Emanuel Donchin - US grants

Ample evidence supports the assertion that the latency of the P300 component of the human event-related brain potential is proportional to stimulus evaluation and categorization processes and is relatively independent of response selection and execution processes. We shall use P300 latency to gain a more detailed understanding of the ubiquitous slowness of behavior in the aged. This project will be conducted in collaboration with Rush-Presbyterian-ST. Luke's Medical Center in Chicago. Such collaboration will bring together the necessary skills as well as the required subject population in which to develop a battery of psychophysiological tests in which P300 latency can be dissociated from reaction time. Each of these tests will assess the duration of a specific phase or stage of huma information processing. The tests will be administered to young adults in Champaign-Urbana and to the "normal control group" of aged individuals in the "normal controals" assembled in the Program Project "Cerebral Decline in Aging" (3P0-AG-00905) at Rush-Presbyterian Medical School in Chicago. The results should allow us to determine which, if any, of the several stages of information processingis specifically affected by aging.

Recent approaches to the study of memory emphasize that memory is the result of the interaction of processes occurring during encoding, rehearsal and retrieval. However, most of these processes lack an overt manifestation and can only be inferred indirectly. Recent evidence indicates that components of the Event-Related Brain Potential (ERP), such as P300 and Slow Wave, can be used to monitor some of these covert processes. In particular, Karis, Fabiani, and Donchin (1984) reported data suggesting that P300 manifests changes in the memory representations of events that facilitate their recall for rote memorizers but not for elaborators. The program we propose is designed to further confirm this observation, test its generality, and elucidate the nature of the change that occurs in the representation concurrently with a P300. We propose experiments testing the generality of the relationship between ERPs and recall in paradigms in oth verbal and non-verbal domains, and investigating the nature of the changes occurring in the memory representation as a consequence of the updating process, and the interactions between the status of the memory representation and subsequent recall performane. All the experiments proposed also address problems related to the role of distinctiveness in memory. The major long term objective of this program is to provide information concerning memory processes while they are occurring, by recording ERPs during and immediately after the presentation of stimuli. The joint analyses of ERPs, conventional performance measures and extended debriefing of the subjects will allow the development and testing of artiulated memory models. We also believe that this approach could be potentially very useful in addressing some of the issues related to the disorders of memory, by allowing the investigator to analyze and decompose the processes involved in the memory function.

The explosive progress in the neurosciences over the last decade has not been matched by an increased understanding of the human brain and its role in human cognition. Progress in cognitive neuroscience has been slowed by obvious ethical constraints on the experimental study of the human nervous system. However, several recent technological developments have made it possible to visualize aspects of brain activity in novel and powerful ways. Of particular importance is the fact that these techniques can be used with awake behaving humans with virtually no risk and practically no inconvenience to the subject. Two classes of techniques are particularly noteworthy, the radiological and the electro-magnetic approaches. Radiological approaches make it possible to visualize metabolic processes in the intact brain. Positron Emission Tomography (PET) and Magnetic Resonance Imaging (MRI) are typical of these approaches, as they allow investigators to determine with very fine spatial resolution the site of brain activity which is invoked in the course of clearly specified cognitive activity. The electro- magnetic approaches monitor either the electrical or the magnetic activity of the brain. These can be monitored with remarkable temporal resolution, but with rather poor spatial resolution. This action will provide NSF's share of the funding for a conference to be held in Carmel, California, during the first week of January, 1990, to discuss ways and means to integrate these novel techniques in order to enhance their power as tools for cognitive neuroscience. For technological and theoretical reasons, each technique has developed its own experimental paradigms. The conditions which are optimal for recording within one approach often conflict with the demands which others impose on an experiment. A necessary condition for the integration is a thorough mutual understanding of the paradigmatic constraints that shape the work in each of the domains. It is for this reason that the conference will bring together investigators who represent each of four main techniques to be considered. The conference will be divided into three parts: For the first day and a half, several prominent investigators will provide tutorials. Then the participants will meet in four smaller panels, each of them receiving a very detailed charge. A day and a half will be devoted to the panel discussions. Then, in the last two days, each panel will report to the assembled group its response to the charge.

biomedical equipment purchase;

The project will evaluate the feasibility of a Brain-Computer Interface (BCI) to be used by disabled individuals communicating with a computer. The BCI's principle of operation is based on the fact that rare, task- relevant, events presented in the so called "Oddball Paradigm", elicit an Event-Related Brain Potential characterized by a P300 component. Farwell and Donchin have used a display of a 6 by 6 matrix of characters, the rows and columns of which were independently and randomly intensified. By detecting, online, the matrix elements that elicited a P300, the identity of the attended cell was determined. This project will be concerned with the application of novel pattern recognition methods, such as wavelets and neural networks, to speed item recognition. The system will be tested with 5 disabled individuals. The demonstration that the system can be used by disabled individuals, at a communication speed of at least 10-12 characters per minute, will lead to a Phase II project in which a commercial version of the BCI will be developed. Such a BCI will meet the communication needs of individuals who are thoroughly paralyzed, and may also turn out to be of use in less severely-restricted cases. PROPOSED COMMERCIAL APPLICATION: There is a large market for assistive, prosthetic, devices that allow an individual to communicate with a data entry device. The system that will be evaluated, and developed, in this project addresses the needs of users who cannot benefit from existing systems because they lack motor controls. There is a significant possibility that the operating mode of the proposed BCI will be more tractable in use than are currently common devices that utilize a head mounted stick, or eye movements, to communicate.

The executive cognitive functions are humans' most advanced component mental operations. Examples of executive functions include manipulating the contents of working memory, inhibiting automatic but inappropriate responses, and flexibly switching between strategies. These functions allow people to make effective decisions, regulate social behavior, and plan ahead. The executive functions are the last to mature in development and are often the first to deteriorate in normal aging as well as mental disorders and neurological injury and disease. Research has yet to fully describe the neural bases of these higher-order mental operations, and understanding the neural bases of the executive functions might provide insight into why some people are better at planning and decision-making than others and allow for more effective intervention and treatment of executive function impairment. The neural bases of the executive functions likely involve networks of neural structures distributed throughout the brain working together in specific temporal sequences, with individual nodes active for only tens of milliseconds. A large literature exists suggesting the prefrontal cortex as a core structure within this network. The research enabled by the dense-sensor array electroencephalography (EEG) system, with extended frontal coverage, investigates the coordinated neural activity underlying the executive functions.

Event-related potentials (ERPs) are scalp recorded neural electrical responses to stimuli and actions embedded in the ongoing EEG. ERPs are an excellent means of assessing the timecourse of functionally relevant neural activity during cognitive processing, and hence a powerful tool for investigating the neural network dynamics that make executive functioning possible. Dense-sensor array EEG, like the 128 channel system here, improves ERP's spatial resolution, allowing better estimation of the neural sources of the scalp-recorded signal. Dense sensor array ERPs allows researchers to examine the neural network activity related to specific cognitive operations. Researchers will use this system to study learning and memory, potentially improving learning strategies, individual differences in risky decision-making (why some individuals make riskier choices than others), attention selection (how the brain may use economic principles, like expected value, to allocate its limited capacity processing resources), and other executive functions, their variability across individuals, and their disruption in neurological injuries (e.g. mild traumatic brain injury) and disease (e.g. Huntington's disease).