Nonlinear Filtering of Athetoid Movement

Principal Investigator: Cam Riviere, Ph.D.

Co-Investigators: Dan Ding, PhD; Rory A. Cooper, PhD; Rosemarie Cooper; MPT, ATP; Annmarie Kelleher, MS, OTR/L, ATP; Emily Teodorski, BS; Sara Sibenaller, BS; Brad Dicianno, MD

2006-2008

The goal is to develop models of the movements of people with athetosis, a movement disorder characterized by writhing motions mainly in the upper extremities, and use these models to develop input filters to assist computer access for persons with athetoid CP. The steps of this process are as follows:

• Subjects with athetoid CP will use a joystick to execute a series of computer tasks without any input filtering. 
• Athetoid movement will be quantified in order to decipher between intended and unintentional movement and determine which variables are most the useful filter inputs during these experiments.   
• Athetoid models will be developed in order to test the effectiveness of filtering algorithms offline. 
• Filtering algorithms will be developed and tested using the athetoid model. 

• The validated models will be implemented online as input filters and will then be field-tested by people with athetoid CP in realistic computer tasks.  

Progress:

To date, fifteen subjects with athetoid CP and ten subjects with spastic CP completed the first phase of the study. The result showed that both groups spent over 60% of the time acquiring the target as compared to 40% of the time moving towards the target. They also had more submovements and pauses than necessary to complete the tasks. Subjects with athetoid CP had lower performance than those with spastic CP as indicated by some of the cursor movement measures. The study will provide insights on optimizing and designing computer access methods
for people with CP.  A master thesis and a few conference papers were resulted from the project. The development of new filtering algorithm is ongoing with our CMU partners.