Analysis of the magnetoencephalograrn background activity in Alzheimer's disease patients with auto-mutual information

Carlos Gomez*, Roberto Hornero, Daniel Abasolo, Alberto Fernandez, Javier Escudero

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract / Description of output

The aim of the present study was to analyse the magnetoencephalogram (MEG) background activity in patients with Alzheimer's disease (AD), one of the most frequent disorders among elderly population. For this pilot study, we recorded the MEGs with a 148-channel wholehead magnetometer in 20 patients with probable AD and 21 age-matched control subjects. Artefact-free epochs of 3392 samples were analysed with auto-mutual information (AMI). Average AMI decline rates were lower for the AD patients' recordings than for control subjects' ones. Statistically significant differences were found using a Student's t-test (p <0.01) in 144 channels. Mean AMI values were analysed with a receiver operating characteristic curve. Sensitivity, specificity and accuracy values of 75%, 90.5% and 82.9% were obtained. Our results show that AMI estimations of the magnetic brain activity are different in both groups, hence indicating an abnormal type of dynamics associated with AD. This study suggests that AMI might help medical doctors in the diagnosis of the disease. (C) 2007 Elsevier Ireland Ltd. All rights reserved.

Original languageEnglish
Pages (from-to)239-247
Number of pages9
JournalComputer methods and programs in biomedicine
Volume87
Issue number3
DOIs
Publication statusPublished - Sept 2007

Keywords / Materials (for Non-textual outputs)

  • Alzheimer's disease
  • magnetoencephalogram
  • auto-mutual information
  • non-linear analysis
  • SCHIZOPHRENIC-PATIENTS
  • EMBEDDING DIMENSION
  • TIME-SERIES
  • EEG
  • COMPLEXITY
  • ENTROPY
  • DEMENTIA
  • STATE
  • ELECTROENCEPHALOGRAMS
  • REGULARITY

Fingerprint

Dive into the research topics of 'Analysis of the magnetoencephalograrn background activity in Alzheimer's disease patients with auto-mutual information'. Together they form a unique fingerprint.

Cite this