Quantifying interhemispheric symmetry of somatosensory evoked potentials with the intraclass correlation coefficient

J Clin Neurophysiol. 2008 Jun;25(3):139-46. doi: 10.1097/WNP.0b013e31817759e2.

Abstract

Although large intersubject variability is reported for cortical somatosensory evoked potentials (SEPs), variability between hemispheres within one subject is thought to be small. Therefore, interhemispheric comparison of SEP waveforms might be clinically useful to detect unilateral abnormalities in cortical sensory processing. We developed and evaluated a new technique to quantify interhemispheric SEP symmetry that uses a time interval including multiple SEP components, measures similarity of SEP waveforms between both hemispheres and results in high symmetry values even in the presence of small interhemispheric anatomic differences. Median nerve SEPs were recorded in 50 healthy subjects (20-70 years) using 128-channel EEG. Symmetry was quantified by the intraclass correlation coefficient and correlation coefficient between global field power of left and right median nerve SEPs. In 74% of subjects left-right intraclass correlation coefficient was higher than 0.60, implying high SEP hemispheric symmetry in terms of shape and amplitude. Left-right intraclass correlation coefficients lower than 0.60 were due to differences in amplitude, unilateral absence of peaks, or shape differences. We quantified SEP waveform interhemispheric symmetry and found it to be high in most healthy subjects. This technique may therefore be useful for detection of unilateral abnormalities in cortical sensory processing.

MeSH terms

  • Adult
  • Aged
  • Algorithms*
  • Brain Mapping / methods*
  • Diagnosis, Computer-Assisted / methods*
  • Dominance, Cerebral / physiology*
  • Electroencephalography / methods*
  • Evoked Potentials, Somatosensory / physiology*
  • Female
  • Humans
  • Male
  • Middle Aged
  • Pattern Recognition, Automated / methods
  • Statistics as Topic