average brain frequency plotted across time

rharding

I want the software to calculate an "average brain frequency".  The method of calculating the "average brain frequency" is not critical.

I want to see a plot of the "average brain frequency" over time (at least 8 hours).

How do I accomplish this?

Please advise me as to what I should purchase.

Is the Ultracortex "Mark IV" EEG the best choice for headset?

The Ultracortex comes in 8-channel and 16 channel versions.  Would I then need either the Cyton Biosensing Board (8-channels) or the Cyton + Daisy Biosensing Boards (16-Channels)?

I presume I need a Rechargeable Battery Pack and a charger (from Adafruit or Sparkfun).

Thank you.

wjcroft

RHarding, hi.

Rather than reducing all the brain activity down to a single number, that changes over time -- instead the spectrogram is used. Which plots time on the horizontal axis and frequency on the vertical axis. And then there would be one of these charts for every scalp 10-20 site electrode position in use.

http://openbci.com/forum/index.php?p=/search&Search=spectrogram

https://www.google.com/search?q=eeg+spectrogram

There is signal processing metric called the spectral centroid, if indeed you just want a single number (per channel), plotted over time. 

https://www.google.com/search?q=spectral+centroid

You can do these measurements with any of the OpenBCI boards (Cyton or Ganglion), with any number of channels. 

For long term recordings, you may want to use a similar type of setup that they use in sleep labs: gold cup electrodes with paste, covered with some type of moisture impermeable barrier that keeps the paste moist over hours. Sleep labs using something (nasty) called colloidion. 

The spikey 'comb' electrodes of the Ultracortex may cause some discomfort if worn for many hours. There are newer softer combs available from FRI.

William

rharding

Thank you sincerely for your input.

My understanding is that I can get a graph where:

--the X-axis is time (covering approximately 8 hours)

--the Y-axis is frequency

--with the spectral centroid, I would see 8 (or 16) different lines running horizontally across the graph.

Is that correct?

I appears that the only equipment I would need is that which I listed in my original post:

--Ultracortex "Mark IV" EEG (8-channel and 16-channel)

--either the Cyton Biosensing Board (8-channels) or the Cyton + Daisy Biosensing Boards (16-Channels)

--Rechargeable Battery Pack and a charger (from Adafruit or Sparkfun)

Is that true?  Do you have a better suggestion?

I do not want electrodes with paste (so I can easily take the headset off and on).

I will try to get the softer spikey combs from FRI.

Thank you.

wjcroft

re: graph. Did you look at the EEG spectrogram examples on image portion of the Google search results? Spectrogram gives MANY times the information of a single value over time. Many of the examples show THREE axes: horizontal is time, vertical is frequency and color (typically cool to warm to hot colors) represents the intensity of the frequency component at that specific frequency x,y coordinate.

So if you had FFT bins that were 1 hz in width, from 1 to 40 hz, you might have 40 different colors per time unit. Cyton sample rate is 250 samples per second. But spectrogram / FFT is usually computed on a courser basis, say 20 or so times per second.

re: spectral centroid. No, spectral centroid will just give you one value per time unit. Which represents a frequency histogram, where it would balance if cut out from paper and suspended with a pin. The pin location is the centroid value. So a spectrum with more slow waves, say 4 to 12 hz, might have a centroid of say 7.x hz. 

re: the Cyton, either 8 or 16 channels would work fine. As would the Ganglion with 4 channels.

William

rharding

Thank you for taking the time to try to educate me.

I did look at Google/image/eeg-spectrogram.  That was not what I had in mind.

I did look at Google/image/spectral-centroid.  Most of the graphs showed frequency on the X-axis and amplitude/magnitude on the Y-axis.

However, many showed time (usually less than 30 seconds) on the X-axis and spectral-centroid on the Y-axis.  That is what I was imaging in my most recent post.

I have had my EEG analyzed by a local EEG specialist.  I am going to call him to see if what I want makes any sense and if so, how it might be accomplished.

I am signing off until I can come back with a better understanding of how I can accomplish what I want to accomplish.

Thank  you again. 

wjcroft

Did your local EEG person put a 19 channel cap on your head, and then do a QEEG analysis?

https://www.google.com/search?q=eeg+qeeg

QEEG samples data from typically two states: eyes closed and eyes open. Each is a separate map. At each 10-20 site, your own amplitude, coherence and phase, is compared to a "normative database", a collection of previously sampled 'normals'. Of the same age and gender.

The map then shows for each site, colors indicating how many standard deviations, your values differ from those of the normative database.

http://scottsdaleneurofeedback.com/services/qeeg-brain-mapping/

rharding

This is not going to be a very informative response.

Yes, as I recall, it was a 19 channel cap (it was in 2009).

I have three reports (which I do not understand).

I have an 18 page LORETA report.

I have a 23 page qEEG report.

I have a 264 page report that is identified as a NG report.  Its images are similar to what you sent.

I can not be more informative.

Since my last post, I sent an email to the local EEG specialist that I referred to.

wjcroft

NG= Neuroguide. A common QEEG processing program. LORETA is a 3D version of the QEEG mapping process, Whereas most QEEG is just done on the surface.

Unfortunately, the packages used to produce QEEG reports and such are only sold to licensed clinical practitioners. And cost multi-thousands of dollars also.

The 8 hour or longer data that you want to gather, while may be somewhat informative -- will be nothing like the information and diagnostic capability in these clinical packages.

rharding

Here is where it gets interesting.

I am experiencing the very earliest deficiencies in very short-term memory.

In December, 2016, an article came out of MIT regarding Alzheimer's.  Although the MIT study involved rats, it has created a huge aftermath (given the brief time period) in the world of human Alzheimer's.  Many people, including the original MIT researchers, are trying to replicate (and monetize) the results of the rat study in human subjects.

The basic idea is that if you stimulate any part of the brain at 40Hz, that frequency will tend to propagate throughout the brain (in particular, to the hippocampus).  That 40Hz frequency will cause the microglia to destroy beta amyloid plaque (a central player in Alzheimer's and other brain pathologies).  Before you dismiss this as patent nonsense, read some of the documents linked to below.

Earlier this week, I was discussing this with a couple of physician friends who asked me to send them some information on the subject.  I put together an email which I sent to each of the parties.  The emails had several pdf's attached.  

Because I do not see how to attach files to this Comment, I replaced the pdf's with internet links.  I have appended the pdf-free email to the end of this Comment.

The 40Hz stimulation can be visual, aural, tactile, or via transcranial electromagnetic radiation (primarily red visual light).  I plan on using the latter approach (i.e., red LEDs mounted in the Ultracortex "Mark IV" EEG Headset)(I have developed a power unit that will drive the LEDs with a square wave of any frequency from 4Hz to 50Hz) .  My only purpose in monitoring my brain frequencies is to monitor the impact the LEDs are having.  When I turn on the LEDs at 40Hz, I should see my overall brain frequency move towards 40Hz.

Here is the article.

https://www.ncbi.nlm.nih.gov/pubmed/27929004


These are summaries of the article.

https://www.ncbi.nlm.nih.gov/pubmed/27929001

https://www.ncbi.nlm.nih.gov/pubmed/29493598

https://www.seeker.com/led-light-therapy-could-radically-change-the-treatment-of-brain-diseas-2294712527.html

https://www.the-scientist.com/?articles.view/articleNo/47682/title/Toward-Treating-Alzheimer-s-Disease-with-Brain-Waves/


Here is an article that came out slightly earlier.  See section 6.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5066074/


The lab that produced the original study has produced some materials aimed at the general populace.

https://www.youtube.com/watch?v=O_p4QWkE2Ls&feature=youtu.be -- If you watch this, I would suggest that you set Speed to 1.5 or 2 and set Subtitles/CC to English.


This is their commercial offshoot:

https://www.cognitotx.com/


Toronto newscast -- watch this from 1:20 to 2:00

https://www.youtube.com/watch?v=HNZTtFTJVxQ&t=81s


The product that the newscast was discussing is described here (I have one).  Look at the Vielight Neuro Gamma (not the Alpha or the 810).

https://vielight.com/neuro-alpha-gamma/


The dark side of medical advancement.

https://www.salon.com/2017/12/13/brain-wave-treatment-for-alzheimers-is-promising-but-the-first-human-subject-is-left-behind_partner/


Here is a sound only study.

https://www.ctvnews.ca/health/the-sound-of-healing-study-says-sound-stimulation-could-help-alzheimer-s-patients-1.2868393

wjcroft

RHarding, hi.

Thanks for those links. I had seen some of the 40 hz articles and research before. And spoke with the Vielight CEO at the Transformative Technology Conference. Was not aware that they added several models to their lineup. 

Cool that you have built your own stimulator, incorporated into the headset.

Note that gamma is sometimes tricky to pickup from the EEG. Because the signal level is so weak, generally below 5 microvolts. And compounding that, the muscles around the head / scalp can also produce gamma-like frequencies via their EMG activity.

What generally gives a stronger effect with gamma, is to add two or more channels together in your signal processing software. Then if the signals are in phase sync, they will additively combine to make a stronger signal. Canceling out any noise components that don't additively combine.

Here is a company making a cheap gamma light stimulator,

https://www.gammalighttherapy.com/

----

Regarding medical research in this area in general, Dr. Mark Hyman put together an excellent documentary series, called Broken Brain. It has been viewable online a couple times already. And will reappear. I have the video mp4 set.

https://brokenbrain.com/trailer/

Hyman is a Functional Medicine practitioner. And much is being discovered as far as the precursors to dementia. And how these relate to inflammatory conditions inside the body. In other words, what you eat, your immune system, your stress level, etc. can all cause excess inflammatory activity that generates some of the precursors. While the 40 hz stimulation approaches have much promise, it's also wise to look at the holistic picture and make changes in life quality / dimensions.

https://www.google.com/search?q=dementia+inflammation

Regards,

William

wjcroft

Hyman did a book a couple years ago, with many of key points covered in the Broken Brain documentary series,

https://www.amazon.com/UltraMind-Solution-Broken-Brain-Healing-ebook/dp/B001NLKU7S/

rharding

Wlliam,

As I said in my previous post, "My only purpose in monitoring my brain frequencies is to monitor the impact the LEDs are having."

If I have the 16-channel Ultracortex "Mark IV" EEG Headset, what is the closest I can come to having a computer screen with 12 hours on the X-axis and having my average brain frequency on the Y-axis.  If I can not have a single plot showing my average brain frequency vs. time, can I have a series of lines or colors from which I could visually estimate and average brain frequency?

I saw your comment about combining two gamma channels to get a more accurate reading.  However, I do not know what that translates to in terms of either hardware positioning or software analysis.

Any comment regarding how to achieve my goal would be greatly appreciated.

Thanks,

Bob

P.S.  Although I imagine having the EEG system functioning whenever I am wearing the Headset, given the other studies that have been done, I expect to have the LEDs functioning only part of the time.  The graphic output from the EEG system would presumably reflect when the LEDs were functioning and when they were not functioning.

wjcroft

Bob, hi.

re: "average brain frequency".  When you asked about this before, I did not know the context. So I suggested spectrogram and spectral centroid. Using those alone on channels connected to various scalp sites, will very likely NOT show any gamma. Because gamma is such a tiny signal compared to the rest of the brainwaves. And gamma is the lowest amplitude brainwave band.

re: wearing the Ultracortex for hours at a time. Neither headsets nor sleep lab EEG type hookups, work very well when there is any motion in the head or body. You have to be relatively still. And as I think I said before, the headset would not be comfortable to wear for many hours. Only possibly and hour or so at a time. Then your head will need a break.

As far as adding EEG channels or various signal processing tasks, this could be done in BrainBay. Here is a tutorial,

https://sites.google.com/site/biofeedbackpages/brainbay-openbci

Regards, William

Billh

I'd be concerned about 40 Hz signal leakage from the LED wiring to the EEG wires. You may need to compare the EEG signal with and without something like black electrical tape over the LED's.

wjcroft

@Billh, good point. I know some of the Vielight equipment IS being used with simultaneous EEG. So some technique may exist. A sine wave modulation of the LED light intensity at 40hz would have less potential for creating interference with the EEG. Than using a square wave, which produces sharp transients and radiates EMF.

Some of the experiments with the dementia and 40 hz light are just going through the eyes, so could be at suitable distance.

rharding

Good point!

rharding

One of the links in the long list of links that I sent a few posts ago was https://www.youtube.com/watch?v=HNZTtFTJVxQ&t=81s (watch from 1:20 to 2:00).

It shows the VieLight sitting over an EEG cap.  It also says that the subjects wore the devices 15 hours per day.

As regards my not being able to wear the cap for long periods and not being able to move when I am wearing the cap, I will simply have to adjust my plans as necessary.  If I only wore it for one hour and had to be perfectly still, that would still allow me to see how my brain was reacting to the LED stimulation (which is the objective).

Regarding sine wave vs. square wave, LEDs (1) are either ON or OFF and (2) they change states virtually instantaneously.  They are not suited to produce a sine wave.

In what I believe was the first human study (see my link https://www.salon.com/2017/12/13/brain-wave-treatment-for-alzheimers-is-promising-but-the-first-human-subject-is-left-behind_partner/), the subjects wore goggles (i.e., their entire visual field cycled at 40Hz), headphones (i.e., their entire aural input cycled at 40Hz), and vibrating gloves (all three stimuli were in synch at 40Hz).

Compare to that, wearing an Ultracortex doesn't seem so bad.

I could test for interference (as you suggested), however, I would like to assume that I could shield both the EEG leads and the LED leads well enough that interference was minimal.

rharding

I installed Brainbay v2.0 and read the two examples.  I am sure that with your experience, this all seem quite elementary.  However, it is greek to me.

The only way I am going to be able to do this is if some kind (or well compensated) soul tells me step by step what to do and how to do it such that I can best achieve my goal.

wjcroft

re: 15 hours a day. They did not say specifically, but my guess this is referring to the Vielight nose unit, or possibly the Vielight headcap. But NOT the EEG cap shown. If you've had any experience wearing an EEG elastic cap or whole head headset, you'd realize that this amount of pressure on the head can only be sustained for limited periods of time and not for hours. It's just too much pressure with the existing systems.

re: LEDs and light modulation. No, LED light output can be actually amplitude modulated by the applied voltage. It is true many LED drivers just use pure pulse width modulation square waves; but the LED diode is capable of operating at various forward currents, not just on or off.

re: googles, gloves, headphones. These do have less potential for messing with EEG.

re: shielding. Not sure. square wave transients generate wide band noise.

re: BrainBay tutorial. The text in the tutorial explains how the signal processing blocks are connected together. In your case you could sum a range of channels, show a spectrogram, etc.

re: "does it work?"  Umm, I'm not trying to dissuade you from setting up your EEG experiment. But in general, these research studies assume there is some entrainment and brain driving going on. I doubt that the EEG equipment you see is being used to merely track whether the brain is following the 40 hz. They already know that is happening. The EEG is likely for doing before / after assessments using QEEG and such. To see the long term before / after effects of the therapy. The goal is not to just generate gamma, but to help heal the dementia, and the EEG could help track that progress over time. In your case with your setup, you would just be confirming that 40 hz is being received. NOT that it is having a specific effect on the longterm brainwave patterns influenced by the treatment.

Regards,

wjcroft

AM vs. PWM LED modulation,

https://engineering.purdue.edu/~ece495/Power_Electronics_Lab/LED_Basics.pdf

rharding

All I can say is that from my reading, the objective is to simply(?):

--stimulate some part of brain at 40Hz

--have that frequency propagate throughout the brain

--have the microglia magically (exact basis of causation not yet determined) become more numerous, become larger, have more processes, seek out beta amyloid plaque more aggressively, and attack beta amyloid plaque more aggressively

Reduction in beta amyloid plaque is the end game.

Significant decreases in beta amyloid plaque can occur in a matter of weeks.  However, it is also true that if the 40Hz stimulation of the brain ceases, the beta amyloid plaque starts to return in a matter of weeks.

    

wjcroft

Would it make sense to try some self stimulation experiments first? Using either your home built unit or ones mentioned in this thead (Vielight or Gamma-Light-Therapy). And see if you sense any changes in your situation? Measuring the entrained EEG is kind of a binary thing: yes or no, gamma is seen. This EEG evidence is not going to track changes over time in your brain performance.