The effects of chemical injury on the ocular surface

Chemical burns of the eye may destroy all of the corneal epithelium and large portions of conjunctival epithelium into the fornices. Restoration of the ocular surface after a chemical burn depends on the centritedal movement of conjunctival epithelial cells and their adherence to the altered corneal stroma. Epithelial movement after a corneal burn is normal for 72 hours after a burn, but persistent epithelial defects thereafter commonly are not resolved until total corneal vascularization occurs. Although this fresh epithelium may be protected by a bandage soft contact lens, more promising and far-reaching approaches may follow epithelial supplementation and even replacement. The link between the health and integrity of the epithelial layer as it relates to the corneal substratum and its cellular constituents remains to be forged.     

Electrochemical sulfur dioxide concentrator for flue-gas desulfurization

A small test cell has been designed to investigate the electrochemical removal and concentration of sulfur dioxide from stack gases. The cell is a high temperature molten-salt cell which uses the ternary eutectic of lithium, potassium and sodium sulfates as the electrolyte. Simulated flue-gas having a composition similar to that resulting from burning 3.5% sulfur coal is fed to the cathode, where sulfur dioxide and oxygen are converted to sulfate ion. At the anode, the reverse reaction occurs resulting in a small volume of gas containing up to 50% sulfur oxides.     

Existence and learning of oscillations in recurrent neural networks

We study a particular class of n-node recurrent neural networks (RNNs). In the 3-node case we use monotone dynamical systems theory to show, for a well-defined set of parameters, that, generically, every orbit of the RNN is asymptotic to a periodic orbit. We then investigate whether RNNs of this class can adapt their internal parameters so as to "learn" and then replicate autonomously (in feedback) certain external periodic signals. Our learning algorithm is similar to the identification algorithms in adaptive control theory. The main feature of the algorithm is that global exponential convergence of parameters is guaranteed. We also obtain partial convergence results in the n-node case.