EVALUATION OF A COMPUTERIZED SYSTEM FOR RECOGNITION EPILEPTIC ACTIVITY DURING LONG-TERM EEG RECORDING

A new method of recognition of epileptic activity using adaptive segme ntation in EEG during long term intensive monitoring was developed in Tampere. The performance of the system was validated and compared to t he commercially available discharge recognition system of Gotman. Twel ve approximately 30 min EEG segments recorded during intensive monitor ing from 6 patients were analysed. On these EEG segments two EEG speci alists marked the occurrence of epileptic activity independently. Late r they re-evaluated any differences in their scoring. This consensus f ile was used as a reference in validating the performance of the two c omputer programs. We found that the program developed in Tampere detec ted discharge activity more often than the Gotman system. Both systems performed poorly in spike recognition. In the specificity of the reco gnized segments, the Gotman system was better.