ERROR-ESTIMATES FOR RESULTS OF NONSTATIONARY NOISE-ANALYSIS DERIVED WITH LINEAR LEAST-SQUARES METHODS

Nonstationary noise analysis of electrophysiological data is applied t o the estimation of the single-channel current, i, and the number of a ctive channels, N-c, whenever they cannot be determined directly due t o limited resolution. Using least squares methods, the accuracy of est imating i and N-c chiefly depends on the statistical error of the ense mble variance. It is shown that if the correlation among the binned da ta points is taken into account correctly, the variability of i and N- c can be remarkably reduced and exact confidence limits of the paramet ers can be calculated. Least-squares methods are introduced which cons ider the measured error-covariance matrix of the binned variance in a model-independent fashion. Employing Monte Carlo methods, it is demons trated that both the error predictions and the confidence limits are c orrect. The method is used to investigate the performance of nonstatio nary noise analysis at low channel open-probabilities. The application of the approach to simulated data as well as to experimental, i.e. no n-ideal, data is discussed. (C) 1997 Elsevier Science B.V.