UC Irvine

High frequency oscillations (HFOs) are a promising new biomarker of epileptogenicity, as they occur more frequently in the seizure onset zone (SOZ) and may aid in the demarcation of the epileptogentic zone. Development of reliable, automatic HFO detection algorithms is necessary for translation into clinical practice. While existing algorithms have demonstrated sufficient levels of sensitivity and specificity on individual data sets, there are currently no standards for their broad application. It is not uncommon for a previously validated algorithm to work poorly when applied to a new data set, and there is no consensus on whether (and how) parameter optimization should be done. Here we evaluate the impact of detector optimization on two independent datasets, consisting of twenty medically refractory epilepsy patients with seizure free surgical outcomes, using a widely cited automatic HFO detector based on the root-mean-square amplitude. We calculated SOZ localization results over a wide range of detection parameters and assessed the variance in results across patients. The optimal parameters were patient-specific, and in some cases, the most accurate localization resulted from detection with unconventional parameters. This suggests that the standard configurations are not suited for all patients. To overcome this obstacle, we suggest a novel method of coalescing the results from multiple parameter sets to isolate robust HFOs of epileptic tissue. This method resulted in localization accuracy that was comparable to the optimal parameter sets, without the difficult task of choosing a single parameter set to rely on. This work has the potential to eliminate per-patient optimization of HFO detection, which will support translation into clinical practice, and it suggests that future studies should continue investigating ways to address patient variability before applying automatic detection algorithms.