Response Dynamics

The principal strengths of neural electromagnetic methods stem from their capacity to define the dynamics of neural population activity. Even a single electrode pasted to the scalp may disclose a complex temporal waveform consisting of a series of peaks and valleys. In some cases, a peak in a waveform at a particular latency is observed across a large subset of the channels in a whole-head sensor array, even though the amplitude or the polarity of the peak changes. This pattern is characteristic of a single anatomical source or a set of sources acting in synchrony. In some cases, a close examination of the waveform montage collected from a sensor array discloses that what appears as a single peak in one waveform can be resolved as multiple overlapping temporal peaks observed at other locations. In such cases, the topographic map of the evoked response typically has apparent features that appear to shift systematically over the course of the response. A simple-minded analysis may suggest that the source is a single focus of activity moving through the brain volume. A more sophisticated source model may allow the same response to be decomposed into two or more component sources with stable NEM topographies and distinct but overlapping time courses.

In the absence of effective source localization techniques, there was a tendency in early work to focus on the peaks in the waveform as the unitary building blocks—the components from which complex event-related responses were built. Components were given names on the basis of the polarity and latency of the waveform peaks: e.g., the N100 (a negative-going response component peaking around 100 msec poststimulus) or the P300 (or P3, a positive peak in the response waveform around 300 msec poststimulus). As the characterization of response components proceeded, descriptions of the component scalp topography were sometimes added to aid identification and discrimination of named components. For endogenous cognitive response components, identifying criteria often included the nature of experimental manipulations required to elicit or enhance a particular peak in the waveform. Whereas such information is critical for investigators attempting to reproduce or extend a particular observation, it complicates the business of component quantification.

In an effort to address this concern, some investigators turned to blind decomposition techniques such as principal components analysis (PCA). PCA is a linear technique based on eigen analysis and singular value decomposition. The method attempts to find a set of basis functions—in this case, field or potential topo-graphies—that can be used to reconstruct the original experimental data. In order to make the decomposition unique, principal components are constrained to be mutually orthogonal. Each principal component has an associated weighting vector that quantifies the representation in the data of each component as a function of time. In some well-behaved cases, principal components correspond to response components identified by other criteria. However, in general, the requirement for orthogonality precludes the proper identification of more than a few components. An alternative decomposition strategy has been developed that appears to hold an alternative decomposition strategy that appears to hold significant promise for functional neuroimaging applications. Independent component analysis identifies a basis set in which components are statistically independent though not necessarily orthogonal. Initial results with the algorithm are promising, although it will certainly be possible to find pathological cases that cause the algorithm to fail. It is not yet clear how effectively the algorithm will identify proper components in routine applications to event-related response data.

In a few cases, the idea that components reflect the successive activation of links in a processing chain appears basically correct. For example, in the auditory brain stem evoked response (ABER), the peaks in the waveform are associated with specific structures in the early auditory pathways, and the waveform morphology can be used to assess the integrity of the relay and processing circuitry. Similarly, the earliest components of somatosensory responses evoked by electrical stimulation appear to be associated with specific anatomical loci. In contrast, for visual evoked responses the situation is considerably more complex. Although some investigators have reported an early, relatively small EEG component (N70) that was identified as the initial activation of cortical V1, most analyses have focused on the more robust P100 component. A variety of source analysis techniques indicate that this component is a complex consisting of temporally overlapping responses from several distinct though nearby visual areas. Although there is an element of sequential processing in the early visual system as activation spreads through the information processing tree, there is also considerable parallel processing. There are also forward and feedback links that skip over portions of the schematic processing hierarchy and delays within areas that can further complicate the simple orderly picture of temporal response dynamics.

Understanding And Treating Autism

Understanding And Treating Autism

Whenever a doctor informs the parents that their child is suffering with Autism, the first & foremost question that is thrown over him is - How did it happen? How did my child get this disease? Well, there is no definite answer to what are the exact causes of Autism.

Get My Free Ebook

Post a comment