Eva Feredoes, Ph.D.
University of New South Wales
My research uses transcranial magnetic stimulation (TMS) and functional magnetic resonance imaging (fMRI) to investigate the neural correlates of working memory processes related to storage and control.
*** News Flash***
Eva is leaving the Postle Lab for Canadian pastures in December. She will be a postdoctoral fellow in the lab of Dr. Todd Woodward at the University of British Columbia working on multivariate approaches to the analysis of fMRI data. You can address any correspondence to her at: evaferedoes(at)gmail.com
Current projects include:
Investigating the mechanisms underlying the resolution of proactive interference in the left inferior frontal gyrus (IFG).
- Invoking a (long-term memory-style) dual-processes framework to investigate the way in which the left IFG may resolve proactive interference. (By the way, check out the PNAS paper in the 2006 publications!).
- There is evidence to suggest that during the early part of the response period, information related to a probe’s familiarity is at its peak (as evidenced by increased false alarms to probes that are misleading because of their familiarity). Once recollection begins to take over from familiarity-based information about the probe, false alarms decrease.
- Applying TMS to the left IFG to coincide with the “early” or “late” part of the probe presentation has shown that TMS has a selective effect on performance for the “early” period only.
- Link to CNS 2007 poster below.
Quantitatively and qualitatively discrepant results yielded by single-subject (SS) and spatially-normalized group averaged (SNGA) approaches to group analyses of fMRI data.
- Different fMRI analysis procedures can produce empirically and theoretically discrepant results.
- Analyses of the same fMRI data set employing a single-subject (i.e., treating each subject like a “case study”) approach yielded topographically variable load-sensitive regions across subjects (but mostly in posterior brain regions). A spatially normalized group-averaged approach, on the other hand, produced a significant load-sensitive region in left prefrontal cortex, incongruent with the single-subject analysis results.
- The topographically varied nature of load-sensitivity across individuals may make this behavior unsuited to a spatially normalized group-averaged analysis approach, that requires the anatomical overlap of load-sensitive regions in the majority of subjects, in order to be detected. (Check out the NeuroImage paper in the 2007 publications!).
- Which analysis procedure is “telling the truth”? TMS was applied to both a “single-subject” and the “spatially normalized group-averaged” regions within individual subjects to investigate which of these makes a critical contribution to the storage of verbal working memory information. (Check out the Journal of Neuroscience paper in the 2007 publications!).
- Follow-up: to use a new multivariate approach to fMRI analysis (constrained PCA) to investigate further load maps derived from single subjects.
Deriving TMS-evoked BOLD HRFs from multiple cortical areas at 3T
- One standard approach to fMRI blood oxygen level-dependent (BOLD) signal evoked by task performance is to linearly convolve regressors with a hemodynamic response function (HRF) estimate.
- Uncertain what the HRF looks like over most brain regions.
- Investigating the interregional variability of the HRF within individuals by applying single TMS pulses over prefrontal and parietal region during fMRI scanning to evoke BOLD HRFs.
- Link to HBM 2007 poster below + see my poster at this year’s SfN meeting in San Diego: #316.13/III7
Link to my CV: