小波分析在早老性痴呆脑电诊断中的应用

AD EEG diagnosis using wavelet analysis

运用小波变换(WT)对老年痴呆(AD)患者与正常老人在安静闭眼时的脑电(EEG)数据进行了多尺度空间能量分布特征提取。初步比较分析表明,与正常人相比,AD患者EEG能量在中高频段α(8～16Hz)显著较低(严重时可低至正常人的5%),而在低频段δ(0～4Hz)的能量却明显偏高(可达正常人的几倍到几十倍)。研究结果提示,可考虑试用小波分析提取脑电能谱分布特征作为AD临床工程诊断的特异性指标。

Wavelet transform (WT) technique was applied to EEG of Alzheimer's Disease (AD) patients and health adult to extract the features of multi-scale space energy distribution. Preliminary analysis results show that the distribution feature of EEG energy is characterized by the evident lack (lower by 95%) of a dominant activity in the α band (8～16Hz) and higher energy (more by several times) in the δ band (0～4 Hz) for most AD patients compared with that of health adults. It's considerable to use WT technique to extract the feature of EEG energy distribution to diagnose AD.