more e-prime/ psychophysiology integration

[Paul Gr]

Dear Ilya,

It’s a bit hard to cover all these topics completely in a single answer, so 
let me focus on a few important issues:

Integrating BioPac and EPrime is not very different from most systems a have 
encountered so far. In most cases two separate computers are being used: one 
for running tests or stimulus software (i.e. EPrime), and one for data 
acquisition (i.e. collecting data from BioPac.) In most cases the only link 
between the two systems is a synchronization connection, which is typically 
realized between the parallel printer port of the EPrime PC and the isolated 
digital input lines of the BioPac hardware (i.e. the STP100C module.) I 
think a suitable interface cable is also supplied with the current (new) 
modules. The link between the two systems is one-way only and can only be 
used to send up to eight digital signals from your EPrime script by using a 
few simple lines of inline script. These 8 signals are available as 8 
separate input signals in BioPac and can be viewed and stored just like any 
other physiological signal. Because there are 8 independent signals 
available it is possible to encode information (such as stimulus type) by 
sending different numerical values to the printer port. In other words: the 
8 lines can be view as a way to send and record event codes. To check for 
any timing errors or delays we performed some tests in our lab. One of these 
tests was a check to test for timing differences between the digital and the 
analog signals. (This is very important for synchronization or event codes!) 
The only thing we found out is that our STP100C (which is not exactly the 
same as currently supplied) had slightly different electrical specs on the 
first digital input line (the one which can also be used as trigger signal.) 
  Although it should be possible to start data acquisition by activating the 
trigger input (one of the 8 digital input lines), we typically start the 
recording manually before starting the EPrime script. While running the 
script, we just send sync. codes at relevant moments to the printer port to 
take care of the event marking. (You could also include one ore more 
instruction screens as part of the EPrime script to instruct how to 
initialize and start a recording.) However, there is one type of event that 
is not very trivial to encode during the recording: subject responses! This 
is because you will have to take care of sending the proper codes at the 
right moments yourself because EPrime currently doesn't support automatic 
forwarding of such events. (Version 2 hopefully does!)

Because the communication between the two systems is limited to a few 
one-way digital lines, it is not possible to use any information about the 
physiological signals at runtime in your EPrime script. (I.e. biofeedback is 
not an option.) You will probably use the digital markers during the 
analysis phase of the physiological data to select relevant fragments of the 
recording. Unfortunately the BioPac analysis software doesn’t have proper 
tools to automate such analysis procedures, so you will have to process a 
lot of the data manually, or convert the data to other analysis packages. 
Anyway, in most cases you will only merge EPrime output and physiological 
data after preprocessing the raw signals to extract the desired information 
(such as average heart rate during a specific period.)


hope this is of any help
Paul Groot


>From: Ilya Korsunskiy <ilyako at eden.rutgers.edu>
>To: eprime at mail.talkbank.org
>Subject: more e-prime/ psychophysiology integration
>Date: Thu, 30 Jun 2005 15:47:39 -0400 (EDT)
>
>Hello e-prime community,
>
>I sent out a query earlier pertaining to contact information from those 
>labs currently or planning to integrate e-prime with psychophysiological 
>recordings (namely through biopac). We are now moving into the initial 
>steps of integration and have come up with the following list of questions 
>to guide our progress:
>?	Do we integrate during the experimental or during the analysis phase?
>?	What kind of inputs would E-prime accept and where would it fit in the 
>software?
>?	How can we synchronize E-prime and biopac to start at the same time on 
>separate computers?
>?	How well can we depend on time precision between two separate operating 
>systems?
>?	How do we convert the biopac data to compare it to the data collected in 
>E-prime (or vice versa)?
>Any helpful answers, questions or comments would be greatly appreciated.
>
>Best
>Ilya Korsunsky
>