The method of source derivation for the EEG.

The technical background to this work was described by Hjorth (1975). The derivation is a practical implementation of the Laplace operator and detects source activity as it appears on the surface level of the scalp. The advantages are improved topographical selectivity as compared to bipolar or referential derivation and simplicity of presentation as the activity at each electrode position is represented by only one channel, without the influence of a monopolar reference. Using the electrode positions in the 10-20 system, the derivations are implemented by means of resistive matrix networks connected between the EEG amplifiers, the source derivation is divided between two programmes, an anterior and a posterior. Personal experience with the method over the last 12 months has been assessed. 500 consecutive and otherwise unselected records were re-examined. Source derivation did not demonstrate an abnormality when the conventional derivations were normal. The principal finding was that localised abnormalities may be more obvious and their location more easily and accurately defined.