Generalized linear least squares algorithm for non-uniformly sampled biomedical system identification with possible repeated eigenvalues.

The recently developed generalized linear least squares (GLLS) algorithm has been found very useful in non-uniformly sampled biomedical signal processing and parameter estimation. However, the current version of the algorithm cannot deal with signals and systems containing repeated eigenvalues. In this paper, we extend the algorithm, so that it can be used for non-uniformly sampled signals and systems with/without repeated eigenvalues. The related theory and detailed derivation of the algorithm are given. A case study is presented, which demonstrates that the extended algorithm can provide more choices for system identification and is able to select the most suitable model for the system from the non-uniformly sampled noisy signal.

[1]  H. Akaike.  A Bayesian analysis of the minimum AIC procedure , 1978 .

[2]  James Hill,et al.  Sampling and data analysis properties of Prony's method applied to model identification† , 1983 .

[3]  Dagan Feng,et al.  An unbiased parametric imaging algorithm for nonuniformly sampled biomedical system parameter estimation , 1996, IEEE Trans. Medical Imaging.

[4]  G. Schwarz.  Estimating the Dimension of a Model , 1978 .

[5]  Zhizhong Wang,et al.  A study on statistically reliable and computationally efficient algorithms for generating local cerebral blood flow parametric images with positron emission tomography , 1993, IEEE Trans. Medical Imaging.

[6]  J. DiStefano,et al.  Optimal nonuniform sampling interval and test-input design for identification of physiological systems from very limited data , 1979 .

[7]  H. Unbehauen,et al.  Identification of continuous systems , 1987 .

[8]  J J DiStefano,et al.  Multiexponential, multicompartmental, and noncompartmental modeling. II. Data analysis and statistical considerations. , 1984, The American journal of physiology.

[9]  David Dagan Feng,et al.  An evaluation of the algorithms for determining local cerebral metabolic rates of glucose using positron emission tomography dynamic data , 1995, IEEE Trans. Medical Imaging.

[10]  C J Thompson,et al.  Dynamic Positron Emission Tomography for Study of Cerebral Hemodynamics in a Cross Section of the Head Using Positron‐Emitting 68Ga‐EDTA and 77Kr , 1977, Journal of computer assisted tomography.

[11]  Lennart Ljung,et al.  The Extended Kalman Filter as a Parameter Estimator for Linear Systems , 1979 .

[12]  Karl Johan Åström.  Maximum likelihood and prediction error methods , 1980, Autom..

[13]  Katsuhiko Ogata,et al.  Discrete-time control systems , 1987 .

[14]  A. Koehler,et al.  A Comparison of the Akaike and Schwarz Criteria for Selecting Model Order , 1988 .

[15]  Joseph J. DiStefano,et al.  Hidden pools, hidden modes, and visible repeated eigenvalues in compartmental models , 1986 .

[16]  H. Akaike.  A new look at the statistical model identification , 1974 .