An Electroencephalograph (EEG) is a system use to register voltage fluctuations caused by the activity of nerve cells in the Cerebral cortex. The recording of neural activity is accomplished by an array of electrodes placed on the head of the subject, each electrode records information in its immediate surroundings. Knowing the exact location of each electrode in relation with the Cerebral cortex is crucial in order to correctly decipher and understand the results of the EEG procedure.
There is a general model which gives the approximate location of each electrode in relation with the Cerebral cortex and so the region from wich it records. This model is sometimes insufficient for medical and research needs.
Finding a model of the electrodes' precise location without external data is very complicated. This is because the human head is a 3D object and therefore it is difficult to estimate the precise location of an electrode on it by measuring the Euclidean distance between all the electrodes, moreover human heads differ widely from one person to another and so calibration is normally required in each session.
This project will offer a solution that incorporates tools from image processing & computer graphics in order to precisely locate the EEG electrodes in 3Dspace. This information together with a calibration of typical points in the human head – corners of the eye, corners of the lips and ears will be able to give a more precise location of the electrodes on the subject's head and increase the effectiveness of such procedures.
The solution is based on taking multiple photos simultaneously. A FUJIFILM-3D camera is used, it has two lenses set apart by a fixed and given distance. This camera was chosen due to the fact that the two resulting images created in one shot are mutually aligned, so easing the process of creating a Disparity map, thus saving time and the error resulting from alignment.