COMMUNICATIONS from (15). The results of the deconvolution are sensitive to the values of {pi} and {vi} and consequently they must be accurately transmitted. CONCLUSIONS Two signals which have been convolved can be separated by the technique outlined in this paper provided that one of the signals is modeled as a sample function from a cyclostationary random process with known or calculable statistics and the other signal is the impulse response of an unknown linear time-invariant system. Since stereotype ECG's can be modeled as sample functions from a CS process, it is conceivable that the technique can be applied to practical ECG's which have been linearly distorted provided an adequate test has been performed to elucidate the parameters required by (15). Results presented show the power of the SDA and the speed with which substantial distortion, even that caused by non-minimum phase systems, can be removed. Abstract-This paper describes a method for the computer reconstruction of surfaces from a sequence of electron micrographs, and a data structuring approach to the problem of representing and analyzing objects of physiological importance. The reconstruction technique involves the following stages: 1) object outlines are traced from each section, 2) the computer chain encodes these outlines, 3) the chain codes are reduced to the minimum number of boundary points which satisfactorily define the boundary, 4) polygons are mapped onto the boundary points between sections to approximate the surface, and 4) color coded, shaded surface views are computed of any subset of objects viewed and illuminated from arbitrary locations. This surface representation provides an ultrastructural data base from which quantitative morphological parameters as well as otherwise impossible visualizations can be computed. Important parameters include surface area, volume, areas of close proximity between surfaces, synaptic area, cross-sectional area, synaptic distribution, and others. This sort of analysis facilitates detailed quantitative correlation of ultrastructure and function in neuronal systems.

A particular time‐varying network consisting of several parallel transmission paths, each containing input and output modulators, is described and analyzed. It is shown that, under certain conditions, the network may be characterized by a transfer function. A particular form of this transfer function yields periodic filtering characteristics over a limited frequency band without employing distributed elements. Techniques are also presented for realizing highly selective band‐pass filters without the use of magnetic elements. Some practical applications are discussed in detail and experimental verification is presented.

No abstract

No abstract

No abstract