An Approach to Efficient On-line Mass Storage of Physiological Parameters in an Intensive Care Environment
From the data processing point of view, monitoring critically ill patients by computer is primarily a problem of reduction of mass data.
This paper describes a way of reducing data for storage purposes which results in a minimum loss of accuracy. The reduction factor is 1:5 on an average for fast varying parameters and up to 1:100 for slowly varying parameters.
Our approach is to store trend information rather than averaged data. The algorithm transforms a continuous input of data in equidistant time intervals to an output in trend format. High compression on quasi-static data input and a fast reaction in case of relevant changes are the main characteristics.
This algorithm is part of a data path in a decentralized minicomputer system for intensive care, the Göttingen Information System for Intensive Care “GISI”.
Unable to display preview. Download preview PDF.
- (1).Drazen, E., Wechsler, A., Wiig, K.: Requirements for computerized patient monitoring systems Computer 1 (1975) 22–26Google Scholar
- (2).Sabean, R.M., Scharnberg, B., Schillings, H. Göttingen information system for intensive care “GISI”. To appear in: Proc. Cönf. Computers in Cardiology Rotterdam 1977. New York: IEEE Computer Society.Google Scholar
- (3).Scharnberg, B., Sabean, R.M., Schillings, H. A data base for intensive care units. In.: Shires, D.B., Wolf, H. (Eds): MEDINFO 77, Vol. 2 p. 1063 Amsterdam: North Holland Publ. Co., 1977Google Scholar