Reduction Of Classification Error Rates And Monitoring System Using An Artificial Class

Abstract

Systems and methods for enhancing the accuracy of classifying a measurement by providing an artificial class. Seizure prediction systems may employ a classification system including an artificial class and a user interface for signaling uncertainty in classification when a measurement is classified in the artificial class.

1 . A method of analyzing a physiological signal using a computing system, comprising:
 receiving at a computing system subject data representing sampled physiological signal measured from a subject;   deriving using the computing system a feature vector from the subject data;   receiving classification information from a reviewer, said classification information identifying an empirical class associated with a segment of the subject data and representing an empirically-determined condition;   generating an artificial class not representing the empirically-determined condition; and   classifying a feature vector from subsequent subject data representing sampled physiological signals measured from the subject into the empirical class or the artificial class.   
     
     
         2 . The method of  claim 1 , wherein said artificial class captures feature vectors atypical of any empirically-determined conditions. 
     
     
         3 . The method of  claim 1 , wherein said artificial class captures feature vectors near a boundary between a first empirical class and a second empirical class. 
     
     
         4 . The method of  claim 1 , wherein said artificial class captures feature vectors associated with an unknown condition. 
     
     
         5 . The method of  claim 1 , wherein said artificial class captures feature vectors determined to be particularly susceptible to misclassification. 
     
     
         6 . The method of  claim 1 , further comprising displaying on a computer user interface the classification of the feature vector from subsequent subject data. 
     
     
         7 . The method of  claim 1 , wherein said sampled physiological signal comprises a sampled neurological signal. 
     
     
         8 . The method of  claim 7 , wherein said empirically-determined condition comprises a seizure state.