[Neuroimaging] Planning for data formats - upcoming journal club

[Qianqian Fang]

Dear list, just want to share this offline self-intro to Matt, following 
the thread of data format discussions - I look forward to the journal 
club discussions tomorrow.



-------- Forwarded Message --------
Subject: 	Fwd: data formats
Date: 	Thu, 11 Nov 2021 15:03:17 -0500
From: 	Qianqian Fang <q.fang at northeastern.edu>
To: 	matthew.brett at gmail.com
CC: 	Guillaume Flandin <g.flandin at ucl.ac.uk>



hi Matt,

Guillaume forwarded me a thread of discussions 
<https://mail.python.org/pipermail/neuroimaging/2021-November/002365.html> 
regarding an upcoming journal club on data formats. I tried to register 
to the mailing list but it has not been approved (the list is moderated).

just a quick self-introduction - I recently received a U24 award 
(NeuroJSON, http://neurojson.org) from the NIH on developing JSON-based 
universal data exchange formats (with NoSQL db support) for the broad 
neuroimaging community. Robert Oostenveld (BIDS), Guilaume Flandin (SPM) 
and a few other project leads (freesurfer, brainstorm, homer ...) are in 
our workgroup.

You can find more info about the NeuroJSON project on our websites and a 
recent poster


https://pbs.twimg.com/media/FCD_JNtWQAgLq6N?format=png&name=4096x4096 (a 
poster at fNIRS2021)
http://openjdata.org
http://neurojson.org  (partially completed)

the idea behind is very simple - we want to make data human-readable, 
universally parsable with minimal overhead, scalable, and flexible for 
future extensions. We have been developing 1-to-1 mapping specifications 
to mirror neuroimaging data files to JSON (JNIfTI 
<https://github.com/NeuroJSON/jnifti/blob/master/JNIfTI_specification.md>: 
nifti->json, JSNIRF 
<https://github.com/NeuroJSON/jsnirf/blob/master/JSNIRF_specification.md#jsnirf-format>: 
SNIRF 
<https://github.com/fNIRS/snirf/blob/master/snirf_specification.md>->json 
etc); the JSON serialization of complex data is based on a lightweight 
annotation specification called JData 
<https://github.com/NeuroJSON/jdata/blob/master/JData_specification.md> 
- supporting N-D strongly-typed binary arrays 
<https://github.com/NeuroJSON/jdata/blob/master/JData_specification.md#annotated-storage-of-n-d-arrays>, 
complex 
<https://github.com/NeuroJSON/jdata/blob/master/JData_specification.md#complex-valued-arrays>/sparse 
arrays 
<https://github.com/NeuroJSON/jdata/blob/master/JData_specification.md#sparse-arrays>, 
trees, graphs, etc. Ultimately, we will map all BIDS folder structure to 
a JSON tree and optionally push to a NoSQL db (mongodb, couchdb ...) for 
scalability and dissemination.

We currently have lightweight parsers/writers to process JData-styled 
annotations 
<http://openjdata.org/wiki/index.cgi?JData/Examples/Basic#2_D_arrays_in_the_annotated_format> 
for Python (pyjdata <https://pypi.org/project/jdata/> and pybj 
<https://pypi.org/project/bjdata/>), MATLAB/Octave (jsonlab 
<http://iso2mesh.sourceforge.net/cgi-bin/index.cgi?jsonlab>), javascript 
(jsdata <https://github.com/fangq/jsdata/blob/master/jdata.js>) and C/C++.

I'd love to participate this discussion, either as part of the journal 
club, or schedule a separate meeting, and present our project 
roadmaps/resources available.

let me know if you are interested.

Qianqian
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.python.org/pipermail/neuroimaging/attachments/20211124/f18570f6/attachment.html>