显著性概率地形图的一些基础问题探讨——少年组谱参量的正态分布转换问题

本文对41例年龄7—12岁右利健康儿童,头部选取5对(10个)样点,对各样点的功率和相对功率Ⅰ,Ⅱ的四个经典频带(δ,θ,α,β)进行几种数学转换,用正态性D检验法观察各参量转换前后的分布状态。结果,转换前,功率分布状态较差;而相对功率的参量原值近正态分布的情况较好,其中相对功率Ⅰ较相对功率Ⅱ更佳。而相对功率Ⅰ的X、log(X)和log(X/1—X)转换为最佳。就功率而言,log(x)转换最好,但仍不能完全满足各频带统计学的要求。

STUDY ON SOME BASIC ISSUES OF SIGNIFICANCE PROBABILITY MAPPING (SPM)-transformations towards the normal distribution of classical band spectral parameters of the EEG in a group of healthy young individuals

The transformations were evaluated by performing D test for normality on four classical band spectral parameters of absolute band power & relative band power I , II, before and after transformation in 41 healthy right-handed volunteers aged 7~12yrs at 5. pairs (10 representative) locations (F3, F4, C3, C4, P3, P4, O1, O2, T3, T4). The results were as follows; Befroe transformation, the distributive situation of power was not so good for absolute powre, but not so bad for relative power (Tab1, 2). The relative power 1 was better than that n , For relative power 1. the transformations of X, Log (x) & log (x/1-x) were almost perfect. Whereas the best transformation for absolute power-log (x) -was not fully satisfactory for all bands.