System identification of neural mechanisms from trial-by-trial motor behaviour: modelling of learning, impairment and recovery. (1st February 2017)

• Record Type: Journal Article
• Authors: Ueyama, Yuki
• Is Part Of: Advanced robotics. Volume 31:Number 3(2017)
• Page Start: 107
• Page End: 117
• Publication Date: 2017-02-01
• Subjects: Reaching -- state-space model -- manipulandum -- rehabilitation -- computational neuroscience; Robotics -- Periodicals; Robotics -- Japan -- Periodicals; Robotics; Japan; Periodicals; 629.89205
• DOI: 10.1080/01691864.2016.1266966
• Languages: English
• ISSNs: 0169-1864

Abstract

Abstract: System identification is an engineering technique used to build mathematical models of dynamic systems from observed data. In neuroscience research, system identification has been used to adapt robotic devices for measurements and to apply perturbations. These techniques have been shown to be useful tools to investigate brain mechanisms of motor learning processes and motor dysfunctions. This paper reviews the application of system identification to motor behavioural data pertaining to the interruption of specific neural mechanisms. This system identification technique could reveal adaptation and generalisation functions in neural mechanisms for motor tasks to fit experimental data measured by the robotic devices in a discrete state-space model. When comparing constructed models between persons with neurological disabilities and healthy control subjects, system identification could be used to identify impaired abilities. Furthermore, a recent study indicated that such a discrete state-space model may predict recovery of motor function after stroke. Therefore, we suggest that the marriage between robotic devices and system identification will comprise a novel approach to clinical assessment and aid in the design of patient rehabilitation protocols. Abstract :