Abstracts

Rationale: Mismatch negativity (MMN) is an objective index for preattentive detection of acoustic changes in human. The goal of this study was to find out the information processing of Japanese vowel-speech and pure-tone changes using MMN in patients with intractable temporal lobe epilepsy. Methods: The subjects consisted of eight right-handed patients with intractable temporal lobe epilepsy (mean age 33.8±11.0 years, 3 females), and nine right-handed healthy controls (34.1±10.4 years, 6 females) . The laterality of epileptic focus was left in 3 patients, right in 1 patient, and unclear or bilateral in the rest. The stimuli of the vowel-speech task were constant Japanese vowel-speech pairs (standard /a/ vs. deviant /o/) and those of the pure-tone pairs (standard 1000Hz vs. deviant 1050Hz). The electrodes were placed at Fz, Cz, F3, F4, and bilateral mastoids. The MMN waveforms were derived by point-by-point subtraction of those for the standard stimulation from those for the deviant stimulation. MMN peak was determined within a window of 50-200 ms. MMN parameters measured were mean amplitudes, peak amplitudes, and peak latencies. Three- and two-way repeated measures analyses of variance (ANOVA) were carried out using Super-ANOVA program. Results: In intractable temporal lobe epilepsy group: The MMN latency for vowel-speech task (86msec at Fz) was significantly shorter than that for pure-tone task (119msec at Fz). MMN amplitudes were similar between the two tasks at all electrodes. No hemispheric preference for processing of vowel-speech or pure-tone was observed. In healthy control group: MMN amplitudes showed a significant main effect of task, indicating that the MMN amplitudes at Fz for vowel-speech task were larger than those for pure-tone task. However, the MMN amplitudes at both mastoids were not different between the two tasks. The MMN latency was similar between the two tasks. No hemispheric preference for processing of vowel-speech or pure-tone was observed. Conclusions: The present results suggest that the processing of vowel-speech sound and pure-tone change may be different between patients with intractable temporal lobe epilepsy and healthy controls.