Pooling functional MRI data from different sites and points in time should allow for increased statistical power, provided that the imaging experiment can be conducted in a known and reproducible way. Unfortunately, this has so far proven difficult, since the details of the vendor-provided fMRI acquisition and reconstruction software are generally not known to the researcher.
This repository contains a fully vendor-agnostic and portable fMRI protocol that ensures identical sequence execution and image reconstruction across different scanners, and across scanner software upgrades. At present, Siemens and GE scanners are fully supported. The acquisition sequence is defined in a simple (and human readable) text file in the Pulseq file format, and sequence execution on real hardware is made possible with sequence-agnostic interpreters that we provide:
- Contact the HarmonizedMRI study team for access to the interpreter for your scanner.
- Ask the study team for the sequence file 'smsepi.seq', or create your own using the 'writeEPI.m' MATLAB function that we provide. The run it on your scanner.
- Reconstruct the data using slice GRAPPA code that we provide.
The ./example/ folder in this repository contains a detailed workflow that implements all steps needed to implement the above diagram (expect for scan instructions which we provide elsewhere).
If you have an interest in multi-site fMRI or want to prototype new EPI fMRI methods,
our Pulseq SMS-EPI fMRI protocol may be for you.
If you'd like to give it a quick try for evaluation purposes,
or are already sure you want to use it for your fMRI studies,
send an email to jfnielse@umich.edu and we will work with you
to install the Pulseq interpreters and the sequence and reconstruction code.
At present, we ask that if you plan to use our Pulseq protocol in your fMRI studies,
your scanner is also equipped with the corresponding SMS-EPI product sequence.