7035999409

Format Abstract »    Send to »

Cc*f Prac EEE Eno M«d Bal Soc 2012 2012 4140-3 «k 10 110ft€U8C 2012 634M78

A neural network-based design of an on-off adaptive control for Deep Brain Stimulation in movement disorders.

ShuMaP¹. Basu I. Graupe D. Turonem D. Staytn KV

$ Author Information

Abstract

The current Food and Drug Admmistration approved system for the treatment of tremor disorders through Deep Bram Stmulation (DBS) of the area of the bram that Controls movement. operates open-ioop rt does not automaticaity adapt to the mstantaneous pafenTs needs or to the progression of the disease This paper demonstrates an adaptive dosed-ioop controiied DBS that. after switchmg off stimulation. tracks few physiological signals to predict the reappearance of tremor before the patient expenences discomfort. at which point it instructs the DBS controller to switch on stmulation agam The core of the proposed approach is a Neural NetworK (NN) which effectivety extracts tremor predictive Information from non-invasiveiy recorded surface-ełectromyogram(sEMG) and accelerometer signals measured at the symptomatic ertremities A srnipłe feed-forward back-propagation NN architecture is shown to successfully predict tremor in 31 out of 33 tnais in two Parkinsona Disease patients with an overan accuracy of 75.8% and sensitiyity of 92 3% This work therefore shows that dosed-ioop DBS control is feasibie m the near futurę and that it can be achieved without modirications of the electrodes imptanted in the brain. i e. is backward compatibie with approved DBS Systems

PUD 23366SJ9 00110.1iretMK.20l2.rafflg

[ndexed lor UEOUNt]

n » ss

Głęboka stymulacja mózgu ułatwiająca życie chorym na chorobę Parkinsona

Roboty medyczne

Odkrywanie nowych leków

Analiza śladów cyfrowych. Marketing celowany
3	laJ		Ustalanie składu mieszania betonowej
■ i	4
n