MEAnalyzer Manual

User Manual:

Open the PDF directly: View PDF .
Page Count: 24

Download
Open PDF In Browser	View PDF

MEAnalyzer User Manual
Contents
Getting Started............................................................................................................................................ 3
Installation ............................................................................................................................................... 3
Navigating MEAnalyzer ......................................................................................................................... 4
Outline of Tabs and Functions ............................................................................................................. 5
Load...................................................................................................................................................... 5
Analyze ................................................................................................................................................ 5
Graph ................................................................................................................................................... 5
Visualize Voltages .............................................................................................................................. 5
Plot Voltages ....................................................................................................................................... 5
Export ................................................................................................................................................... 5
Histograms .......................................................................................................................................... 6
Loading and Saving Spike Trains ............................................................................................................ 6
Export to HDF5 from Data Manager.................................................................................................... 7
Opening .H5 file ...................................................................................................................................... 7
Importing Data from CSV ...................................................................................................................... 8
Importing with No Electrode Layout .................................................................................................... 8
Opening file Saved by MEAnalyzer ..................................................................................................... 9
Editing Parameters............................................................................................................................. 9
Insert Metadata ................................................................................................................................. 10
Spike Train Calculations ......................................................................................................................... 10
Burst Calculations ................................................................................................................................ 11
Raster Plot ............................................................................................................................................. 11
Binned Plot ............................................................................................................................................ 11
Exporting Binned Raster Plot ............................................................................................................. 12
Spike Calculations................................................................................................................................ 12
Periodicity Analysis .................................................................................................................................. 13
Autocorrelation...................................................................................................................................... 13
Welch’s Periodogram........................................................................................................................... 13
Calculating Peak................................................................................................................................... 14
Histograms ................................................................................................................................................ 14

Inter-burst Interval ................................................................................................................................ 15
Burst Length .......................................................................................................................................... 16
Number of Spikes in Bursts ............................................................................................................ 16
Voltage Histograms .......................................................................................................................... 17
Exporting Histograms ...................................................................................................................... 17
Connectivity Analysis ............................................................................................................................... 18
Graph Selection .................................................................................................................................... 18
Graph Visualization .............................................................................................................................. 18
Graph Math ........................................................................................................................................... 20
Graph Fullness ................................................................................................................................. 20
Degree ............................................................................................................................................... 20
Cluster Coefficient ............................................................................................................................ 20
Visualizing Voltages .............................................................................................................................. 20
Plotting Voltages .................................................................................................................................... 21
Exporting Data ........................................................................................................................................ 22
Exporting to Excel ................................................................................................................................ 22
4D Spike Visualizations ....................................................................................................................... 23
Connectivity Graph Movie ................................................................................................................... 23

Getting Started
Installation
MEAnalyzer was compiled on MATLAB 2019a and requires 64-bit Windows with at least
8GB RAM. The first time installing MEAnalyzer will require prior installation of MATLAB
runtime, a standalone set of shared libraries that enables the execution of compiled
MATLAB applications without MATLAB or a MATLAB license. If MATLAB runtime is not
present, MEAnalyzer will automatically download and install it. Future updates and
installations of MEAnalyzer will not repeat this step. If the runtime does not install
properly, you can also access it here: https://www.mathworks.com/products/compiler/matlabruntime.html

Navigating MEAnalyzer

MEAnalyzer has six main tabs: Load, Analyze, Graph, and Visualize Voltages, Plot
Voltages, Histograms, and Export. Due to the orientation of graphical displays in
MEAnalyzer, window sizes are not adjustable. If the software is not fully viewable on
your monitor, you may need to increase your screen resolution.

To the right of the tabs is a status bar that indicates the current task that MEAnalyzer is
performing. In the center is a stop button that will allow you to interrupt ongoing
functions.
Outline of Tabs and Functions
Load
The load tab allows for the loading, and customization of spike train data. Spike train
time stamps can be loaded from an HDF5 or SPK file exported from MC_Rack,
Experimenter, or Axion, a csv file, or a file previously loaded and saved by MEAnalyzer.
On this tab, file metadata can be viewed and edited. The time segment of the
experiment can be adjusted, as well as the Bin Size for visualizations and subsequent
analyses. Labeled electrodes are displayed in their relative locations on the array, and
any number of individual electrodes can be removed from analysis. This tab also
displays two plots: A raster plot of the selected data and a binned plot.
Analyze
The Analyze tab includes a spreadsheet of electrode-specific calculations (spike rate,
burst rate, percent of spikes in bursts) that can easily be copied to the clipboard and
pasted into any spread sheet or statistical software. It also contains panels for
periodicity analysis, including Autocorrelation and Welch’s periodogram that can be
calculated based on Network Bursts, Binned Spike Rate, or Binned Bursts.
Graph
The Graph tab contains functional connectivity analysis options as well as a
topographically correct display of electrode locations and multiple calculation and
visualization options.
Visualize Voltages
The visualize voltages tab is optional if voltage data is loaded (not recommended). It
allows the user to visualize voltages and corresponding spikes.
Plot Voltages
The Plot Voltages tab is optional if voltage data is loaded (not recommended). It is
similar to the load tab that displays binned plots and raster plots, but can be color coded
to correspond to voltages identified at spike times.
Export
The export tab allows the user to select desired quantitative calculations and export
them to excel spreadsheets. It also allows for the creation of movies that represent
spike rates or connectivity graphs of overlapping bursts.

Histograms
The histogram tab displays fully customizable histograms of inter-spike intervals, interburst intervals, burst lengths, and the number of spikes in bursts.
Loading and Saving Spike Trains
Before loading a file, select the Electrode layout that matches your experiment:

MEAnalyzer is equipped with a variety of electrode layouts to match your desired
experiment:

MEAnalyzer allows data to be imported from an HDF5 file exported from an MC_Rack
or Experimenter file, a csv file containing spike trains, or a file previously analyzed and
saved with MEAnalyzer:

After loading a file, MEAnalyzer will display the full raster plot and binned spikes.
As this initial loading may take some time, MEAnalyzer will display popups to let you
know that it is still calculating

Note that on some systems the GUI popup for opening a file may hide behind the main
MEAnalyzer window instead of layering on top. If you do not see a popup, please check
behind this window.
Export to HDF5 from Data Manager
To open a file from MC_Rack you will need to convert it to an HDF5 file. You can do this
using MCS DataManager, which is available free of charge at:
http://www.multichannelsystems.com/software/multi-channel-datamanager
Make sure to export the Spike Time Stamps. Exporting voltage data is optional and will
make the file size considerably larger. Plotting Voltage data will also slow down
MEAnalyzer, but the option is there if desired.
Opening .H5 file
MEAnalyzer will display a dialog box where you can select the file you want to load. If
Voltage data is attached, MEAnalyzer will present you with the option of loading the
data. Remember that loading Voltage Data does not affect spike analysis, and will slow
down MEAnalyzer.

Importing Data from CSV
Spike Trains may be imported either from a CSV file exported from Data Manager.
CSV files may also be loaded from other programs as along as the format is the same.
Data should start on the 7th row, with the 7th row corresponding to electrode labels and
each column containing the time (in µs) of each detected spike.

Importing with No Electrode Layout
Data may also be imported from any spike train not associated with an electrode layout.
Some MEAnalyzer functionality that is dependent on topography, such as the 3-D spike
videos. However, as everything else is a mathematical calculation, they will all work the
same. Connectivity graphs will be displayed with an arbitrary layout, with all the same
visualization options. For this type of format, the csv file must contain spike
timestamps in 10^6s units. Each column will represent one spike train. The first row
will be interpreted as node labels.

Please note that under this setting, there is no option to remove electrodes from
analysis, so this should be done before importing the CSV file.

Opening file Saved by MEAnalyzer
Files previously saved from MEAnalyzer as a .MAT file may also be opened to the same
settings
Editing Parameters

Electrode Removal
The orientation of the electrode layouts is displayed so you
can easily select the electrodes to include or remove.
Reasons for excluding electrodes from analysis may
include that the electrode has high noise levels, where
there is no activity, or where the user is aware that spike
detection was erroneous.
The user may also use the “Remove electrodes” button to
automatically remove inactive electrodes, as defined by
electrodes that do not exceed the threshold of spikes or
bursts desired to be considered active.
Cropping Time Segment
The user may type in the time segment of the original recording that they wish to
analyze.

Adjusting BinSize
BinSize defines the length of time over which spikes are summed in the Binned plot.
Binsize is also used on other tabs to calculate periodicity and network connectivity.

Insert Metadata
For experimental purposes you may want to type in metadata such as the number of the
MEA plate, the title or experimental condition, and any notes about the experiment.
When saving the MEAnalyzer file or exporting results to excel these Metadata will also
be included.
Spike Train Calculations
After changing spike parameters it will be necessary to redo all spike calculations by
clicking on the “Calculate and Graph” button. All calculations will be computed with the
previously defined parameters and electrodes.

Burst Calculations
Bursts are calculated by user-defined parameters of the minimum
number of spikes in a duration of time

Raster Plot
The raster plot is presented as vertical black lines. The x-axis corresponds to the
electrodes in the order they were loaded and they-axis corresponds to the time. For
each electrode a vertical black line will be plotted at each time point where a spike was
detected. The time axis will be adjusted to start at 0.
Note that that if many electrodes are included, not all electrode labels will be shown in
order to prevent overcrowding.

The bursts may also be shown on the raster plot if desired. It may be
displayed as a red line on top of the spikes or a blue box overlaying the
spikes.

Binned Plot
The binned plot is displayed on top of the raster plot, and is
calculated according to the BinSize as defined by the user. During
the first data plotting the y-axis will be chosen automatically.
However, you can adjust the y-axis as well as the type of data that is
plotted. The binned data plotted can be: Spikes/Time/Electrode,
Percent of Electrodes Spiking, and Percent of Electrodes Bursting.
Traditionally, the percent of electrodes spiking can be interpreted as network bursts if
they exceed a certain percentage. The data from the selected binned plot can be copied
to the clipboard for easy transfer to spreadsheet software.

Exporting Binned Raster Plot
The Binned Spike and Raster plot can easily be exported in publication-ready format by
clicking the “Export Figure” button.

You can then select which part of the figure you want to save and in what dimension.

Spike Calculations
Spike calculations will be displayed on the analyze tab. This allows for a quick scan to
aid in the inclusion or removal of electrodes from analysis.

Periodicity Analysis
Periodicity analysis can be used to identify repetitive oscillating behavior. Periodicity is
shown on the analyze tab and can be calculated using autocorrelation or periodicity.
The user can select the type of binned data that is used for the calculations: Network
bursts (% of electrodes spiking), binned spike rate, or binned bursts (% bursting).

All plots can be copied to the clipboard for easy transfer to spreadsheet software and
further analysis. Quick rudimentary analysis options attempt to identify the peak, and
the associated periodic interval by finding the highest local maxima that is greater than
2 standard deviations of the mean.
Autocorrelation
Autocorrelation measures the similarity between a signal and a copy of itself that has
been shifted by a time lag. If the signal demonstrates periodic behavior there will be a
peak at the lag that represents the cycle length. The normalized autocorrelation function
𝑟(𝜏) measures the probability that the next event will occur at time 𝑡 + 𝜏.
𝑟(𝜏) =

∑𝑇−𝜏
̅)(𝑦𝑡+𝜏 − 𝑦̅)
𝑡=1 (𝑦𝑡 − 𝑦
; 𝜏 = 0, ±1 ∗ 𝑓𝑠, , ±2 ∗ 𝑓𝑠, ….
(𝑇 − 1)𝑉𝑎𝑟(𝑦)

where fs is the sampling frequency (binSize-1) and 𝜏 is the time lag.

Welch’s Periodogram
Welch’s power spectral density estimates correction instead of a standard power
spectrum. This approach reduces the variance of the periodogram by breaking the time
series into overlapping segments, computes a modified periodogram for each segment,

and then averages the segments to estimate power spectral density (PSD). The PSD
reports how much of expected signal power is at each frequency, and is expressed as a
1
−𝑗2𝜋𝑚𝑓/𝑓𝑠
function of the frequency by: 𝑃𝑥𝑥 (𝑓) = 𝑓 ∑∞
. Peaks in the power
𝑚=−∞ 𝑅𝑥𝑥 (𝑚)𝑒
𝑠

spectrum correspond to the repeating periodic intervals. Calculating periodicity in this
manner allows for identification of multiple periodic frequencies.

Calculating Periodicity Peaks
The data can be copied and pasted into Excel for plotting purposes and for identifying
peaks. MEAnalyzer can help with peak detection if the “Calculate Peak” button is
selected. Only one peak will be identified, and it will be the highest value between the
Min and Max Interval that is at least two standard deviations above the mean in that
interval. The data in each plot can be copied directly into excel or other programs for
custom peak selection methods.

Histograms
The histogram tab allows for the creation of fully
customizable histograms of inter-spike intervals, inter-burst
intervals, burst lengths, and the number of spikes in bursts.
If voltage segment data is available, histograms can also be
created of various voltage parameters at identified spikes.
These histograms can be used to analyze data sets, or to
inform parameter choice for other analysis methods. Idea
axis limits and bin widths can be automatically chosen or
can be manually specified.

The inter-spike interval
The inter-spike interval (ISI) option allows for the traditional display of time between
spikes.

Inter-burst Interval
The inter-burst interval option creates a histogram of the time between when one burst
ends, and the following burst begins. While this is one potential option for describing
periodicity, please refer to the periodogram or autocorrelation options on the analyze
tab for more advanced methods.

Burst Length
The burst length option creates a histogram of the length (in seconds) of each burst.
This can also be used to inform burst detection parameters.

Number of Spikes in Bursts
The histogram of the number of spikes in each burst can be used to inform burst
detection parameters or to compare changes in activity patterns.

Voltage Histograms
If voltage segment data is available for
the spike trains, MEAnalyzer can
calculate a histogram for the maximum
voltages, the absolute value of the
minimum voltages, or the difference
between the two.

Exporting Histograms
Similar to exporting the binned spike and
raster plot, Histograms may be exported
according to user-defined dimensions

Connectivity Analysis
Graph Selection
MEAnalyzer provides options to create functional connectivity graphs based on a variety
of different measures. Each electrode is a node, and an edge is created if it displays
correlation based on the selected measure.

Spike Correlation, and Burst Correlation look at the cross-correlation between two
electrodes at the time lag of 0 based on either the binned spikes or binned bursts. An
edge is created between two electrodes if their mean cross-correlation at a time lag of 0
exceeds the user-defined threshold. As an example, a Spike Correlation of 0.5 would
approximate a 50/50 probability of the two electrodes showing spike activity at the same
time.
Graph Visualization
MEAnalyzer includes options to change the size and color of nodes and edges to
represent different variables. The Node and Edge size and color can be adjusted, along
with their own individual colormaps. Whenever a visualization option is selected the
values are scaled to match the range designated by the “max” numerical field.

Node Size and color can be changed to represent standard graph math metrics. For
example, they can be set to represent node degree or the cluster coefficient. Node size
and color can also be changed to represent spike train metrics, such as spike rate or
burst rate.

If a cross-correlation method was used to create the graph, then the edges have a
weight property that corresponds to the average cross-correlation of two nodes. The
visualization can be adjusted to let the color or width of each edge represent the edge
weight.
The following color maps are available:

Graph Math
Graph Fullness
Graph fullness is calculated as the percent of possible edges that are connected
𝑛𝑒𝑑𝑔𝑒𝑠
(𝑛𝑒𝑙𝑒𝑐𝑡𝑟𝑜𝑑𝑒𝑠 )(𝑛𝑒𝑙𝑒𝑐𝑡𝑟𝑜𝑑𝑒𝑠 − 1)
2
Degree
In Graph Theory, node degree represents the number of other nodes that a node is
connected to. In MEAnalyzer , degree is normalized so that it represents the percentage
of other nodes that an individual node is connected to.
Cluster Coefficient
Cluster Coefficient essentially measures the connectivity density of each graph18 where
2𝑛𝑖
each node 𝑖 in graph G is defined as 𝐶𝐶𝑖 = 𝑘 (𝑘 −1)
and 𝑛𝑖 is the number of connections
𝑖

𝑖

between a node and its topographical neighbors, and 𝑘𝑖 is the node degree.
Visualizing Voltages
MEAnalyzer does not perform spike detection, but if Voltage Data is available it may be
useful to visualize voltages. This can help the user evaluate which electrodes are
appropriate for exclusion from analyses, either because of high electrode noise or
inappropriately detected spikes. Three electrode voltages can be visualized at once.
The user may decide to display spikes overlaying the voltage plots.

Please note, it is often standard practice when saving MEA data during recording to
save the raw voltage data and spike time stamps and segments. As spikes are often
calculated based on filtered data, there may be some mismatch between spike
segments that have been stored and the raw voltage data. For this reason, MEAnalyzer
has increased the time span usually included in spike segments and highlights them in
the raw voltage.
Plotting Voltages
If spike segment data is available, MEAnalyzer
offers additional visualization options for the
voltages. This is the same data used in the
histograms tab. This tab allows you to plot
pseudo-colored raster plots and binned values
where the color corresponds to the voltage value of
your choosing. Additionally, an electrode specific
view is available to display average electrode
activity. Please note that these data represent the
desired metric from spike waveforms and are not
an average of all voltage measurements.

Exporting Data
Exporting to Excel
Under “Spike Calculations” and “Connectivity Graph Calculations” the user may select
the desired metrics to export. Upon clicking the “Save Calculations” button MEAnalyzer
will present a dialog box that allows the user to define the location and file to be created.
In each case, an excel file of multiple sheets will be created, and the first sheet will
contain the metadata and calculation parameters for the Spike Trains.

4D Spike Visualizations
Exporting the data as a 4D Spike Visualization creates a
movie file where each frame consists of a 3D bar graph.
Any electrodes removed from analysis will not be
represented by bars in the movie file. The x-y- location of
each bar will correspond to the topographical location of
that electrodes on the plate. The height and color of each
bar will correspond to the total number of spikes by that
electrode in the time interval defined by the user. If a title
is given it will be displayed at the top of the video. To be
able to create comparable movie files between different
plates, the maximum number of spikes can be adjusted to control the graph
visualization.

Connectivity Graph Movie
Connectivity graphs based on overlapping bursts may not be
appropriate for longer experiments, as longer time lengths will
have a higher probability of overlapping bursts and may result in a
full graph that would prevent comparisons between different
experiments. To represent longer time lengths with the
overlapping burst method, the connectivity graph movie option
was created.
This will create a movie file where each frame consists of a raster
plot and connectivity graph. The connectivity graph will be the
graph based on overlapping bursts for a subset of the full time, as defined by the Time
Interval. The raster plot will have a shaded region that shows the time interval for which
the connectivity is being calculated.



---

Source Exif Data:

File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.7
Linearized                      : No
Page Count                      : 24
Language                        : en-US
Tagged PDF                      : Yes
XMP Toolkit                     : 3.1-701
Producer                        : Microsoft® Word for Office 365
Creator                         : Raha Dastgheyb
Creator Tool                    : Microsoft® Word for Office 365
Create Date                     : 2019:04:25 15:14:23-04:00
Modify Date                     : 2019:04:25 15:14:23-04:00
Document ID                     : uuid:D6218EE1-6F9E-4454-9102-A4120BA7E725
Instance ID                     : uuid:D6218EE1-6F9E-4454-9102-A4120BA7E725
Author                          : Raha Dastgheyb

EXIF Metadata provided by EXIF.tools