Spike
Detection Toolbox: Overview
Spike Detection Toolbox
The Spike Detection toolbox is the first of
four NeuroMAX Toolboxes, which also includes Spike Sorting,
Spike Train Analysis, and Display Tools. Each toolbox contains
a variety of tools for specific applications. NeuroMAX lets
you select multiple tools and link them together in a powerful
chain, or Workspace. The Workspace can be any combination
of packaged NeuroMAX tools and your own custom MATLAB-based
tools.
Ideally, intra- and extra-cellular recordings consist of
clearly distinguishable spike trains recorded on individual
channels. In reality, spike data is often contaminated with
noise and interference from multiple neurons. Before any
relevant information can be extracted from this data, the
firings of individual neurons must be separated from each
other and from background information, using tools from
the Spike Detection toolbox.
The first step in the spike detection process is a Pre-Process
tool, used to remove unwanted frequencies from the raw data.
One of the Fragmentation tools can then be used to segment
the data into potential “spike trains” for further
analysis. A summary of each tool is included, below. For
information on a particular PreProcess or Fragmentation
tool, click on that link in the list of tools, left.
Once the individual spikes are detected, they can be sorted
into unique groups with the Spike Sorting tools, and specific
data (such as peak height and area) can be extracted with
tools in the Spike Train Analysis toolbox.
PreProcess Tools: Overview
PreProcess tools provide general purpose signal processing
algorithms to filter multi-channel data as the first step
in spike detection.
UpsampleAnalog: This tool increases the original sampling
rate of the analog input data by an integer factor using the
MATLAB interpolation routine "interp."
UpsampleSegment: This tool increases the original sampling
rate of a data segment by an integer factor using the MATLAB
interpolation routine "interp."
FilterTool: This tool provides a standard Infinite Impulse
Response (IIR) Butterworth filter. You can select High Pass,
Low Pass, Band Stop, or Band Pass, along with the appropriate
cutoff frequency.
Fragmentation Tools: Overview
After separating unwanted noise and frequencies from neuronal
data with one of the PreProcess tools, select a tool in the
Fragmentation category to locate sections of data that may
contain spikes. NeuroMAX returns these fragments for additional
processing. The Fragmentation tools use unique sets of user
selected criteria to locate potential spikes.
BasicThresh: This is a traditional threshold routine that
finds and stores spike train fragments based on user selected
search criteria, including threshold value, the number of
points to store before and after a threshold exceedence, and
the number of points to skip between spike train fragments.
WaterfillThresh: Detects spikes by first finding the largest
data point in the entire data set. A user-specified number
of points before and after this point are selected as the
first spike. A user-selected number of points is skipped between
spikes to remove the risk of creating a second spike around
the same high data point. An automatic 3 Sigma or 2 Sigma
threshold is also available. |
