Congrats to Brian Rivera for receiving the 3rd place prize in the 2018 UA STEM Forward Conference! This is the third award Brian got in a month.
See http://training.ua.edu/stem/stem-poster.php for details.
Our two undergraduate RAs presented their posters in the 2018 Undergraduate Research & Creative Activity (URCA) Symposium on March 28, 2018. Congrats to both!
Rachel Remmes's poster is titled "Structural Brain Features Associated with Autism in Children: A VBM Study" and she received a second-place award in the education section. Double congrats to Rachel for getting this award for two years in a row!
Jakub Denkiewicz's poster was titled "Structural Brain Features Associated with Mild Cognitive Impairment in the Elderly: A VBM Study"
Here is a very simple Python script to extract behavioral data from an Elist.txt file, produced by ERPLAB. The output is a tab delimited txt file which can be opened in Excel and used to create a pivot table/chart. Click the link below to download the file (change the extension to .py). The script is also copied below.
Public Dataset: The Differential Relationship Between Finger Gnosis, and Addition and Subtraction: An fMRI Study
We have publicly shared the data from a recent study(Soylu, Raymond, Gutierrez, & Newman, 2017) through the Harvard Dataverse. This dataset includes raw fMRI data from 24 second and third graders, and the analysis files (durations and the second-level unthresholded & thresholded spmT files). The tasks include single-digit addition and subtraction. Finger gnosis scores are also included to be used as a covariate of interest.
Link & Citation for the Dataset:
Soylu, F. & Newman, S.D (2017), Public Dataset: The Differential Relationship Between Finger Gnosis, and Addition and Subtraction: An fMRI Study", Harvard Dataverse, doi:10.7910/DVN/I7KP3V
Soylu, F., Raymond, D.R., Gutierrez, A.M., & Newman, S.D. (2017). The differential relationship between finger gnosis, and addition and subtraction: an fMRI study. Journal of Numerical Cognition, 3(3), 694–715. doi:10.5964/jnc.v3i3.102
Switching gears, after posting scripts for fMRI data analysis in the last two posts, in this post I will share a MATLAB script I developed for ERP (Event-Related Potentials) analysis, using ERPLAB. This script is based on example scripts posted by Steve Luck's lab (see https://github.com/lucklab and http://www.erpinfo.org/erplab.html).
For the script to work, the main data analysis folder needs to be structured in a specific way. You can either use this structure or change the paths in the script. Here is how one of my data analysis folders look like:
The script file should go under "scripts" (well, obviously). Not all of the folders listed above are used by the script. You only need the subject folders, "AveragedFigures" (guess for what), "BinFiles" (check the ERPLAB documentation for how to create bin descriptions and bin equations files), and "GrandAvgERPs" folders. Before you run the script ensure that you run EEGLAB/ERPLAB once on the console and then close it (not MATLAB, just EEGLAB). Again, you won't be able to run this script as is, since the parameters in this script is set for a specific study. But, this can give you a starting point for building a script that involves all stages of ERP data analysis. If you are using ICA you will need to first run it and visually inspect the components that you want to get rid of, and enter those components in the script, separately for each subject. This is also true for interpolating electrodes (e.g., if you have a bad electrode for a specific subject, enter the electrode number in the list location for that subject). Feel free to post any questions in the comments section. Good luck!
If you are using SPM for fMRI data analysis and entering onsets and durations for each condition in the first level manually, then you should consider using the "multiple conditions" option, and input a .mat file that has all the onsets and durations, separately for each subject. To do this you can use a script that takes the log files (from your stimulus presentation software), extracts the onsets and durations, and outputs a txt file (with .m extension), formatted in a way that Matlab can read and convert into a .mat file.
Here I present a Python script I put together to extract the onset and duration data from PsychoPy logfiles and to create the .m files formatted for Matlab. Even if you don't use PsychoPy you might still benefit from the parts of the script that creates the .m file.
You can download the Python script from the link at the bottom (change the extension to .py once you download). You will need to tweak it a bit before it can work with your logfiles.
This is how the output .m file (for one subject) should look like. There are three conditions (2 task, 1 baseline) and for each condition you have a separate sublist of names, onsets, and durations.
I developed a set of MATLAB scripts to analyze the fMRI data for a project I am working on. Instead of postponing sharing this on the blog until I cleaned it up and make it more presentable, I decided to post it in its raw form. While developing these scripts I took advantage of countless online posts from good samaritans to figure out numerous problems and bugs I ran into. They are too many to list, but I am thankful to all who take the time to share their fMRI analysis wisdom online.
For each level of analysis (preprocessing,1st level, 2nd level) the main scripts call sub-scripts that have the code specific for each step of analysis. The scripts require a certain folder structure to run. You can either structure your folder similarly, or change the paths in the scripts.
This is how the folder are structured in my analysis:
Feel free to use, change, share, distribute any material presented in this blog.