HCC: Medium: Haptic Simulation Design for Motor Rehabilitation and Skill Training

North Carolina State University, Raleigh NC

Investigators

Abstract

The PI seeks to design and investigate novel features and strategies for use of virtual reality based haptic simulations to retrain impaired motor functions and train new fine motor skills in veterans (medics) with traumatic brain injury (TBI). Previous approaches to haptic simulation design have typically focused on completeness of presentation relative to reality. However, simulators may be used for many different tasks, so completeness may not ensure usefulness. The PIs argue that the role of the simulation for a user and its required functionality are the more important design questions, and that consideration of critical applications like motor skill training in haptic simulation design will promote effective design from a human performance perspective. Thus, they propose a cognitive-oriented approach to haptic simulation design and evaluation. In this project, the PIs will test their ideas by designing and prototyping advanced VR haptic simulators for drawing and surgical tasks. The simulations will be defined in terms of the type and resolution of (medical) data sources used for virtual object modeling, the type of visual and force presentation for motor skill and brain function assessment, and the approach to graphic and haptic rendering of the simulation. Through human factors experimentation with the simulations, the PIs will also assess interventional strategies for motor development, including virtual haptic aids (e.g., force boundaries and potentials) and force graduation across rehabilitation trials. Finally, they will validate the effect of the haptic simulations on neurocognitive and motor performance using behavioral indices and advanced magnetic resonance imaging (fMRI). Cognitive tasks analyses will be conducted with expert neuropsychologists and surgeons on psychomotor task performance and surgical operations to inform the simulation design process. Simulation design manipulations will include source data reduction through approximation of point-cloud data using graphical methods, development of a haptic-based motor skill training workstation with an effective human-computer interface, and optimization of visual and haptic representation using GPU-based graphics and a smoothed particle hydrodynamics (SPH) model for reduction of CPU overhead in haptic object rendering with force feedback. The simulation design will also reflect results on baseline motor performance and VR drawing and surgical simulator practice. This testing will be followed by fMRI of subjects in motor tasks in order to examine activation of brain regions mediating motor control. Subsequently, a series of motor training sessions will be conducted using the haptic simulators. Subjects will be exposed to the various settings of the simulation design parameters along with the virtual haptic aids and gradual reduction of force feedback, relative to nominal forces, across sessions. Post-therapy motor and simulator tests, as well as follow-up fMRI scanning, will be conducted. Motor recovery and neuroimaging of brain regions mediating motor performance will provide evidence of the effectiveness of the simulation design and rehabilitation efficacy. The PIs' hypothesize that experience with a haptic simulation design based on motor skill training demands and human performance metrics will accelerate skill development relative to a fidelity-centered approach to design. They also expect that haptic-simulator experience will improve fine motor control and motor planning (praxis), that experience on the drawing-simulation device will generalize to improved performance on the surgical-simulation device, and that brain blood flow will increase in regions mediating motor control. Broader Impacts: This work will make contributions to the design of computer graphics and haptic rendering, in terms of better understanding of optimal computational modeling for VR haptic simulation. It will also advance the state of the art in computer-based therapeutic approaches to motor skill development with haptic simulation, and enhance our understanding of brain-behavior relationships governing motor output and the nature of neural recovery following motor rehabilitation. Improvements in existing VR-based rehabilitation strategies for motor and praxis impairment in individuals suffering from TBI will be identified; rehabilitation treatment regimens will be identified that may have implications for various populations suffering from brain injuries (e.g., stroke patients).