Activity Number:
|
333
- SPEED: Biopharmaceutical Statistics, Medical Devices, and Mental Health
|
Type:
|
Contributed
|
Date/Time:
|
Tuesday, August 1, 2017 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Section on Medical Devices and Diagnostics
|
Abstract #324714
|
|
Title:
|
Developing a Performance Sustaining Decoder for a Brain Computer Interface Controlled Neuroprosthetic Device
|
Author(s):
|
David Friedenberg* and Mingming Zhang and Michael Schwemmer and Nick Annetta and Marcia Bockbrader and Chad Bouton and Ali Rezai and W. Jerry Mysiw and Herbert Bresler and Gaurav Sharma
|
Companies:
|
Battelle Memorial Institute and Battelle and Battelle and Battelle and Center for Neuromodulation, The Ohio State University & Department of Physical Medicine an and Battelle (currently at Feinstein Institute for Medical Research) and Center for Neuromodulation, The Ohio State University and Department of Physical Medicine and Rehabilitation, The Ohio State University and Battelle and Battelle
|
Keywords:
|
brain computer interface ;
machine learning ;
wavelet analysis ;
neural network ;
deep learning
|
Abstract:
|
Brain computer interfaces (BCIs) coupled with neuroprosthetic devices have demonstrated great promise for restoring movement to those paralyzed as a result of a spinal cord injury(SCI). As part of an ongoing clinical study of a BCI controlled neuroprosthetic, we examined the properties of neural recordings from an intracortically implanted multi-electrode array in a quadriplegic human over an extended period of time. We examined different types of neural features including novel wavelet based features for classification. We explored different classification paradigms including support vector machines and deep neural networks to determine which combinations led to optimal performance when retrained daily. Additionally, we adapted these algorithms and demonstrated performance sustaining characteristics such that regular retraining could be minimized, which is vital to transitioning this type of BCI technology from the lab to a device that can improve the day to day lives of paralyzed patients.
|
Authors who are presenting talks have a * after their name.