Contest Submitter Feedback
Invent-O-Doc
Posts: 768
This topic will be updated regularly to provide feedback on submissions coming in for the contest.
For the contest so far, we've got over 10 submission and more than 75% are for physiology monitoring. My feedback for all of you is that remote monitoring of heart rate, etc, is probably not enough.
What do you want to monitor vital signs for??? Are you making a group exercise monitor that looks for people to be in the 'optimal training zone' for heart rate? Are you making a seizure detector? Be creative!
For the contest so far, we've got over 10 submission and more than 75% are for physiology monitoring. My feedback for all of you is that remote monitoring of heart rate, etc, is probably not enough.
What do you want to monitor vital signs for??? Are you making a group exercise monitor that looks for people to be in the 'optimal training zone' for heart rate? Are you making a seizure detector? Be creative!
Comments
And in any case, if I can think of something else that I want to do as well, I'll do it. It's just that I have the good fortune to be able to avoid hospitals and the sick, so I have almost no contact with them and don't know what the current problems are.
How about a circadian rhythm monitor/adjuster? People experiencing jet lag, especially those who need to respond quickly to emergencies, earthquake rescue, etc. need to be as alert as possible, and optimizing circadian rhythm can be important. Such rhythms might be monitored via body temperature and heart rhythms. Adjustments can be made to the body by instructing the person when to eat a bolus of protein. Or perhaps even provide artificial sunlight via LEDs, etc. Emergency management of people could include identifying people based on their natural alertness levels - have the night owls working at night, the super early risers doing the early morning shift, etc. as opposed to a rigid schedule structure that forces night owls to work in the morning and morning people to sleep in the day, etc. Such a monitor might also be good at detecting when people are getting fatigued, burnt-out in the ER, etc. so that people can be rotated into and out of critical functions in a more rational way. The sensor part might be easy for this, but I'm guessing the software/algorithms would be the bear to work out.
Im not sure if this fits the contest though.
The contest page says "project or product for medical simulation (training) or a prototype medical device that performs a useful medical function"
What kind of electronic medical device can there be that isn't doing physiology monitoring?
Medical simulation would be a good area to submit to, I think. Those CPR dummies must have some kind of microcontroller in them...
I think his point is that doing monitoring by itself is probably not enough. I'm guessing your device needs to do something with those measurements, integrate them into some sort of actionable intelligence. That's why contestants would probably do well to enlist the help of some EMTs or nurses, etc. for developing concepts, people who are experienced at making split-second decisions in real time.
Rwgast - fitness apps are a cool idea too.
Yep, an I am guessing things like alarms for various things similar to what the big OEMs offer in their devices like the Philips MP 70s etc...... That will be the thing that will make or break our teams entry, can the prop handle actually analyzing ecg waveforms well enough to accurately alarm on other than significant changes in heart rate. SPO2 should be pretty easy to monitor and alarm on dropping below a setpoint
The easy part is making the Propeller into a data acquisition system as I've already put an accelerometer and SD data logger on one of the Quickstart boards and was walking around with it today. *Note: don't forget decoupling capacitors and effects of powerline spikes as finally figured out this stupid error at 06:00 today when I was searching Spin code for the source of the problem. *Hopefully will get the Polar monitor interfaced later this evening.
Where things get interesting is when one wants to perform data analysis on the Propeller system. *32 bits really helps and, to get practice with the limitations of this type of fractional integer math, all of my photoplethysmography analysis in VB6 was done with fractional integers. *That was before I finally dove in and started doing everything on the Propeller and the VB6 application acquired data from a Phidgets 8/8/8 board. *What that showed me is that one can expect very noisy data.
It's probably better to indicate "data too noisy" than to try to analyze it. *I routinely get calls from the nurses at the hospital where I work about a patient's SpO2 being 75%. *I now check for myself using a $35 Walgrens finger SpO2 analyzer. *Invariably when I put this cheap device on the patients finger, it refuses to give a reading. *That's because the patient is peripherally shut down and one needs either a blood gas or a better perfused limb. *The expensive hospital pulse oximeters provide a reading that is likely accurate but clinically irrelevent aside from making one aware that there is decreased peripheral perfusion. *One can do a simple experiment of putting a pulse oximeter probe from a medical grade oximeter on ones finger and then lying on the arm and watching the SpO2 drop into the 70's as one arm starts to go to sleep. *I've found that my cheap pulse oximeter is superior to the hospital ones because it gives me that "data no good" indication.
What's going to be fun in this contest is that I get to reuse all of my neat hacks from my electrophysiology research days when we'd analyze data on as primitive a machine as a PDP-8 with 12 Kw of core. *The Propeller's raw processing performance is considerably in excess of the PDP-11/23 system that I last used. *If I have time, will try to include some other photoplethysmographic functions other than just simple pulse oximetry. *There's a huge number of parameters that one can measure just from shining various frequency lights through the skin such as HGB concentration, CO-HGB levels (that would be a useful function) as well as tissue O2 uptake. *My plans are to trim down the really neat physiologic possibilities to something more practical by talking to our local ER docs and country docs that work in rural hospitals where they're essentially on their own overnight. *The huge medical potential of photoplethysmography is going to be a big danger with this project as I've downloaded and read dozens of papers on photoplethysmography in the last month alone.
Awesome! Keep us posted on how you're doing. I can't wait to see how this turns out.