Hosted by Carboncopies.org @ 6:30pm - 9:30pm, October 16, 2012
in the "Bridge" room of the Hilton New Orleans Riverside.
The last few years have seen a rapid increase in the interest in detailed large scale reconstructions of neuronal circuitry identified in human or animal brain tissue. On the structural side, this development is directly related to advances in connectomics, the acquisition of large-scale and high-resolution neural structure and neural connectivity data (e.g., see projects listed below). On the functional side, recent developments are indicating that neuroscience is ready for a new approach to data acquisition that would provide functional recordings at high spatial and temporal resolution from tens of thousands to millions of neurons.
Together, the data that these developments seek to collect should enable the functional reconstruction of meaningful pieces of biological neuronal circuitry or even Whole Brain Circuit Reconstruction. In the big picture, a functioning reconstruction of this sort may be considered a testable goal or major milestone of Neuroscience.
Such reconstructions can be used to test hypotheses and to properly ground our theories about operations carried out in brain regions. By providing a virtual brain laboratory, they will also improve our understanding of brain disorders and deficits, and may aid the development of treatments, neural interfaces, neuroprostheses and artificial intelligence. Ultimately, reconstruction of specific brain circuitry will lead to Whole Brain Emulation with all that that entails.
For these reasons, large scale or whole brain neuronal circuit reconstructions should be explicit goals of the neuroscience community and its funding organizations. During the symposium, Todd Huffman of 3Scan will address the rationale of this approach. The four pillars of neuronal circuit reconstruction, namely hypothesis testing, platform development, structural connectomics and functional characterization are introduced by Randal Koene of the non-profit carboncopies.org, who will also moderate the symposium/workshop.
Before any of these fantastic results can be achieved, we need to build the tools, grasp the System Identification problem at the foundation of brain circuit reconstruction and deal with the integration of the different technologies and methods that together solve the System Identification problem. Likely, this will be an iterative process. To date, some of the best results in this field include the ground-breaking 2011 Nature publications by Brigmann et al and Bock et al.
The format of the event emphasizes dialogue. Participants engaged in projects will have the opportunity to introduce those projects briefly. The workshop portion of the event will address the questions of Integration or cooperation between projects, demonstration of neuronal circuit reconstruction, improvements in funding, and the pros and cons of different approaches taken in the projects. Results of the workshop and participants' responses to the main questions will be used to compile a joint white paper containing an updated road map for whole brain circuit reconstruction.
A number of projects exist today to address the requirements for obtaining a detailed connectome. There are also several projects that are attempting to produce tools for high spatial and temporal resolution recordings, which can provide the necessary functional reference points for tractable System Identification. And there are projects that study the matter of reconstruction, how to create representations, assign parameter values and carry out emulations of identified neuronal circuitry. We will look at ways to streamline cooperation between these projects so that an integrated solution is possible. Presented and discussion led by Randal Koene (carboncopies.org).
To date, there has been no demonstration that the interesting content (e.g. memories) of a neural system can be captured and recognized through the proposed detailed data acquisition and reconstruction. To elicit true enthusiasm in the scientific community and in funding bodies, it is essential to demonstrate that the approach is in fact a practical possibility. Presented and discussion led by TBA.
There are a number of projects with great intrinsic value and significant implementation knowledge that must not be lost! Several avenues to guarantee support and progress for all parts of the puzzle, and as a group effort rather than an individual struggle will be presented. The integrated results, which lead to a virtual brain laboratory should be embraced as big picture goal of neuroscience. Presented and discussion led by Todd Huffman (3Scan).
Ways to compare different proposals and projects, specific road-blocks and metrics will be introduced. Projects are briefly introduced (5 mins) by participants. Moderated by Randal Koene (carboncopies.org).
Representative projects to be discussed (and where possible introduced by the researchers involved) include:
In addition to these topics, there is always room for general brainstorming within this interest-driven group, as well as an opportunity to learn about the goals and next steps of the supporting Foundation 2045, carboncopies.org and the GF2045 Congress in New York City that is strongly associated with this symposium.
Time: 6:30pm - 9:30pm, October 16, 2012
Place: "Bridge" room, Hilton New Orleans Riverside, Society for Neuroscience Annual Meeting 2012
(Please note that the Symposium may not appear on the SFN2012 Satellite Events page until the final program is printed.)
To participate, simply show up or contact Randal.A.Koene@carboncopies.org. While RSVPs are not strictly necessary, they are much appreciated, because RSVPs allow us to better prepare the event for the participants!