EP component identification and measurement by principal components analysis.

Between the acquisition of Evoked Potential (EP) data and their interpretation lies a major problem: What to measure? An approach to this kind of problem is outlined here in terms of Principal Components Analysis (PCA). An important second theme is that experimental manipulation is important to functional interpretation. It would be desirable to have a system of EP measurement with the following characteristics: (1) represent the data in a concise, parsimonous way; (2) determine EP components from the data without assuming in advance any particular waveforms for the components; (3) extract components which are independent of each other; (4) measure the amounts (contributions) of various components in observed EPs; (5) use measures that have greater reliability than measures at any single time point or peak; and (6) identify and measure components that overlap in time. PCA has these desirable characteristics. Simulations are illustrated. PCA's beauty also has some warts that are discussed. In addition to discussing the usual two-mode model of PCA, an extension of PCA to a three-mode model is described that provides separate parameters for (1) waveforms over time, (2) coefficients for spatial distribution, and (3) scores telling the amount of each component in each EP. PCA is compared with more traditional approaches. Some biophysical considerations are briefly discussed. Choices to be made in applying PCA are considered. Other issues include misallocation of variance, overlapping components, validation, and latency changes.

Word meaningfulness and event-related potentials during phrase processing.

Evoked Potentials (EP) were recorded to visual presentation of words in phrases. Three EP components, P250, N400, P560, and their dependence on the role that each word plays in the phrase were studied. Subjects were requested to perform a congruity/incongruity task after the presentation of short phrases of two-types that differed in the position of meaningful keywords. We show how these keywords contribute substantially to N400 and possibly to P560 but do not reliably affect P250.

Principal components analysis of sources of variability in retinal ganglion cell responses.

An approach to the functional organization of retinal ganglion cell processing in terms of correlation matrices (Levine and Shefner, 1975; 1977a, b; Shefner and Levine, 1979) is extended by applying Principal Components Analysis. This analysis reduces each correlation matrix to a few components which are implicit in the data. The component loadings describe properties of the system in terms of loadings (correlations) of time bins on the underlying components. Each component is identified by the experimental conditions associated with the highest loadings. Mixed conditions are quantitatively interpreted as weighted contributions from the various identified components. This has led to new interpretations of existing data. Several properties of ganglion cell inputs are analyzed in this manner, including ON and OFF processes, center and surround mechanisms, rod and cone inputs, and spatially distinct areas within the receptive field center. Although the details vary, generally one of the components is highly associated with ON processes and the other with OFF and/or MAINTAINED processes. several advantages may be realized through the use of Principal Components Analysis: (1) all of the data contribute to the analysis, (2) the number and relative importance of contributing processes may be assessed, (3) the relative contribution of underlying processes to mixed responses may be assessed, and (4) the most parsimonious representation of the data is obtained.

Short-term memory: the "storage" component of human brain responses predicts recall.

An evoked potential component with a poststimulus peak at about 250 milliseconds is related to the storage of information in short-term memory. This storage component was found in an investigation of brain potentials in relation to a number and letter comparison task. In replications of this experiment at three different light intensities spaced 1.0 log unit apart, the component had essentially the same waveform and pattern of scores. The memory storage interpretation was confirmed in a behavioral experiment that probed short-term memory. Recall was predicted by the magnitude of the storage component.