Prefiltering Of Data Research Articles

Review, Discussion, and Summary of Epidemiological Studies

This paper reviews and summarizes the epidemiological studies presented at the Symposium on the Health Effects of Acid Aerosols. Two studies of acute episodes examined different indicators of respiratory morbidity before, during, and after the January 1985 air pollution event in western Europe. In the U.K. no increase in respiratory morbidity, as reported by a group of general practitioners, was observed, but measured concentrations of air pollutants failed to substantiate the existence of an identifiable episode. In the Federal Republic of Germany, the air pollution episode was documented and was associated with a 10 to 25% increase in several indicators of respiratory and cardiovascular morbidity, but could not be attributed to acidic aerosols as such. In two further studies, investigators related day-to-day variations in air pollution with admissions to acute care hospitals in southern Ontario for respiratory disease over a 9-year period, and with daily mortality in London from 1963 to 1972. In the study of hospital admissions, significant correlations were observed with sulfate, ozone, and SO2 pollution, but the data were insufficient to isolate the separate or combined effects of these pollutants. In the London mortality analysis, the strongest correlations were observed for sulfuric acid levels of the prior day, but prefiltering of the mortality data may have dampened the true relationship, and age- and cause-specific analyses would have been desirable. Finally two reports on chronic effects of residence in high air pollution areas have thus far made little contribution to the evidence for an adverse effect of specific pollutants.

Review, discussion, and summary of epidemiological studies.

This paper reviews and summarizes the epidemiological studies presented at the Symposium on the Health Effects of Acid Aerosols. Two studies of acute episodes examined different indicators of respiratory morbidity before, during, and after the January 1985 air pollution event in western Europe. In the U.K. no increase in respiratory morbidity, as reported by a group of general practitioners, was observed, but measured concentrations of air pollutants failed to substantiate the existence of an identifiable episode. In the Federal Republic of Germany, the air pollution episode was documented and was associated with a 10 to 25% increase in several indicators of respiratory and cardiovascular morbidity, but could not be attributed to acidic aerosols as such. In two further studies, investigators related day-to-day variations in air pollution with admissions to acute care hospitals in southern Ontario for respiratory disease over a 9-year period, and with daily mortality in London from 1963 to 1972. In the study of hospital admissions, significant correlations were observed with sulfate, ozone, and SO2 pollution, but the data were insufficient to isolate the separate or combined effects of these pollutants. In the London mortality analysis, the strongest correlations were observed for sulfuric acid levels of the prior day, but prefiltering of the mortality data may have dampened the true relationship, and age- and cause-specific analyses would have been desirable. Finally two reports on chronic effects of residence in high air pollution areas have thus far made little contribution to the evidence for an adverse effect of specific pollutants.

Instrumental Variable Methods for Closed Loop Systems

Instrumental variable (IV) methods are normally designed for systems operating in open loop. However, if there is a measurable external signal like a setpoint variation, then IV methods can be constructed also for systems operating under feedback control. Some ways to do this are described. Recent theory of instrumental variable methods shows that the estimates are asymptotically Gaussian distributed. The instruments and data prefilters influence the covariance matrix of the estimates. They can in particular be chosen so that an optimal accuracy is obtained. In the paper this result is extended to the case of closed loop operation. It is also described how an optimal IV method can be designed on a min-max basis. Minimization is then with respect to the instruments and maximization with respect to the disturbance properties. Some simulated examples are included as well for illustration of the theoretical results.

Optimal instrumental variable estimation and approximate implementations

The accuracy properties of instrumental variables (IV) methods are investigated. Extensions such as prefiltering of data and use of additional instruments are included in the analysis. The parameter estimates are shown to be asymptotically Gaussian distributed. An explicit expression is given for the covariance matrix of their distribution. The covariance matrix is then taken as a (multivariable) measure of accuracy. It is shown how it can be optimized by an appropriate selection of instruments and prefilter. The so obtained optimal instrumental variable estimates cannot be used directly since the true system and the statistical properties of the disturbance must be known in order to compute the optimal instruments and prefilters. A multistep procedure consisting of three or four simple steps is then proposed as a way to overcome this difficulty. This procedure includes modeling of the disturbance as an ARMA process using a statistically efficient method such as a prediction error method. The statistical properties of the estimates obtained with the multistep procedure are also analyzed. Those estimates are shown to be asymptotically Gaussian distributed as well. The covariance matrix of the estimation errors is compared to that corresponding to a prediction error method. For some model structures these two approaches give the same asymptotic accuracy. The conclusion is that the multistep procedure, which is quite easy to implement and also has nice uniqueness properties, can be viewed as an interesting alternative to prediction error methods.

Application of a non-bayesian approach to computer-aided diagnosis for upper abdominal pain

Two non-Bayesian predictor methods of automatic diagnosis were applied to a data base of 300 gastro-enterological patients known to have one of six diseases determined by radiological diagnoses. It was found that the efficiency of the automatic diagnosis may be as high as 92 per cent. The calculations involved are relatively simple in that the major computation is inversion of a matrix whose coefficients depend on the symptoms and diseases of a previous set of patients. Each subsequent automatic diagnosis requires only a summation of appropriate terms. It was found that some pre-filtering of data might prove useful in increasing diagnosis efficiency.

A PDP-8 computer control system

We describe a control system assembling a data acquisition unity on-line with a PDP-8 computer as feedback element of external devices. The acquisition is performed by means of a matrix ordering; some “status words” are provided for a data prefiltering.