论文部分内容阅读
早期的研究表明,舒张期的心音包含着检测冠状动脉闭锁的有用信息。在本研究中,采用了包含自适应的递归最小平方格构(recursive least—square lattice—RLSL)和梯度格构(gradient lattice—GAL)预测器在内的自回归法(autoregressive—AR)对舒张期心音段的记录建立模型。Akaike标准的使用给出了要完全描述一个心舒张期需要5～15个自回归系数才行。反射系数、预测系数、反向滤波器多项式
Earlier studies have shown that diastolic heart sounds contain useful information to detect coronary atresia. In our study, autoregressive-AR with adaptive recursive least-square lattice (RLSL) and gradient lattice-GAL predictor was used to predict diastolic Period of heart sound recordings to establish models. The use of the Akaike standard gives us 5 to 15 autoregressive coefficients to fully describe a diastolic period. Reflection coefficient, prediction coefficient, inverse filter polynomial