RE: Open-source MRI hardware initiative project
Dear all,
Thanks to all your positive answers!
I did take some time to process the data received and to answer back (my apologies): this is a project I would really like to make happen, but at present I have no resources to work on it so only little time to manage it.
I realise that Debian Med is doing packaging work, so some of my questions may not have been adapted to you. However I had some very interesting ideas from several members, here is a summary here for those interested. I am also thinking to regroup all this into the community portal once it is created.
Regarding the community portal:
- I think I will start with Github to host the project, as it is a well-established company that offers good help (especially if I can interest the Github team).
- A wiki is a good idea to collect data (already existing projects, ideas, etc...) and a mailing list for contacts.
- Readthedocs.org certainly seems a good site, I will have a look at it too!
- Unfortunately my University won't help on this, the IT services are already saturated...
For the system requirements, I gathered the following:
- compatibility with PACS and DICOM. I have been following Orthanc for a while now, and I am currently thinking to integrate it to our current MRI scanner.
- adding export options and aim for modularity
- real-time acquisition capabilities: this will depend on the console development. I am currently in touch with a group at the MIT who is developing an open-source MRI console, I need to get back to them regarding this.
- simulators: I knew of ODIN, and a few others such as spindynamica. I keep an eye on these but I think that developments in these directions will require additional workforce. I agree that it is a good idea to include these into the workflow.
- There already are some very good OS software for image processing (https://mrirecon.github.io/bart/) and pulse sequence programming environment (http://wisdmhub.org/dev2/apps/sequencetree/)
About the project management side:
- start small and fast. Let's have this portal done this week!
- The short-term objective for now is to regroup research units interested in MRI hardware development to help them build their scanners.
- The mid-term objective is to build an open-source MRI platform ready for research use, with tools that easily allow collaborative work (sharing pulse sequences, data, etc...)
- The long-term objective is to make this platform clinical and to have it produced by manufacturers (this takes typically 10 to 20 years with proprietary developments, just to give you an idea)
- clearing out license issues as soon as possible: I need to talk to someone specialised in open source licenses for this, as I know that there are issues specific to open-source hardware.
- pay particular attention to public domain hardware: I already know of several sources of hardware designs that are in the public domain, and can be re-used. I am not aware of the licensing issues for the designs that are produced this way.
- I received an interesting book reference, that I plan to read: Karl Fogel, Producing Open Source Software: How to Run a Successful Free Software Project. http://www.producingoss.com/
- involving companies will be required at some point if we are to make clinical devices. Research scanners can be build and used without too many problems (at least in the UK) but clinical certifications are required for medical use, and they cost a leg.
- involving clinicians early may help to focus on the aspects most needed for the developments. I already have several clinicians interested in my work here, so this could be a bonus if I can have them to check the project progress on occasions.
- it certainly looks like there would be some interest in this initiative in the neuroscience world, though this implies high-field MRI, and therefore supraconductor magnets, which I have little experience with. Therefore I would need some help to explore that route.
- organise 'sprints' to push project development on practical work
On the money side:
- INCF seems interesting, though this is too early a stage to really push forward a specific application, unless I manage to convince fMRI people to join in. to be confirmed.
- I had a look at EPSRC, the British public agency for physics research. They may help to start a small-scale work, which would be a good start.
- I may also benefit from various European COST actions, I have to check which ones are relevant.
About community-building:
- I had very interesting info about community building as well: it seems like the most helpful members in open-hardware projects are not always professional, nor experts in the field. This came to me as a surprise, and definitely has to be taken into account.
- It looks like there already are people working on open-source MRI at a 3D-printable scale (https://hackaday.io/project/5030-low-field-mri) and other interesting projects to look at. I shall start by listing these on a wiki.
- Advertising at FOSDEM seemed a good idea, I will aim to make a presentation next year. I can also put something for ISMRM or ESMRMB, depending on how things go.
Ok, time to act now! Thanks again for all the help, I will send a message to let you all know the community portal once it is done.
All the best,
Lionel
The University of Aberdeen is a charity registered in Scotland, No SC013683.
Tha Oilthigh Obar Dheathain na charthannas clàraichte ann an Alba, Àir. SC013683.
Reply to: