[4eyes] FW: [Faculty] Seminar: April 23, 2009: Dianna Han

[Matthew Turk]

Please attend this talk if you are able. Dianna is now a Four Eyes Lab
student, albeit spending most of her time in our "satellite UCLA lab."

	Matthew

-----Original Message-----
From: faculty-bounces at lists.cs.ucsb.edu
[mailto:faculty-bounces at lists.cs.ucsb.edu] On Behalf Of Tiffany Sabado
Sent: Wednesday, April 22, 2009 11:15 AM
To: Typical faculty; grads at lists.cs.ucsb.edu; research at lists.cs.ucsb.edu;
colloquia at lists.cs.ucsb.edu
Subject: [Faculty] Seminar: April 23, 2009: Dianna Han

Thursday, April 23, 2009
CS Conference Room
2 PM

Title: Cognitive State Detection using Functional MRI
Speaker: Dianna Han, PhD student, Computer Science, UCSB

Abstract:

This talk will give an overview of cognitive state detection using fMRI.

Magnetic Resonance Imaging (MRI) is a medical imaging technique based on 
the physics of nuclear magnetic resonance. It has been widely used in 
radiology to visualize the internal structure and function of the body. 
Functional MRI (fMRI) is an indirect measure of neural activity changes 
in the brain that uses the MRI equipment to detect changes in cerebral 
metabolism. The most popular technique employed in neuroscience studies 
utilizes blood oxygenation level dependent (BOLD) contrast. Although the 
precise nature of the coupling between neural activity and the BOLD 
signal is still an ongoing research subject, it has been observed that 
increased neural activity causes an increased amount of oxygenated 
hemoglobin relative to deoxygenated hemoglobin in the active region, 
which leads to a signal increase detectable by the scanner.

It has been proven that fMRI can be used to detect and characterize 
changes in cognitive state. Given the high dimensionality of the data 
dimension reduction is an essential step in fMRI data analysis. Common 
approaches include creating statistical summaries of individual voxels 
using t-test and correlation test, isolating 'regions of interests' 
(ROIs), and applying classic dimension reduction methods such as 
principle component analysis (PCA). A more attractive approach uses 
independent component analysis (ICA), which separates the multivariate 
signal into components most effective in classifying the sources by 
maximizing the statistical independence among them. Classification 
techniques then can be applied to the reduced search space to 
discriminate different states.

Functional MRI and its applications of detecting cognitive states have 
drawn much research interest. In March 2009 a lie-detection test based 
on fMRI was submitted in court as evidence in San Diego. Other 
applications such as craving state detection for drug rehabilitation and 
diagnosis of mental diseases are also being explored. Efforts have also 
been made to improve its temporal resolution (only on the order of 
seconds) by combining fMRI with other techniques of higher temporal 
resolution such as EEG.

Dianna Han is a Ph.D student in the Department of Computer Science, 
UCSB. Her research interest lies in medical image analysis, more 
specifically, cognitive state detection using functional MRI and EEG. 
Her research advisor is Professor Matthew Turk. She is currently working 
on a collaborative research project led by Professor Mark Cohen in UCLA 
Center for Cognitive Neuroscience.

Host: Matthew Turk

-- 
Tiffany Sabado
Assistant to the Chair
Computer Science Department
Phone (805) 893-2207
Fax (805) 893-8553

_______________________________________________
Faculty mailing list
Faculty at lists.cs.ucsb.edu
https://lists.cs.ucsb.edu/mailman/listinfo/faculty