Acquiring, modeling, and predicting intracranial pressure in the intensive care unit

被引：0

作者：

Tsui, Fu-Chiang ^{[1
]}

Li, Ching-Chung ^{[1
]}

Sun, Mingui ^{[1
]}

Sclabassi, Robert J. ^{[1
]}

机构：

[1] Univ of Pittsburgh, Pittsburgh, United States

来源：

Biomedical Engineering - Applications, Basis and Communications | 1996年 / 8卷 / 06期

关键词：

Adaptive algorithms - Data acquisition - Feature extraction - Intensive care units - Least squares approximations - Neural networks - Neurosurgery - Signal filtering and prediction - Wavelet transforms;

D O I：

暂无

中图分类号：

学科分类号：

摘要：

Signal processing in the intensive care unit (ICU) is discussed in this paper. The data-acquisition layout for the computerized ICU is briefly reviewed. The prediction of intracranial pressure (ICP) is investigated based on artificial neural networks (ANNs). The biorthogonal wavelet transform is employed to extract features from ICP signals to facilitate long-term prediction by using wavelet/scaling coefficients at a coarse resolution as the input and the output of a neural network. An adaptive algorithm is utilized in the ANN for updating and prediction of the ICP signal. A comparison between the ANN predictor and the least-meansquare (LMS) adaptive predictor is studied. The results demonstrate that the ANN predictor has better performance than the linear LMS predictor.

引用

页码：566 / 578