AR model orders of lung sounds

Earlier studies of our group have demonstrated that autoregressive (AR) models clearly depend on flow and volume. In those studies the AR model order was fixed at 6. An evaluation has been made of the applicability of seven order selection criteria. The Finite sample Information Criterion (FIC) appears to be the best choice. The results also indicate that the true model order lies between 2 and 6. However, in general there are no major decreases in criterion values above order 3. This implies that it is justified to take a fixed order of 3 or 4.

[1]  J. Vanderschoot,et al.  Flow And Volume Related AR-modelling Of Lung Sounds , 1991, Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society Volume 13: 1991.

[2]  Piet M. T. Broersen,et al.  The prediction error of autoregressive small sample models , 1990, IEEE Trans. Acoust. Speech Signal Process..

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

[4]  S.M. Kay,et al.  Spectrum analysis—A modern perspective , 1981, Proceedings of the IEEE.

[5]  J. Rissanen,et al.  Modeling By Shortest Data Description* , 1978, Autom..