Endoscopic sensing of alveolar pH.

Previously unobtainable measurements of alveolar pH were obtained using an endoscope-deployable optrode. The pH sensing was achieved using functionalized gold nanoshell sensors and surface enhanced Raman spectroscopy (SERS). The optrode consisted of an asymmetric dual-core optical fiber designed for spatially separating the optical pump delivery and signal collection, in order to circumvent the unwanted Raman signal generated within the fiber. Using this approach, we demonstrate a ~100-fold increase in SERS signal-to-fiber background ratio, and demonstrate multiple site pH sensing with a measurement accuracy of ± 0.07 pH units in the respiratory acini of an ex vivo ovine lung model. We also demonstrate that alveolar pH changes in response to ventilation.

A new method of spike modelling and interval analysis.

Here we develop a new model of spike firing, based on the leaky integrate and fire model, modified to simulate afterpotentials. We also develop new analysis techniques, applying these to recorded and model generated data in order to make a comparative analysis and develop the model as a hypothesis for the functional components of the neuron. The model is based in this first instance on hypothalamic oxytocin neurons. We demonstrate how model parameters and cell properties relate to features observed in inter-spike intervals histograms, and the limits of these in being able to detect patterning features in spike recordings. A new technique, spike train analysis, is able to detect previously unobserved patterning, showing a dependence of spike intervals on previous firing activity. This effect is reproduced in the model by adding the small amplitude but long lasting after hyper-polarising potential (AHP). A fit measure based on log likelihood is used to compare model generated data to recorded spike intervals, taking account of interval dependence on previous activity. This measure is used with the simplex multiple parameter search algorithm to develop an automated method for fitting the model to recorded data.

Products of gaussians and probabilistic minor component analysis.

Recently, Hinton introduced the products of experts architecture for density estimation, where individual expert probabilities are multiplied and renormalized. We consider products of gaussian "pancakes" equally elongated in all directions except one and prove that the maximum likelihood solution for the model gives rise to a minor component analysis solution. We also discuss the covariance structure of sums and products of gaussian pancakes or one-factor probabilistic principal component analysis models.