No abstract

The effects of noise autocorrelation on neural waveform recognition (detection, classification, and superposition resolution) are investigated in this study using microelectrode recordings from the cortex of a monkey. Optimal waveform recognition is accomplished by passing the data through a whitening filter before matched filtering for detection or template matching for classification and superposition resolution. Template matching without whitening requires about 40% higher signal-to-noise ratio than template matching with whitening for comparable classification and superposition resolution. The comparable difference for detection is 15%.

No abstract

To determine the influence of changes in nasal pressure (Pn) on airflow mechanics in the upper airway, we examined the effect of elevations in Pn on upper airway resistance and critical pressure (Pcrit) during stage I/II sleep in six patients with obstructive sleep apnea. When Pn was elevated above a Pcrit, periodic occlusions of the upper airway were eliminated and inspiratory airflow limitation was demonstrated by the finding that inspiratory airflow (VI) became maximal (VImax) and independent of fluctuations in hypopharyngeal pressure (Php) when Php fell below a specific Php (Php′). As Pn was elevated, VI vs. Php demonstrated 1) marked decreases in early and late inspiratory resistances from 75.9 +/- 34.7 and 54.6 +/- 19.0 to 8.0 +/- 1.7 and 7.6 +/- 1.6 cmH2O.l-1.s (P less than 0.05), respectively, and 2) increases in early and late inspiratory Php′ to levels that exceeded Pcrit by 3.0 +/- 0.6 and 3.1 +/- 0.7 cmH2O, respectively, at the highest level of Pn applied (P less than 0.01). This latter finding suggests that elevations in Pn result in increases in Pcrit. We suggest that elevations in Pn produce distinct alterations in upper airway resistance and collapsibility, which may influence oppositely the level of airflow through the upper airway during sleep.

No abstract

No abstract

The main difficulties in reliable automated detection of the K-complex wave in EEG are its close similarity to other waves and the lack of specific characterization criteria. We present a feature-based detection approach using neural networks that provides good agreement with visual K-complex recognition: a sensitivity of 90% is obtained with about 8% false positives. The respective contribution of the features and that of the neural network is demonstrated by comparing the results to those obtained with i) raw EEG data presented to neural networks, and ii) features presented to Fisher's linear discriminant.

Atomic force microscopy (AFM) allows rapid, accurate, and reproducible visualization of DNA adsorbed onto solid supports. The images reflect the lengths of the DNA molecules in the sample. Here we propose a solid-state DNA sizing (SSDS) method based on AFM as an analytical method for high-throughput applications such as finger-printing, restriction mapping, +/- screening, and genotyping. For this process, the sample is first deposited onto a solid support by adsorption from solution. It is then dried and imaged under ambient conditions by AFM. The resulting images are subjected to automated determination of the lengths of the DNA molecules on the surface. The result is a histogram of sizes that is similar to densitometric scans of DNA samples separated on gels. A direct comparison of SSDS with agarose gel electrophoresis for +/- screening shows that it produces equivalent results. Advantages of SSDS include reduced sample size (i.e., lower reagent costs), rapid analysis of single samples, and potential for full automation using available technology. The high sensitivity of the method also allows the number of polymerase chain reaction cycles to be reduced to 15 or less. Because the high signal-to-noise ratio of the AFM allows for direct visualization of DNA-binding proteins, different DNA conformations, restriction enzymes, and other DNA modifications, there is potential for dramatically improving the information content in this type of analysis.