You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hello,
I've encountered similar problems as other people regarding multiband acquisition data. I checked past forum entries/issues and it seems like this hasn't been discovered yet. I'm using version R2022a-v8.1.0, so if this has been fixed already kindly forgive my ignorance.
Here is the bug I discovered:
I have 72 slices with an acceleration factor of 8. As such slices 0, 9, 18, ..., 63 were acquired at the same time (at the first point in time) so they are the first slices listed in the log-file. The toolbox instead assumes that slices 0-8 were acquired first. (This also applies to all following slices.) As such the calculated acquisition times are incorrect and accordingly the delays for different onset_slices are also incorrect. I have modified the source code to fit my needs (see appended files, mainly tapas_physio_create_scan_timing_from_tics_siemens) and to fix this bug. All modifications are marked with the original code also still visible. If you have comments/questions I'm happy to answer.
I thought you might find it interesting.
Kind regards,
Clemens
The text was updated successfully, but these errors were encountered:
poeplau
changed the title
Bug with multiband acquisition data (Siemens ..._Info.log)
Bug with multiband acquisition data (Siemens ..._Info.log) in PhysIO toolbox
Jun 26, 2023
Thank you for your contribution. In order to understand it better, would you be willing to share an example dataset and the script/SPM job you used to process it with PhysIO?
For sure! I have attached example log-files and the script I use (individual_regressors.m) aswell as some other modified scripts that are used by the main script (*_mod.m).
The main script (individual_regressors.m) is a modified version of tapas_physio_main_create_regressors.m. All modifications should be marked but they are completely irrelevant to the issue at hand. (They serve to output slicewise RETROICOR regressors as nii-volumes, oversampled versions of the physiological regressors at the original sampling rate and the raw cardiac/respiratory phase.)
The only relevant modifications are in the scripts I included in my first message, so you can safely ignore the rest. A standard example with the attached logs should run with the first round of modified scripts.
Hope this helps! If you have questions regarding any of my modifications feel free to ask!
Hello,
I've encountered similar problems as other people regarding multiband acquisition data. I checked past forum entries/issues and it seems like this hasn't been discovered yet. I'm using version R2022a-v8.1.0, so if this has been fixed already kindly forgive my ignorance.
Here is the bug I discovered:
I have 72 slices with an acceleration factor of 8. As such slices 0, 9, 18, ..., 63 were acquired at the same time (at the first point in time) so they are the first slices listed in the log-file. The toolbox instead assumes that slices 0-8 were acquired first. (This also applies to all following slices.) As such the calculated acquisition times are incorrect and accordingly the delays for different onset_slices are also incorrect. I have modified the source code to fit my needs (see appended files, mainly tapas_physio_create_scan_timing_from_tics_siemens) and to fix this bug. All modifications are marked with the original code also still visible. If you have comments/questions I'm happy to answer.
I thought you might find it interesting.
Kind regards,
Clemens
tapas_physio_create_scan_timing_from_tics_siemens.txt
tapas_physio_crop_scanphysevents_to_acq_window.txt
tapas_physio_get_cardiac_phase.txt
tapas_physio_get_sample_points.txt
The text was updated successfully, but these errors were encountered: