Use the Filemenu to open or close a file. After the file has been opened,
the signaldialog will appear.
You can also use "drag and drop" to open a file.
Use the Filemenu to open a stream. The stream must be a "growing" file.
This mode is used when you have an acquisitionprogram running that writes EDF or BDF.
While the acquisitionprogram writes the data to the file, you can use EDFbrowser to watch
(follow) the actual data. EDFbrowser will regularly check the actual filesize and display the
last part of the file i.e. the most actual/recent data.
The default intervaltime between the updates is 500 milliSeconds and can be changed in the settingsmenu.
In order for EDFbrowser to be able to start a video, the VLC mediaplayer must be installed on your system.
On windows: make sure VLC is installed in: C:\Program Files\VideoLAN\VLC\ or C:\Program Files (x86)\VideoLAN\VLC\
Open an EDF or BDF file and select/setup your montage as usual. Then press Ctrl-Shift-v (or go to File -> Start video)
to select a video. After the video has started, you can use the slider on the bottom to change the file position or
use page-up/page-down.
Also, the startdate and starttime needs to be set in the filename of the video (for synchronization).
The following text is copied from https://www.edfplus.info/specs/video.html :
Video filenames of the same patient must start with the same patient identification,
followed by the start-date and -time of the video, as follows:
NL_012348168_03-MAY-2013_14h45m49.013s_Video.ogv
in which case the video apparently starts on May 3, 2013 at 0.013 seconds after 14:45:49hr. To be more precise,
the patient identification is followed by an underscore (_), followed by the video startdate DD-MMM-YYYY,
followed by an underscore, followed by the starttime HHhMMmSS.XXXXs. The startdate days, DD, are 01, 02, 03, ... 30 or 31.
The months, MMM, are JAN, FEB, MAR, APR, MAY, JUN, JUL, AUG, SEP, OCT, NOV or DEC.
The years, YYYY, obviously are 2013, 2014 and so on. The starttime is based on the 24h clock with HH ranging from 00 till 23,
so midnight is coded as 00h00m00s. The decimal fraction of a second (noted here by .XXXX) can have any length and can also be omitted
(for example in 06h37m12s). The addition _Video is not obligatory and neither do we standardize any video format such as ogv.
So, a perfectly OK video filename would also be:
NL_012348168_03-MAY-2013_14h45m49.013s.mpeg
Note: EDFbrowser will ignore the part of the filename directly before the startdate and directly after the starttime.
If the filename of the video does not contain the startdate and starttime, you can still continue to use the video but
in that case EDFbrowser will assume that the starttime of the video coincides with the starttime of the EDF file.
You can resize the video window by dragging the bottom-right corner with the mouse.
All other operations like stop, pause, etc. must be done by using the videoplayer controls (buttons) in EDFbrowser.
Do not use the control interface of VLC, it will interfere with EDFbrowser.
Use the Signalsmenu to add or remove signals. If you want to remove only one particular signal,
leftclick on the signallabel of the signal you want to remove. A small dialog will appear and gives
you the possibility to remove that particular signal from the screen. See also signaldialog.
Use the Timescalemenu to choose the pagetime (the amount of seconds that must be displayed on the screen).
"30 mm/sec" will adjust the pagetime so that one second equals 30 mm horizontally. In case you get a different
result, calibrate your screen in Menu -> Settings -> Calibration.
The timescale can also be changed with the keyboard shortcuts Ctl++ and Ctl+-,
or by turning the mousewheel while keeping the Ctl-key pressed.
The resolution of the timescale indicator (milliSecond or microSecond) can be selected in: Settings -> Options -> Annotation editor.
Use the Amplitudemenu to adjust the amplitude (size/gain/sensitivity) of all signals on the screen.
The values in the dropdown list are expressed in units per cm.
If you want to adjust the amplitude of one particular signal, use Signalproperties.
"Fit to pane" will adjust the height and size of all signals so that they will fit on the screen.
"Adjust offset" will adjust the height only (of all signals so that they will fit on the screen).
"Offset -> 0" will set the offset on the screen of all signals to zero.
Use the "+" or "-" keys on your keyboard to increase or decrease the amplitude of all signals.
The signallabel is the name of the signal which is printed on the screen at the start of the trace.
This little piece of text has some "hidden" functions. When you leftclick on it, the Signalproperties
dialog will be opened. You can also drag this label with the left or right mousebutton to change
the offset or amplitude of the signal, see also Adjusting the position and size of the signals.
The alias, if used, is an alternative signallabel which is printed on the screen at the start of the trace.
Aliases can be created in the Signalproperties dialog and will be stored in your montage.
This mode activates a colorschema and grid boxes typical for plotting ECG waveforms.
The ECG waves are recorded on special graph paper that is divided into 1 mm2 grid-like boxes. The ECG paper speed is ordinarily 25 mm/sec.
As a result, each 1 mm (small) horizontal box corresponds to 0.04 sec (40 ms), with heavier lines forming larger boxes that include five
small boxes and hence represent 0.20 sec (200 ms) intervals. On occasion, the paper speed is increased to 50 mm/sec to better define waveforms.
Each large box is therefore only 0.10 sec and each small box is only 0.02 sec.
Vertically, the standard calibration is 10 mm (10 small boxes), equal to 1 mV. On occasion, particularly when the waveforms are small,
double standard is used (20 mm equals 1 mv). When the wave forms are very large, half standard may be used (5 mm equals 1 mv).
The ECG view mode can be activated by going to Settings -> Options. In the Colors tab, in the Colorschema section, click on the "ECG" button.
It is advised to enable the vertical scrollbar (in the menubar go to View -> Vertical scrollbar) and set the number of traces to 4 in
Settings -> Options in tab Other (look for the setting with the text "Enable a vertical scrollbar and show max.").
Set the timescale to 50 mm/sec. and the amplitude to 500 uV (0.5 mV).
In order for the gridboxes to be calibrated, it is necessary that the physical dimension (unit) of the ECG waveform is expressed in either
uV, mV or V. If this is not the case, the gridboxes are uncalibrated and they cannot be drawn correctly.
Use the signaldialog to add one or more signals to the screen. Go to Signals -> Add.
On top you will see a list of opened files.
Select (highlight) the file from which you want to add signals. At the left part of the dialog you see a list of
all the signals which are in the selected file. Select one or more signals and click on the "Add signal(s)"
button. Now you will see the selected signals on the screen. You can add more signals afterwards,
just go to Signals -> Add.
When you want to make a combination (derivation) of two or more signals, do as follows. Open the signalsdialog.
Select the file from which you want to add signals. In this example we choose an EEG file which contains
the signals "P3" and "C3" and we want to subtract "C3" from "P3".
- Select (highlight) the signal "P3".
- Click on the "Add" button.
- Select (highlight) the signal "C3".
- Click on the "Subtract" button.
- Click on the "Make derivation" button.
- The result of "P3" minus "C3" will appear on the screen.
Now you can add more combinations or simply close the dialog.
note: It is only possible to make combinations with signals who:
- are in the same file
- have the same samplerate
- have the same physical dimension (e.g. uV)
- have the same resolution (e.g. uV/bit)
In case the signals in your file don't have the same resolution, you may use the tool
Unify resolution to correct this.
note 2: If you want to invert the signals (which is common in EEG) do as follows:
instead of P3 - C3, use -P3 + C3 (or C3 - P3).
You can also invert a signal afterwards, go to Signals -> Organize.
You can view or adjust the properties of a specific signal by leftclicking on the signallabel.
A small dialog will be opened and let you change the amplitude and color.
You can also remove filters (if any) or remove the signal from the screen.
You can also enter an alias for the signalname.
Another option is to activate one or more crosshairs for precise measurements.
Another way to view/change the properties of a signal is to go to Signals -> Properties.
This will show you a list of all signals which are on the screen. Simply click on the
signal of interest.
Leftclick on the signallabel and move the mousepointer up or down while keeping the
left mousebutton pressed. This will move the signal up or down on the screen i.e. it
changes the offset of the signal.
Rightclick on the signallabel and move the mousepointer up or down while keeping the
right mousebutton pressed. This will increase or decrease the amplitude (size/gain) of
the signal on the screen.
You can organize the signals via Signals -> Organize.
Here you can change the order of the signals on the screen, remove or invert signals.
When you want to visualize many signals, the screen can become crowded and hard to read.
There are two solutions for this:
- divide the signals you want to visualize in smaller groups and use montages to quickly switch
between these groups (using the Fn keys).
- enable a vertical scrollbar for the signal window so that the signals will be distributed over more
vertical space. The vertical scrollbar can be enabled in Settings -> Options -> Other.
Navigation through a file can be done by using the PageUp/PageDown buttons on your keyboard.
Use the left/right arrowkeys to shift the screen one tenth of the pagetime.
The mousewheel can be used to scroll horizontally. The stepsize of the mousewheel can be set in the settings menu.
A value of zero disables mousewheel scrolling.
Keep the middle mousebutton pressed to drag horizontally.
There's also the file position slider at the bottom of the screen.
Pressing Ctl+Home or Ctl+End will jump to the start or the end of the recording respectively.
In case multiple files are opened, the file position slider, the "Home" and "End" keyboard shortcuts and
the file position indicator, are referred to the file that is the "reference". For more info and how to change
the reference file, have a look at Timelock and File synchronisation.
The resolution of the fileposition indicator (milliSecond or microSecond) can be selected in: Settings -> Options -> Annotation editor.
You can also zoom-in.
When a file contains annotations (sometimes called events), the annotation window will appear.
Just click on one of the annotations to jump to that position in the file (use Ctl+N to jump to the next annotation or Ctl+B for the previous annotation).
The annotation marker will appear either at the start or the middle of the page,
depending on an option that is in Settings -> Options -> Other.
The annotation window is dockable. You can change the size of the window, but also the position.
You can make the window floating as well by dragging it with the mouse. After closing
the annotationwindow, you can make it appear again via Window -> Annotations.
Een annotation/event has three properties:
If an annotation has a duration, a semi transparent colored overlay will be applied for that duration.
In Settings -> Options -> Colors you can assign your own colors to max. eight different annotation descriptions.
The transparency of the overlay color is set using the alpha channel: 0 = completely transparent, 255 is completely opaque.
The default value for the alpha channel is 32.
An annotation can be linked to a specific signal. This is done by adding a suffix to the description,
for example: spindle@@C3 - A1
The string immediately after the "@@" must be one of the signal names.
Annotations can be created, modified or deleted using the Annotation editor
By right-clicking on an annotation in the annotationlist, you can perform several operations on the annotations.
It's possible to filter groups of annotations. You can select if these filter operations take place on the annotations list
only or also on the waveform window. This option is in Settings -> Options -> Other.
Right-click on an annotation marker in the plotting window in order to select it (the color will change) in the annotation list,
or to perform waveform averaging.
If you like, you can enable "auto scroll" for the annotations window so that, when browsing through the recording, the annotation window's
scroll position is synchronized with the fileposition of the plotting window. This option is in Settings -> Options -> Other.
The resolution of the onset time (milliSecond or microSecond) can be selected in: Settings -> Options -> Annotation editor.
The annotations list can be filtered based on the maximum and minimum interval time
between two consegutive annotations with the same name/label.
Right-click on the annotations list and select "Filter Interval Time".
A dialog will appear where you can set the filter parameters.
The annotation editor can be used to create, edit or delete annotations.
Start the editor via menu -> Window -> Annotation editor.
To create a new annotation, write the text of the annotation in the description window and set
the onset time. You can set the duration as well. If the duration is unknown or not
applicable, set it to -1. Next, simply click on the "Create" button.
Instead of manually entering the onset time, you can use a crosshair.
Drag and drop the crosshair and the onset time will be automatically adjusted.
You can use a second crosshair to adjust the duration of the event.
To modify an existent annotation, click on the annotation in the annotationlist.
Adjust the text and/or time manually or drag and drop the annotation marker with the mouse.
To delete an annotation, click on the annotation in the annotation window and click
on the delete button (or press the Delete key on your keyboard).
Pay attention, there's no "undo".
When you have finished editing annotations, save the file in menu -> File -> Save.
Your file will not be altered, instead a copy of your file with the edited annotations
will be made.
When a crosshair is activated, the onsettime field will or will not be updated when navigating through the file,
depending on a setting in the options menu ("Auto update annotation-editor onsettime").
You'll find this setting in: Settings -> Options -> Annotation editor.
Keeping the shift-key pressed when left-clicking on an annotation in the wavefom window,
will prevent the annotation from moving (i.e. the onset time will not be modified).
The annotation will only be selected and made visible in the annotations list.
It's possible to enable and configure up to eight user buttons which can be used to quickly create annotations.
This can be done in: Settings -> Options -> Annotation editor.
You can freely assign any name to a user button. The name will be used as the description of the annotation.
You can setup which actions will be performed when clicking on a user button:
- set the description
- set the onset time
- set the duration
- advance the viewtime (file position) with the stage /epoch time
Keyboard shortcuts for the user buttons are: '1', '2', '3', etc.
Note: It's possible to let the viewtime automatically stay on an integer multiple of the stage /epoch time when using the user buttons to create annotations.
I.e. the viewtime and pagetime will always be at the epoch boundary. This can be setup in: Settings -> Options -> Annotation editor.
Note: It's possible to let advance the viewtime automatically when using the user buttons to create annotations.
I.e. the viewtime will advance with the selected stage / epoch time when clicking on a user button. This can be setup in: Settings -> Options -> Annotation editor.
Keeping the shift-key pressed while hitting a user button will cause the viewtime to jump backwards.
Signal linked annotations can be created by using a suffix "@@" followed by the signal label in the annotation description.
e.g.: "sleep spindle@@C3-A2"
To quickly create a signal linked annotation by drawing a rectangle with the mouse, keep the Ctrl key pressed when releasing the left mouse button.
(The rectangle drawn must have a size of at least 15 x 15 screen pixels. If not, the action will be ignored.)
After drawing the rectangle over the part of the signal of interest, a menu will popup which will allow you to select one of up to eight
predefined annotation descriptions. After selecting one of them, a signal linked annotation will be created with the specified description,
onset and duration will be set according to the rectangle drawn.
Instead, if the Shift key is being pressed, a normal (non signal linked) annotation will be created.
You can setup the predefined descriptions in: Settings -> Options -> Annotation editor.
Alternatively, press 'a' when one or two crosshairs are enabled (on the same signal) in order to create a signal linked annotation.
The second crosshair will be used to set the duration of the event. If not present, an annotation without duration will be created.
When the Control key is pressed when selecting the annotation (after having pressed "a"), the delta (difference) of the amplitude of
the two crosshairs will be printed into the annotation's description.
Note: At default, the resolution of the onset-time of the annotations is 1 milliSecond.
If desired, this can be changed to 1 microSecond in Settings -> Options -> Annotation editor.
This affects also the resolution of the crosshairs, the viewtime and pagetime indicators and the annotation markers.
Use the filtermenu to add or remove filters. First select the type of filter: highpass, lowpass, notch, bandpass or bandstop.
Now select the model: Butterworth, Chebyshev, Bessel or "moving average" (a notchfilter has resonator model only).
Then set the frequency(s) of the filter (or number of samples in case of a "moving average" filter).
Depending on the type and model of the filter, you can adjust the order, Q-factor and passbandripple.
The curve will show the frequency and phase response of the filter (except for "moving average" filters).
Now select one or more signals to apply to and click the "Apply" button.
Reviewing the filters can be done in the Montage -> View this montage menu.
Algorithm for the moving average filters:
LPF: the mean of the last n data samples.
HPF: last-n/2 sample minus the mean of the last n data samples.
You can adjust or finetune the filters. Left-click on a signallabel and select "Adjust filter".
You can adjust or finetune the frequency, order or Q-factor of a filter and watch the result at the same time.
Use this dialog when you want to remove a filter.
To view the powerspectrum of a signal, leftclick on the signallabel and choose "Spectrum".
The Power Spectral Density of the corresponding signal will be shown (uV)2/FFT-resolution.
The amount of datasamples used to perform the FFT equals the data which is
shown on the screen. Increasing or decreasing the FFT blocksize affects the FFT resolution.
So, if the samplefrequency of the selected signal is 256Hz, the timescale is set to 10 seconds and
the FFT-blocksize is 512, the number of FFT's performed is (10 seconds x 256 Hz) / 512 = 5.
The output of the five FFT's are averaged.
In case you set the FFT blocksize equal to timescale x samplefrequency, exactly one FFT will be performed.
Note 1: You can freely customize the block size up to 1000 samples. Above that, only selected (optimized)
values can be choosen.
Note 2: The maximum block size possible equals the number of samples actually on the screen.
It's possible to let the FFT blocks overlap each other.
This powerspectrum is static and will not update when you change the fileposition (pressing PgUp/PgDn etc.).
This way you can compare multiple powerspectra from different signals and/or filepositions.
There is also a powerspectrum window available that automatically updates when you walk through a file.
Go to Window -> Power Spectrum.
A small dialog will be opened which shows a list of signals which are on the screen.
Click on a signal of interest and a docked Powerspectrum (FFT) will be opened.
This Powerspectrum will be automatically updated when you navigate through the file. Clicking on the
"ctls" button toggles the controls. Here you can switch-on colorbars to highlight different frequency-
regions. The properties of the colorbars (number of colorbars, frequency, color, etc.) can be adjusted
in the Settings menu. The height of the colorbar is relative to the sum, peak or average of the the
power of all the frequency-bins in that region, according to the settings in the Settings menu.
Use the sliders to zoom into a particular area.
When you check the "Amplitude" checkbox, the amplitude spectrum will be shown.
Printing can be done by clicking on the printbutton. You can print to a printer, file (PDF or Postscript)
or as an image. It is also possible to "print" to a textfile. It exports the data to text (ASCII).
Clicking on the cursorbutton will display a cursor. Use the mouse to drag the cursor.
Click again on the "ctls" button to hide the settings and make room for the FFT plot.
The algorithms for the window functions are taken from
Spectrum and spectral density estimation by the Discrete Fourier transform (DFT), including a comprehensive list of window functions and some new at-top windows
from the Max Planck Institute.
The Color Density Spectral Array (CDSA) can be enabled by leftclicking on a signallabel and selecting "CDSA", or via the Window menu.
The default settings are usually fine for EEG. Adjust the min/max levels for optimal results.
The CDSA will appear as a docked window on the bottom but can also be dragged to the top or can be put in detached mode (floating) where ever you want on your screen.
Adjust the size of the CDSA window as you prefer.
- Segment length:
The length of one measurement. The number of segments in the CDSA is: recording duration / segment length.
All segments together will present a horizontal array of segments.
- Block length:
The block length is the window size of the FFT. If the segment length is 30 seconds and the block length is 2 seconds,
15 FFT's will be performed in one segment (if the overlap is set to 0%). The output of these 15 FFT's will be averaged.
The block length affects the FFT resolution: FFT bin size (Hz) = samplerate / block length.
Higher block lengths increases the FFT resolution. Lower block lengths increases the signal to noise ratio.
- Overlap:
Percentage of an FFT block that will overlap the next FFT block. Increasing this value increases the number of FFT's that will be performed
and increases the signal to noise ratio but increases also the computational effort.
- Window:
Window is the kind of smoothing (taper) function used to prepare the data before the FFT is performed in order to lower the spectral leakage of the FFT.
- Min. / Max. frequency:
This can be set to only display the frequency range of interest. Maximum frequency is limited to half the samplingrate (nyquist frequency)
- Max. level:
Maximum level equals white. Adjust and experiment with this setting until you find the optimal setting.
- Min. level:
Minimum level equals black. Adjust and experiment with this setting until you find the optimal setting.
- Logarithmic:
If checked, it will apply the base-10 logarithm of the output of the FFT in order to increase the dynamic range.
- Power:
If checked, display the power instead of the voltage.
- Export data:
If checked, the data generated by the FFT will be exported to a text/CSV file.
The decimal separator for the numbers in this file is a dot. The column separator is a comma.
The hypnogram can be enabled via the "Window" menu. A dialog will appear where you can setup the labels for the different
sleep stages and the mapping to the annotations in the file. If you don't have the sleep annotations already in your file,
then you can import them using the tool "Import annotations/events" in the "Tools" menu.
When using the Annotation editor, the hypnogram will be updated realtime when adding, moving or deleting annotations.
You can select whether you want the hypnogram draws from onset to the next onset or according to the duration of the annotation.
(check or uncheck "Use annotations' duration for epoch length")
When you select the latter, any gaps / overlaps between consegutive annotations greater than 0.1 second will be highlighted red in order
to assist you with scoring the file.
In addition, there's the possibility to create colored overlays using user defined annotations.
Simply check the option "Add overlays" and insert the annotation(s). Click on the colored buttons in order to modify
the color. The transparency of the overlay color is set using the alpha channel: 0 = completely transparent, 255 is completely opaque.
The default value for the alpha channel is 32.
Right-click on the hypnogram and select "settings" in order to adjust the height or select "close" in order to close it.
Amplitude integrated electroencephalography (aEEG) or cerebral function monitoring (CFM) is a technique for monitoring brain function.
The implementation was inspired by "Calculation of compact amplitude-integrated EEG tracing and upper and lower margins using raw EEG data"
(Zhang, Dandan & Ding, Haiyan).
The aEEG can be enabled by leftclicking on a signallabel and selecting "aEEG", or via the Window menu.
Once the aEEG plot has been opened, you can right-click on it to close it or to adjust the height of the plot.
Averaging waveforms (multiple parts of a signal) can be done by rightclicking on an annotation (trigger) in
the annotationlist. A contextmenu will be opened, now click on "Average".
A dialog will be opened where you can choose to which signals you want to apply the averaging.
If necessary, change the "From"- and "To"-times in order to select a particular group of triggers.
Only the triggers with an onsettime that lies between the two timevalues will be used for the averaging.
Select the ratio of the time before and after the triggerpoint.
Now click on "Start" and a new window(s) will be opened which will show the result of averaging
all signalparts of all annotations (triggerpoints) with that specific name.
The yellow colored, dashed, vertical line is the triggerpoint, i.e. the onsettime of the annotations (triggers).
Use the sliders to zoom into the signal. The timescale is relative to the trigger (onset) point.
The result can be printed or exported as a new EDF/BDF file.
You can also use external triggers in ASCII(*.csv or *.txt) or XML format.
See Import annotations/events how to import triggers/events/annotations.
Note: It is also possible to change the selection of annotations/triggers for averaging, by editing the annotationlist first.
The ECG powerline interference subtraction filter is inspired by the subtraction method described by:
- Subtraction Method For Powerline Interference Removing From ECG
Chavdar Levkov, Georgy Mihov, Ratcho Ivanov, Ivan K. Daskalov Ivaylo Christov, Ivan Dotsinsky
- Removal of power-line interference from the ECG: a review of the subtraction procedure
Chavdar Levkov, Georgy Mihov, Ratcho Ivanov, Ivan Daskalov, Ivaylo Christov and Ivan Dotsinsky
- Accuracy of 50 Hz interference subtraction from an electrocardiogram I. A, Dotsinsky I.K. Daskalov
- Dynamic powerline interference subtraction from biosignals
Ivaylo I. Christov
The subtraction method extracts the powerline interference noise (50/60Hz and harmonics) during a
a linear region between two consecutive QRS complexes and stores it in a buffer.
The reference noise from the buffer is used to subtract it from the signal outside
the linear region i.e. during the QRS complex.
This method only works correctly when the samplefrequency of the ECG recording is an
integer multiple of the powerline frequency.
In case they are synchronized, this method will remove also the harmonics of the
powerline frequency. In that case extra notch-filters for the harmonics are
not necessary. The advantage of this method is that it will not cause ringing or other distortion
in the waveform of the QRS complex (like notch-filters do).
The following rules apply:
- The samplefrequency of the ECG recording must be at least 240 Hz
- The samplefrequency of the ECG recording must be an integer multiple of the powerline frequency
- The physical dimension (units) of the ECG signal must be expressed in uV, mV or V.
- The resolution of the ECG signal must be better than or equal to 10 uV/bit.
Note: Acceptable results will be achieved when the accuracy of the recording sampleclock is <=100ppm
and the powerlinefrequency is within +/-400ppm (+/-0.02Hz).
A perfect result will be achieved when the recording sampleclock is synchronized with the
powerlinefrequency in hardware (e.g. using a Phase Locked Loop in the acquisition hardware).
The EEG powerline interference subtraction filter is inspired by the subtraction method described by:
- A simple method for the removal of mains interference from pre-recorded electrophysiological data (W.J. Heitler)
The subtraction method extracts the powerline interference noise (50/60Hz and harmonics) from the last 0.5 second
and stores it in a buffer.
The reference noise from the buffer is used to subtract it from the signal.
This method only works correctly when the samplefrequency of the EEG recording is an
integer multiple of the powerline frequency.
In case they are synchronized, this method will remove also the harmonics of the
powerline frequency. In that case extra notch-filters for the harmonics are
not necessary. The advantage of this method is that it will not cause ringing/distortion
(like notch-filters do).
The following rules apply:
- The samplefrequency of the EEG recording must be at least 100 Hz
- The samplefrequency of the EEG recording must be an integer multiple of the powerline frequency
Note: Acceptable results will be achieved when the accuracy of the recording sampleclock is <=100ppm
and the powerlinefrequency is within +/-400ppm (+/-0.02Hz).
A perfect result will be achieved when the recording sampleclock is synchronized with the
powerlinefrequency in hardware (e.g. using a Phase Locked Loop in the acquisition hardware).
The quickest way to run the QRS detector and to import the beats as annotations, is to use the QRS detector in the Tools menu.
You will be able to select one of the signals on your screen as the source for the QRS detector.
The selected signal must have a samplerate of at least 200 Hz and must have a physical dimension (unit)specified in uV, mV or V.
The detector will process the whole file and it will import the R-peaks as annotations.
The label "R-peak" is used by default for the annotations but this can be changed in the settings (Settings -> Options).
Also, the heart rate will be plotted in a separate window. Right-click on the plot to access some settings for the plot.
The annotations can be exported to different formats using the Export annotations tool.
Otherwise, toggling the "Heart Rate" button toggles on or off the detection and calculation of the heartrate expressed in beats per minute (bpm).
The "Heart Rate" button can be found in the Signalproperties dialog.
This tool will create a new trace on the screen that represents the beats per minute.
The algorithm used for the QRS detection is derived from "A Real-Time QRS Detection Algorithm" by Jiapu Pan and Willis J. Tompkins.
https://courses.cs.washington.edu/courses/cse474/18wi/labs/l8/QRSdetection.pdf
The R-peaks and/or RR-interval times (beat to beat) can be exported to an ascii-file for analysis in other programs.
After you switched on the Heart Rate detection, go to "Tools -> Import/Export ECG RR-interval".
Only the part of the signal that is visible on the screen will be exported, except when you check the "Whole recording" checkbox.
There is also the possibility to import the RR-intervals as annotations.
Some statistics about the Heart Rate (variability) are available here.
When the ECG QRS detector is enabled, you can view some statistics about the
Heart Rate (variability) in the Signalproperties. Once you see the Signalproperties dialog,
click on the "Statistics" button. A brief description of the statistics:
- Mean, the average of the RR-intervals.
- SDNN, the standard deviation of RR-intervals.
- RMSSD, the square root of the mean squared difference of successive RR's.
- NN20, the number of pairs of successive RR's that differ by more than 20 ms.
- pNN20, the proportion of NN20 divided by total number of RR's as a percentage.
- NN50, the number of pairs of successive RR's that differ by more than 50 ms.
- pNN50, the proportion of NN50 divided by total number of RR's as a percentage.
The data used for the statistics is derived from the timewindow that is visible on the screen.
In case you want to see this info permanently at the bottom of the screen during navigation through the file,
use the following procedure.
First, import the output of the ECG QRS detector as annotations as described here: Export/Import ECG RR-interval
Now right-click on an annotation named "R-peak" and select "Heart Rate Variability".
A new window with the statistics will appear. Close it.
From now on, after changing file position e.g. PgUp/PgDn, you will see the updated statistics at the bottom of the screen.
This tool can be used to export the output of the ECG QRS detector,
or
it can be used to import the output of the ECG QRS detector as annotations.
After having imported the ECG QRS detector output as annotations, more options
are available by right-clicking on one of the annotations in the annotationlist.
The Custom FIR filter can be used to create a customized Finite Impulse Response (FIR) filter.
There are three methods to design/specify a FIR filter:
If you already have the filter coefficients (so-called "filter taps") obtained from another source, simply copy and paste them into
the Filter taps list and select the signals to which the filter needs to apply to (select/highlight them with the mouse).
The second method is to use the FIR filter designer that creates a phase-linear, Windowed Sinc FIR filter with flat passbands,
or a phase-linear filter with a user specified passband and ripple and stopband attenuation, using the Parks-McClellan algorithm.
Specify filter type (lowpass, highpass, bandpass, bandstop), samplerate of the selected signal(s), the cutoff frequency and
and the transition bandwidth (steepness).
Specify the pass- and stopbands with their respective start- and stopfrequencies, gain and ripple.
A passband usually has a gain of 1 (it must be more than zero).
A stopband must have a gain of zero. Bandwidth must be at least 5% and the transition between bands must be
at least 5%. After entering the filter specifications press the "Design" button.
Examples:
For a lowpass filter, specify a passband and a stopband (2 bands).
For a highpass filter, specify a stopband and a passband (2 bands).
For a bandpass filter, specify a stopband, a passband and a stopband (3 bands).
Click on the "No predefined" combobox to select one of the predefined filtertypes to help you get started.
Note: The Parks-McClellan algorithm cannot find solutions for all possible filter specifications.
In case it's not able to create the filter, try changing the filter parameters or use the Windowed Sinc filter instead.
Resuming:
- Select the filter design method by selecting the tab "Windowed Sinc" or "Parks-McClellan".
- Adjust the parameters for the filter.
- Click on the "Design" button.If there are no errors in the parameters, the coefficients will be calculated
and shown together with the filters' response curve.
- Select and highlight with the mouse the signals which you want to apply the filter to.
- Click on the "Apply" button.
Note: A filter "tap" is simply a coefficient/delay pair. The number of taps is an indication of:
- the amount of memory required to implement the filter
- the number of calculations required
- the amount of "filtering" the filter can do; in effect, more taps means more stopband attenuation, less ripple, narrower filters, etc.
Warning: using a FIR filter with many taps can slow down the processing considerably.
The Spike filter can be used to filter out spikes, fast transients, glitches or pacemaker impulses.
Velocity is expressed in units (e.g. uV) per 0.25 milli-Seconds. If the filter detects two fast transients
with opposite polarity and within 3 milli-Seconds, it will consider it a spike which will be suppressed.
The transient is measured for every sample with a delta t of 0.25 milli-Seconds.
(The sample will be compared with an older sample 0.25 milli-Seconds before.)
Setting the value of velocity too high, will cause the spikes not to be detected.
Setting the value of velocity too low, will cause false triggers.
Hold-off is used to prevent a re-trigger of the filter within the hold-off period after the last trigger.
The Spike filter cannot be used with samplerates lower than 4000 Hz.
Signals with lower samplerates will not be visible in the Spike filter dialog.
The Spike filter dialog has no apply button. Activating or de-activating the spike filter is simply done
by selecting or de-selecting one or more signals.
Changing the value of velocity or holdoff will cause an immediate update of the screen.
There are multiple ways to zoom-in:
- By pressing (several times) Ctrl++ or Ctrl+- (this will zoom-in/-out horizontally only i.e. the timescale changes).
or
- By using the mouse wheel while keeping the Ctrl key pressed (this will zoom-in/-out horizontally only i.e. the timescale changes).
or
- You can zoom into a trace on the screen by drawing a rectangle with the mouse.
Keep the left mousebutton pressed and move the mousepointer in the direction of the
right lower corner. When you release the left mousebutton, the content of the rectangle
will be expanded to the whole screen. (It will zoom-in horizontally and vertically) You can repeat this step and zoom in again.
(The rectangle drawn must have a size of at least 25 x 25 screen pixels. If not, the action will be ignored.)
Use the backspace button on your keyboard to zoom out and restore the previous settings.
After using the backspace button, you can zoom in again by pressing the insert button on
your keyboard.
You can use a floating ruler for measurements. Leftclick on the signallabel of the signal
of interest. A small dialog will appear. Click on "Ruler". A floating ruler will appear.
Drag and drop the ruler with the left mousebutton. Rightclick to remove the ruler or press the "Esc" key.
In Settings -> Options -> Other you can enable or disable a property of the ruler that will adjust the width
of the ruler automatically in such a way that only integer numbers for "Hz" will be shown.
You can use one or two crosshairs for precise measurements. Press Alt+Shift+C or Leftclick on the signallabel
of the signal of interest. A small dialog will appear. Click on Crosshair. A crosshair will
appear. Drag and drop the crosshair with the left mousebutton. Now you can add another crosshair.
The second crosshair will show you the differences (delta) in time and value of the signal(s).
Pressing repeatedly Alt+Shift+C will move the crosshairs to the next signal.
Rightclick to remove the crosshair(s) or press the Escape key.
The looks of the crosshairs can be changed in the settings menu. You can enable/disable a horizontal line
and/or a centered dot.
The resolution (milliSecond or microSecond) can be selected in: Settings -> Options -> Annotation editor.
Some statistical info about a signal can be found in the Signalproperties.
and clicking on the "Statistics" button.
It will show the number of samples showed on the screen, the sum, mean, RMS (root mean square), and MRS
(mean root square i.e average of rectified values) value of the samples showed on the screen.
In addition, the number of zero-crossings (i.e. switching from positive to negative or vice versa)
and the frequency of that part of the signal will be shown.
Only the data from the page visible on the screen is used to calculate the statistics.
Montages can be used to store information about which signals should be shown on
the screen, what combinations (derivations) should be made and other properties
like filters, amplitude (sensitivity), offset, alias, color and pagetime (timescale).
Once you have loaded a file for review and you have added all the signals you want
(and/or all the combinations/derivations) and you have adjusted amplitude and colors
and/or filters, you can save them as a montage.
Go to Montage -> Save. A dialog will be opened where you can choose a directory
and filename to store your montage.
Next time when you load a file, you can load a montage via Montage -> Load.
You can view the properties of a montage via Montage -> View saved montages or
Montage -> View this montage.
note:
There are two possible methods to identify the signals in a montage, label based and index based.
The default method is label based which uses the name of signal to identify to which signals the
montage settings need to be applied.
The second method uses the index (order of appearance in a file) of a signal. When in doubt about
which method to use, use the default (label based).
You can bind the F1 to F12 keys to a maximum of twelve different montages. This way you can quickly switch
between different montages by pressing key F1, F2, etc.
Go to Montage -> Edit key-bindings for montages.
It is possible to create keyboard shortcuts (bindings) for predefined montages.
Go to Montage -> Edit key-bindings for montages. Click on the first row (F1);
Now you can select a montage. You can assign a maximum of twelve different
montages to keys F1 - F12.
Now you can quickly switch between predefined montages by pressing one of these keys.
You can save an EDFbrowser session for later use. This is particularly useful when you need to open
multiple files and want to keep the state of the montages (including filters), file position, timesync mode,
hypnograms, CDSAs, video, etc.
When loading a previously saved session, all files opened (if any), will be closed. Then the file(s) which were
saved in the session file will be opened. Then, the montage/signals will be added to the screen including any filters.
The session file will have the extension .esf (EDFbrowser Session File).
When saving a session, absolute paths or relative paths can be used.
The default is to use the absolute paths but you can change this in Settings -> Options -> Other
This setting is evaluated only when saving a session, not when loading a session.
So, changing the setting between absolute and relative paths does not affect previously saved sessions.
Pay attention, both ways of storing the paths have their advantages and disadvantages. If you want to share your files
including the session file, with other people, than you should use relative paths and make sure that all files are
on a local drive and on the same partition and/or station. Ideally, put all files in the same directory.
In case you don't want to share your session file, it's recommended to use the absolute paths which will work
with all file locations you have access to, including network drives. Also, when using absolute paths, you can move the
.esf file to another directory (on the same computer!) without breaking anything.
Note: The settings in Settings -> options are not in included in the session file.
Changing one of those settings might still affect a session in some way.
Alternatively, you can use a montage instead of a session. Montages can be applied to different
recordings (of the same type).
If you want to check the details of a saved session, you can do this by using: Session -> View saved sessions
It will show you the settings stored in the session file. It also checks if the file paths are still valid.
Go to the Settings menu to change the colors like you prefer.
In the Settings menu go to the Calibration tab to calibrate your screen.
This will achieve that the amplitude expressed in units/cm will have
the correct height/value.
When you open multiple files, you can choose to "timelock" all files with
each other. This means that when you navigate (by pressing PgUp/Pgdwn)
this will affect all signals on the screen. When you choose to "unlock" the files,
only the signals of the "reference" file will be moved on the screen. You can
change which file should be the reference in the Timesync menu.
There are three kinds of timelock:
- Offset
The start of the recording of all files will coincide with each other.
- Absolute time
The clock/date-time of the different files will coincide with each other.
(this is only useful when the period of registration overlap each other or when they are consecutive)
- No timelock
Only the signals from the file that has the "time reference" will move when you navigate.
Use this option when you want to align two or more files manually.
- Userdefined synchronizing
Files are timelocked with a custom offset. Usually after synchronizing the
files manually.
The time shown in the leftbottom corner represents the time from the
file that acts as the "reference". You can change which file should be
the reference in the Time menu.
You can adjust the horizontal position (time) of two different files by using
two crosshairs. Put one crosshair at a signal of the first file and put the second
crosshair at a signal of the second file. Now click on Time -> synchronize by crosshairs.
Now the position of the two crosshairs (and that particular position of the two files) will
coincide with each other.
Or you can drag a trace horizontally by keeping the middle mouse button (wheel) pressed.
To view the content of the header of a file (subject, recording, signalproperties like samplerates etc.),
go to file -> info and select the file of interest.
This tool can be used to edit (or anonymize) the header (file info) like subject, birthdate, etc.
You can also edit labels (signalnames), physical dimension, pre-filter and transducer names.
If the file you want to edit has been opened in EDFbrowser already, close it first before starting
this tool.
This tool can also be used to repair a file that cannot be opened with EDFbrowser because of an incompatibility.
Errors that can be repaired with this tool:
- invalid characters in the header (the EDF format allows (7-bit) ASCII characters only)
- a wrong or missing timestamp (format) in the header (e.g. a colon instead of a dot between the numbers)
- a wrong or missing datestamp (format) in the header (e.g. a dash instead of a dot between the numbers)
- a wrong number of datarecords written in the header
- a wrong filesize (file is not ending at the boundary of a datarecord)
- a broken number i.e. a number with decimal(s) in the digital maximum/minimum field
- the value for digital maximum is lower than or equal to digital minimum
- the value for physical maximum equals physical minimum
- illegal or extraneous data in the reserved field
Do as follows: In EDFbrowser go to Tools -> Header editor. Select the file. Now click on the
"save" button (you don't need to edit the content manually). Close the header editor.
Now open the file in the usual way.
This tool converts all the signals in an EDF or BDF-file to a plain ASCII text-file.
It supports different samplerates between the signals.
Four textfiles will be created:
- EDFfilename_header.txt contains the general header of the file.
- EDFfilename_signals.txt contains all signal headers of the file.
- EDFfilename_data.txt contains the data of all signals.
- EDFfilename_annotations.txt contains all annotations.
filename_data.txt contains a separate line for each sampletime that occurs.
Note to windows-users: these lines are separated by a linefeed only,
so the file does NOT look OK in Notepad, use a real editor instead.
Each line contains the comma-separated values of the sampletime and of all
samples that were taken at that time.
Sampletime is expressed in seconds, while sample values are expressed in the
physical dimensions that are specified in filename_signals.txt.
All values have a dot (.) as decimal separator.
Different sampling frequencies are allowed in the file. In that case,
not all signals are sampled at each sampletime.
Those sample values are simply left empty, but the comma-separator is there.
note:
All comma's in headertext and annotations will be replaced by
single quotes (') because they interfere with the comma-separated
txt-files.
Unicode/UTF-8 decoding in annotations is not supported because of
the limitations of the ASCII txt-files.
This tool can be used to reduce the number of signals in a file and/or to reduce the duration (length) of a file
and/or to reduce the samplerate of one or more signals.
First, open the file in the usual way in EDFbrowser. Then start the tool and select the file.
Uncheck all the signals you do not want into the new file. If you want to shorten the file, choose the first
datarecord (starttime) and the last datarecord (endtime).
If you don't want to downsample, set the samplerate divider to 1 (default). Otherwise select one of the possible
divider values. Those values depend on the actual samplerate and the datarecord duration.
The anti-aliasing filter order can be chosen as well. The default is 4th order. This prevents practically
any chance of aliasing.
Note: if the samplerate divider is set to 1, no downsampling and no filtering will be performed.
This tool can be used to export the signals which are visible on the screen to a new EDF/BDF file
including any applied filters and or derivations.
This tool exports the annotations/events from an EDF+/BDF+ file.
They can be saved as CVS, EDF+ or XML.
If you have opened multiple files, you can choose to export all
annotations of all opened files into one new file. The header of the file which acts as
the "reference" will be copied into the new EDFplus file. The "reference" file can be selected
in the menu Time -> Timereference.
This tool imports annotations/events. The inputfile can be ASCII (CSV), XML, EDF+, BDF+ or MIT/WFDB format.
It's also possible to use a triggersignal in a DC-coupled channel/signal.
Note: You have to open an EDF/BDF file first in order to be able to import annotations/events.
An ASCII-file must be organised in rows and columns.
The decimal separator of numbers must be a dot.
Grouping characters like the comma in for example "12,345.678" are not allowed.
The example must be written as "12345.678".
Line endings must use the "newline" or "linefeed" character (Unix style)
In the import dialog, set the separator character (the character that separates the columns),
for example: "tab" (tabulator),"," (comma) or ";"(semicolon).
Set the column number of the onset time and the description/label.
If the file starts with a header, adjust the "Data starts at line" spinbox in order to skip these lines.
If there's only one column in the ascii/txt file (with onset time values), check the
"Manual description" button and enter a description. This description will be used for all
annotations that will be imported from that ascii/txt file.
The onset time encoding can be expressed as:
XML-files need to be encoded either in ISO-8859-1 or UTF-8 character encoding.
The root element name must be "annotationlist", child elements must be named "annotation".
Inside the element "annotation" are the elements "onset", "duration" and "description".
Onset is expressed in the ISO date-time format "yyyy-mm-ddThh:mm:ss" or "yyyy-mm-ddThh:mm:ss.xxx".
Duration is expressed in seconds "145.23" or empty.
Example:
<?xml version="1.0" encoding="UTF-8"?> <annotationlist> <annotation> <onset>2006-11-14T09:14:31.0000000</onset> <duration></duration> <description>REC START IIB CAL </description> </annotation> <annotation> <onset>2006-11-14T09:14:31.0000000</onset> <duration></duration> <description>PAT IIB EEG </description> </annotation> <annotation> <onset>2006-11-14T09:14:32.0000000</onset> <duration></duration> <description>A1+A2 OFF </description> </annotation> </annotationlist>
The onset time of the annotations in MIT/WFDB format are expressed in samples offset from the start of the recording.
In case your file contains signals with different samplerates, you need to specify which samplerate should be used to
calculate the onset time of the annotations.
MIT/WFDB annotation files can have various extensions like *.ari *.ecg *.trigger *.qrs *.atr *.apn *.st *.pwave
A "DC-event" consist of a positive impulse (startbit) followed by eight databits and a stopbit (lowlevel).
This signal is recorded onto a dc-coupled inputchannel of the hardware during acquisition.
The eight databits are used to identify a maximum of 256 different types of events/triggers.
The startbit has always a high level and the stopbit has always a lowlevel.
At idle the signal should stay at low level.
The following example shows a dc-event with decimal code 170:
onsettime of the event | bittime | |<->| v | | 1 Volt +---+ +---+ +---+ +---+ +---+ | | | | | | | | | | | | | | | | | | | |<--triggerlevel = 0.5 Volt | | | | | | | | | | 0 Volt ----------+ +---+ +---+ +---+ +---+ +---+-------------- ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ | | | | | | | | | | | idle | | | | | | | | | stopbit, always low | | | | | | | | db7 | | | | | | | db6 | | | | | | db5 | | | | | db4 | | | | db3 | | | db2 | | db1 | db0 startbit, always high
The binary code in the example is 01010101 which is 170 decimal.
The rising edge of the startbit represents the time of the event (for example, an evoked potential).
The voltage levels can be freely chosen. In the importdialog adjust the triggerlevel to a level that is equal
to: (lowlevel + highlevel) / 2.
Also, set the bittime used for the DC-event.
Note that the intervaltime between two consegutive events must be more than ten times the bittime.
It is possible not to use the code but just only the startbit. In that case all triggers will have the code "0".
In that case you can use a simple on/off pulse as a trigger.
Note: The signal which contains the DC-event must be put on the screen in order to be available in the selectionscreen.
Also, the signal must be unipolar, i.e. not a derivation. Next, the detectionalgorithm of the triggers in this signal will bypass
any filters.
To use the Z-EEG measurement, leftclick on the signallabel and choose "Z-EEG".
Z-ratio = ((Delta + Theta) - (Alpha + Beta)) / (Delta + Theta + Alpha + Beta)
Z-ratio, therefore, is a measure of the relative contribution of slow to fast EEG activity within a
2-second epoch. It ranges from -1 to +1, with negative values indicating a predominance of fast
activity and positive values indicating a predominance of slow activity.
A per page score for z-ratio was derived by averaging the majority of 15 2-second z-ratios.
The average of the majority is the z-page.
The algorithm is derived from
https://pubmed.ncbi.nlm.nih.gov/8746389/
This tool converts an ASCII-file (.csv or .txt file) to EDF or BDF. The following rules apply:
- the data in the ASCII-file must be presented in their physical dimension (i.e. uV, bpm, mmHg, etc.)
- the data in the ASCII-file must be organised in columns and rows
- a row must contain values from different signals/channels acquired at the same sampletime
- a column must represent a continuous sampled signal/channel
- all signals/channels must have the same samplefrequency
- the timeinterval between two consecutive samples of the same signal/channel must have a stable value (fixed samplerate)
- the decimal separator (if any) must be a dot, not a comma
Enter the following parameters:
- the character that separates the different columns (i.e. "tab" ","(comma) ";"(semicolon) " "(space) etc.)
- the number of columns
- the line at which the data starts. some ASCII-files contain headers, so use this entry
to tell the converter to skip n-lines before interpreting the data
- samplefrequency
- subject (the name of the subject)
- recording (a description about the type of recording/acquisition)
- startdate and time of the recording/acquisition
- signals: for every column in the ASCII-file, there is a row in the signals-table.
every row has a checkbox and four inputfields:
- label (name of the signal/channel i.e. FP1, SaO2, Heartrate, etc.)
- physical maximum (the maximum physical value that can occur, i.e. the maximum inputlevel
of the acquisition equipment. A common value in EEG applications is 3000 uV)
Physical minimum will be equal to physical maximum. For example, if you enter 3000, the
range will be from +3000 to -3000
- physical dimension (units, e.g. uV, mV, %, mmHg, bpm)
- multiplier is normally 1.0. Some programs exports their ascii-data expressed in Volts while
the signal is in the microVolt range. By changing the multiplier you can add gain to the signal before converting.
For example, if the ascii-data is expressed in Volts and the signal is in the microVolt range (EEG),
set the physical maximum to 3000, set the physical dimension to "uV" and set the multiplier to 1000000.
It is possible to exclude columns, by unchecking the corresponding row in the signals-table.
This can be useful when a column contains a time/datestamp or when you don't want to include
a particular signal in the EDF/BDF-file.
Click on the startbutton to start the conversion.
Click on the savebutton to save the entered parameters in a template.
Click on the loadbutton to load parameters from a template.
Note: It's important to enter a correct value in the "physical maximum" field.
If you to choose this value too small, the top of the signal will clip at the peaks.
If you choose this value too big, small values of the signal will not be visible
and the signal will look coarse.
The ideal value for the "physical maximum" is just above (or equal to) the maximum value that can occur,
i.e. the maximum possible output of the sensor or ADC, either positive or negative.
Note 2: The maximum physical value that the converter can handle, is limited to 9999999. Values above
will be clipped to 9999999 and values below -9999999 will be clipped at -9999999.
If you expect higher values in your ascii-file, use the multiplier. For example,
set the multiplier to 0.1, 0.01 or 0.001, etc.
Note 3: If you think that, later, you want to make derivations (for example P4 - C4 in EEG applications),
make sure that the signals have the same value in the physical maximum field.
Also, do not set the checkbox "Auto detect physical maximum".
This tool converts EEG-files in Nihon Kohden format to the European Data Format
(EDF or EDF+ with annotations/events).
Three files are needed for a conversion to EDF+:
- nihonkohdenfile.eeg
- nihonkohdenfile.pnt
- nihonkohdenfile.log
In case you have only the nihonkohdenfile.eeg you cannot convert to EDF+.
It is still possible to convert to EDF. In that case there will be no annotations/events and patientinfo.
The Marker/Event channel will be decoded on the fly when opening the new created EDF+ file with EDFbrowser.
If the screen becomes too "noisy" because of drawing too many annotations, you can switch them of by
unchecking the "Read Nihon Kohden Trigger/Marker signal" checkbox in the Settings-menu.
The new settings will be effective after closing and re-opening the file.
Also, you can selectively hide annotations by rightclicking on an annotation in the annotationwindow.
Note: In case you changed the default electrodenames in the Nihon Kohden acquisition software,
you need also the *.21e file which contains the electrode names (labels).
This tool converts the waveform data and events/triggers from a collection of Unisense data files to EDF+.
Limitations of the converter:
- the Unisens XML-file must be encoded in UTF-8 or ISO-8859-1.
- Broken numbers for samplerate >= 1 Hz will be rounded to the nearby integer e.g. 2.75 Hz will be rounded to 3 Hz.
Samplerates below 1 Hz will be rounded to 1/2, 1/3, 1/4, etc. til 1/1000 Hz.
Lowest samplerate is 0.001 Hz
- Output format of the converter is EDF+ (16-bit) when all files have a samplesize of 16-bit or lower.
When one or more files have a samplesize of more than 16-bit, output format of the converter will be BDF+ (24-bit).
An extra feature of the converter which is not in the official Unisens data format specification:
- support for binary sampletypes: int24 and uint24.
This tool converts an SCP-ECG (EN 1064:2005+A1:2007) file to EDF+.
Before conversion it performs a CRC-check on the file.
Bimodal compression is supported.
Reference beat subtraction is supported.
Pacemake spike measurements, if present, are imported as well.
Encoding using the default Huffman table is supported.
Customized Huffman tables are not supported.
Character-encoding other than ASCII/Latin-1 is not supported.
This tool converts the waveform data and annotations from the MIT-format to EDF+.
It's used for files from PhysioBank at physionet.org.
At this moment it supports the Formats "212", "16" and "61".
Most of the signal files in PhysioBank are written in format 212.
Annotations (in the files with extensions *.atr, *.ari, *.ecg, *.trigger, *.qrs, *.apn, *.st)
will be read and converted to EDF+ as well.
In case the file that contains the annotations has a different extension, just rename the
extension to .atr. Otherwise the converter will not read the annotations.
This tool converts a Manscan MICROAMPS binary data file to EDF+.
Two files are needed for the conversion, one with the extension ".mbi" and one with the extension ".mb2".
This tool converts a Biosemi data file to BDF+ and makes it possible to visualize the
trigger interrupts and the "new epoch" in EDFbrowser.
The Status signal of the Biosemi datafile which contains the input triggers will be
converted to annotations in the BDF+ format. The original file will not be altered.
Instead, a copy of the file will be made in the BDF+ format which will contain annotations
extracted from the Status signal (which contains the input triggers).
The original Status signal will be preserved in the new file.
You'll find the Biosemi to BDF+ converter in the Tools-menu.
BDF+ files are backwards compatible and can be read and processed with older BDF software.
BDF+ files can be converted to EDF+ while preserving your annotations.
Use the BDF to EDF converter in the Tools menu.
Limitations:
- The maximum number of annotations (input triggers) in the new file is:
duration of file in seconds multiplied by 32. If the original file contains more trigger input events, they will be ignored.
If, for example, the duration of a file is 5 minutes, the first 9600 input triggers will be converted to BDF+ annotations.
- The resolution of the annotations is limited to 100 microSeconds (0.0001 sec.).
- Only the 17 lowest bits of the Status signal will be evaluated and converted.
This tool converts the ASCII-output (*.txt) of the Finometer (Beatscope) to EDF.
Finometer is a non-invasive stationary blood measurement and beat to beat haemodynamic monitoring system
made by Finapres Medical Systems.
Start Beatscope and open the file of interest.
Go to File -> Export.
- set export type to textfile
- set timeformat to seconds
- check output: header
- set averaging to none
- check datacolumnwidth: delimiter
- set delimiter to semicolon ";"
- check export beats and/or events: beats only
- select the channels you want to export
Now click on the Export-button to create the file.
In EDFbrowser, you'll find the Finometer to EDF converter in the Tools-menu.
In order to create a correct EDF-file header, enter the subject and a description about
the type of recording/acquisition.
Adjust the startdate and starttime to the date and time when the Finometer registration took place.
Click on the "select file" button to select the Finometer/Beatscope file.
This tool converts a Waveform Audio File (*.wav) to EDF.
The converter accepts 8, 16 and 24 bit wave files.
It accepts 1 to 256 channels and samplerates from 1 Hz to 500000 Hz.
8 and 16 bit wave files will be converted to EDF. 24 bit wave files will be
converted to BDF. Compressed files are not supported.
This tool converts an audiorecording containing an FM modulated ECG waveform to EDF.
The carrier centerfrequency of the FM signal is 1900Hz. Frequency deviation is 100Hz
per milli-Volt. This technique is used to upload ECG recording via POTS (Plain Old Telephone Service).
The converter has been tested with audio recordings from Cardiette Microtel but should work
also with devices from Aerotel HeartView.
The converter accepts recordings stored in WAV format with 16-bit resolution.
The recording must contain only one channel (mono) and the samplerate must be
one of the following: 8000, 11025, 16000, 22050, 32000, 44100, 48000, 88200 or 96000 Hz.
This tool converts the ASCII-output (*.csv) of the Nexfin (FrameInspector) to EDF.
Nexfin is a non-invasive stationary blood measurement and beat to beat haemodynamic monitoring system
made by Bmeye B.V.
Start FrameInspector (it can be downloaded from the Nexfin website), open the file of interest and
choose "Generic export".
Now click on the Export-button to create the csv and bin files.
In EDFbrowser, you'll find the Nexfin to EDF converter in the Tools-menu.
In order to create a correct EDF-file header, enter the subject and a description about
the type of recording/acquisition.
Adjust the startdate and starttime to the date and time when the Nexfin registration took place.
Select which file you want to convert: beat-to-beat data, blood pressure waveform or the analog inputs.
(for analog inputs: select "4 inputs" in case AO option is configured, else "1 input")
Click on the "select file" button to select the Nexfin/FrameInspector csv or bin file.
For beat-to-beat data you need to select the file that ends with "_100.csv".
For blood pressure waveform you need to select the file that ends with "_103.bin".
For the analog inputs you need to select the file that ends with "_032.bin".
This tool converts an Emsa eeg-file (*.PLG) to EDF+. If the corresponding *.LBK file is
present, the annotations/events will be exported to EDF+ as well.
This tool converts a CardioTrak EDF file to EDF+ and extracts non-compliant events to EDF+ annotations.
This tool converts an EDF+D file to multiple EDF+C files. An EDF+D file can contain
interrupted recordings. This converter export every uninterrupted recording to a new
uninterrupted EDF+C file.
This tool can be used for BDF+D files as well.
This tool converts a BDF file to an EDF file. A BDF+ file will be converted to
EDF+, including the annotations/events.
It offers a first order highpass-filter to remove the DC-offset of the signal in order
to make the 24-bits samples fit into the 16-bits samples of EDF.
The cutoff-frequency of the highpass-filters can be adjusted for every signal.
You can also select which signals have to be converted into the new EDF(+) file.
Increasing the divider lowers the resolution but increases the range (physical
maximum and minimum)
This tool checks if a file is EDF(+) or BDF(+) compliant.
It checks the data and all the annotations and timestamps.
When it finds an error, it tells you what the error is and where it is in the file.
A description of the EDF fileformat can be found at:
https://www.teuniz.net/edfbrowser/edf format description.html
A description of the EDF+ fileformat can be found at:
https://www.edfplus.info/specs/edfplus.html
A description of the BDF fileformat can be found at:
https://www.biosemi.com/faq/file_format.htm
A description of the BDF+ fileformat can be found at:
https://www.teuniz.net/edfbrowser/bdfplus format description.html
note:
A compatibility check of the header of the file is always done when you open an
EDF(+)/BDF(+) file. When a file appears to be incompatible it will not be opened,
you will be informed about the cause of the incompatibility instead.
Note: If a file cannot be opened with EDFbrowser because of invalid characters in the header
(the EDF format allows 7-bit ascii characters only), you can repair the non-compliant file with the header editor.
In EDFbrowser go to Tools -> Header editor. Select the file. Now click on the "save" button
(do not edit the content!). Close the header editor. Now open the file in EDFbrowser.
This tool converts the binary (raw) (ADC-)data from a file to a new EDF-file.
This is a kind of "last resort" measure when there are no other tools or EDF-converters available.
You need to figure out how the data in your file is organised and you need to have a good knowledge of
how computers store data in memory (file). This tool is for experts only.
All signals must have the same samplerate and the data-samples in the file must be stored sequentially:
sample 1 ch.1, sample 1 ch.2, sample 1 ch.3,
sample 2 ch.1, sample 2 ch.2, sample 2 ch.3,
sample 3 ch.1, sample 3 ch.2, sample 3 ch.3,
sample 4 ch.1, etc.
Set the samplefrequency and the number of signals (channels).
The samplesize is the size of a digital sample expressed in bytes.
Encoding is the binary system used to present a digital value:
value bipolar offset binary 2's complement
15 1111 0111
14 1110 0110
.... .... ....
2 1010 0010
1 1001 0001
0 1000 0000
-1 0111 1111
-2 0110 1110
.... .... ....
-15 0001 1001
-16 0000 1000
See also: Coding Schemes Used with Data Converters
Most files contain some kind of header. Adjust the Offset to the point where the data starts.
When a digital sample consists of multiple bytes, and the LSB (least significant byte) comes first,
select little Endian. If the MSB (most significant byte) comes first, select big endian.
Data blocksize is usually 0. Used only when some bytes must be skipped in a regular interval.
Skip bytes is the number of bytes that must be skipped (ignored) after every Data blocksize number of databytes.
Skip bytes is ignored when Data blocksize is set to 0.
You can print/export to PDF in File -> Print -> to PDF
You can print/export to an image in File -> Print -> to Image
The option "Print to EDF" creates a new EDF-file which will contain the signals
(and annotations) as they are on your screen. What you see is what you get,
like when you print to a printer, PDF-file or image. You can also open multiple
files, shift them in time (by using the different timelock options in the timemenu)
and "print" the result as one new EDF-file. The result will contain the time-shifted
signals.
It is also possible to extract one or more signals from one (or more) files into a new file.
This function is located in File -> Print menu.
The following rules apply:
- the length (duration) of the new file will be equal to the selected displaytime (timescale)
- combined signals (derivations) will become one signal like they are on your screen
- in case of multiple files, the new file will get the subject, recording, date and time
of the file that has the reference at the moment of printing.
You can select the reference in the Time-menu
- if (one of) the file(s) is of type EDF+, the outputfile will be an EDF+ file as well,
otherwise EDF
- when selected multiple files, either the datarecord duration of these files
needs to be the same or an integer multiple of each other,
or the samplerates of the signals must have integer values.
- if you selected any filters, these will be applied as well
The option "Print to BDF" creates a new BDF-file which will contain the signals
(and annotations) as they are on your screen, in the same way like the Print to EDF function,
except that it is also possible to print EDF(+) files to BDF. You can even mix
EDF and BDF files onto your screen and print them to a new BDF file.
This tool can be used to give a selected group of signals the same resolution.
Here, resolution means units per least significant bit e.g. uV/bit.
The selected signals must have equal physical dimension (unit) e.g. uV, mV, etc.
The tool creates a copy of the file and corrects the resolution for the selected signals.
It's advised to set the value for the new physical maximum equal to or just above the highest
physical maximum/minimum found in the group of selected signals.
Typically, physical maximum is set to 3000uV (3mV) for EEG, 6000uV (6mV) for ECG, in case of EDF (16 bit).
For BDF (24 bit) on the other hand, 300000uV (300mV) is a typical value for ExG.
edfbrowser [--rc-host-port=<portnumber>] [[--stream] datafile.edf [mymontage.mtg]]
It is possible to start the program from the commandline:
edfbrowser
this will start the program.
or
edfbrowser myfile.edf
this will start the program with the file myfile.edf opened.
or
edfbrowser myfile.edf mymontage.mtg
this will start the program with the file myfile.edf opened and using the montage mymontage.mtg.
or
edfbrowser --stream myfile.edf mymontage.mtg
this will start the program with the streaming file myfile.edf opened and using the montage mymontage.mtg.
EDFbrowser can be controlled by another program using a direct TCP connection (aka Telnet) using a loopback port.
Start EDFbrowser with the option --rc-host-port=<portnumber>.
Portnumber is the TCP port and must be free and between 1024 and 65535.
(You can use the command ss -lntu to check for ports already occupied on your system, use a port number that is not in that list.)
The following rules apply:
- Max one connection at a time.
- Every command/query must end with a newline character.
- The decimal separator (if any) must be a dot.
- Commands are not case sensitive
- Commands may be abbreviated as long as they can be uniquely identified and as long as they contain at least three characters.
After starting EDFbrowser, you can test the control port by using the command telnet 127.0.0.1 <portnumber> (or use Putty)
In case of connection problems, check your firewall settings.
By default, the remote control port will be configured as a loopback port and will not be accessible from your network.
You can make the remote control port publicly accessible by enabling it in Settings -> Options -> Other.
However, this is not recommended and can be a potential security issue.
Command list: parameters in angle brackets are mandatory and must be replaced with actual values
Command | Description |
---|---|
*LIST? | Returns a list of commands |
*IDN? | Identity query, returns the program name and version |
*QUIT | Closes EDFbrowser |
*RST | Closes all open files |
*CLS | Clears the error queue |
*OPC? | Operation complete: returns 1 if it has finished the operation, otherwise it returns 0 |
FILe:OPEn <path> | Opens a file |
FILe:CLOse:ALL | Closes all open files |
FILe:PLAy | Toggles play/pause |
FILe:STOp | Stop playback |
STReam:OPEn <path> | Opens a streaming file |
MONtage:LOAd <path> <file number> | Loads a montage file and applies it to the file "file number", file number indicates which one of the opened files, must be in the range 1 - 32 File numbering order equals the order in which they were opened |
SESsion:LOAd <path> | Loads a session file |
SIGnal:ADD:LABel <label> <file number> | Adds the signal "label" from the file "file number" to the screen, file number indicates which one of the opened files, must be in the range 1 - 32 File numbering order equals the order in which they were opened |
SIGnal:AMPlitude:ALL <units> | Sets the amplitude for all signals (expressed in units/cm), units can be an integer or a real number and must be in the range +1e-7 to +1e6 |
SIGnal:AMPlitude:LABel <label> <units> | Sets the amplitude for signal "label" (expressed in units/cm) e.g. SIGNAL:AMPLITUDE:LABEL F3 100 Units can be an integer or a real number and must be in the range +1e-7 to +1e6 |
SIGnal:AMPlitude:FIT:ALL | Adjusts the amplitude for all signals automatically (fit to pane) |
SIGnal:AMPlitude:FIT:LABel <label> | Adjusts the amplitude for signal "label" automatically (fit to pane) |
SIGnal:OFFset:ADJust:ALL | Adjusts the vertical offset relative to the baseline for all signals automatically |
SIGnal:OFFset:ADJust:LABel <label> | Adjusts the vertical offset relative to the baseline for signal "label" automatically |
SIGnal:OFFset:ZERo:ALL | Sets the vertical offset relative to the baseline for all signals to zero |
SIGnal:OFFset:ZERo:LABel <label> | Sets the vertical offset relative to the baseline for signal "label" to zero |
SIGnal:INVert:ALL <0|1|2> | Display all signals inverted if value is 1, display all signals normally if value is 0, if value is 2: toggle (invert if not inverted, display normal if inverted) |
SIGnal:INVert:LABel <label> <0|1|2> | Display signal "label" inverted if value is 1, display signal label normally if value is 0, if value is 2: toggle (invert if not inverted, display normal if inverted) |
SIGnal:REMove:LABel <label> | Removes signal "label" from the screen |
SIGnal:REMove:ALL | Removes all signals from the screen |
TIMEScale? | Returns the timescale in seconds |
TIMEScale <seconds> | Sets the timescale, seconds can be an integer or a real number and must be in the range +1e-5 to +6048e3 |
TIMEScale:MMSecond? | Returns the timescale in millimeters per second |
TIMEScale:MMSecond <millimeters> | Sets the timescale in millimeters per second, can be an integer or a real number and must be in the range 1e-3 to 1e6 |
VIEwtime? | Returns the viewtime in seconds relative to the start of the recording |
VIEwtime <seconds> | Sets the viewtime relative to the start of the recording, seconds can be an integer or a real number and must be in the range -3e1 to +6048e3 |
TIMELock:MODe? | Returns the Timelock / File synchronisation mode, 0 = offset, 1 = absolute, 2 = no locking, 3 = user defined locking |
TIMELock:MODe <0|1|2|3> | Sets the Timelock / File synchronisation mode, 0 = offset, 1 = absolute, 2 = no locking, 3 = user defined locking |
TIMELock:REFerence? | Returns the file number of the file that acts as the reference, file numbering order equals the order in which they were opened |
TIMELock:REFerence <file number> | Sets the file number of the file that will act as the reference, file number indicates which one of the opened files, must be in the range 1 - 32 File numbering order equals the order in which they were opened |
SYStem:ERRor? | Queries the last entry in the error queue and deletes it, if there are no more errors in the queue, it returns 0 |
SYStem:LOCk? | Returns 1 if the system is locked, otherwise 0 |
SYStem:LOCk <0|1> | If value is 1, the GUI (graphical user interface) will be locked, otherwise it will be unlocked If the GUI is locked, EDFbrowser will ignore all user input (mouse, keyboard) |
Error | Description |
---|---|
0 | There are no errors in the error queue |
101 | Cannot open a second file |
102 | Cannot open a second file |
103 | Attempt to open too many files |
104 | File has an unknown extension |
105 | Cannot open file for reading |
106 | File is not EDF or BDF compliant |
107 | File is discontiguous (EDF+D or BDF+D) |
108 | File has a formatting error |
109 | Cannot open a streaming file when there's already a file open |
201 | Command syntax error |
202 | Unknown command received |
203 | Invalid parameter received |
204 | Missing parameter |
205 | Operation not possible |
206 | Malloc error |
207 | Item not found |
208 | Parameter value out of range |
301 | Busy, try again later, use the command *OPC? to check if EDFbrowser has finished the operation |
Q. How do I add signals to the screen?
A. Use the signalsdialog.
Q. How can I change the amplitude of a signal?
A. Leftclick on the signallabel of that signal. A small dialog will appear. There is a
spinbox that will let you change the amplitude of the signal.
Another way is to rightclick on the signallabel and move the mousepointer up or down
while keeping the right mousebutton pressed. This will increase or decrease the amplitude
of the signal on the screen.
Q. How can I see the value of a signal?
A. Use a crosshair.
Q. I added signals from multiple files but some of them are not visible on the screen.
A. Go to the "Time" menu and set it to "Synchronize start of files".
Then select Time -> Go to start of file.
Now select Amplitude -> Fit to pane.
Q. I want to open two files in different windows.
A. Start the program twice and you can open files in different windows.
Q. When I try to open a file, it says that the file is not EDF(+)/BDF(+) compliant.
What do I do?
A. Try to find out what program created that file. Write the maker/developer of that
program and tell him/her that the program does not produce valid EDF(+)/BDF(+) files.
Ask him/her to fix that program.
If you get the message that a character in the header is not a valid ASCII-character,
use the header editor to repair your file.
Q. How can I combine multiple EDF/BDF files?
A. Use the Print to EDF function.
Q. What is the difference between EDF and BDF?
A. The difference is the resolution. In EDF the maximum resolution is 16 bits, so the ratio
between the smallest and the largest value is 1 to 2^16 or 1:65536.
In BDF the resolution is 24 bits, so the ratio between the smallest and the largest value is
1 to 2^24 or 1:16777216. In other words, the dynamic range of a BDF file is much higher.
BDF is the 24-bits version of EDF, invented because more and more ADC's/acquisition systems use
24-bits analoog to digital converters. When you try to store the data from a 24-bits ADC into EDF,
you will lose information.
Q. When I print a file, the borders are cut away (the first characters of the signallabels are missing).
A. In the printerdialog, choose "landscape" and the right papersize (A4) and
adjust the margins in the printerproperties.
For example, set top and bottom margins to 0.85 cm and the left and right margins to 1.27 cm.
Q. When I print to a PDF file, the traces look grey instead of black when I open the file
in Adobe Reader 8.x.
A. In Adobe Reader, go to Edit->Preferences->Page Display and uncheck "Smooth line art".
Q. Why are the annotationmarkers not visible?
A. Go to the Settings menu. Check "Annotation marker" and make sure that it
has a different color than the background color.
Q. The amplitude settings are not correct. For example, 100 uV/cm is 100 uV/1.5 cm.
A. This can happen when the DPI-settings of the operatingsystem are wrong. For example,
when EDFbrowser cannot read your monitorsettings or when the monitorsettings are wrong.
You can solve this as follows. Go to the settingsmenu and open the Calibration tab and
check the "Manually override automatic DPI settings" checkbox. Now measure the two black
rectangles, enter the values and click on the apply button.
Q. What are the system requirements?
A. There are no special requirements. However, the faster your CPU and the more memory,
the faster you can navigate. For example, a large pagetime (five minutes or more) will make
the responsetime of the program longer, specially when you add a lot of signals and/or
when the signals have a high samplerate.
When you open a file with a size of hundred megabytes and you choose to make the whole
recording visible on the screen, the program will load the whole file (hundred megabytes)
into memory (RAM).
Q. Which operatingsystems are supported?
A. Linux and windows 2000/XP/Vista/7/8/10. The Mac is not supported but some people reported that
they have successfully compiled it from source on a Mac.
Q. Which license is used for the program?
A. The GNU General Public License version 3. For more info go to Help -> About EDFbrowser.
Q. I think I found a bug. Where do I report it?
A. First check if you are using the latest version of EDFbrowser.
If the bug persists, send an email to teuniz@protonmail.com
Q. Why is EDFbrowser so strict about formatting and refuses to open files that are not completely EDF compliant?
A. EDFbrowser has a zero-tolerance policy regarding non-compliant files. This is done in order to make people aware of it and
to push them to contact the producer of the software that created the non-compliant file and file a bugreport there.
It's the opinion of the author of EDFbrowser that problems with non-compliant files should be solved at the producer,
not at the viewer.
In practice, if EDFbrowser should allow minor formatting errors, nobody will file a bugreport because of reasons like:
"Yes, I'll file a bug report later when I have time..."
"Why should I file a bugreport if there's already a solution at hand? Time is money!"
EDFbrowser offers a tool that can repair minor errors in non-compliant files: the header editor.
Simply open the file in the header editor and press "Save". That's all.
In case you have a whole bunch of non-compliant files, you can use this script: https://www.teuniz.net/edf-hdr-repair/
Direct your complains to the producer of your non-compliant files.
Q. Why can't EDFbrowser accept extended-ASCII characters in the EDF header?
A. Officially, there doesn't exist a so-called "extended" ASCII set. The ASCII character set is a 7-bit only character set and
and does not have any characters like for example with accents.
Unofficially, there exists many extended character sets. So, which one should be chosen? And why that particular one and not another one?
Also, there's no method in EDF to indicate which extended ASCII set is in use. So, a character that looks fine when opened using a computer
with a certain country setting (localization setting), can look wrong when opened using a computer with a different country setting.
There's no way to guarantee that the information is correct when using extended ASCII.