Computational Neuroimaging

Psych204B: Human Neuroimaging Data Analysis Methods

This course  will introduce the student to data analyses and design considerations of human neuroimaging experiments. Emphasis is on building deep understanding of the underlying signals and computational approaches. The goal is to not only gain mathematical understanding of data analysis choices but also understand their consequences on interpreting brain function.

The course format will include directed readings, lectures & discussions related to the reading,  hands-on matlab tutorials, and fMRI data analyses on example dataset. 

The course will begin with fundamental topics including understanding the nature of fMRI signals, temporal and spatial resolution; signal to noise in fMRI and preprocessing; general linear models; statistics of fMRI. Then, the course will introduce advanced approaches including  multivoxel pattern analyses, representational similarity analyses, as well as decoding and encoding algorithms. Finally, we will discuss experimental design considerations and how to combine data across participants 

Meetings: Monday 3-5:50pm. 

Given the online format we will have weekly zoom meetings;  Each meeting will have 2 parts with a 10 minute break in between.

Instructor: Prof. Grill-Spector

Contact: kalanit@stanford.edu; zoom office hours: schedule by email

TA: Lester Tong.

Contact: lctong@stanford.edu; zoom office hours: schedule by email

Prerequisites: Psych 204a

Recommended: Cognitive Neuroscience;  Matlab.

Textbook:  Functional Magnetic Resonance Imaging   Huettel et al., FMRI (Third Edition, Sinauer).  The book can be rented or purchased.

Roadmap for readings and assignments:

The course is organized by modules, where each module contains the relevant reading, discussion questions, and matlab tutorial/data analyses.

Modules for each week:

Week 1: From Neural Responses to fMRI Signals

Week 2: Spatial and Temporal Resolution of BOLD Responses

Week 3: Signal and Noise in fMRI: quantification and preprocessing

Week 4: Linear Systems Approach and the General Linear Model (GLM)

Week 5: Statistics for fMRI Data Analyses

Week 6: Multivoxel Pattern Analyses and Decoding

Week 7: Encoding Approaches for fMRI Data

Week 8: Experimental Design and Group Analyses

 

Required software and environment:

1) Linux Machine and/or Mac

2) Matlab 2016 and higher

3) VistaSoft: download here

If you do not have (1) or (2) email me.

Spring 2020 updates

Spring 2020 is a unique quarter.  Due to the covid-19 situation, the course this year is going to be on zoom rather than in person. I have adjusted the course to fit this format. For example, typically, we would have gone to the CNI to scan an example dataset for the class. Instead this year, we will  use an example dataset that I have collected in the past. I believe that  despite the online format we will accomplish our learning mission. I look forward to an interesting quarter.