Electrical source imaging of sleep spindles

Clin EEG Neurosci. 2014 Jul;45(3):184-92. doi: 10.1177/1550059413497716.

Abstract

To identify and compare cortical source generators of slow and fast sleep spindles in healthy subjects, electroencephalographic (EEG) signals were obtained from 256 channels, and sources on neuroanatomical Montreal Neurological Institute (MNI) space estimated with low-resolution brain electromagnetic tomography analysis (LORETA). Spindle activity was recorded in 18 healthy volunteers during daytime napping. Because of lack of sleep or excessive artifacts, data from 13 subjects were analyzed off-line. Spindles were visually scored, marked, and bandpass filtered (slow 10-12 Hz or fast 12-14 Hz). EEG was segmented on the marker, and segments separately averaged. LORETA projected cortical sources on the MNI brain. Maximal intra- and inter-individual intensities were compared using the Wilcoxon test (P < .05) and cortical sources distribution compared using a chi2 test. Two to three slow spindles generators were consistently identified in frontal lobes, with additional sources in parietal and limbic lobes in half cases. Fast spindles had multiple temporo-parietal sources, with an inconstant frontal source. Inter-individual (P = 0.44), and intra-individual (P = 0.09 slow and P = 0.10 fast spindles) source intensities were comparable. Slow spindles sources were preferentially concentrated over frontal cortices in comparison with fast spindles (P = 0.0009). Our results demonstrate multiple, synchronous, and equipotent spindles cortical generators in healthy subjects, with more anterior generators for slow spindles.

MeSH terms

  • Adult
  • Brain Mapping / methods*
  • Cerebral Cortex / physiology*
  • Electroencephalography / methods*
  • Female
  • Frontal Lobe / physiology
  • Humans
  • Magnetoencephalography / methods*
  • Male
  • Middle Aged
  • Observer Variation
  • Reference Values
  • Signal Processing, Computer-Assisted*
  • Sleep Stages / physiology*
  • Wavelet Analysis*