In Attendance

• Alastair Knowles (AK)
• Leslie Smith (LSS)
• Simon Schultz (SS)
• Tom Jackson (TJ)
• Liz Williams (LW)
• Jennifer Simonotto (JS)

Apologies 

• Nathan Lepora (NL)

Previous Meeting Report

The report from the previous meeting was approved as a true and accurate record.

Progress Reports 

WP0 (TJ): Usability testing was conducted on the CARMEN portal. The session was coordinated by Liz Stuart (Plymouth), with user groups from WP5 and WP6 at Newcastle. A number of specific bugs, misconceptions and dislikes were identified and have been documented in a report, which has been circulated within WP0. Others wishing to view the report may contact TJ. The interface for entering metadata was identified as a key failing. Two alternative interfaces have been generated by Mark Jessop and are under review within WP0. A second usability testing session will be delivered to the wider consortium as part of the services workshop in Newcastle on 31^st July and 1^st August.

The system design phase has been completed, and internal system interfaces have been generated from the design documentation. The build environment is also in place. A system release is scheduled for Q3/Q4 2008 and will focus on improved support for data upload, download and searching. Dan Swan is also engaging with users to convert analysis code into preliminary web services. These will also be made available. The subcomponent that is being used to handle metadata, Symba, has proven problematic and discussions have taken place to identify potential solutions, including alternative systems.

The group has also had a review paper accepted for a special issue of Neural Networks on neuroinformatics. Further to this, CARMEN has been accepted for a demonstration slot at the INCF Congress on Neuroinformatics (Stockholm) from 7^th – 9^th September 2008.

WP1 (LSS): Juan Martinez‑Gomez has been working with Dan Swan to convert the Wave_clus toolkit into services. LSS is working with Shahjahan to quantify the precise differences between their spike detection and sorting algorithms. A paper on newly developed techniques has been resubmitted to the Journal of Neuroscience Methods. LSS has also had papers accepted for the MEA Meeting, where he will be presenting an overview of CARMEN in addition to a poster on WP1 with Shahjahan Shahid.

WP2 (SS): Robin Ince is currently working on his PhD transfer report with Stefano Panzeri. He is also keen to deploy his algorithms as services and has been having early discussions with Dan Swan. A paper has also been submitted recently on information extraction techniques to quantify sensitivity to different variables in data. Work continues on the development of information extraction techniques for analysis of local field potentials (i.e. continuous signals). Simon Schultz and Phillip Bream are continuing with the acquisition of calcium signal data from cerebellum. Work has also been published on entropy estimation techniques applied to legacy imaging data.

WP3 (NL): Report circulated after the meeting. Script has been developed allowing: (a) Matlab to modify channel conductances in the model neuron in GENESIS, and; (b) GENESIS to output the results of the simulation back to Matlab. A model neuron has been developed in GENESIS and will be used to test the search algorithm. Latterly, in vitro data will be fitted to this model. Immediate future goals are to refine the existing algorithm for effectiveness and robustness, and to apply the algorithm to legacy in vitro data. Efforts are also being made to produce usable code in time for the user training workshop on July 31^st and August 1^st 2008.

WP4 (TJ): Bojian Liang has been working on extensions to the SDE to accommodate 16‑bit data resolution generated by new data acquisition system employed by Evelyne Sernagor (12‑bit currently the norm). Also working on refinements to the new visualisation interfaces demonstrated at the April 2008 consortium meeting. Developing translation services and APIs for the CARMEN intermediate data format. Preliminary solution exploits the NeuroShare DLLs to convert the data to static files coregistered with the source data. Anticipating delivery in 2‑3 months. Dynamic translation (e.g. to eliminate duplication of files) to be explored in the medium‑to‑long term.

WP5 (LW): Stuart Baker and LW took part in the usability study, and met with Graham Pryor to discuss data curation issues related to CARMEN. Interviews also took place for the unfilled RA permission. A strong candidate has been identified. LW continues her work on surrogates for testing the significance of patterns in data. AK asked LW to obtain progress updates from Warwick and Plymouth for the next meeting.

WP6 (JS): Evelyne Sernagor has been reviewing the newest version of the SDE. She sends positive feedback but is not ready to update in detail. Anne Smith, Tom Smulders and Christoph Echtermeyer are implementing a framework to assess the learning capability of their Bayesian methods. Stephen Eglen has been contacted by Georgios Pitsilis to authorise release of his software under a different open source licensing framework and requests clarification on the reasoning behind this. AK suggested JS should ask Phil Lord to telephone Stephen Eglen, to provide further clarification.

JS is working on analysis of high‑frequency oscillations in epilepsy data with Mark Cunningham. A paper has been submitted to the Journal of Neuroscience Methods.

Any other Business

A discussion took place on involving representatives from each of the 11 CARMEN partners in the monthly teleconferences. It was decided this should not go forward. AK suggested including a personal update from each of the project investigators in the newsletter (issue 1 to be released soon). It was agree this should go ahead.

Next Teleconference

2pm 14^th July