Saturday, May 4, 2013

Crowdsourcing my next fMRI task

For those of you who have been following my self-tracking study, I am happy to announce that we are once again collecting data as of April 30.  I had a followup audiometry exam this week which showed that the poor reading in early March appears to have been a fluke, and that my hearing does not seem to have changed since the beginning of the study.

We have decided to make a number of changes to the MRI protocol based on our analyses of the first 6 months of data. In particular, we are going to reduce the frequency of the structural and DTI acquisitions (because, as expected, we don't see much in the way of changes over time).  In addition, after acquisition of 14 sessions on a working memory task, I have decided to reduce its frequency and add another task fMRI paradigm that will be collected once a week.

My question for you is: What should I do next?  I have several things in mind, but I want to see if the community can do even better.  If you have an idea for task fMRI paradigm where a large number of sessions from a single individual would be useful, please email me (poldrack at utexas dot edu) with a brief proposal that outlines specifically why it would be interesting to perform the task repeatedly on a single person and how many sessions it would require.  I will work with the winner to implement the task, and that person will of course be offered coauthorship on any papers that include those data.  There are two requirements:

  • Each session of the task must be accomplished within a single 10-minute imaging run.
  • The task cannot require any special response or stimulation devices (for stimulation we have video projection and headphones; for response, we have several button boxes available as well as eye tracking, noise-cancelling microphone, and physiological monitoring).
I look forward to hearing your ideas! 


  1. Any tasks about what internal thoughts you are having and how they are affecting your well being?

  2. Of course it's most interesting to see things that change. I can think of a few: day to day stress levels, performance/structure/function with learning a challenging task, long term atrophy (hopefully not too much!), perhaps hematrocrit - but's that's hard to get a sample for each day. For the extra task, I might perform a difficult continuous attention task that will show day to day variation and that has enough dynamic range in performance to show changes for up to a year. From heart rate (measured during the scan as well?) you can get heart rate variability - a sensitive indicator of stress levels.

    1. Peter - thanks for the HRV suggestion, that's a good one. We have not been collecting HR because of the logistic hassles of setting it up, but I think it would be a good idea.

  3. This is an amazingly interesting study. Do you encounter also for annual fluctuations, since particularly in males hormone levels etc. underlie an annual rhythm?

    Also I would like to suggest a task: alternate between sessions: During the first session study a poem, during the second recall it. Two reasons for that: both tasks are hippocampus dependent, but might integrate different other areas as well and blood flow might differ strongly between study and recall sessions, 2nd it is known that learning increases grey matter in the hippocampus over time, but it has never been demonstrated in such a direct way longitudinally.

  4. Dear Russ,

    If you are still looking for ideas, - how about something from the other side of brain - something from neuro regulation

    For example, spent 10 min doing some sort of Vagal maneuver, choose any method that works best for you and the situation..maybe breath-holding ..

    measure your blood pressure/pulse before and after