Method and apparatus for detecting seizures

Abstract

A method of detecting seizures may comprise receiving an EMG signal and processing the received EMG signal to determine whether a seizure characteristic is present in the EMG signal during a time window. An apparatus for detecting seizures with motor manifestations may comprise one or more EMG electrodes capable of providing an EMG signal substantially representing seizure-related muscle activity; and a processor configured to receive the EMG signal, process the EMG signal to determine whether a seizure may be occurring, and generate an alert if a seizure is determined to be occurring based on the EMG signal.

1 - 78 . (canceled) 
     
     
         79 . An apparatus for detecting seizures with motor manifestations, the apparatus including:
 one or more EMG electrodes capable of providing an EMG signal substantially representing seizure-related muscle activity;   a processor configured to receive the EMG signal, process the EMG signal to determine whether a seizure may be occurring, and generate an alert if a seizure is determined to be occurring based on the EMG signal;   said processor being configured to detect bursts of EMG signal, assign certainty values to individual burst members among said detected bursts, and weight the number of said detected bursts as a function of the certainty values assigned to said individual burst members;   wherein said certainty values are determined based on a comparison of how said individual burst members compare to a reference burst in terms of one or more burst characteristics selected from the group of characteristics including burst signal-to-noise ratio, burst width and burst amplitude;   said processor being further configured to identify the presence of a plurality of bursts over a time window, determine the periodicity of bursts over said time window, and determine a periodicity contribution to seizure detection; and   said processor being further configured to use a supervisory algorithm to determine a seizure detection value using the certainty value weighted number of detected bursts and the periodicity contribution, and compare said seizure detection value to a threshold seizure detection value suitable to indicate if said seizure is occurring.   
     
     
         80 . The apparatus of claim  1 , further including one or more of an ECG electrode, a temperature sensor or an accelerometer. 
     
     
         81 . The apparatus of claim  1 , wherein the one or more EMG electrodes are mounted to one or more of an arm band, adhesive tape, or item of clothing so as to allow positioning of the one or more EMG electrodes over a muscle. 
     
     
         82 . The apparatus of claim  1 , wherein the one or more EMG electrodes are differential bipolar electrodes. 
     
     
         83 . The apparatus of claim  1 , including two EMG electrodes capable of being associated with an agonist/antagonist muscle pair, wherein one EMG electrode is associated with an agonist muscle, and the other EMG electrode is associated with its antagonist muscle. 
     
     
         84 . The apparatus of claim  5 , wherein the agonist/antagonist muscle pair includes the triceps brachii and biceps brachii. 
     
     
         85 . The apparatus of claim  1 , further including a transceiver for transmitting the alert. 
     
     
         86 . The apparatus of claim  1 , further including a base station in communication with the processor for receiving the alert. 
     
     
         87 . The apparatus of claim  8 , wherein the base station further includes an I/O device capable of allowing manual adjustment of alert settings and visually displaying the EMG signal or data based thereon. 
     
     
         88 . The apparatus of claim  1 , wherein said bursts are qualified against a minimum threshold duration and a maximum threshold duration. 
     
     
         89 . The apparatus of claim  1 , wherein the determining of said periodicity contribution includes calculating an average deviation for times between bursts included among said plurality of bursts and identifying if said average deviation is less than or greater than a threshold average deviation. 
     
     
         90 . The apparatus of claim  11 , wherein the processor is further configured to negatively weight said periodicity contribution against seizure detection if said average deviation is less than a threshold average deviation. 
     
     
         91 . The apparatus of claim  1 , wherein the processor is further configured to eliminate bursts from said plurality of bursts if the bursts among said plurality are too close together or too far apart; and
 wherein the determining of said periodicity contribution includes comparison of the periodicity of burst characteristics over said time window to a minimum uniformity threshold and a maximum uniformity threshold.   
     
     
         92 . The apparatus of claim  1 , further including one or more leads-off detectors configured to indicate whether one or more of the EMG electrodes is sufficiently close to a muscle to provide a substantially accurate EMG signal representing activity of the muscle. 
     
     
         93 . The apparatus of claim  1 , wherein the one or more EMG electrodes and processor are packaged as a single unit mountable to a human body.