摘要
本文应用自回归(AR)模型方法于异常呼吸音的分析,AR模型阶次由阿凯克信息准则确定。实验结果表明:间断性音的功率谱是连续谱,而连续性音的功率谱为离散线状谱。
In our study, the abnormal breath sounds were analysed by means fo autoregressive model, and the AR model order of abnormal breath sounds were computed by AKaike information theoretic criterion(AIC). The experimental results indicate that the power spectra of discontinuous breath sounds are continuous spectra, while the power spectra of continuous breath sounds are discrete line spectra.
出处
《生物医学工程学杂志》
EI
CAS
CSCD
北大核心
1996年第1期42-46,共5页
Journal of Biomedical Engineering
关键词
自回归模型
信号处理
功率谱
异常呼吸音
Autoregressive model Signal processing Power spectrum Abnormal breath sounds