Children who have serious speech or writing impediments struggle every day with the simplest of communication tasks most of us take for granted. Over 500,000 children in the US require assistive technology to communicate and 20-60% of these children may require special-purpose keyboards, screens, or software to communicate with computers and computer-controlled assistive devices. For children with dyskinetic cerebral palsy (CP and severe arm movement disorders, devices such as adapted keyboards, programmable touch-screens, and joystick interfaces may be their most important communication link to their families, friends, teachers, and caregivers. However, these devices are so difficult to use that communication is often extremely slow and laborious and often results in high levels of frustration for the children using them. Additionally, because every child with CP presents differently, no one device is optimized for a particular child's impairments. Therefore, our goal for this study is to address the critical need for improved assisted communication devices and to optimize the interface design for each individual child. Two of the most common computer-based interface devices are touch- screen and joystick controllers. We propose to use a set of recently-developed techniques based on information theory and Bayesian signal processing to optimize the design of touch- screen interfaces and joystick controllers for children with dyskinetic cerebral palsy that impairs hand movement. Optimization of the touch-screen device will be based upon measurement of the speed, accuracy, and error rate of individual children as they reach to targets of varying size, spacing, and number. Optimization of the joystick device will make use of the complete path of the joystick as the child approaches the target. We have developed a new algorithm that can efficiently use Bayesian nonlinear filtering to estimate the intended target of the joystick early in the movement. This estimation significantly improves performance, but it introduces a tradeoff between speed and accuracy when children have variable movements. Therefore we must measure speed, accuracy, and error rate as a function of target size and spacing for the joystick as well. Bradykinesia (slow movement) and Hyperkinesia (increased variable movement) are common impairments of children with dyskinetic cerebral palsy that will affect performance on both devices. We will quantify these impairments and determine their affect on communication device use and optimization. Finally, we will test whether implementation of improved interfaces can increase the rate of communication. This study will provide essential quantitative links between disorders of movement and disorders of assisted communication for children with the dyskinetic form of cerebral palsy.

Public Health Relevance

Up to 60% of children with dyskinetic cerebral palsy depend on computer-based assistive communication devices to talk with parents or friends and to participate in school. There is currently no reliable method for determining the best touch-screen layout for individual children. We will perform detailed testing of individual children and combine the results with a model of information processing in order to find the best possible design of a touch-screen or joystick interface that will maximize the speed of communication in children with dyskinetic cerebral palsy.

National Institute of Health (NIH)
National Institute of Neurological Disorders and Stroke (NINDS)
Research Project (R01)
Project #
Application #
Study Section
Motor Function, Speech and Rehabilitation Study Section (MFSR)
Program Officer
Chen, Daofen
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Southern California
Biomedical Engineering
Schools of Engineering
Los Angeles
United States
Zip Code