The Get Your Brain Straight hackathons bring together neuroimage data generators, image registration researchers, and neurodata compute infrastructure providers for a hands-on, collaborative event. This community collaboration aims to create reproducible, open source resources that enable discovery of the structure and function of brains.
There are three components to the hackathon. First, the primary goal of each hackathon is the generation of a Reproducible Resource for registration and analysis of a specific brain imaging modality. Tutorial sessions share how to work with open source registration tools, open access datasets, or neurodata archives. Birds-of-a-Feather (BOF) Breakout sessions enable participants interested in collaborating to work on relevant topics.
Example ways to participate:
- Contribute a registration pipeline resource that can be deployed on Brain Imaging Library (BIL) resources to register the challenge dataset.
- Give a tutorial about your registration tools. A pre-recorded or live presentation, along with example code and recipes, can teach a data analyst how to run your tool and use the output on the challenge dataset and/or another open dataset.
- Come and learn about registration from experts in the field.
-
Dates: Monday, April 4th - Thursday, April 7th, 2022
-
Location: The first hackathon will be online, held on Zoom videoconferencing, Image.sc Island Gather.Town virtual space, and Image.sc Zulip Chat.
-
Registration: Fees: none, it's free! Use this form to register. Deadline: April 2nd.
-
Communication: to receive information about this and future events please join the Hackathon Mailing List.
- Register for the event. There is no cost.
- Sign up for the mailing list.
- Set up an account at the Brain Image Library (BIL).
- Optionally prepare a tutorial or organize a Birds-of-a-Feather (BoF) Breakout, described below.
The week will start 8 AM Pacific Time, 11 AM Eastern Time, Monday, April 4th in an introductory all-hands videoconference. recording
Following the introduction, participate in the Reproducible Resource Challenge, join the tutorials, and participate BoF breakouts.
On Thursday, 11 AM Pacific Time, 2 PM Eastern Time, participants will delegate one member to present their registration processing pipelines, results, and discuss lessons learned. Recording
Get Your Brain Straight hackathons are open to all and publicly advertised. Email announcements are sent to the mailing list.
How to add this calendar to your own?
This aim of this hackathon is to generate reproducible pipelines to register whole-brain microscopy image data to the CCFv3. Two datasets are provided, each with their unique quirks. You may work on either dataset during the hackathon.
In order to work with the neuroimage data generators, these pipelines will take a standardized input without assumptions of directory structures, filenames, etc and generate standardized outputs. Expected outputs include: resampled brain, spatial transformation, and a manifest of outputs. The processing pipelines should be designed to executed in independently in parallel. The output should be a resampled image with the same size, orientation, and origin as the provided CCFv3. The output should include an affine transformation file, and a deformation field transformation file to transform SWC and/or annotation files from the challenge dataset image space into the CCFv3 space.
Criteria for inclusion in a summary paper:
- Open source with an OSI-approved license
- The code can be executed in the future
- Researchers can understand what the code is doing
- Researchers can extend or fix as needed
- Works on open standard data formats used by data providers and consumers
- The provided input NIFTI or OME-NGFF images
- Provide outputs in open standard formats on the BIL at
/hackathon/2022_GYBS/output/<team-name>/<dataset>/*.
- Deployable
- Can be executed across many environments
- Provided in a published singularity image
- Can be executed by an independent analyst on the BIL
The primary goals for this hackathon is to ensure that everyone's code can run on the dataset provided and can be replicated.
In future hackathons, we will focus on:
- Combine registration results and methods
- Quantification and characterization of deformation patterns in fMOST imaging
- Identify biologically relevant regions where improvements need to be made
- Focus on accuracy quantification leading to potential improvements
How to add a new reproducible registration processing pipeline?
The fMOST brain volumes and CCFv3 atlas for the hackathon are available on the BIL here. The input is a single fMOST NIFTI brain volume. Details about this dataset can be found here.
The Light-sheet imaged brain volumes are available
on the BIL here. For this challenge
set, the data and atlas is provided for a single brain hemisphere (for size considerations and to
mimic most real world experiments). Multiple data replicates are provided. We provide one
dataset subject0.zarr in full imaging resolution for those seeking a bigger challenge. As a
convenience we provide the challenge sets at approximately atlas size as
Nifti(subjectN_25.nii.gz)
or OME-NGFF (subjectN_25.zarr). Participants may use either format but
OME-NGFF is preferred because it allows more efficient memory usage for larger datasets.
Details on working with the data can be found here
Tutorial sessions share how to work with open source registration tools, open access datasets, or neurodata
- Brain Imaging Library (BIL), 1 PM-3 PM ET, (Ivan Cao-Berg, Greg Hood, Alex Ropelewski) recording
- Get Your Brain Pipelined, 3 PM ET, (Jeff Duda, Min Chen, Jim Gee) recording
- About the Challenge Dataset, 3:30 PM ET, (Lydia Ng) recording
- Registering Cleared Tissues, 4 PM ET, (Ricardo Azevedo) recording
- ITKElastix Image Registration Tutorial, 10:30-11 AM ET (Viktor van der Valk, Matt McCormick) recording
- Image Registration with a Maximization Minorization (MM) Optimization Algorithm, 12:30-1 PM ET (Daniel Tward, Gary Zhou, Ken Langea) recording
- Metadata Preservation for Image Registration, 1-2 PM ET, (Matt McCormick, Lydia Ng, Dženan Zukić) recording
- Your Transform Object Needs Metadata, 4-4:45 PM ET (Hastings Greer) recording
- OME-NGFF: Towards a community standard image file format for sharing big image data in the cloud, 8-11 AM ET, (Christian Tischer, Josh Moore) - Sister event -- requires additional registration Recordings: Introduction to OME-NGFF next generation file format, Practical part 1, Practical part 2
- Mapping Micro to Macro Scale Anatomy with LDDMM, 12-1 PM ET, (Kaitlin Stouffer, Bryson Gray, Michael Miller, Daniel Tward) recording
Birds-of-a-Feather (BOF) breakout sessions enable participants interested in collaborating to work on relevant topics.
To lead or join a Birds-of-a-Feather (BoF) breakout session, create or join a topic in this spreadsheet. During the BoF, find the leader by clicking on their name in the Image.sc Island Gather.Town and moving towards their avatar with the keyboard arrow keys. When you are close to their avatar in the virtual space, you will be able to see, hear, and talk to each other.
If notes are taken during the BoF, please add them to the BoF breakouts folder. We recommend HackMD for collaborative, well-formatted notetaking.
Participants and contributors are expected to adhere to the ITK Code of Conduct.
Registered participants:
- Yufei Chen, AHU
- oylei, AHU
- Alice, Allen Institute for Brain Science
- Pamela Baker, Allen Institute for Brain Science
- Lydia Ng, Allen Institute for Brain Science
- Rachel Dalley, Allen Institute for Brain Science
- Li, Anhui University
- Yuxiao Zhang, Anhui University
- tingtinghan, Anhui University
- Tingting Han, Anhui University
- Jesse, Anhui University
- Liyunayuan, Anhui University
- Iana Vasylieva, Center for Biologic Imaging, University of Pittsburgh
- Adrian Arias Abreu, Centre for Genomic Regulation
- Stuart gano, Google
- John Bogovic, HHMI Janelia
- Bin Duan, Illinois Tech
- Can Ceritoglu, JHU Center for Imaging Science
- Kaitlin Stouffer, Johns Hopkins University
- Thomas Athey, Johns Hopkins University
- Brock Wester, Johns Hopkins University Applied Physics Laboratory
- Erik Johnson, Johns Hopkins University Applied Physics Laboratory
- Dzenan Zukic, Kitware
- Tom Birdsong, Kitware
- Will Schroeder, Kitware
- Matt McCormick, Kitware
- Paul Elliott, Kitware
- Ebrahim Ebrahim, Kitware
- Viktor van der Valk, LKEB - LUMC
- Brian Eastwood, MBF Bioscience
- Kara, N/A
- Fae Kronman, Penn State University
- Yongsoo Kim, Penn State University
- Art Wetzel, Pittsburgh Supercomputing Center
- Mariah Kenney, Pittsburgh Supercomputing Center
- Luke Tuite, Pittsburgh Supercomputing Center
- Greg Hood, Pittsburgh Supercomputing Center / CMU
- Alexander Ropelewski, PSC
- Xin Wu, RTI International
- Zhixi Yun, SEU
- Adam Aji, SonoVol
- Yufeng Liu, Southeast University
- Tony Reksoatmodjo, Translucence Biosystems
- Kate Lawson, UC Irvine
- Ricardo Azevedo, UC Irvine
- Negin, UC Irvine
- Jaclyn Beck, UC Irvine
- Zhongkai Wu, UC San Diego
- Christopher Choi, UCLA
- Daniel Tward, UCLA
- Ian Bowman, UCLA BRAIN
- Luis, UCLA BRAIN
- Karl Marrett, UCLA Computer Science, VAST lab
- Guorong Wu, UNC Chapel Hill
- Hyejin Yang, UNC Chapel Hill
- Sarah Khan, UNC Chapel Hill
- Marc Niethammer, UNC Chapel Hill
- Carolyn McCormick, UNC Chapel Hill
- Min Chen, University of Pennsylvania
- Jeffrey Duda, University of Pennsylvania
- Alan Watson, University of Pittsburgh
- Nick Tustison, University of Virginia
- brian, University of Virginia
This hackathon is supported by the National Institute of Mental Health (NIMH) of the National Institutes of Health (NIH) under the BRAIN Initiative award numbers 1RF1MH126732, 1U19MH114830-01, 5R24MH114793-02, 1U24MH114827-01 and the BICCN.