System and method for processing brain signals in a BCI system

1. A computer-implemented method of processing brain signals in a Brain Computer Interface (BCI) system, the method comprising the steps of:

providing a pre-defined set of stimuli and repeatedly activating the stimuli in rounds, wherein in each of the rounds each of the stimuli is activated once;

detecting the respective brain signals in response to each of the stimuli being activated during each of the rounds;

determining whether a subject intends to use the BCI system based on the detected respective brain signals using a processor; and

processing the brain signals for command recognition if it is determined that the subject intends to use the BCI system;

wherein the determining whether the subject intends to use the BCI system comprises:

after k collected number of rounds, determining all the possible signal windows covering l consecutive rounds, from a most recently collected round to a round that was collected l rounds before the most recently collected round, where L<l<k, L is a predefined minimum value, and 3<L<k;

determining the signal window from among said possible signal windows which gives a highest posterior probability that a control pattern was received in said signal window; and

checking whether said highest posterior probability is greater than a predefined first threshold.

2. The method as claimed in claim 1 , wherein the determining whether the subject intends to use the BCI system further comprises collecting an additional round of the respective brain signals if the highest posterior probability is not greater than the first threshold.

3. The method as claimed in claim 1 , wherein the determining whether the subject intends to use the BCI system further comprises calculating a risk of making a decision on whether the subject intends to use the BCI system based on the signal window which gives the highest posterior probability.

4. The method as claimed in claim 3 , further comprising collecting an additional round of the respective brain signals if the risk is not lower than a second threshold.

5. The method as claimed in claim 1 , wherein the processing the brain signals for command recognition is performed over said signal window which gives a highest posterior probability.

6. The method as claimed in claim 1 , wherein the determining whether the subject intends to use the BCI system comprises determining whether for one of the stimuli the brain signal corresponds to a P300 signal.

7. The method as claimed in claim 6 , wherein the processing the brain signals for the command recognition comprises applying a classification method which determines a maximal averaged Support Vector Machine (SVM) score amongst the stimuli for having an associated brain signal corresponding to the P300 signal.

8. The method as claimed in claim 1 , wherein the stimuli comprise one or more of a group consisting of a visual stimulus, an auditory or a somatosensory stimulus.

9. A system for processing brain signals in a Brain Computer Interface (BCI) system, the system comprising of:

a stimulation unit comprising a set of stimuli, wherein the stimulation unit repeatedly activates the stimuli in rounds, such that in each of the rounds each of the stimuli is activated once;

a detection unit for detecting the respective brain signals in response to each of the stimuli being activated during each of the rounds and for determining whether a subject intends to use the BCI system based on the detected respective brain signals; and

a recognition unit for processing the brain signals for command recognition if the detection unit determines that the subject intends to use the BCI system,

wherein the detection unit, after k collected number of rounds, determines all the possible signal windows covering l consecutive rounds, from a most recently collected round to a round that was collected l rounds before the most recently collected round, where L<l<k, L is a predefined minimum value, and 3<L<k;

determines the signal window from among said possible signal windows which gives a highest posterior probability that a control pattern was received in said signal window; and checking whether said highest posterior probability is greater than a predefined first threshold.

10. A non-transitory data storage medium having stored thereon computer code means for instructing a computer system to execute a method of classifying brain signals in a Brain Computer Interface (BCI) system, the method comprising the steps of:

providing a pre-defined set of stimuli and repeatedly activating the stimuli in rounds, wherein in each of the rounds each of the stimuli is activated once;

detecting the respective brain signals in response to each of the stimuli being activated during each of the rounds;

determining whether a subject intends to use the BCI system based on the detected respective brain signals; and

processing the brain signals for command recognition if it is determined that the subject intends to use the BCI system;

wherein the determining whether the subject intends to use the BCI system comprises:

after k collected number of rounds, determining all the possible signal windows covering l consecutive rounds, from a most recently collected round to a round that was collected l rounds before the most recently collected round, where L<l<k, L is a predefined minimum value, and 3<L<k;

determining the signal window from among said possible signal windows which gives a highest posterior probability that a control pattern was received in said signal window; and

checking whether said highest posterior probability is greater than a predefined first threshold.