Obviously, you have already figured that things are not so easy. To discriminate between different brain patterns associated with actual movements or imagined movements, often quite complicated signal processing algorithms are necessary. Discriminating between all the tasks you suggested seems to be impossible to me. Therefore, I suggested you should better start with something simpler than discriminating these five tasks.
Further, I really advice you to start with reading some relevant papers, because I have the suspicion you haven't done that yet. Here are some quick answers: Reference placement is usually not so critical. With monopolar recordings, the earlobe or the mastoid (behind the ear) is often used. With bipolar recordings you do not have to worry anyway. Avoid any tasks that require muscle activity close to recoding locations, otherwise you will most likely discriminate myographic activity and not brain activity. Spatial resolution of the EEG is poor (caused e.g. by volume conduction), that is, you do not only measure the activity below a particular electrode. Spatial filters like Laplacian can improve that a little bit. C3 and C4 correspond to the hand areas (note the association is contralateral, i.e. C3 = right hand, C4 = left hand). Best, Bernhard Graimann Anish Mittal wrote: > Dear Prof Bernhard Griamann > > > I thank you for your valuable information. > > > I would like to ask you a few more things.I hope you will accept my > humble request and help me a little more > > > *do you want to detect these movements or do you also want to > discriminate between these movements? * > > I want to detect them as well as discriminate them as i want five type > of different signals for cursor control > > > *1) You have to be very careful that you will not actually use > myographic > activity instead of brain activity (I assume you want to use brain > activity only). Especially with movements that require muscle activity > of muscle groups close to the EEG electrodes.* > > > How can i differentiate whether the signals are myographic or brain ones?? > > *2) The SNR of the brain patterns associated with movements is usually > very well correlated with the size of the corresponding cortical area. > That is, finger movements or tongue movements produce better (easier to > detect or discriminate) patterns than shoulder or elbow movements, since > the latter two have much smaller cortical representations in the motor > cortex. > > 3) Because of the previous two points I suggest you should focus on > discriminating the following tasks: (right finger(s), left finger(s), > foot movement (both at the same time). For electrode locations, I > suggest to cover the motor-sensory area as good as it is possible with > your number of electrodes and your electrode caps available - including > the positions C3, C4, and Cz.* > > > ---------When i placed the electrodes on the motor cortex the was unable > to figure out the difference between the left and right movements.They > were coming in the same manner at same electrodes with out any change > One more problem is to select the reference electrode?? > Where should i place it > I placed it between eyebrows > --------And my EEG machines measure biploar EEG like between 2 > electrodes like Cz-Pz,Fp2-F4. > I am not able to measure the EEG of single electrode with respect to > reference. > > --------I have 19 electrodes which are supposed to be placed in 10-20 > system format,but i placed all on motor cortex as electrodes only > measure the potential at where they are placed.Am i correct in this > respect?? > *---------*Which type of signals i will get by placing electrodes on > C3-C4.Is it that of fingers?? > -------My machine software is giving visual (which can taken as print) > or .eeg file as output.How i process the .eeg file.Are there some > standard softwares available for this?? > > > Awaiting reply > > > Regards > > Anish Mittal > > > ------------------------------------------------------------------------ > > _______________________________________________ > BCI-info mailing list > [email protected] > https://mlist.tugraz.at/mailman/listinfo/bci-info _______________________________________________ BCI-info mailing list [email protected] https://mlist.tugraz.at/mailman/listinfo/bci-info
