====== Body MRI data ======


In the following wiki page, you can find information about the local body MRI acquisition and processing and the link to the **CoMorMENT study**, including a description of datasets and information relevant for researchers at NORMENT.

==== Internal NORMENT procedures for working with body MRI data ====


=== Background regarding the processing of body MRI data ===

We started processing body MRI data for derived tabulated output of various body fat and muscle tissue metrics with AMRA/AMBRA in 2019 under the "OUH001 - OUH - psychiatric study" agreement. As of October 2021, AMRA/AMBRA process all our body MRI data under the CoMorMent contract (project "OUH003 - OUH - CoMorMent"). We also obtain pdf reports of all participants. For the batch analyzed under the CoMorMent agreement, we should also receive videos showing all participants' segmentations. 
AMRA is a partner of the EU CoMorMent project, and we analyze the body MRI data with AMRA/AMBRA under an agreement between CoMorMent/NORMENT and AMRA.

//Please note: Reanalyzing missing data in the future may be costly. Thus, we must secure this derived data for the future as we cannot easily regenerate it.//

=== Acknowledge CoMorMent on publications ===

Please note, since we analyze the body MRI data under the EU CoMorMent project, you should add the following acknowledgment to your publication:
European Union's Horizon2020 Research and Innovation Programme (CoMorMent project; Grant #847776).

=== Overview over datasets with body MRI data at Ullevål - status October 2021 ===
  * TOP (actively acquiring new data).
  * Stroke-MRI (not actively acquiring data). 
  * BRAINMINT (actively acquiring new data)
  * MoBaBRAIN (not actively acquiring data)

==== Contact persons ====

=== AMRA/AMBRA ===

  * AMRA/AMBRA: Clinical Project Manager, Anneli Tinnerholm: anneli.tinnerholm@amramedical.com

=== Internal MRI coordinators ===
  * TOP: Synne Øiestad
  * BRAINMINT: Henning Hoel Rise / Ingvild Sandø Lofthus (study coordinator).
  * MoBaBRAIN: Henning Hoel Rise / Ingvild Sandø Lofthus (study coordinator).
  * Stroke-MRI: No current study coordinator.

=== REDCap Database ===
  * Tiril P. Gurholt (t.p.gurholt@medisin.uio.no; Database specific things).
  * Sandeep Karthkeyan (sandeep.karthikeyan@medisin.uio.no; Technical aspects). 
  * Oleksandr Frei (oleksandr.frei@medisin.uio.no; Technical aspects).


**Please note: REDCap Database is work in process**

//Note: The idea is that REDcap gives an instrument for data management and for making it available to others. This is handled by the individual projects (i.e., here body MRI). Oleksandr/Sandeep will help by maintaining backups and upgrading the software and general advice on how to use it - but they will not manage the data for individual projects.//

==== Relevant REC Permissions ====

The relevant REC Permissions regarding storing of data & combining data with other datasets are as follows:
  * TOP has permission to combine data with StrokeMRI (REC 2009/2485, revised), MoBaBrain, and BRAINMINT.
  * Stroke-MRI has permission to combine data with TOP, MoBaBrain, and BRAINMINT. The data storage plans have been reviewed and approved by the Regional Committee for Medical and Health Research Ethics (REC), South East division in Norway (2014/694).
  * MoBaBrain & BRAINMINT have permissions in place to combine data with TOP and Stroke-MRI. The following is stated in their permissions: "In order to strengthen the scientific and clinical value of the study, we will, according to relevant approvals and consents, harmonize and analyze data from the current project with existing databases and cohorts and other ongoing projects with overlapping aims (REK 2019/345, 2016/1226, 2016/1756, 2015/1282, 2014/694, 2009/2485, 203645 [LoTECA]), which are all relevant in terms of overlapping aims and methods."

To Conclude: We have REC permissions to combine TOP, Stroke-MRI, MoBaBrain, and BRAINMINT AMRA-derived data in the same database.

==== Data storage ====

=== Data storage raw body MRI data ===

All projects (TOP, BRAINMINT, MoBaBRAIN, and Stroke-MRI) have access to storing their data on the same TSD-project (p33). The imaging data is being backed up on TSD. 

== Raw body MRI data: TOP == 
The TOP MRI coordinator is responsible for uploading the data to TSD. Additionally, the raw DICOM data is stored on CDs in the archive of building 48 (before scanner upgrade) & 49 (after scanner upgrade). 

The raw body MRI DICOMs is stored in the same fashion as the brain MRI DICOMs on TSD. The DICOM data are stored as follows:   

  * Before scanner upgrade (2020): 
    * On CDs locally archived at Ullevål hospital (building 48). 
    * On Wahanda (DICOMs lie unsorted) and converted to //nifti// (sorted). 
    * **Transfering the raw data to TSD is work in process.** 
  * After scanner upgrade (2020): 
    * On CDs locally archived at Ullevål hospital (building 49) 
    * On TSD: durable/groups/imaging/TOP/data/archive (upload in process).
    * //Note: MRI Caput & AMRA DICOMs are stored on separate CDs - both need transferring to TSD.//   

== Raw body MRI data: Stroke-MRI ==
The raw DICOM data is stored on 
  * On CDs locally archived at Ullevål hospital (building 48)
  * at TSD /cluster/projects/p33/wahanda/syn01/strokemri/raw  

== Raw body MRI data: BRAINMINT & MoBaBrain ==
The raw MRI data, including body MRI data, for BRAINMINT and MoBaBrain is stored at: /tsd/data/durable/groups/imaging/BRAINMINT/data/archives/ . 

=== Data storage AMRA derived data ===
Derived tabulated, pdf, and video data from AMRA
All body MRI data is currently being processed by AMRA/AMBRA. We have created one folder for storing the derived tabulated data in one REDCap database for TOP, Stroke-MRI, MoBaBRAIN, & BRAINMINT. 

It is as follows:
''/cluster/p/p33/cluster/groups/imaging/BodyMRI/''

Sub-folders will hold downloaded excel-spreadsheets from AMRA, and project-wise pdf reports for all participants, and video data showing the segmentations for data analyzed under the CoMorMent agreement ("OUH003 - OUH - CoMorMent").

Storage of ("OUH001 - OUH - psychiatric study") tabulated data & PDF reports is as follows (October 2021):

  * **TOP Excel file & PDF reports**
    * Path: ''/cluster/p/p33/cluster/groups/imaging/BodyMRI/TOP''

  * **Stroke-MRI Excel file & PDF reports** 
    * Path ''/cluster/p/p33/cluster/groups/imaging/BodyMRI/StrokeMRI''

==== Data uploaded to AMRA/AMBRA ====

=== "OUH001 - OUH - psychiatric study" agreement ===
Under this study agreement the local body MRIs were uploaded to the AMBRA portal via CDs stored in building 48 by:

  * Ingvild S. Lofthus uploaded the TOP body MRI data.
  * Dani Beck uploaded the Stroke-MRI body MRI data.

=== "OUH003 - OUH - CoMorMent" agreement ===
Under this study agreement, the local body MRIs were uploaded to the AMBRA portal by: 
  * Ingvild S. Lofthus is in charge of uploading the BRAINMINT body MRI.
  * Ingvild S. Lofthus is in charge of uploading the MoBaBRAIN body MRI data (status Jan 2022, not yet started). 
  * Synne Øiestad is in charge of uploading the TOP body MRI data. 


==== How to upload/download data to/from the AMBRA portal ====
BodyMRI data is uploaded to AMBRA through an online portal on the AMBRA website (https://access.ambrahealth.com/). A comprehensive guide detailing this process with images can be found here: https://ambrahealth.com/self-guided-demo/upload-a-cd-study/.

Briefly, you must first gain access to the online portal by creating an account. This is something AMRA Medical can help you with (so just contact an administrator). After you have an account, you log in and click on the "add study" tab, followed by "Upload studies" in the sub-tab. This takes you to the upload portal, where you are asked to insert a CD or choose a file. To elaborate, you can either insert the CD that includes the body MRI data or upload data from an external disk containing body MRI data (not including the brain MRI data). 

After choosing the files to upload, AMBRA scans through the contents and allows you to preview the data before uploading. If you're happy with the previews, you can proceed with "Upload selected studies". AMBRA will tell you if the upload has been successful, whereby you can go back to the overview panel of the portal to see the newly uploaded study in a work-list. Here, you can also edit the study to provide the study with information on ID, sex, DOB, and description of the scan. You can also go on "view study in viewer" to look at the body MRI images.

After AMBRA segments the images/data, you are provided with a PDF report for each study/subject added to the work-list. These PDF reports can be downloaded directly from the portal (and should thereafter be uploaded to TSD). An excel sheet with the segmented results of all the subjects in the work-list can/should be requested from AMBRA in order to get data on specific variables/measurements of interest used for analysis (liver fat, muscle, and fat tissue volume, etc.).

AMBRA will also provide you with a final excel summary, upon request. It will go through their internal quality control procedures before delivery, which takes at least 20 business days to complete.
