Technical Details
Cardiorespiratory Recording
Data for respiration and electrocardiogram can be recorded with a LabQuest device from Vernier (www.vernier.com). Software for graphing, interpreting and converting data is available for free (Logger Lite); a more full functioned version, Logger Pro 3 can be purchased. I have been using version 3.8.7.
After saving a recording in the native format, .cmbl, I open the file in Logger Pro 3 and export it as a text file. The text file can be opened by my program Cardiorespiratory Viewer, or alternatively converted to another format which can be opened by my Physiology Viewer.
Using a program I wrote in Python called Cardio_generate_h5.py I open the text file and save the data in two files in the .h5 format, for example rec23.breath.h5 and rec23.cardio.h5 , which correspond to recording #23. These files are used in turn by my Physiology Viewer program which can display graphs of breath and heart signals alongside EEG signals.
EEG Recording
The Muse SDK from InteraXon is available at www.choosemuse.com/developer-kit/. This free download Includes muse-io.exe, muse-player.exe and MuseLab.exe, essential for recording brainwave data on Macintosh or Windows computers.
The InteraXon Developer Forum is at forum.choosemuse.com/forum/developer-forum
To make a recording of brainwaves, be sure the Muse headband is fully charged (via USB port on computer). Make sure that Muse headband has been turned on is paired with a computer via Bluetooth.
My laptop computer is a Samsung NP900X3C, which came originally with Windows 7 installed. Unfortunately, I was unable to establish a Bluetooth connection between my computer and the Muse headband, even after considerable troubleshooting by Microsoft. After I upgraded to Windows 8.1, the Bluetooth connection worked just fine. However, after upgrading to Windows 10, Bluetooth no longer functioned, so I had to revert back to Windows 8.1.
I have also been able to use my Dell workstation (Windows 7) equipped with a Kinivo Bluetooth Adapter in a USB port to successfully link up with my Muse headband. Bluetooth on this computer continues to work after my upgrade to Windows 10.
At the start of a recording session, open a Command window and run muse-io with the following parameters:
muse-io –osc osc.udp://localhost:5001,osc.udp://localhost:5002, –osc-timestamp
(I keep this command line in a notepad document, copy it to the clipboard with Ctrl-a, Ctrl-c. I then start a command window, right-click and Paste into the command window.)
Press enter and look for “Connected to paired device: “Muse”: open!
Bits/second: 8380 receiving at: 220.00Hz
Next, while muse-io is running, start the program Muse Lab.
Under OSC, Incoming, select UDP and enter 5001 in Opened Ports, and click on Open Port.
Under Incoming Messages, you should see a list of Address Patterns.
The horseshoe graphic in the upper right hand corner should show five solid-colored patches. If it does, then you have good contact with all four sensors of the Muse. If one of the patches is not filled it, then adjust the contacts on the headband, wait a few seconds and try again. Make sure there is no hair between the contact and the skin at the sensor locations.
To do a recording, select Recording from the dropdown list box. It is not necessary to display graphs when doing a recording—that’s optional. I save all recordings in a top-level directory, C:\MuseRec.
Under Location: be sure to change from the default location, C:Program Files (x86)\Muse\recordings to your own directory, e.g., C:\MuseRec. Otherwise recordings will not be saved.
Under Filename:, give your file a meaningful name. I use recN.muse, where N is an integer that increments for each recording.
Click Record to begin the recording. I find that the size of the .muse file for a 25-minute recording is about 130 MB.
If, alternatively, you want to view real-time EEG graphs, select Visualizers from the dropdown list box. Click for example, New… Scrolling Line Graph. You should see a list of Signals under Scrolling Line Graph Options and also a list of Signal Group Options below. Try clicking on of these, such as /muse/elements/alpha_relative. This will result in four lines appearing to the right, one for each of the four Muse sensors. With the default scaling of the y axes, you won’t see any curves, but if you go to Settings and under Amplitude, change the Max Amplitude to 2.0 and the Min Amplitude to -2.0, you should start to see scrolling curves. Try closing your eyes and relaxing for several seconds. You will probably see your alpha values go up.
After the recording, I do a conversion from the .muse format to a .CSV file format. To do this, I run the Muse Player in a command window with the parameters,
Muse-player -f C:\MuseRec\rec1.muse -C C:\MuseRec\rec1.CSV
where rec1.muse and rec1.CSV are replaced with the file names I want to convert.
The resulting .CSV files can be opened by a program I wrote called EEG_generate_h5.py which converts the data into files in the .h5 format, for example rec23_eeg_raw.h5 (raw EEG signal), rec23_eeg_rel.h5 (relative power values for delta, theta, alpha, beta and gamma frequency bands), rec23_eeg_abs.h5 (absolute power values as base 10 logarithms for delta, theta, alpha, beta and gamma bands) and rec23_eeg_user.h5. (jaw clench, eye blink, concentration and mellow). The resulting files are used by Phys_Viewer.py, the Physiology Viewer to display various graphs and charts of brainwaves and associated data.
The programs Phys_Viewer.py, EEG_generate_h5.py and Cardio_generate_h5 were written in Python 3.4 using the Anaconda Spyder 2.3.1 development environment from Continuum Analytics.
Note: Source Code availability
Source code for the Physiology Viewer and associated accessory programs are now available at,
https://github.com/davidtro/physiology-viewer/
Next page: Research Literature