Spike sorting based upon machine learning algorithms (SOMA)

We have developed a spike sorting method, using a combination of various machine learning algorithms, to analyse electrophysiological data and automatically determine the number of sampled neurons from an individual electrode, and discriminate their activities. We discuss extensions to a standard unsupervised learning algorithm (Kohonen), as using a simple application of this technique would only identify a known number of clusters. Our extra techniques automatically identify the number of clusters within the dataset, and their sizes, thereby reducing the chance of misclassification. We also discuss a new pre-processing technique, which transforms the data into a higher dimensional feature space revealing separable clusters. Using principal component analysis (PCA) alone may not achieve this. Our new approach appends the features acquired using PCA with features describing the geometric shapes that constitute a spike waveform. To validate our new spike sorting approach, we have applied it to multi-electrode array datasets acquired from the rat olfactory bulb, and from the sheep infero-temporal cortex, and using simulated data. The SOMA sofware is available at http://www.sussex.ac.uk/Users/pmh20/spikes.

Applications of multi-variate analysis of variance (MANOVA) to multi-electrode array electrophysiology data

We have developed an adaptation of multi-variate analysis of variance (MANOVA) to analyze statistically both local and global patterns of multi-electrode array (MEA) electrophysiology data where the activities of many (typically >100) neurons have been recorded simultaneously. Whereas simple application of standard MANOVA techniques prohibits extraction of useful information in this kind of data, our new approach, MEANOVA (=MEA+MANOVA), allows a more useful and powerful approach to analyze such complex neurophysiological data. The MEANOVA test enables the detection of the "hot-spots" in the MEA data and has been validated using recordings from the rat olfactory bulb. To further validate the power of this approach, we have also applied the MEANOVA test to data obtained from a simple computational network model. This MEANOVA software and other useful statistical methods for MEA data can be downloaded from http://www.sussex.ac.uk/Users/pmh20

Update on hospitalized pesticide poisonings in South Carolina, 1983-1987

Three hundred twelve hospitalizations for pesticide poisoning occurred in South Carolina during 1983-1987. This represents a twenty percent decline from an average of 78.5 cases hospitalized per year (1979-1982) to 62.4 cases hospitalized per year currently. Non-occupational poisonings accounted for one-half of the hospitalizations while 20% were related to occupation. Intentional poisonings represented 27% of the total. Two deaths as a result of suicide occurred during the four year period giving a case fatality of less than 1.0%. This five year update reenforces the need for continued education and prevention efforts.