Classification of patient conditions using known and artificial classes
View Patent ↗Methods of classifying a subject's condition are described. The method includes: receiving measured signals from the subject; processing the measured signals using a computing device to identify a class associated with an identified condition of the subject; introducing an artificial class, the artificial class being associated with an unknown condition of the subject; classifying a feature vector from the subject into the identified class or the artificial class; and generating a signal in response to classifying the feature vector. The measured signals from the subject may include at least one signal extracted from brain activity of the subject.
1. A method of classifying a subject's condition, the method comprising:
receiving measured signals from the subject;
processing the measured signals using a computing device to identify a class associated with an identified condition of the subject;
introducing an artificial class, the artificial class being associated with an unknown condition of the subject;
classifying a feature vector from the subject into the identified class or the artificial class; and
generating a signal in response to classifying the feature vector.
2. The method as recited in claim 1 , wherein:
said processing the measured signals to identify the class associated with the identified condition of the subject comprises processing the measured signals to identify a plurality of classes associated with a corresponding plurality of identified conditions of the subject; and
classifying the feature vector from the subject comprises classifying the feature vector from the subject into one of the plurality of identified classes or the artificial class.
3. The method as recited in claim 2 , wherein:
said receiving measured signals from the subject comprises receiving a training data set of measured signals from the subject;
said processing the measured signals to identify the class associated with the identified condition of the subject comprises processing feature vectors from the training data set to associate each processed feature vector with one of the plurality of identified classes.
4. The method as recited in claim 1 , wherein introducing the artificial class comprises:
generating artificial feature vectors; and
associating the generated artificial feature vectors with the artificial class.
5. The method as recited in claim 4 , wherein generating artificial feature vectors comprises randomly generating artificial feature vectors.
6. The method as recited in claim 5 , wherein randomly generating artificial feature vectors comprises randomly generating artificial feature vectors according to a specified probability distribution.
7. The method as recited in claim 1 , wherein the identified condition of the subject comprises a neurological condition.
8. The method as recited in claim 1 , wherein the measured signals from the subject comprise at least one signal extracted from brain activity of the subject.
9. The method as recited in claim 1 , wherein said generating the signal in response to classifying the feature vector comprises providing an output to the subject indicative of classification into the identified class or the artificial class.
10. The method as recited in claim 1 , wherein the identified class comprises a class associated with a pre-ictal condition or a pro-ictal condition.
11. The method as recited in claim 1 , wherein the identified class comprises a class associated with a contra-ictal condition.
12. The method as recited in claim 1 , wherein the identified class comprises a class associated with an inter-ictal condition.
13. The method as recited in claim 1 , wherein said receiving measured signals from the subject comprises receiving signals measured using a device implanted in the subject's body.
14. The method as recited in claim 1 , wherein said receiving measured signals from the subject comprises receiving signals measured using a device implanted in the subject's head.
15. The method as recited in claim 1 , wherein said receiving measured signals from the subject comprises receiving signals measured using a device implanted in the subject's head.