Diagnosis value of EEG power spectrum changes to cerebral vascular stenosis in patients with MCI

Xiaohong Zhao,Shuqin Zhan
DOI: https://doi.org/10.3969/j.issn.1004-0188.2015.03.002
2015-01-01
Abstract:Objective To explore the diagnosis value of electroencephalogram (EEG) power spectrum changes to cerebral vascular stenosis ( CVS) in patients with mild cognitive impairment ( MCI) .Methods Total 100 patients with MCI were divided into non-CVS group, mild CVS group, and severe CVS group;the conventional EEG of all patients were recorded , and the relative power ofα,θ, and δfrequency range and α/θvalue of the three groups were compared and analyzed;EEG power spectrum analysis was made in order to diagnose the severity of CVS .Results The values of relative power and α/θ in frequency range αdecreased with the severity of CVS, and those in frequency ranges θandδincreased with the severity of CVS;the coincidence rate of EEG analysis with the diagnosis of non-CVS, mild CVS, and severe CVS was 87.18%, 81.25%, and 96.55%, respectively, with an overall coincidence rate of 88.00%, so the coincidence rate with diagnosis is good (kappa=0.819, P<0.05).The area under the ROC curve to diagnose whether there was CVS was 0.905 (P<0.05).Conclusion The change of the frequency range of α,θ, andδandα/θvalue may be taken as an observation index indicating the severity of CVS;the EEG power spectrum changes has good diagnosis value to CVS in patients with MCI .
What problem does this paper attempt to address?