In the last few years, a significant new development in fMRI is the introduction of event-related techniques that are also known as single-trial or time-resolved approaches. By providing temporally-resolved fMRI response evoked by individual events, event-related fMRI can elucidate temporal profiles of events taking place in the neural circuit and provide trial-specific information. Therefore, event-related techniques represent a major breakthrough in fMRI, hold a great potential for imaging the function of brain with an added dimension and have become widely utilized since its introduction. However, since event-related studies differ from those based on the traditional block design, and the associated temporal response is often small and unknown a priori, technical difficulties are still present. The first part of this application is thus aimed at developing a set of tools to improve event-related fMRI. Specifically, the applicants propose to develop and validate: 1) methods for improving the signal-to-noise ratio in event-related fMRI data; 2) methods for time course feature extraction and activation identification in event-related fMRI; and 3) analysis techniques for characterizing event-related fMRI time courses. In addition, the propose to also develop 4) adaptive imaging techniques for event-related fMR] with improved temporal resolution. After their development and validation, the techniques will be further demonstrated with two cognitive tasks that have not been previously examined with event-related fMRI. The first paradigm is use-word generation task that was previously demonstrated by PET studies to involve spatially separate areas that were shown by ERP to exhibit unique temporal relationships. The applicants propose to use event-related fMRI to investigate these areas and compare the timing derived from event-related fMRI with that reported by ERP to test the hypothesis that event-related fMRI can ascertain temporal differences in neuronal response to the use-word generation in the domain of hundreds of milliseconds. The second task to be studied is a delayed non-match to sample (DNMS) task which is thought to involve both cortical areas and medial temporal lobe, in a temporally differential manner. Specifically, the applicants propose to examine a modified DNMS paradigm using event-related fMRI to ascertain the activation in areas involved in the task and to elucidate temporal profiles of the activity in these areas. The hypotheses are that the medial temporal lobe is involved, differences exist in activation profiles of different areas, and the differences can be elucidated with event-related fMRI to aid in the understanding of the roles played by these areas.
Showing the most recent 10 out of 69 publications
