A methodology for modeling the distributions of medical images and their stochastic properties

The probabilistic distribution properties of a set of medical images are studied. It is shown that the generalized Gaussian function provides a good approximation to the distribution of AP chest radiographs. Based on this result and a goodness-of-fit test, a generalized Gaussian autoregressive model (GGAR) is proposed. Its properties and limitations are also discussed. It is expected that the GGAR model will be useful in describing the stochastic characteristics of some classes of medical images and in image data compression and other applications.     

Composition of acid tars from production of oils

Translated from Khimiya i Tekhnologiya Topliv i Masel, No. 2, pp. 32–35, February, 1990.     

Creation of progressive technology for production of lubricating oils

Translated from Khimiya i Tekhnologiya Topliv i Masel, No. 5, pp. 17–20, May, 1990.     

Training for an agricultural discrimination task

Skeleton weed, centaurea juncea, is a declared weed in Western Australia because it competes with grain crops for nutrients and moisture. When it is found during harvesting, mechanised teams search and eradicate it. In an earlier report of field trials (Hartley et al, 1989) it was reported that search teams' detection rate was poor and since search teams had usually never seen skeleton weeds, visual discrimination learning was to be expected and observed during searches. The present study investigated the nature of this discrimination learning in a laboratory by developing a training programme of colour photographic slides of weeds in stubble. Subjects receiving specific training with feedback on their performance compared with those receiving pseudo-training showed a significant improvement in detections. Subsequently the benefit of the programme was validated in a field trial.     

An accumulating interpreter for cognitive vision production systems

Knowledge-based recognition and analysis of high dimensional data such as aerial images often has high computational complexity. For most applications time and computational resources such as memory are limited. Therefore approximately correct interpreters with any-time capability are proposed. In this contribution a special software architecture is published, which can handle the administration of complex knowledge-based recognition and analysis in a tractable manner.     

Computer simulations of time-dependent suppression of EOF by polymer coatings

We use molecular dynamics simulations in order to investigate the time evolution of the effect of adsorbed polymer coatings on the electro-osmotic flow (EOF) in a capillary. Weakly adsorbed coatings show no time-dependent performance, but they do not strongly reduce the EOF. On the other hand, strongly adsorbed coatings made of longer polymer chains are often quenched in non-equilibrium conformations that can strongly reduce the EOF over extremely long periods of time. For intermediate adsorption strengths, we observe that the EOF increases as a function of time due to the relaxation of the coating layer. The concentration of polymers in solution and the length of the polymer chains also affect the time-dependence of the EOF. These results show that the quality of electrophoretic separations can depend on the waiting time between the formation of the coating and the beginning of the separation. We conclude by suggesting experimental tests of our predictions.     

The piscine bioconcentration characteristics of cyclic and linear oligomeric permethylsiloxanes

Using a conventional "resaturation" method whereby aquarium water was continuously passed through a column containing sand or fine glass beads coated with cyclic and linear permethylsiloxanes, their uptake levels by rainbow trout and fathead minnows have been compared. Because of the uncertainty associated with defining the actual aqueous concentrations of such poorly soluble substances, this study was focused on defining the "attainable uptake" levels from saturated solutions rather than precise definition of actual bioconcentration factor values. Although cyclic Me2SiO-oligomers accumulated to a greater extent in fish than did comparable linear oligomers, uptake decreased sharply with increasing molecular weight. Thus, in the cyclic series (Dx), order of magnitude decreases were observed for each incremental molecular weight increase; i.e., for the compounds D4, D5, and D6 uptake levels of approximately 200, 20 and 2 ppm, respectively, were observed. Uptake of D8 was below our detection limit of 300 ppb. In the linear series, uptake of the tetramer MD2M was an order of magnitude less than observed for D4 and little or no uptake (i.e., less than 0.5 ppm) was observed for MD3M, MD4M and MD7M. The branched oligomer M3T exhibited levels comparable to its unbranched isomer MD2M, while M4Q was more comparable to the D6 uptake of 1-2 ppm. Very similar uptake levels of D5 resulted with and without a surfactant, even though the surfactant afforded a 20-fold increase in the D5 content of the water. This suggests that bio-availability is defined by the amount present in true solution as individual molecules and is not affected by the presence of aggregates or micelles. The highly inverse relationship observed in this study between uptake and molecular weight is strongly supportive of earlier estimates of a limiting molecular weight of about 600. These findings also strongly contradict a recent Japanese study, which concluded that bioconcentration not only occurred but actually increased with molecular weight in a series of commercial polydimethylsiloxane fluids. Also contrary to a recently published inference of biotransformation in fish, no evidence for such phenomena was observed in this study.