Here are reports by two participants from Neuroinformatics Japan Center, RIKEN BSI, Yoshihiro Okumura and Alexander Woodward below.
Report by Yoshihiro Okumura
Report by Yoshihiro Okumura (See in Japanese)
Report by Alexander Woodward
The annual INCF congress, Neuroinformatics 2017 was held at the DoubleTree by Hilton Hotel, Kuala Lumpur, Malaysia from August 20-21. A number of interesting talks managed to be fit into the two days in which the congress took place. Some of the highlights were as follows:
Satrajit S. Ghosh commented that science is like manufacturing; we are transforming data to give an end product. This product is currently in the form of a publication, but he suggests the future will be the entire (scientific) process, i.e. reproducible research is the code plus the data.
Chris Gorgolewski (See Ref) talked about the importance of big data and how there is a long history of progress being made by the analysis or reuse of data collected many years prior. Data can drive progress and he suggested ways to promote data sharing by offering incentives to research groups such as ‘badges’, citations for opening their datasets, or analysis of their data. He presented NeuroVault (Ref) as an example of the work he is doing to provide a public repository of brain data.
Lalet Scaria presented an interesting talk on reproducibility and how to quantify the effect that a change of operating system (O.S.) has on the results of different neuroscientific software packages. It was found that different O.S. versions give different results for tools such as Freesurfer. In the future there will be a need to identify and fix the numerical stability of pipelines and perhaps standardize compiler options.
Randy McIntosh presented The Virtual Brain (TVB), a community project for brain simulation that can model intrinsic activity (resting state), brain stimulation and diseases such as stroke. It is also being used for epilepsy simulation and soon for surgical planning. TVB focuses on large-scale brain connectomics and the simulated dynamics that the tool can generate has given insight into optimal coupling strengths for information transfer, network formation and dissolution, and the effect of time delays. Relevant to TVB are questions such as how to scale between different resolutions of models and how do you embody a brain simulation.
Viren Jain (see Ref) gave an impressive talk on the work his lab has been doing for processing electron microscopy (EM) at large-scale. He described Flood-Filling Networks as a state-of-the-art image segmentation algorithm and its application to EM data.
At the conclusion of the congress, attendees were treated to a cultural Malaysian performance of song and dance. This was very exciting, displaying the rich history and blending of cultures that make up the country. Overall the congress was a success with a lot of opportunity for knowledge sharing. The venue was very comfortable and we were well taken care of by our Malaysian node hosts.
- INCF Congress Neuroinformatics Montreal (2018-02-09 18:31:13)
- NIJC demo at the 80th National Convention of Information Processing Society of Japan (IPSJ) (2018-02-08 10:43:40)
- Call for Participation in J-Node Hackathon 2018 March (2018-02-06 09:34:53)
- 第17回日中韓合同ワークショップNBNI2017開催報告 (2018-01-04 15:09:26)
- AINI 2017 Organization Report (2017-11-24 12:25:37)
- INCF Congress 2017 participation reports (2017-11-02 18:59:08)
- INCF Japan Node Presentation schedule at Neuroscience 2017 Washington D.C. (2017-11-01 16:08:14)
- AINI2017: “Program” released (2017-10-24 16:03:58)
- Advanced Scientific Programming in Python course at Melbourne (2017-10-12 11:11:45)
- Abstract Submission and Registration Open for NBNI 2017 (2017-10-02 16:36:00)