prism adaptation. The Kalman filter better explained internal model dynamics with uncertain visual feedback compared to a linear model.
Area of Science:
- Neuroscience
- Cognitive Science
- Robotics
Background:
- Prism adaptation involves recalibrating visuomotor transformations.
- Understanding internal models of space is crucial for motor control.
- Information processing under visual uncertainty influences adaptation.
Purpose of the Study:
- To develop a predictive model for state-space estimation in prism adaptation.
- To identify information processing requirements for external space identification.
- To compare the efficacy of Kalman filtering versus linear models in representing internal dynamics.
Main Methods:
- Fifty-seven healthy students performed reaching movements under prism conditions with varying visual feedback.
- A linear parametric model and a Kalman filter (state estimation model) were employed.
- Goodness of fit was evaluated using the Akaike Information Criterion (AIC).
Main Results:
- The Kalman filter demonstrated a better fit (lower AIC values) than the linear model across conditions.
- Kalman gain varied, indicating reliance on prior estimates under uncertainty.
- The Kalman filter effectively simulated state estimation based on visual feedback reliability.
Conclusions:
- The Kalman filter, a Bayesian approach, accurately models internal dynamics during visuomotor adaptation with uncertain feedback.
- Probabilistic estimation models are suitable for simulating state estimation based on feedback reliability.
- This research provides insights into sensorimotor integration and predictive coding.
