| The acoustic research of voice often focuses on the stable segment of a long vowel sound trying to analysis some acoustic parameters, such as NNE, Jitter, Shimmer. In this study, we are going to do some researches on starting stage of the voice using the new glottal parameters:OQ,SQ,STF,AQ getting by a new inverse filtering technology. We also do a case study of patient with voice disorders.The content and results of the acoustic characteristics on pathological voice's starting stage are divided into the following areas:1,A study on materials of the acoustic parameters test in the starting stage. We have two groups of subjects. First group is normal voice group, we get 38 subjects from 2010,2011 undergraduates and graduate students of East China Normal University by random extraction, including 20 boys and 18 girls. Second group is abnormal voice group. We get 52 subjects including 23 male and 29 female with laryngitis, vocal cord polyps, vocal nodules, vocal cord paralysis from People's Hospital Haimen, Jiangsu Province and Shanghai Yueyang Hospital. We get the data from their standardized recording and do the analysis by four experiments for 4 parameters. We use the the 2*2*4 three-factor experiment design by four times. Results:(1) The language materials all have significant different on the parameters. And there are no significant interaction between language materials. (2) The language material/(?)/is extremely significant different on OQ. This indicates that/(?)/is very sensitive to identify the subjects of different voice. The other three materials/i/,/u/,/(?)/are not significant on those four parameters. So, they have little sensitiveness on identifying the subjects different voice type.2,The analysis of objective parameters on voice starting stage of different voice disease. We choose 52 subjects with laryngitis, vocal cord polyps, vocal nodules, vocal cord paralysis from People's Hospital Haimen,Jiangsu Province and Shanghai Yueyang Hospital, including 23 male and 29 female to do the experiment. This is a 4*2 two-factor experimental design. The first independent variable is the type of disease. It is a between subjects variable which has 4 levels. They are laryngitis, vocal cord polyps, vocal nodules and vocal cord paralysis. The second independent variable is gender. It is also a between subjects variable which has 2 levels:male and female. The dependent variables is OQ. Result:In differentiating four types of voice disorders: laryngitis, vocal cord polyps, vocal nodules, vocal cord paralysis, the parameter OQ has no significant difference. This shows that acoustic parameters of OQ can not identify different categories of disease. |