Abstract:
ERP’s (Event-related Potentials or Evoked Potentials) have demonstrated both research and clinical value in providing neurophysiological information relevant to assessment, diagnosis, and monitoring treatment progress. Despite this value, one barrier that remains is the time required to measure a typical ERP, which can range from 1 to many minutes.
This is required due to the large number of responses required in order to create a reliable estimate by averaging. Moreover, different platforms provide different degrees of flexibility and programmability, ranging from predetermined, hard-coded procedures, to toolkits that provide extensive configurability. Some systems do not provide an ERP estimate until all trials are complete, and some also require a further step of postprocessing, with specialized software which may be proprietary, or may be open- sourced, but must be installed on the user PC. This requirement is a concern for clinical assessment use, and also stands in the way of real-time recording, for faster measurement, and possibly for neurofeedback or brain dynamical process study. This talk will present alternative ERP methods that provide reliable ERP estimates in shorter times, and, in particular, in real time, using both surface EEG and sLORETA-derived brain current- source activation data. The ability to process single trials in real-time will be demonstrated, providing the capacity to determine during testing whether sufficient samples have been acquired. Thus, the test is actually completed in real time, and results are visible even before the stimulation is stopped. Recorded raw EEG as well as ERP data can be stored and postprocessed for further analysis, and for cross-platform confirmation. Applications to single-subject controlled designs will be demonstrated. Examples will include exposure to faces with gender and racial content, smoking cessation, and forensic applications. Criteria for determining the validity and completion of event-related potential data will be introduced and demonstrated, using multiple methods of analysis.
Presented by: Thomas Collura, Ronald Bonnstetter, Estate Sokhadze, Carlos Zalaquett, Rhea Banerjee, Michael Russo
