CSL 865 Special Topics in Computer Applications: Introduction to Neuroimaging Methods and Analysis

Course Description


Brain is the most mysterious and complex organ in the human body. Still, the modern science has very limited knowledge about its working. Until late 1990, the methods to image the brain function were very limited. The most effective methods required physical access to the brain. Such an access was only possible from patients with critical disorders such as epilepsy, tumours, stroke etc., which led to slow progress in the field.

With the recent invention of functional MRI (fMRI) and other imaging methods, the situation has been drastically changed.  Functional magnetic resonance imaging (fMRI) provides a non-invasive method to peek into human brain to identify patterns of brain activity. While, this science is still relatively new and developing, scientists and researchers are already reporting a number of fascinating results using fMRI. The changes in the activity of brain in several neurological disorders such as anxiety, depression, chronic pain, phantom limbs are already being studied. In addition, several novel experimental designs such as brain activity while watching Alfred Hitchcock movies or brain activity of meditating Buddhists are also being studied. Finally, novel machine learning-based data analysis methods are making inroads into the field and opening doors to novel science. One of the famous experiment being the one by Tom Mitchell demonstrating that human thought can be predicted using fMRI (see the links below).

This course will provide an introduction to brain imaging methods to people who are fascinated by the human mind and would like to know more. The course will cover:

·         Basic mathematics, statistics and signal processing required for the course (linear algebra, probabilities, statistics, signal and image processing)

·         Basic neuroanatomy

·         An overview of various brain imaging modalities (EEG, PET etc.)

·         Basics of MR physics and fMRI image acquisition

·         Conventional statistical fMRI data analysis methods

·         Advanced fMRI data analysis methods (network analysis and machine learning based methods such as ICA, MVPA, NMF, Granger causality analysis)

The course is open to all streams. It is aimed at students interested in studying and understanding the human mind. It will enable students to develop new methods for advancing the current state-of-the-art in neuroimaging and data analysis so that they can take it up for higher studies. Students interested in doing a research project in this area are especially encouraged to take this course. The course will be a mix of lectures and a large number of assignments for practice. There will be a course project / term paper at the end (to be done in teams).

Prerequisites


Necessary:
Linear algebra, probability, statistics, signal processing, programming skills, data structures.
Desirable: Machine learning, algorithms, linear programming, C/C++, Matlab and/or R.

Interesting links

1.      “My Stroke of Insight”, Jill Bolte Taylor had a massive stroke, and watched as her brain functions as the stroke developed. Watch her Ted talk at: http://www.ted.com/talks/jill_bolte_taylor_s_powerful_stroke_of_insight

2.      “3 clues to understanding your brain” Vilayanur Ramachandran tells us what brain damage can reveal about the connection between cerebral tissue and the mind, using three startling delusions as examples. Watch the Ted talk at:  http://www.ted.com/talks/vilayanur_ramachandran_on_your_mind?language=en

3.       Predicting thought using fMRI, by Tom Mitchell, http://www.psc.edu/science/2010/brainactivity/

 

About the instructor

                Rahul Garg is has recently joined the department of CSE as a Professor after fifteen years of industrial experience at various places, including IBM T J Watson Research Center. His research interests are in the areas of Brain Imaging, Neuroscience, Machine Learning, Big Data Analytics and IT for Society. For more details, visit http://www.cse.iitd.ac.in/~rahulgarg