An evaluation of kurtosis beamforming in magnetoencephalography to localize the epileptogenic zone in drug resistant epilepsy patients

Clin Neurophysiol. 2018 Jun;129(6):1221-1229. doi: 10.1016/j.clinph.2017.12.040. Epub 2018 Mar 9.

Abstract

Objective: Kurtosis beamforming is a useful technique for analysing magnetoencephalograpy (MEG) data containing epileptic spikes. However, the implementation varies and few studies measure concordance with subsequently resected areas. We evaluated kurtosis beamforming as a means of localizing spikes in drug-resistant epilepsy patients.

Methods: We retrospectively applied kurtosis beamforming to MEG recordings of 22 epilepsy patients that had previously been analysed using equivalent current dipole (ECD) fitting. Virtual electrodes were placed in the kurtosis volumetric peaks and visually inspected to select a candidate source. The candidate sources were compared to the ECD localizations and resection areas.

Results: The kurtosis beamformer produced interpretable localizations in 18/22 patients, of which the candidate source coincided with the resection lobe in 9/13 seizure-free patients and in 3/5 patients with persistent seizures. The sublobar accuracy of the kurtosis beamformer with respect to the resection zone was higher than ECD (56% and 50%, respectively), however, ECD resulted in a higher lobar accuracy (75%, 67%).

Conclusions: Kurtosis beamforming may provide additional value when spikes are not clearly discernible on the sensors and support ECD localizations when dipoles are scattered.

Significance: Kurtosis beamforming should be integrated with existing clinical protocols to assist in localizing the epileptogenic zone.

Keywords: Beamforming; Epilepsy; Kurtosis; MEG; Neuroimaging.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Brain / diagnostic imaging*
  • Brain / surgery
  • Brain Mapping
  • Drug Resistant Epilepsy / diagnostic imaging*
  • Drug Resistant Epilepsy / surgery
  • Humans
  • Magnetoencephalography
  • Male
  • Middle Aged
  • Neuroimaging / methods
  • Retrospective Studies
  • Seizures / diagnostic imaging*
  • Seizures / surgery
  • Young Adult