Activities and Seminars

Cesar Caballero, University of Nottingham
Date: Aug 08, 2011

What: Novel methods and analysis techniques for BOLD fMRI: Paradigm free mapping and information theoretic analysis for concurrent EEG-fMRI.

When: Monday, August 8th, 4:00pm

Who: Cesar Caballero, University of Nottingham

Where: BCBL, Paseo Mikeletegi 69, Floor 2.


The high contrast to noise ratio of the BOLD effect available at high field MR scanners (3T and beyond) makes possible the study of the haemodynamic response without trial averaging: single trial BOLD fMRI. Whereas conventional single-trial fMRI analyses heavily rely on the knowledge of the timing of the events, this information can sometimes be imprecise or unavailable. In the first part of this talk, I will introduce a novel fMRI analysis framework which enables to detect single-trial BOLD responses without prior timing information: Paradigm Free Mapping. Validated on comprehensive simulations and experimental data from a visuomotor paradigm, this technique provided further insight on the relationship between transient, spontaneous BOLD events and slow-frequency fluctuations in resting state fMRI data, and to detect the BOLD signal changes related to interictal epileptic discharges without (completely) relying on the EEG recordings. In the second part of this talk I will present an information theoretic method for the analysis of concurrent EEG-fMRI data. When applying this method, we focus on the investigation of the location and dynamics of epileptiform networks associated with interictal epileptic discharges. Based on the mutual information between the EEG signal and the fMRI data, the most prominent feature of this approach is that it balances both imaging modalities. In addition, the approach does not require a-priori models for the haemodynamic shape nor does it assume a linear relationship between the spiking activity detected on the scalp-EEG and BOLD responses, which is relevant in clinical cases, such as infants or patients with brain lesions, where the neurovascular coupling might deviate from its canonical form.