PDA

View Full Version : Tri Coder X-Prize contest win 10 million thats alot of Propeller chips.



jdolecki
01-13-2012, 10:16 AM
Build a Tricorder like the one in Star Trek and win money

http://www.qualcommtricorderxprize.org/

prof_braino
01-16-2012, 02:51 PM
Build a Tricorder like the one in Star Trek and win money

http://www.qualcommtricorderxprize.org/

I want to start a team to work on this. Anybody else interested? Reply or PM me. The odds of winning this competition are a millions times greater than winning a state lottery.

ElectricAye
01-16-2012, 09:38 PM
I want to start a team to work on this....

Did you have a particular disease or technology in mind?

You might want to contact Tymkrs http://forums.parallax.com/member.php?65253-Tymkrs. I think she's a cardiac person and she and her buds do some wetware hacking, etc. I don't get the impression there are very many other biology types on this forum. Doggiedoc, maybe?

I'm always impressed at the number of people who go after X-prizes. To me the winnings seem kinda paltry compared to the feat they expect you to accomplish. For example, 20 million dollars sounds like a lot money, but 20 million dollars to get a robot to drive around on the moon??? Looks to me like you'd have to spend 200 million to win 20. And 7 million dollars sounds like a lot for a tricorder, but when it comes to testing things on humans, I could see how you could burn through 7 meg in a short amount of time. Is there something about the economics of these kinds of competitions that I don't know about? Does simply belonging to a team somehow make your project a magnet for feverish investors ready to bury you in cash? Of course, if you were going to build such a device anyway, then I guess it would make sense to apply for it. But, otherwise, how do the numbers add up?

prof_braino
01-18-2012, 08:23 PM
Did you have a particular disease or technology in mind?

No, the parameters say 15 diseases, I bet we could find 15 easy ones. You are right, winning comes from progress on the task, and the offered money has little relevance. But this is actually very easy compared to getting to orbit. "Diagnosing" diseases does not require prescribing treatment for the diseases, so we don't don't have to worry about liabilities etc. The diagnoses don't even have to be accurate, only consistent with a commonly accepted interpretation of the observed symptoms. So the question is, what symptoms can we detect?

I should PM Tymkrs...

ElectricAye
01-18-2012, 09:57 PM
No, the parameters say 15 diseases, I bet we could find 15 easy ones. ...

One tricorder must be capable of diagnosing 15 diseases? How different do the diseases have to be? For example, some fungal skin infections will glow under UV. But does having the ability to detect fungus abc count as one disease, and then having the ability to detect fungus xyz count as a separate disease, disease number 2?

I would think you'd have to cobble together whatever you're doing from expired patents that have already been very well clinically tested, otherwise you would be facing lots of groundbreaking testing with willing patients and their docs. (And what doc wants to let you strap on an untested wig of sensors to her patient?) Not to mention the lawsuits from patent infringement if you try to adapt a technique that's relatively new.

A lot of neurological disorders are diagnosed by having patients move their limbs, follow objects with their eyes, etc. There are also devices that test nerve responses by essentially shocking part of a limb and measuring the muscle response, etc. Perhaps observing gait abnormalities can indicate a disease, too.

There are myriad things that can be measured, oxygen, body temperature, the sounds your heart, lungs and intestines make, skin resistance, pulse, the sounds your joints make if there's a problem with arthritis, etc. I think your biggest challenge would be trying to boil down all the sensor measurements to a single diagnosis while avoiding the temptation to enshroud your patient in a shining suit of sensor armor bristling with wireless antenna.

prof_braino
01-19-2012, 02:50 AM
One tricorder must be capable of diagnosing 15 diseases? How different do the diseases have to be? ...

Yes. Didn't say. Maybe, plan is to try for one and continue from there. At present, I plan on just the tricorder, and will not be asking to remove patients' limbs or eyes, expired or otherwise. Nor do I plan administering electrical shocks, that I believe is a Borg technic, and only useful with Hirogen communication arrays, and so out of scope for this effort. Although it all sounds like fun. But seriously;.

http://en.wikipedia.org/wiki/Tricorder - "The word "tricorder" is a portmanteau of "tri-" and "recorder", referring to the device's three default scanning functions: GEO (geological), MET (meteorological), and BIO (biological). ... The ship's doctor uses a variant of this model with a detachable "medical scanner" stored in the bottom compartment when not in use."

That's all we got to go on. I'm NOT thinking "wig of sensors"; I'm thinking android phone, and in addtion to the gyro and accellerometer, etc; add a micro with a bunch more sensors for oxygen, CO2, IR, and simple stuff like that, all "in the box"; prop is ideal for this. Data is collected, preprocessed, and sent over the phone to a remote computer for processing and analysis. MAYBE there is a second part that contacts directly to a patient, like the red LED oxygen sensor they stick on one's finger at the hospital; but nothing that costs me more than a few quid, or I won't be able to affort it.

I would suggest just figure out what DATA can be collected (with a set of parts I can afford), and then figure out if any disease can be indicated from that data. Probably not the best way to do it, but its a start.

ElectricAye
01-19-2012, 03:57 AM
...

I would suggest just figure out what DATA can be collected (with a set of parts I can afford), and then figure out if any disease can be indicated from that data. Probably not the best way to do it, but its a start.

That's why I kinda suggested you take a look at neurological disorders. The application of gyros and accelerometers seems to be more up the alley of the peeps on this forum, so you guys might be more at home measuring things like gait and motion and eye movement abnormalities rather than struggling with two-photon fluorescence and superparamagnetic immunoprecipitation and such.

Or, at least look at diseases that can present themselves as neurological disorders, then maybe zero in using other measurements such as body temperature, retinal weirdness of one kind or another, and pee-on-a-probe tests, etc.

rod1963
01-19-2012, 05:39 AM
I can see a established medical equipment company doing it on the side such as Space Labs. Or maybe a medical electronics guru doing some midnight engineering.

Better bring a good doctor on board to verify your diagnostic criteria - "not some commonly accepted interpretation" that's a cop-out and set you up for a lawsuit when your machine misdiagnoses someone. Doubly so for neurological conditions. This is not something you can get from reading a book on or a doing a wikipedia lookup.

And good luck getting any doctor to let you hook up any sort of powered device to a patient that hasn't be certified for use on people. That means probably bringing on-board a Medical instrumentation design engineer to verify design and implementation and not fellow hobbyists.

Even if you did have access to patients, you'd have to have rather large pool of patients to work from to make sure the machine and expert system s/w is at least 99% accurate. Remember this device is meant to be used in the field by people who are probably near technologically and medically illiterate and they will depend on the machine to render a accurate diagnoses. So bring on a expert on human interface design, this is one design where it's going to be critical. Here's the thing, modern medical instrumentation has come a long way in the user interface dept. Windows, GUI, touch screens, etc. Take a look at patient monitors or even dialysis machines, they are set up so even a non geek or even a patient in the case of say a dialysis machine can be taught to operate them.

Personally i think it's a non-starter. The companies that have the expertise and tech realize this device is a legal time bomb because it forces them into the business of diagnosing patients via their 'tri-corder'. Normally they just supply the instrumentation and doctors do the diagnoses from the results.

jazzed
01-19-2012, 06:10 AM
McCoy's medical tricorder never touched the patient. It was a "proximity" scanner.
It was probably also geologic to a point since he cured that silicon based life-form.

prof_braino
01-19-2012, 02:32 PM
"not some commonly accepted interpretation" that's a cop-out and set you up for a lawsuit when your machine misdiagnoses someone.


Nope, you got the wrong idea, its quite the opposite. This is NOT to be thrown into thousands of hospitals on day one, and expected to allow Amateurs to perform major surgery (although that sounds like the basis for a cool B-movie). NO diagnosis will be made by the device itself, doctors will only LOOK at the device, and maybe play with it, and any diagonosis will be made BY the doctor. Nothing in the rules says actual doctors have to be removed from the loop.

The function of the tricorder is to COLLECT RAW DATA, mildy pre-process it, and make it available to an EXPERT to INTERPRET; perhaps do minimal display. The device ONLY collects and records data. That is tricorder function.

Recall "sensors indicate..." each time they look at the tricorder. The device GATHERS DATA, the TRAINED SCIENTIST interprets it. (That was constatntly reinforced in the show, to keep the "human" part in the forefront). The device does NOT attempt interpretation of the data. This distinction is critically imprtantm, and makes the device definable, and therefore possible (though still unlikely).


Personally i think it's a non-starter.

You're off the team until you change your mind. :)

@ElectricEye - Tracking gait and all that is a good idea, and easy, particularly if the individual was video'ed walking, and a remote computer aids in analysis of the gait. The unit could simply record a video of the target, and send it off for processing, or simply have a doctor look at it. Since this is already built into an android phone, I count this as completed.

Think KISS, think easy.

Easy stuff: body temperature, environment temp, heat rate, blood oxygen, room oxygen, CO2, video, audio, breathing rate, body resistance, what else can we do right off the bat for cheap?

ElectricAye
01-19-2012, 03:19 PM
...
Think KISS, think easy.

Easy stuff: body temperature, environment temp, heat rate, blood oxygen, room oxygen, CO2, video, audio, breathing rate, body resistance, what else can we do right off the bat for cheap?

Well, I don't know about "easy" but I imagine there might be lots of things you could do with urine testing, possibly a crude spectrophotometer made from LEDs (how come LEDs never come in yellow?), especially if you get bold enough to add reagents of some sort. For example, there are fluorescence tests, etc. and there are magnetic bead immunoassays, perhaps, but that's fairly new stuff so you'd have to look at the patents. I'm guessing most electronics people will want to avoid getting involved with bodily fluids, but I bet if you don't get involved with urine somehow, your competition will blow you out of the water.

prof_braino
01-19-2012, 06:37 PM
Okey dokey, you can handle the urine if thats your thing; personally, I avoid bodily fluids whenever possible as a matter of policy. If anybody else has experience with sensors beyond non-contact, we can work them in as well. I can do sensors and drivers, but I'm not the guy for wet-work. Maybe Tymkrs will chime in with her doctor buddies and drive those aspects of the project.

Also, there is no "competition", I refuse to compete; this is not a zero sum game. I will participate, and if anybody makes progress, we all win.

This need not be a sport or a war where for one to win all others must be defeated. Science can be more like music, where the orchestra is the team, and if they do well, we are happy, but they don't have to beat up another orchestra to prove it. Although that might make opera more tollerable. That's actually quite a good idea, but off topic....

ElectricAye
01-19-2012, 07:01 PM
... personally, I avoid bodily fluids whenever possible....

I kinda figured that. It's why there's still a techno-psychological gulf between biologists/medicine and physicists/engineering. And I confess, all that gooey icky smelly stuff isn't something I find all that attractive, either, but after having raised infants and toddlers, you learn to soldier through it all to get the job done. After dealing with an exploded diaper or having an infant power puke into your face, what's a drop of urine in a little test tube, dude? Think of all those brave EMTs and ER nurses and docs who've got to deal with people after major car accidents, gurgling gun shot wounds, hunting trips with Dick Cheney, etc. As they say, if you can't stand the heat....

Maybe something a little more antiseptic for you would be sleep disorders, monitoring snoring, etc. And there's a reason docs have stethoscopes, too: there's a lot to be learned by listening to the body: heart, lungs, blood flow in the neck, joint problems. Perhaps you can make something to analyze the noises people make when moving their joints, or snoring, or breathing or even talking. Even the intestines make noises which might be analyzed. Or perhaps incorporate some speech recognition software so when your patient is screaming "My arm! My arm! Get off my freakin arm, you idiot!" your little LCD screen/annunciator can light up and go ....DING.... Condition localized to one of two arms.... DING....

jazzed
01-19-2012, 11:16 PM
... after having raised infants and toddlers, you learn to soldier through it all to get the job done. After dealing with an exploded diaper ....

Just make sure the baby is finished doing it's job before changing that smelly diaper. Otherwise be forever reminded of the Play-Doh Fun Factory extrusion kit in action ....

ElectricAye
01-19-2012, 11:57 PM
Just make sure the baby is finished doing it's job before changing that smelly diaper. ....

There's your first billion dollars if you can figure out a device that alerts you to the probability of getting pooped on during a diaper change. At present the only defense is reactive armor and a hazmat suit.

RDL2004
01-20-2012, 12:00 AM
Lol, first time I've read this thread, not what I expected. Thought it was about ASM, C++, and BASIC...

Tymkrs
01-23-2012, 04:20 AM
Hey! Got a couple PMs seeing if we'd be interested in potentially joining the discussion. Seems like a cool thing to try out though I imagine I'd be better as a health consultant over programmer as I'm nowhere near as capable in programming as most of you...(trust me, listen to http://firstspin.tv where I'm still learning :p)

Looks like the requirements are <5lbs, ability to store and share their information, which must be accessible remotely via the Internet. No harm from electrical energy, thermal energy, chemical exposures, needles, lancets, and infection. (This rules out blood sugar/INR testing). Needs to be able to diagnose 15 diseases and provide a good user experience.

Suggested systems to use: Wireless sensing, imaging diagnostics, lab-on-a-chip, and molecular biology
Population: "A set of 15 distinct diseases in a group of 15-30 people in three days" - seems that they haven't stated what the diseases are.
Timeline is about a year for the final guidelines, and 2 years for development.

So off the bat, the information that you could get easily would be:
Heart Rate, ECG, O2 levels, Temperature, Respirations, auscultation probe (to listen to lung sounds/heart sounds, bowel sounds)

You could analyze urine - they have those dipsticks that can already tell you a number of things and only require the stick and urine. So maybe someone puts some urine in a cup, you dip the stick, the stick gets placed into the machine, the machine stores/uploads the information from it. Oh and if you can analyze urine, then you can tell if someone's pregnant (though I don't think that's a disease :p).

If the machine is to give information to doctors rather than diagnose, you could potentially have:
Ultrasound technology, and if you can do ultrasounds, you could do ALL sorts of internal scans - this could tell you how your heart was doing, arteries (carotids), whether you have any DVTs, emboli in general, bladder scans, uterine scans.

X ray technology (though....obviously you need someone who knows what they're doing with this).

Doppler Technology - we use this in some of our heart patients who don't have effective blood flow - and you can tell how well blood is flowing through their outer extremities.

Just some ideas, what do you guys think?

Tymkrs
01-23-2012, 04:40 AM
I brought this up on our IRC (https://qwebirc.afternet.org/?channels=tymkrs) and it seems that there's a company out there that already can do ultrasound in a tiny package:

http://www.sonosite.com/products/nanomaxx
http://www.geekwire.com/2011/mobisante-release-smartphonebased-ultrasound-system
(http://www.geekwire.com/2011/mobisante-release-smartphonebased-ultrasound-system)
http://ocw.mit.edu/resources/res-ll-003-build-a-small-radar-system-capable-of-sensing-range-doppler-and-synthetic-aperture-radar-imaging-january-iap-2011/

Seems pretty awesome. This could help doctors diagnose atherosclerosis, valve issues, bladder issues, vascular issues. Hmm!

rod1963
01-23-2012, 05:35 AM
Interesting.

Cost wise the getting the various devices that measure human vitals isn't that expensive Heart Rate, ECG, O2 levels, Temperature, Respiration, auscultation probe. All of these are COTS items and can be had for $300.00 - $400.00, at least the cheaper versions.

With COTS devices you either have to adopt current designs to fit your tricoder or failing that design these sensors from the ground up or cannibalizing parts from existing ones.

Ultrasound; haven't seen one that isn't the size of a full sized laptop and then some. If you know of one that that can fit in the palm of your hand and is open source you're stuck with the task of designing and building one from the ground up and then there is the software issue. So far the cheapest device I've seen is around $10k.Used versions that are 15+ years old can be had for $1500 or so.

So the more bells and whistles you want on a tricoder the more it's going to cost. Keep it with the initial set of sensors, it's doable. Incorporate Ultrasound like you see in the hospitals, you better be GE.

jazzed
01-23-2012, 05:54 AM
Parallax sells the TSL235 and TSL237 light-to-frequency sensors.

It is claimed here (http://www.parallax.com/Portals/0/Downloads/docs/prod/audiovis/604-00084medicalapps.pdf) that such sensors can be used in heart rate and other applications.
How exactly would this be done?

rod1963
01-23-2012, 06:03 AM
...
Easy stuff: body temperature, environment temp, heat rate, blood oxygen, room oxygen, CO2, video, audio, breathing rate, body resistance, what else can we do right off the bat for cheap?

Not easy unless you're a expert at interfacing all these COT devices. Are you capable of tearing down a commercially available heart rate monitor or ECG machine and hooking it up to a Android? A cheapo ECG(reads off the thumb) can be had for around $200.00. That would be a nice starter device to interface and proof of concept.

Tymkrs
01-23-2012, 06:38 AM
Ultrasound; haven't seen one that isn't the size of a full sized laptop and then some. If you know of one that that can fit in the palm of your hand and is open source you're stuck with the task of designing and building one from the ground up and then there is the software issue. So far the cheapest device I've seen is around $10k.Used versions that are 15+ years old can be had for $1500 or so.

So the more bells and whistles you want on a tricoder the more it's going to cost. Keep it with the initial set of sensors, it's doable. Incorporate Ultrasound like you see in the hospitals, you better be GE.

Well I'm thinking this won't be able to actually DIAGNOSE, just act as a portable information getter for doctors. I'm just initially thinking about this practically, not in terms of the contest.

In terms of the ultrasound, the links I show show them being iphone size: http://www.sonosite.com/products/nanomaxx
http://www.geekwire.com/2011/mobisante-release-smartphonebased-ultrasound-system

And let's say people want to enter this contest. I'd assume people would put money into the development because ROI would be /amazing/ even if you didn't win the contest (assuming of course a good product was made). Whoever made that small iphone ultrasound thing is selling it for near 8-10k each.

For diagnoses, vitals don't tell you THAT much about a variety of disease states or what disease they may have. A low blood pressure could mean sepsis, or low fluid volume, or gross dilation of blood vessels.

jdolecki
01-23-2012, 10:11 AM
Parallax sells a X-band motion sensor it says it can detect movement throught walls whats the difference between a wall and your chest?

Could be used to look at the chest and measure heartbeats based on heart movement?

I may just order one to check it out.


http://www.parallax.com/Store/Sensors/ObjectDetection/tabid/176/CategoryID/51/List/0/SortField/0/catpageindex/2/Level/a/ProductID/606/Default.aspx

prof_braino
01-23-2012, 07:10 PM
Parallax sells a X-band motion sensor it says it can detect movement throught walls whats the difference between a wall and your chest? Could be used to look at the chest and measure heartbeats based on heart movement?

As sold, the sensor only detects objects that are moving, but doesn't give any other info.

Folks wrote about detecting a person moving on the other side of a brick wall.

So, out of the box, probably not; but I think these could be modified to detect internal motion, and could be used for imaging. But haven't started on that part yet, so its still open as to HOW MUCH work would be involved. This is definitely on my list of possibilities.

prof_braino
01-23-2012, 07:14 PM
light-to-frequency sensors.. used in heart rate and other applications.
How exactly would this be done?

It looks like certain diagnoses are based on reflected light color, others on reflects light amount. So if I get this correctly, the amount of light varies with the volume of blood in the veins, and this varies each time the heart beats. Can anybody confirm this?

ElectricAye
01-23-2012, 07:21 PM
... I think these could be modified to detect internal motion, and could be used for imaging.....

If I were you guys, I would stick to working with only "passive" sorts of sensors. The instant you start pumping energy into a human being via ultrasound or electromagnetics, you enter a whole new ball game with respect to clinical cooperation, safety, liabilities, human testing regs, etc. Have somebody sweat or pee on a stick and you can bombard that stick with whatever your hearts' desire, but aim something at a human being, and you'll have hell (manifesting itself in the form of lawyers) to pay.

tritonium
01-23-2012, 07:51 PM
Interesting project
I found this...
http://www.ebay.co.uk/itm/Olimex-SHIELD-EKG-EMG-electrocardiography-electromiography-arduino-shield-/270896950685?pt=UK_BOI_Electrical_Components_Suppl ies_ET&hash=item3f12b76d9d (http://www.ebay.co.uk/itm/Olimex-SHIELD-EKG-EMG-electrocardiography-electromiography-arduino-shield-/270896950685?pt=UK_BOI_Electrical_Components_Suppl ies_ET&hash=item3f12b76d9d)

The circuit diagram is available from seller
A medical tricorder is HARD but a multimeter sized box with lots of leds, cheap colour phone display and sensors would make a nice toy (dont forget the sound effects)
temperature, altitude, magnetic direction, humidity, time-date etc, gps position, distance to neaby object (rangefinder), obstacle detector for moving in the dark (oops a torch!) I mean how many could you cram in?
There must be LOTS more without having to invent anything but perhaps re-use an everyday device in a new way.
OOps sorry wandering and getting off topic - perhaps new thread? (I should have looked).
Dave

frank freedman
01-23-2012, 07:59 PM
If I were you guys, I would stick to working with only "passive" sorts of sensors. The instant you start pumping energy into a human being via ultrasound or electromagnetics, you enter a whole new ball game with respect to clinical cooperation, safety, liabilities, human testing regs, etc. Have somebody sweat or pee on a stick and you can bombard that stick with whatever your hearts' desire, but aim something at a human being, and you'll have hell (manifesting itself in the form of lawyers) to pay.

Actually the lawyers will be lined up, or shall I say circling above the FDA inspectors who are there to padlock and seize the whole operation if it does not submit proper paperwork for just about anything you could possibly associate with the project, get all possible clearances and inspections, comply with the current flavor of GMP, comply with whatever parts of 21CFR apply to their device in question...........

Uhm, Mr. Eye, may we see your clearances and approvals, processes and inspections on that "electric eye"? You don't have what? You want us to turn a BLIND eye????

Frank Freedman, CRES

frank freedman
01-23-2012, 08:01 PM
It looks like certain diagnoses are based on reflected light color, others on reflects light amount. So if I get this correctly, the amount of light varies with the volume of blood in the veins, and this varies each time the heart beats. Can anybody confirm this?

This is how the SpO2 sensors work, using a ratio of colors to determine O2 sats and pulse rate. Can;t tell you detail, my specialty is X-ray systems.

Frank Freedman, CRES

erco
01-23-2012, 09:45 PM
http://www.rdmag.com/News/2012/01/General-Science-Physics-T-Rays-Could-Help-Develop-Star-Trek-Style-Handheld-Medical-Device/?et_cid=2438662&et_rid=54725820&linkid=http%3a%2f%2fwww.rdmag.com%2fNews%2f2012%2f 01%2fGeneral-Science-Physics-T-Rays-Could-Help-Develop-Star-Trek-Style-Handheld-Medical-Device%2f

jdolecki
01-23-2012, 09:49 PM
I think the Tri Coder is for the future where there are no more lawyers only healthy happy people.

Or Borg

jdolecki
01-23-2012, 09:58 PM
Some of the ideas behind the Contest

Address the dual challenge of a) transforming healthcare by turning the "art" of medicine into a science, and b) making health a willing part of individuals' daily lives.

Tackle needed breakthroughs in sensing technologies, technological integration, regulatory acceptance, and the perception that healthcare should be controlled by only a few skilled parties.

Incentivize teams to focus on a) precision diagnosis and measurement that is independent of healthcare institutions/professionals and on b) consumer needs/adoption.

Facilitate necessary partnerships and regulatory pathways required for this major transformation.

Inspire a future where consumers demand the tools to assess and manage their health independent of a hospital or doctor's office.

Spark the creation of markets and products that offer medical detection, prevention, and management, as well as more complete diagnostics.

jazzed
01-24-2012, 12:57 AM
It looks like certain diagnoses are based on reflected light color, others on reflects light amount. So if I get this correctly, the amount of light varies with the volume of blood in the veins, and this varies each time the heart beats. Can anybody confirm this?
Would this qualify as lawyer-nip?

From Wikipedia (kind of):

Nepeta lawyaria (and lawmints (http://en.wikipedia.org/wiki/Nepeta)) are mostly known for the behavioral effects they have on lawyers, not only on domestic lawyers but also large enterprise lawyers. N. lawyaria is used as a recreational substance for lawyers' enjoyment, and lawyer-nip and laced-lawyernip products designed for use with domesticated lawyers are available to consumers. Not all lawyers are affected by lawyers-nip.The common behaviors when lawyers sense the bruised leaves or stems of lawyers-nip are rubbing on the plant, rolling on the ground, pawing at it, licking it, chewing it consuming much of the plant followed by drooling, sleepiness, anxiety, leaping about and purring. Some will growl, meow, scratch, or bite the hand holding it. Some l (http://en.wikipedia.org/wiki/Felidae)awyers will eat dried lawyer-nip. Often, eating too much can cause lawyers to be overtly aggressive, typically making them hiss.[citation needed (http://en.wikipedia.org/wiki/Wikipedia:Citation_needed)]