Hi all

I want a EROS (Event Related Optical Signal) system for brain.
I found Imagent (http://www.iss.com/Products/imagent.html), but it's
too expensive (US 110,000).

So i wanna build a cheaper EROS system by myself. What components
should i use ? My optimum budget is US$250, maximum budget is US$750.

Thanks very much for help.

Since you are willing to dosh out $30, this is a serious question.
Simple answer, unless you are a profound expert in electronics,
programming and reasonably expert in physiology, you won?t be able to
do it. The machines are created by design groups of diverse and
disparate expertise. There are good reasons why they are so expensive,
and development costs would be extreme.

There?s also the ethical code to consider. Medical examination and
experiments, especially invasive, on humans and animals are strictly
controlled. The penalties are quite severe -- kill joys.

But if you want to indulge in slightly simpler, purely passive brain
examination, you could buy or design an alpha (delta etc) wave
detector. Alternatively, an electroencephalogram (EEG). You might even
find designs on the web, although most electronic circuits published
by self professed ?electronic experts? are redolent with ignorance.

You ask the question so you?re not an electronics designer? If this be
so, I?m afraid that you will experience frustration and difficulty
with both an alpha wave detector and EEG. Don?t let me put you off
though; the world needs more decent analogue engineers. Wishing you
good experimenting.

Best

I'm going to guess if you can't spell it, you probably shouldn't build it.

To guzzi

Thanks very much for your reply and comment.

Actually i wanna build a ciruit for brain to control Walkman speaker
volume (by the busy level of logical thinking, more logical thinking,
smaller volume, vice versa) or other brain computer interface
applications.

My university supervisor also suggested EEG b4, but i refused coz it
needs gel and so not user-friendly for normal users.

So i changed the approach to use near-infrared red (NIR)
http://www.acad.polyu.edu.hk/~02471629d/nir.pdf to capture the status
of brain.

It has features of noninvasive, no interferece with EMG, ECG, and the
most important, no need gel.

I think there is a chance to lower the price coz it seems that the
main component of the machine (700/880nm LED, modal 5210, Hamamatsu
C5460-01) is lower than US110,000, I can find modal 5210 and 700/880nm
LED lower than $US625, Hamamatsu C5460-01 is unknown coz its company
don't reply our request :(, but it should be lower than $US 109,375.
right ?

The other reason i think it is possible to build a cheap EROS coz i am
concentrating on slow EROS (detect brain oxygen supply) rather than
fast EROS (detect scattering properties of neurons) which is done by
Imagent.

I am not expert in electronic coz my degree is concentrating on
Bioengineering rather than EE, my supervisor is expert in EEG and
biosignal but he never tounch NIR EROS b4 and can't give too much help
to me. That's why i tried Google Answer and hoped that there are some
smart Researcher who are familiar in electronic components can give
some help in it and provide a list of components which can meet
optimal budget of US$250 or maximum budget of US$750.

Any researchers can give help ?

Thank very much for help. I am willing to pay up to US$40 if anyone
can really suggest a component list at budget around US$250. Size of
components is not important, coz we just need to build a prototype to
show the workable of the idea.

Thanks.

--------------------------------

To quibbs0, 

i think u get some misunderstandings. EROS stands for "Evoked-Related
Optical Signals" but it can also be "Event-Related Optical Signals",
they are the same. Feel free to go to Google Scholar and take a look.
Thanks.

Yea, that?s reasonable. Please excuse my previous caution. I used to
know a bike breaker (the clunky, not the owner) who wanted to build
cheap kidney dialysis machines. Think he intended using old oil
filters. I?d assumed you wanted to build a comprehensive 3D machine
with lots of sensors but I guess if you can find the ?hot spot? it
might just be possible on the cheap. My experience (on the electronics
side) was haemoglobin and plasma absorption characteristics but we
used a red laser. Can?t remember if it was detecting oxygen uptake. We
also did cuvette sampling with RGB LEDs so maybe there is a cheap
venue. Certainly a little more practical for cheap research than
neurone activity :-)

Couldn?t get your link to download -- but finally. So haven?t studied
it fully ?cos it?s 3am and I?m knackered. Could you send any links you
have for the photo devices and I?ll see if I can suggest something
cheaper.

Found the link below. Shows a circuit. Pretty dire I?m afraid, but
electronic engineers do awful biology.

http://www.opticsexpress.org/view_file.cfm?doc=%23(L%2F)%0A&id=%23)%3CK%26%0A

You?ve probably come across this one.
http://www.ebire.org/hcnlab/papers/Rinne_NeuroImage1999.pdf

Best

Hi ?Ding? -- you?re in Hong Kong?

My brain was in error, we used an IR laser not red for haemoglobin.
That?ll make more sense.

Anyway, the problem with APDs (avalanche pin diodes) is that I don?t
recall ever seeing them available to the experimenter. Usually they
are directly incorporated into units such as the Hamamatso. You might
get away with a photodiode (preferably PIN) in a transimpedence
amplifier if it was designed properly with a synchronous detector. At
least you don?t need much speed so it would be pretty easy to design
-- for an electronic designer of experience. On that link I sent
you?ll see that a much cheaper detector was used so all you can do is
try.

As for the LEDs, 880nm are a dime a dozen. For both LEDs you should
take the beam angle into account, and whether or not they are
diffused. The 700nm (deep red) is a bit of a problem though because
these were the first to be made on account of the technology 35 years
ago. Not very eye efficient, so have been superseded by slightly
longer wavelength. I have catalogues which list the occasional 700,
some from Kingbright but they don?t appear on the parametric search
page which jumps from 660 to 850.

http://www.us.kingbright.com/productsearch.asp?CategoryAdvance_name=Through-hole%20LED

However you could download the catalogue (10 megs) which is probably
sensible ?cos the site is dumb.

http://www.us.kingbright.com/Products/eCatalog.asp

Ideally with a big budget you would use lasers. LED?s will be OK I
guess, especially if you collimate the light to localise. Make up an
arrangement with surface silvered mirrors and lenses to combine the
beams perhaps. Bear in mind that there is slight frequency shift with
heating though so let things stabilise before you test.

Another way is with an incandescent light source and two detectors,
each with a narrow window filter but unless you have a convenient
optics lab this?ll be expensive. Rather than narrow band you could use
a low pass and high pass if you?ve got access to sharp ones.

Any help so far? I could do more LED hunting if you want. Just to
sicken you, it would cost me nothing to build the machine from my
junk. I'm thinking of building a hanger! Whether or not it would work
to theory is of course a different matter :-)

Best

Hi, 

Why don't you check out this:

http://www.opticsexpress.org/abstract.cfm?URI=OPEX-13-12-4525

Best,
Gerwin Schalk