A novel high data compression of ECG data in the measurement space by clipping the signal and using the zero-crossing intervals as features. This yields low-dimensional features sufficient and efficient for screening and some diagnostics. It is validated by experiment using ECG of an in vitro heart of a rat. This new idea is built around a Z-80 microprocessor.
Get full access to this article
View all access options for this article.
References
1.
BallG. H.HallD. J. (1965) ISODATA, a novel method of data analysis and pattern classification. NTIS Report 699616, Institute of Menlo Park, California.
2.
GonzalezR. C.HowingtonL. C. (1976) Machine recognition of abnormal behaviour in nuclear reactors. PATREC, IEEE 3rd International Conference, Coronado, CA, pp 8–16.
3.
HenryR. M. (1979) An improved algorithm allowing fast on-line polarity correlation by microprocessor or mini computer. IEEE Colloquium on Correlation Processing, London, May 1974, pp 3/1–3/4.
4.
IbiyemiT. S. (1982) Zero-crossing pattern monitoring of continuous periodic wareforms. PhD thesis, University of Bradford.
5.
KittlerJ.YoungP. C. (1973) A new approach to feature selection based on the Karhunen-Loeve expansion, Pattern Recognition, 5, 335.
6.
PietyK. R.RobinsonJ. C. (1976) An on-line reactor surveillance algorithm based on multivariate analysis of noise, Nucl. Sci. Engng, 59, 369–380.
