DF Sittig, KH Cheung (1992) "A parallel implementation of a multi-state Kalman filtering algorithm to detect ECG arrhytmias", International Journal of Clinical Monitoring and Computing 9 , 13--22  Home/Search   Document Not in Database   Summary   Related Articles   Check

.... a number of alternative models of the input data (trends ) a priori estimates of the likelihood of each model, an estimation of the variance of the input data, the filtering method itself, and a method of choosing the best model based on the derived a posteriori probabilities of each model [Sittig 92a] A parallel implementation of multi state Kalman filters using the process trellis [Factor 92] has been put forward in the ICM project [Sittig 90b, Sittig 92a] It appears that Kalman filtering is particularly good at very early detection of a trend (or, rather, deviation from a trend) a ....

.... data, the filtering method itself, and a method of choosing the best model based on the derived a posteriori probabilities of each model [Sittig 92a] A parallel implementation of multi state Kalman filters using the process trellis [Factor 92] has been put forward in the ICM project [Sittig 90b, Sittig 92a] It appears that Kalman filtering is particularly good at very early detection of a trend (or, rather, deviation from a trend) a property it has in common with more primitive forecasting methods such as cumulative sums [Allen 83, Avent 90] Other trend detection methods such as the ones based ....

[Article contains additional citation context not shown here]

DF Sittig, KH Cheung (1992) "A parallel implementation of a multi-state Kalman filtering algorithm to detect ECG arrhytmias", International Journal of Clinical Monitoring and Computing 9 , 13--22

.... [33] and structural instability in buildings [4] or to optimize more routine tasks such as the control of manufacturing plant processes [18] electric power delivery systems [26] and freeway traffic flow [42] Medical examples of alarms include those that assist in the cardiopulmonary [13, 49] or anesthesia management [14, 50] of critically ill inpatients; medical alarms also faciliate the management of outpatients with disorders such as sleep apnea [5] or diabetes mellitus [30] In economics and finance applications, alarms assist in making timely security investment decisions based ....

Dean F. Sittig and Kei-Hoi Cheung. A parallel implementation of a multistate Kalman filtering algorithm to detect ECG arrhythmias. International Journal of Clinical Monitoring and Computing, 9:13--22, 1992.

....and IRI 9311950: 8] 18] 38] 47] 48] 3 Objectives and Outline of the Proposed Research The need to make decisions in a dynamically changing environment is a recurring problem that spans many disciplines. Medical examples include the management of cardiopulmonary abnormalities [25, 57] or anesthesia settings [26, 63] in critically ill inpatients, as well as the monitoring of outpatients with sleep apnea [12, 59] or diabetes mellitus [2, 44] Engineering applications include monitoring for catastrophes such as earthquakes [45] and structurallyunstable buildings [7] or more ....

Dean F. Sittig and Kei-Hoi Cheung. A parallel implementation of a multi-state kalman filtering algorithm to detect ECG arrhythmias. International Journal of Clinical Monitoring and Computing, 9:13--22, 1992.

....trend detection based on fuzzy courses puts forward a graphical approach, where specification of a trend is intuitive and highly flexible at the same time. Kalman filters and their multi state extensions have also successfully been employed to detect trends in biomedical signals, for example in [7, 16, 26, 29, 30]. It appears that Kalman filters are particularly good at very early detection of a trend (or, rather, deviation from a trend) a property they have in common with more primitive forecasting methods such as cumulative sums [2, 3] This makes them particularly suited for the critical care ....

D.F. Sittig, K.H. Cheung, A parallel implementation of a multi-state Kalman filtering algorithm to detect ECG arrhytmias, International Journal of Clinical Monitoring and Computing 9 (1992) 13--22.

.... to identify temporal developments are mostly rooted in statistics and time series analysis: regression analysis, adaptive forecasting and Kalman filtering have successfully been applied to the interpretation of clinical time series [Allen 83, Gordon 86, Gordon 88, Avent 90, Challis 90, Sittig 90, Sittig 92] However, like the temporal abstraction methods mentioned above these methods also rely on a minimum number of available samples. This is particularly untoward in their prospective employment where a diagnosis must be made after each sample including the first, a situation for which none of the ....

DF Sittig, KH Cheung (1992) "A parallel implementation of a multi-state Kalman filtering algorithm to detect ECG arrhytmias", International Journal of Clinical Monitoring and Computing 9 , 13--22

....by introducing probability distributions or fuzzy sets. DIAMON 1 in turn would profit from explicitly and separately maintaining alternative hypotheses instead of tracking them all with one automaton. Sittig et al. have worked on the detection of trends using Kalman filtering and linear regression [5, 6]. However, based on statistical models of the problem domain these approaches lead to results with an entirely different meaning [3] DYNASCENE [1] an earlier project in which Kalman filtering was planned to be integrated, also implemented states called clinical scenes . It is DIAMON 1s most ....

Sittig, D.F. and Cheung, K.-H. (1992) "A Parallel Implementation of a Multi-State Kalman Filtering Algorithm to Detect ECG Arrhytmias", Int J Clin Mon Comp 9 13--22

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC