Parallel path consistency

Filtering algorithms are well accepted as a means of speeding up the solution of the consistent labeling problem (CLP). Despite the fact that path consistency does a better job of filtering than arc consistency, AC is still the preferred technique because it has a much lower time complexity. We are implementing parallel path consistency algorithms on multiprocessors and comparing their performance to the best sequential and parallel arc consistency algorithms.^(1,2) (See also work by Kerethoet al. ⁽³⁾ and Kasif⁽⁴⁾) Preliminary work has shown linear performance increases for parallelized path consistency and also shown that in many cases performance is significantly better than the theoretical worst case. These two results lead us to believe that parallel path consistency may be a superior filtering technique. Finally, we have implemented path consistency as an outer product computation and have obtained good results (e.g., linear speedup on a 64K-node Connection Machine 2).     

Oxygen content of the fixative is important in the interpretation of the ultrastructure of ischaemic myocardium

Isolated rat hearts were subjected to 15, 45, or 60 minutes of global ischaemia and then fixed by perfusion at 37°C with glutaraldehyde containing various amounts of oxygen. This either had been bubbled with 100% oxygen (PO₂ 620 mm Hg) or with 100% nitrogen (PO₂ 40 mm Hg) immediately before use, or it had been routinely prepared and stored exposed to atmospheric oxygen (PO₂ 245 mm Hg). The ultrastructure of myocytes and endothelial cells subjected to 15 minutes of ischaemia was not affected by the treatment of the fixative. However, when the tissue subjected to longer periods of ischaemia was fixed with routinely prepared or oxygen-bubbled glutaraldehyde, ultrastructural changes characteristic of reoxygenation damage were uniformly evident in both the microvasculature and myocytes. These qualitatively distinct changes included mitochondrial swelling, cell swelling, endothelial bleb formation, and narrowing of capillary lumina. These abnormalities were not observed in tissue fixed with nitrogen-bubbled glutaraldehyde. These findings indicate that deliberate steps should be taken to reduce or eliminate dissolved oxygen from the fixatives used to study ischaemic tissues. Otherwise artefactual reoxygenation damage in vitro may occur and make valid ultrastructural interpretation difficult or impossible.     

Optimizing Router Performance in TCP/IP Networks

This article reviews the steps necessary to configure and implement a complex router internetwork. This article will specifically concentrate on Cisco routers, and since they support Transmission Control Protocol/Internet Protocol (TCP/IP) natively, it will concentrate on a backbone network employing that protocol.     

Two representational systems in dynamic visual identification.

How does the visual system retain and combine information about an object across time and space? This question was investigated by manipulating the spatiotemporal continuity and form continuity of 2 perceptual objects over time. In Experiment 1 the objects were viewed in central vision within a single eye fixation, in Experiment 2 they were viewed across a saccadic eye movement, and in Experiment 3 they were viewed at different spatial and retinal locations over time. In all 3 experiments some information about the object was found to be linked to its spatiotemporal continuity, and some information was found to be independent of spatiotemporal continuity. Form continuity was found to produce no effect. The results support a theory of dynamic visual identification according to which information is maintained over time by both episodic object representations and long-term memory representations, neither of which necessarily code specific sensory information. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Roles of object-file review and type priming in visual identification within and across eye fixations.

Dynamic visual identification was investigated in 4 experiments. In Exps 1 and 2, 2 perceptual objects (2 frames, each containing a letter or 1 containing a letter and the other a plus sign) were previewed in the periphery. A saccade brought these objects to central vision. During the saccade the display was changed so that 1 frame contained a letter and the other a plus sign, and the S identified the letter by naming it aloud as rapidly as possible. In Exp 3, the retinal events of Exps 1 and 2 were simulated. In Exp 4, both the preview and the target were presented centrally within a single fixation. In all experiments both object specific and nonspecific preview benefits were observed. These results support a theory in which the preview benefits observed during visual identification arise from 2 processes, object file review and type priming. (PsycINFO Database Record (c) 2010 APA, all rights reserved)