Local institutional MRI database for reproducible research

Hello, I am a neurologist seeking advice from the neuroscientists’ community.

My hospital is contemplating a local MRI register (database) to facilitate research. The idea is to design a “standard protocol” that is extended/tailored for computational analysis. Currently, all patients go through a clinical protocol for ~15 minutes, and it does not utilize, e.g., 3D FLAIR crucial for MS lesion segmentation. At this starting point, I want it to be done from a reproducibility standpoint. For sure, I am aware of BIDS, but before storing the data, I should identify if there are any guidelines/consensus on particular MRI acquisition protocols.

We identified several core pathologies that we work with: demyelinating disorders (MS, NMO), movement disorders (parkinsonian syndromes), chronic cerebrovascular disease (small vessel disease), and Alzheimer’s disease. Every pathology would receive its own “standard protocol.”

Existing guidelines on, i.e., multiple sclerosis imaging, are clinically based. I found information about HCP protocols and was curious if there are any pathology-specific MRI protocols with computational postprocessing in mind. I understand that there are no solutions that fit all purposes.

I will also be thankful for any information on “How to set up a local MRI database for dummies.”

Yours faithfully,