Rheological Evaluation of the Isothermal Cure Characteristics of Medical Grade Silicone Elastomers

Silicone elastomer systems have been shown to offer potential for the fabrication of medical devices and sustained release drug delivery devices comprising low molecular weight drugs and protein therapeutics. For drug delivery systems in particular, there is often no clear rationale for selection of the silicone elastomer grade, particularly in respect of optimizing the manufacturing conditions to ensure thermal stability of the active agent and short cycle times. In this study, the cure characteristics of a range of addition-cure and condensation-cure, low-consistency, implant-grade silicone elastomers, either as supplied or loaded with the model protein bovine serum albumin (BSA) and the model hydrophilic excipient glycine, were investigated using oscillatory rheology with a view to better understanding the isothermal cure characteristics. The results demonstrate the influence of elastomer type, cure temperature, protein loading, and glycine loading on isothermal cure properties. By measuring the cure time required to achieve tan delta values representative of early and late-stage cure conditions, a ratio t(1)/t(2) was defined that allowed the cure characteristics of the various systems to be compared. Sustained in vitro release of BSA from glycine-loaded silicone elastomer covered rod devices was also demonstrated over 14 days. (C) 2010 Wiley Periodicals, Inc. J Appl Polym Sci 116: 2320-2327, 2010

Inter-Frame Contextual Modelling For Visual Speech Recognition

In this paper, we present a new approach to visual speech recognition which improves contextual modelling by combining Inter-Frame Dependent and Hidden Markov Models. This approach captures contextual information in visual speech that may be lost using a Hidden Markov Model alone. We apply contextual modelling to a large speaker independent isolated digit recognition task, and compare our approach to two commonly adopted feature based techniques for incorporating speech dynamics. Results are presented from baseline feature based systems and the combined modelling technique. We illustrate that both of these techniques achieve similar levels of performance when used independently. However significant improvements in performance can be achieved through a combination of the two. In particular we report an improvement in excess of 17% relative Word Error Rate in comparison to our best baseline system.