Abstracts

Rationale: A series of pediatric patients were evaluated with simultaneous hemodynamic (NIRS) and EEG during focal seizures. Patients were selected on the basis of demonstrating asymmetric electrical activation of left or right hemispheres during an ictus. The hypothesis tested was that there is decreased cereberal oxygnation/ blood flow during an ictus, as reflected by NIRS values, and these changes would be reflected in background slowing of the EEG, and that asymmetry in interictal NIRS will be correlated to asymmetries in the delta component of the backgrond EEG. Methods: EEG electrodes were affixed using standard methods and locations (10-20 system). Responses were recorded for at least one full day , including ictal, preictal, post-ictal and interictal with six replications for each of these conditions. The NIRS recording patches (INVOS) were placed in a fronto-parietal location over each hemisphere and there was a control patch over the flank. NIRS values wwere collected and displayed every five seconds and the EEG was recorded continuously although calculations of the specta were performed on 9 second segments. EEG segments were chosen so as to be artifact free, free of intervention with infused anticonvulsants and lastng more than 10 seconds for the ictus. Spectral analysis (FFT) was performed on a nine second segment just prior to,during and following the seizures and at least five minutes away from the seizure. There were six replications of each of these conditions. The EEG data was analyzed for the power in the conventionally defined delta, theta, alpha and beta frequency bands. A repeated measures within subject ANOVA was performed. Results: During baseline recordings lower NIRS values were associated with increased absolute and relative delta frequency power but had only a modest relationship to frequency bands. During the ictus there were only modest changes in the NIRS values but highly significant increases in the delta component ipsilateral to the side of the electrographic ictus. Post-ictus the changes in the delta contribution persisted, even when the NIRS value had returned to baseline. At times the increase in the delta power ictally and post-ictally exceeded 100%. In two of the patients PET scan changes of hypometabolism agreed with NIRS depressed levels. Conclusions: Simultaneous, temporally sychronized NIRS readings and EEG recordings can be recorded using conventional technology. There is a clear relationship bertween interictal lower NIRs values and increased delta activity in the EEG. During an ictus, however, there is a complex relationship between changes in the NIRS values and the spectral content of the EEG. EEG changes are both more consistent and larger than NIRS changes. It will be necessary to record more prolonged seizures to determine if the NIRS changes will be more dramatic in status epilepticus. Within subject designs are likely to be most useful in these studies given the large inter-individual differences compared to the within subject variability.