Page 1 of 1
Artifact rejection/detection settings
Posted: 08 Apr 2008, 02:35
by emily
Hello,
Can anyone tell me where the artifact rejection/detection settings are?Where you can set voltage levels and cursor colour change for eg. to notify participant when they are using jaw clenches etc to control the cursor.
thanks
artifact rejection ...
Posted: 08 Apr 2008, 20:39
by gschalk
Emily,
If it were that easy
There is currently no reliable way to detect whether somebody uses muscle activity instead of brain activity. While in theory there clearly are differences when activating different muscle groups compared to brain signal changes (e.g., see I.I. Goncharova et al., 2003, EMG contamination of EEG: spectral and topographical characteristics), it is difficult to detect these changes on a single trial basis, in particular when they are small contractions. Thus, because nobody really knows how to do this, there is currently no artifact detection system in BCI2000.
The Gerv
artifacts
Posted: 08 Apr 2008, 20:47
by gschalk
Emily,
To expand on this issue: the best thing you can do is to analyze data after each feedback session, and to use this information to determine whether the activations that you see (e.g., the patterns that discriminate the two feedback conditions) are consistent with brain signal changes, i.e., located over sensorimotor cortex, and at least somewhat sharply focused in the frequency band between 10-30 Hz, or with muscle-based changes, i.e., peripheral activation and broad spectral changes in large frequency bands.
The Gerv