continuous or regular analysis of the human body and or biofeedback

[Noah Gift]

One other thing I forgot to mention that I am interested in is
continuous, or regular analysis of the human body. I am about to go
for a 20 mile run today, and it would be interesting to have my heart
rate sent asynchronously to something that would analysis it. Maybe
at mile 10, my "bot" would come back and mention my pace was too easy,
based on the last years worth of data it had to analyze. This is very
doable, especially with things like the iPhone.

About 10 years ago I had the idea for an exercise device that would
learn about the person that used it, and then would continuously adapt
the workout based on the users historical data, their athletic
potential, and data from all over the world.

Another idea, is it seems silly that we can't just monitor our bodies
24/7 for glucose, cholesterol, heart rate, liver function, etc, and
then have a machine let us know that we might die in 2 years, because
it sensed we drank too much Pepsi. The downside of this is that we
would need a tap into the bloodstream, which could cause problems. At
the very least, we could tap into our bloodstream once a week, and
then have that data analyzed.

--
Cheers,

Noah

[Tennessee]

What can be cheaply monitored?

I believe eeg style devices are not prohibitevly expensive /
difficult... blood pressure is also possible. What devices exist for
non-intrusive realtime body monitoring, say in a workplace/office?

-T

[Noah Gift]

On Wed, Apr 22, 2009 at 11:53 PM, Tennessee <tleeuw...@gmail.com> wrote:
>
> What can be cheaply monitored?
>
> I believe eeg style devices are not prohibitevly expensive /
> difficult... blood pressure is also possible. What devices exist for
> non-intrusive realtime body monitoring, say in a workplace/office?

Not entirely sure yet. I am happy to make myself an experiment and
post data for a couple weeks, for example. I do know heart rate
monitors are pretty easy and not invasive, as you just strap them to
your chest. The iPhone also support tethering, which makes it a nice
choice for potentially more invasive devices.

I would be curious about other portable devices like Nokia phones,
Android, etc, as well.

>
> -T
>
> On Apr 12, 9:26 am, Noah Gift <noah.g...@gmail.com> wrote:
>> One other thing I forgot to mention that I am interested in is
>> continuous, or regular analysis of the human body.  I am about to go
>> for a 20 mile run today, and it would be interesting to have my heart
>> rate sent asynchronously to something that would analysis it.  Maybe
>> at mile 10, my "bot" would come back and mention my pace was too easy,
>> based on the last years worth of data it had to analyze.  This is very
>> doable, especially with things like the iPhone.
>>
>> About 10 years ago I had the idea for an exercise device that would
>> learn about the person that used it, and then would continuously adapt
>>    the workout based on the users historical data, their athletic
>> potential, and data from all over the world.
>>
>> Another idea, is it seems silly that we can't just monitor our bodies
>> 24/7 for glucose, cholesterol, heart rate, liver function, etc, and
>> then have a machine let us know that we might die in 2 years, because
>> it sensed we drank too much Pepsi.  The downside of this is that we
>> would need a tap into the bloodstream, which could cause problems.  At
>> the very least, we could tap into our bloodstream once a week, and
>> then have that data analyzed.
>>
>> --
>> Cheers,
>>
>> Noah
> >
>

--
Cheers,

Noah

[Guy K. Kloss]

On Thu, 23 Apr 2009 00:24:43 Noah Gift wrote:
> I do know heart rate
> monitors are pretty easy and not invasive, as you just strap them to
> your chest.

I could imagine that it's not too hard to derive another sensor by adding a
kind of a "stretch sensor" to that strap that then can pick up expansions of
the upper body due to breathing, so the breathing rate could be easily picked
up.

Guy

--
Guy K. Kloss
Institute of Information and Mathematical Sciences
Te Kura Pūtaiao o Mōhiohio me Pāngarau
Room 2.63, Quad Block A Building
Massey University, Auckland, Albany
Private Bag 102 904, North Shore Mail Centre
voice: +64 9 414-0800 ext. 9585 fax: +64 9 441-8181
eMail: G.K...@massey.ac.nz http://iims.massey.ac.nz

[Tennessee Leeuwenburg]

My wife has used such a heart rate monitor, and they're pretty bodgy devices -- they only work intermittently. Basically you need to strap them on, then get to the point where you're sweating like a horse before they take vaguely reliable readings. I mean, noisy data is better than no data, but still...

A stretch sensor is a good idea.

It would be neat if we could come up with a list of low-cost, DIY capable biometric sensors than anyone interested in the idea could put together.

-T

--
--------------------------------------------------
Tennessee Leeuwenburg
http://myownhat.blogspot.com/
"Don't believe everything you think"

[Rick Thomas]

GPS gives you pace, plus a stride counter gives you stride length.

Add a topo map and you get grade, thus work and calories spent.

Probably need to monitor sleep and diet too.

What kind of intelligence are we talking here? Expert trainer to give workout advice? Life coach?

What kind of sources for rules? Explicit training rules, medical statistical studies, personal workout successes...?

How is the knowledge applied? Rule engine, neural net? Does it learn across a bunch of users?

RT

[Eric]

Take a look at the Arduino (http://arduino.cc) platform. The Arduino
itself doesn't use Python (instead, you write "sketches" for the board in
the C-like Processing language), but it's very easy to communicate with the
boards using the pySerial module. You can buy or hack many different types
of sensors, and I've come across at least one project that connects an
Arduino board and an iPhone (the one that comes to mind is an
iPhone-controlled robot).

Eric

[Noah Gift]

On Thu, Apr 23, 2009 at 11:46 AM, Rick Thomas <even...@gmail.com> wrote:
>
> GPS gives you pace, plus a stride counter gives you stride length.
> Add a topo map and you get grade, thus work and calories spent.
> Probably need to monitor sleep and diet too.
>
> What kind of intelligence are we talking here? Expert trainer to give
> workout advice?

This is something I am interested in. There is obvious a commercial
market for something like this as well.

Life coach?

This is also a good idea. I wonder how much different we would act
with real time access to information such as this. I think caloric
intake and and exercise are two particularly interesting problem that
the world is searching for a solution for. Maybe the answer is real
time, continuous monitoring...

> What kind of sources for rules?
Explicit training rules,

I think the rules are fairly straightforward and understood for
developing different forms of fitness. Raw cardiovascular performance
has rules, overloading the anaerobic system has rules, and so does
muscle growth.

medical statistical
> studies,

Some of these rules could put into place as well.

personal workout successes...?

This would need to be a component because everyone has specific
genetic potential. If you ever run track in college, it becomes
pretty obvious, but a clever machine could figure out a way to
constantly adjust your load to challenge you. This is one of the core
problems in exercise training. In order to get really fit, it is
actually really simple, you start off slow at some form of training,
and then you gradually adapt the load once your body adapts. Humans
happen to be pretty horrible at figuring this out themselves without a
guide.

There are many stories of Native Americans, take Apache warriors, who
could run 70-100 miles in a stretch. They had good reasons to do it,
so their bodies adapted. Currently we adapt our bodies at sitting all
day :)

> How is the knowledge applied?

I think the simplest form would be to have small iterations, say
weekly, in which your exercise is analyzed. This data would go into
the system would provide feedback either in real time, or before a
workout, that would advise on the best course of action for that
weekly goal. For example, you might have only run 30 K total for that
week, and it could calculate you should do 30K run in one shot the
next week.

Rule engine, neural net?

I don't know enough yet to answer.

Does it learn across
> a bunch of users?

It seems like tapping into this would be useful.

> RT
>
> My wife has used such a heart rate monitor, and they're pretty bodgy devices
> -- they only work intermittently. Basically you need to strap them on, then
> get to the point where you're sweating like a horse before they take vaguely
> reliable readings. I mean, noisy data is better than no data, but still...
>
> A stretch sensor is a good idea.
>
> It would be neat if we could come up with a list of low-cost, DIY capable
> biometric sensors than anyone interested in the idea could put together.
>
> -T
>
> On Thu, Apr 23, 2009 at 7:37 AM, Guy K. Kloss <g.k...@massey.ac.nz> wrote:
>>
>> On Thu, 23 Apr 2009 00:24:43 Noah Gift wrote:
>> > I do know heart rate
>> > monitors are pretty easy and not invasive, as you just strap them to
>> > your chest.
>>
>> I could imagine that it's not too hard to derive another sensor by adding
>> a
>> kind of a "stretch sensor" to that strap that then can pick up expansions
>> of
>> the upper body due to breathing, so the breathing rate could be easily
>> picked
>> up.
>>
>> Guy
>>
>
> >
>

--
Cheers,

Noah

[Noah Gift]

On Thu, Apr 23, 2009 at 11:45 AM, Eric <er...@noticings.com> wrote:
>
>
> Take a look at the Arduino (http://arduino.cc) platform.  The Arduino
> itself doesn't use Python (instead, you write "sketches" for the board in
> the C-like Processing language), but it's very easy to communicate with the
> boards using the pySerial module.  You can buy or hack many different types
> of sensors, and I've come across at least one project that connects an
> Arduino board and an iPhone (the one that comes to mind is an
> iPhone-controlled robot).

That does look pretty nifty.

>
> Eric
>
>
> On Thu, 23 Apr 2009 00:24:43 +1200, Noah Gift <noah...@gmail.com> wrote:
>>
>> On Wed, Apr 22, 2009 at 11:53 PM, Tennessee <tleeuw...@gmail.com>
>> wrote:
>>>
>>> What can be cheaply monitored?
>>>
>>> I believe eeg style devices are not prohibitevly expensive /
>>> difficult... blood pressure is also possible. What devices exist for
>>> non-intrusive realtime body monitoring, say in a workplace/office?
>>
>> Not entirely sure yet.  I am happy to make myself an experiment and
>> post data for a couple weeks, for example.  I do know heart rate
>> monitors are pretty easy and not invasive, as you just strap them to
>> your chest. The iPhone also support tethering, which makes it a nice
>> choice for potentially more invasive devices.
>>
>> I would be curious about other portable devices like Nokia phones,
>> Android, etc, as well.
>
>
> >
>

--
Cheers,

Noah

[Noah Gift]

Ivan Krstic posted about "Mind Hacks":
http://radian.org/notebook/mind-hacks It seems like this could also
fit into a continuous monitoring stack where you could have chemicals
injected into your bloodstream based on responses you have programmed,
although this is a bit more scifi and invasive...

[Graham Higgins]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 26 Apr 2009, at 14:23, Noah Gift wrote:

>
> Ivan Krstic posted about "Mind Hacks":

> although this is a bit more scifi

Frankly, I question whether this approach will ever enter the realm of
practicality. There is a meta-level representation issue to be solved,
not unlike the one that scuppered David Li's Gaussian copula function
for default [1]:

"The development of the model had, ironically, changed the nature of
the reality it was modelling."

I strongly encourage you to read the accompanying PDF that Ivan made
available. [2].

Cheers,

Graham

[1] http://www.ft.com/cms/s/2/912d85e8-2d75-11de-9eba-00144feabdc0.html
[2] http://radian.org/notebook/wp-content/uploads/2009/04/brain-botox.pdf

-----BEGIN PGP SIGNATURE-----

iEUEARECAAYFAkn0ZM0ACgkQOsmLt1NhivzUVQCeI2uZ2Ql/BQClJXX3Oj968DBW
t9AAmM+HOXvQH6n3WbMNnqZ7IZtw+SOJAJUCBQFJ9GTNWetZVntpcPUBAlteBADJ
96RoBpV2DLeOZBq8u18NA/FDOQqbINsyF/cHQyM9T635wLgGW5p+HNpabXqEfsSY
U2RQ/b/MIKmUl2QWUjQY+JDrGsstM3cRggwQAQvUiS5Yz+UFeSi1z6M6qI8LY/KR
7LqMksjavCTlN94HZMpKub04H83IYdW16ehuOSaEJg==
=hTKj
-----END PGP SIGNATURE-----

[noah.gift]

On Apr 27, 1:42 am, Graham Higgins <gjhigg...@googlemail.com> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On 26 Apr 2009, at 14:23, Noah Gift wrote:
>
>
>
> > Ivan Krstic posted about "Mind Hacks":
> > although this is a bit more scifi
>
> Frankly, I question whether this approach will ever enter the realm of  
> practicality. There is a meta-level representation issue to be solved,  
> not unlike the one that scuppered David Li's Gaussian copula function  
> for default [1]:
>
> "The development of the model had, ironically, changed the nature of  
> the reality it was modelling."

I agree that I am also skeptical of there ever being a "free lunch"
with chemicals.