I am trying to implement a realtime fMRI neurofeedback application for the 7T Siemens (Terra.X) machine at my current institution. The scan sequence is a built-in Siemens simultaneous multi-slice (SMS) protocol. We are exporting the EPI data using a BOLD add-in feature (syngo MR XA30 platform) to our processing computer with AFNI (v24.1.06 April 26, 2024) running Ubuntu.
I am building off of the scripts that I successfully implemented in 2018 on the 3T Siemens Prisma at the NIH Clinical Center (with Vinai Roopchansingh). Everything seems to be working properly up to a point (AFNI opens in RT mode, computer receives DICOMS, Dimon (v.4.32) sends incoming DICOMS to AFNI). As far as I can tell, AFNI seems to be interpreting each volume (76 slices, TR = 1) as 76 individual volumes instead of slices. I have attached an image of the real time plot AFNI generates by default. The Dimon output below lists the volumes correctly, but lags (does not move to the next volume until AFNI finishes plotting the 76 slices).
Dimon output:
ART: comm link to afni established at
-- scanning for first volume
++Data detected to be oblique
**CID: have non-mosaic <full_file_path>.dcm with 76 images
*+ WARNING: Bad DICOM header - assuming oblique scaling direction!
**CID: have non-mosaic <full_file_path>.dcm with 76 images
no more warnings will be printed...
-- first volume found (1 slices)
-- scanning for additional volumes...
-- run ##: 1 2 3 4 5 6 7 8 9 10 11 12
Note that the script I run uses Dimon as follows:
Dimon -rt -quit -inf
I tried modifications to Dimon (below), and in non-real time mode, which produced the same behavior as described above.
"-num_slices 76"
"-assume_dicom_mosaic"
Is there something else I should implement so that AFNI parses the incoming data appropriately, or is this novel behavior? Thanks for your help.
-Samantha Fede