Magnetoencephalographic Recordings in Infants Using a Standard-Sized Array: Technical Adequacy and Diagnostic Yield.

Abstract

Although a few magnetoencephalography (MEG) systems have been built specifically for infants (<2 years), most children are routinely examined using MEG systems designed primarily to accommodate adult subjects. The practicality of the adult MEG device has not been systematically reviewed in infants with epilepsy. The aim of this study is to investigate whether infant epilepsy patients, whose heads are small relative to the sensor helmet, can have clinically successful MEG recordings using the conventional adult-size MEG device. We reviewed our database of 964 patients who were referred for routine MEG epilepsy examination between January 2008 and May 2015. We reanalyzed the original data of the infant patients to investigate whether epileptiform discharges of the usual amplitudes could be detected. Nine patients (about 1%) were infants. All the infant MEG records were processed with temporally extended signal space separation to remove environmental artifacts. Seven of the nine were processed with movement compensation processing using continuous head position monitoring. Magnetoencephalography captured epileptic abnormalities in all the infants: interictal in nine and ictal in five. The source amplitudes of 10 representative interictal dipoles from each of the infants ranged from 46.0 to 437.2 (mean: 180.1, SD: ±92.5) nA-m. Despite increased brain to sensor distance, the conventional MEG device can detect epileptic abnormalities in infants, facilitated by monitoring head position and postprocessing with temporally extended signal space separation, with or without movement compensation. Magnetoencephalography is useful for managing infants with epilepsy using the conventional adult machine, without special adaptations for small heads.