Abstracts

Rationale: An integral part of assessment of many candidates for epilepsy surgery is the IAP. The IAP yields lateralized language and memory scores that provide both a risk assessment and some prognostic data prior to epilepsy surgery. Our aim is to investigate if mathematical algorithms performed with SEEG data correlate with the memory results of IAP. Methods: All patients who have had SEEG and IAP were retrospectively reviewed. For the IAP, the memory scores were analyzed for degree of memory failure and for symmetry between hemispheres. Scores lower than 50% were considered deficient.  A difference equal or greater than 25% was considered significant asymmetry. For the SEEG data, a 10 minute segment of quiet wakefulness was analyzed for all patients.  Mean spectral power, phase amplitude coupling and phase locking value were calculated for each electrode in the amygdala, the hippocampus, and in the temporal lobe in total. The results were compared to the values for extratemporal structures. The SEEG calculations were performed by an epileptologist blinded to the IAP.  Each value was correlated with IAP memory failure and symmetry between hemispheres to establish convergent validity. Results: To date, there are 9 patients that have had both IAP and SEEG.  In the IAP, 2 patients had memory failure on the right and 4 on the left. Of these, 2 patients had bilateral failure.  A total of 3 patients had significant asymmetry in their IAP. The SEEG calculations have not yet been performed. Conclusions: A computational algorithm using SEEG data has the potential to provide similar prognostic information to the IAP.     Funding: None