Society for Neuroscience Annual Meeting. www.sfn.org
October 18-21, 2009
Chicago, IL
The Society for Neuroscience (SfN) annual meeting has
been the showcase for R.C. Electronics’ electrophysiology
data acquisition and analysis products for over 20 years.
We’re proud to announce our continued presence at
the SfN annual meetings, and to tell
you about the results of the 2008 show in Washington, D2007
show in San Diego.
2008 SfN Annual Meeting
Washington, DC
We were pleased to present the latest NeuroMAX status at the 2008 show in Washington, DC. The following 2 posters were presented. You're welcome to contact us and request a copy.
Poster: A plug-in software architecture for spike sorting and
analysis tools.
This poster presented a MATLAB-based software architecture that
lets user-developers
embed a developed spike sorting/analysis module into a
framework for access from a larger toolbox. The developer
specifies the inputs and output data types for a tool
either broadly, in terms of an already
existing data class, or in terms of a
novel developer-designed class when needed. Parameters
are specified in terms of a default setting for
a given set of input data. A GUI template
and GUI widget library assists the developer in designing
a GUI for the tool if desired. We developed a framework
that will deliver data and parameters
to the computational wrapper module while hiding
the details of the data class implementation. The tool
can leverage toolbox utilities, data visualization
tools, and data handling
capabilities. A registered tool can be added to
a
workspace toolchain, letting a user
link tools for data analysis in a flexible feed-forward
manner. The tool is accessible from the main workspace
GUI and has its own GUI for modifying
parameters and visualizing data.
Poster: Assessment of the current state of the Neuroshare data interoperability
standards.
This poster presented the usability of the Neuroshare API
(with associated vendor supplied libraries) as a universal
interface for seamless access to electrophysiology data
acquired from differing data acquisition hardware platforms.
We developed a MATLAB-based utility that uses the Neuroshare
MATLAB import filter and the existing Neuroshare DLLs.
The utility allows users to select and open a file, list
the file’s entities by type (Event, Analog, Segment,
Neural Event), and access the informational and data fields.
We attempted to open and access all vendor sample data
sets on the Neuroshare web site.
Overall, the libraries worked seamlessly. We used the same
MATLAB code to access data and information independent
of vendor format.
As advertised, the DLLs successfully handled these details.
When data was inaccessible, the problem was traced to changes
in a vendor’s
acquisition format without an update to the corresponding
DLL. We suggest that a process
be implemented to regularly update DLLs as a vendor data
format changes.
2007 SfN Annual Meeting
San Diego, CA
R .C. Electronics Inc. presented the following posters
at the Society for Neuroscience show, November 3-7, 2007,
in San Diego,
California.
If you’d like a copy of these posters, please
contact us.
Poster: Object oriented architecture for a spike sorting
and analysis software toolbox simplifies addition of new
tools
This poster presented a software architecture for a MATLAB-based
spike sorting and analysis toolbox. This architecture was
built to address the need of the neuroscience community
for an easily modifiable and expandable set of tools for
performing operations for spike sorting and analysis. The
“toolbox” concept allows a software developer
to implement a specific sub-operation of spike sorting and
analysis, leveraging the existing tools for other parts
of the analysis and visualization, saving time and effort.
Additionally, the standardized format of the toolbox modules
facilitates sharing of new algorithms.
Poster: MATLAB-based object oriented architecture for data
handling based on NeuroShare data format
This poster presented a MATLAB-based, object oriented framework
for data handling within a spike sorting and analysis software
toolbox. The basis for the data handling classes is the
NeuroShare data format. We utilize existing NeuroShare libraries
to load data into MATLAB. We have created a set of MATLAB
data classes that are based on the NeuroShare basic data
types. The key advantage of using MATLAB’s class constructs
for managing data is that the inherent complications in
the NeuroShare data format can be hidden and data fields
can be accessed in a more controlled manner. The analysis
result is a NeuroShare compatible data that can be easily
shared or ported to another application.
