Speech and non-speech processing in hemispherectomised children: an event-related potential study.

Although hemispherectomy is now used as a radical treatment for intractable seizures in a number of centres, there have been limited electrophysiological studies investigating post-procedure auditory-speech processing and recovery or reorganisation. We therefore recorded auditory event-related potentials (ERPs) evoked by pure tones and syllables employing a 51-channel electrode array concentrated over the functional hemisphere in 17 patients (nine males, mean age 14.2 years) who had undergone hemispherectomy for intractable seizures; eight of the patients had congenital brain damage and nine had sustained their initial insult at an age of 1 year or older. For comparison, recordings were made from 10 controls (five males, mean age 13.5 years). Responses from patients consisted of five small amplitude components restricted to the functional hemisphere whose scalp polarity was opposite to the ERPs observed in control subjects. The topography of the auditory ERPs in the patients was localised to the centro-temporal regions of the functional hemisphere in comparison to the midline distribution observed in control subjects. As in the control subjects, the latencies of the components evoked by tones were shorter than those evoked by syllables in both left and right hemispherectomised children. In addition, further analysis of the N1 and P2 component revealed increased syllable latencies in left hemispherectomised children only. The results of this study suggest that the left hemisphere alone may be more efficient at processing both pure tones and syllables compared to the right or both together.