EEG Data Analysis Of Memory Impairment At Rest And Cognitive States

Progressive memory loss is considered one of the earliest signs ofAlzheimer's disease(AD). AD has high morbidity rate, it's treatment andearly diagnosis is very difficult. A wealth of evidence now suggests that:in the early stage of AD, amyloidβprotein interrupt the function ofsynapses that help encode new declarative memories. This stage maybe themost meaningful to intervent AD progress. In this study, we try to findthe cerebral function changes at different level of memory decline.In thisstudy, 12 AD, 11 memory impairment (MI) and 11 normal subjectswere involved. All the subjects underwent EEG recording both in rest (eyeclosed) and cognitive load states. The EEG data was analyzed by spectralestimate based on FFT transform and Chaos analysis based on the wavelettransform. Then, all the subjects were evaluated byneuropsychological(MMSE and SECF). At last, we analysis the relationshipbetween EEG analysis and neuropsychological results.The result shows thatγband relative power percentage increasedfrom rest to cognitive state in normal group, but declined in MI group(P＜0.05). The Dc value of AD group were lower compared with that in normalgroup (P＜0.05). In rest state, the Dc value in high frequency increasedin many areasln normal group compared with that in cognitive performancestate from rest to cognitive state (P＜0.05). But in MI group fewer areasshowed this result. As to neuropsychological test, the short time recallsubtest played a good role to distinguish MI from normal. Our conclusionis that the subtle changes of memory function in the progress of AD canbe perceived by EEG, especially by cognitive performing EEG. Furthermore,EEG components of wavelet transform could give more information on thedifference among the different memory condition.