Smart Thermometer Probes First, Asks Questions Later

As flu season encroaches upon the northern hemisphere, doctor’s offices and walk-in clinics will be filled to capacity with phlegm-y people asking themselves that age-old question: is it the flu, or just a little cold? If only they all had smart thermometers at home that can tell the difference.

Typically, a fever under 101°F (38.5°C) in adults and 100.4°F (38°C) in children is considered low-grade, and thus is probably not the flu. But who can remember these things in times of suffering? [M. Bindhammer]’s iF°EVE is meant to be a lifesaving medical device that eliminates the guesswork. It takes readings via 3D printed ear probe mounted on the back, and then asks a series of yes/no questions like do you have chills, fatigue, cough, sore throat, etc. Then the Teensy 3.2 uses naive Bayes classification to give the probability of influenza vs. cold. The infrared thermometer [M.] chose has an accuracy of 0.02°C, so it should be a fairly reliable indicator.

Final determinations should of course be left up to a throat swab at the doctor’s office. But widespread use of this smart thermometer could be the first step toward fewer influenza deaths, and would probably boost the ratio of doctors to patients.

12 thoughts on “Smart Thermometer Probes First, Asks Questions Later

  1. The problem with a lot of people is that they go to the doctor for practically everything…
    Instead of using some common sense and prior experience.

    So a meter able to fill in the gap and aid people might be a decent thing. Though, would also be nice if people were able to think rationally to begin with.

    Though, even the influenza can express fairly mundane symptoms, it all depends how one’s immune system reacts and how good it is at keeping things in check. After all, some people doesn’t even really get sick from it. In these cases it can be hard to even know that one had it.

    1. Yeah exactly, the question that needs answering is “How ill are you?” and it’s secondary why you are that ill when there are no specific treatments like the flu/cold and other viral illnesses.

      1. That is why you need a “third party” to evaluate how really really ill you are.

        Let’s face it if you think you are ill no way you’ll say: no, I am not ill.

        A third person will help.

      2. The specific case of influenza is just about the only common viral infection that _does_ have a somewhat effective treatment: oseltamivir (aka Tamiflu™) which is somewhat effective in shortening and lessening the course of symptoms, but more importantly can be used prophylactically if there are at risk people you can’t avoid being around (i.e. young children, the elderly, anyone with a compromised immune system) Oh and 100% agree with comments down-thread, get your shots, especially if you are in this situation (care giver for elderly parents or small children, for example)

        In fact, in order to get an official doctor’s opinion that “yup, you have influenza” (actually, the lab test’s opinion) so that a prescription for a family member who _really_ needed to not get sick could be issued, has been the only time in the last decade I’ve gone to the doctor for acute care.

  2. This is ridiculous. A complete waste of effort and resources. The reason why is laid out in this article if you read carefully. Although I don’t suspect many will have trouble identifying the obvious reasons this is complete rubbish.

  3. I made my own thermometer using a PT-100 probe in a stainless steel sleeve. It’s so hard to get decent thermometers without spending an arm and a leg, and you never know what you’re getting. My little device is great since i KNOW it’s accurate and calibrated. I also have a graph so you can see the reading settle over time. I can usually take an educated guess within 20 seconds just by looking at the curve of the graph.

  4. You’ll need to have the whole device in a smooth package that’s easy to wipe down and sterilize. As is, there are too many surfaces to be cleaned. Otherwise it’ll become a vector for the virus when you pass the device to other people.

  5. Sigh… I feel like the difference between *accuracy* and *resolution* should be beaten into people’s heads. Imagine I make crappy plastic rulers. I make a 30cm ruler with 1mm markings. However, since I don’t actually have a good reference standard, my rulers end up being off by 1cm. My *accuracy* is therefore (30cm – 29cm) / 30 * 100 = 3% (or just “1cm error over fullscale range”) off. My *resolution* is still “1 mm” however.

    The *resolution* of this sensor is 0.02degC. The *accuracy* of this sensor is 0.5degC.

  6. First step toward fewer influenza deaths – get a Flu shot.
    Make a lot more difference than this would.

    (If you are sick enough to be wondering is it flu or cold – stay home.
    Don’t go spreading it around at work/school/etc.)

    Also, note that 98.6, 101, etc. degrees – are averages over lots of people.
    They may not be that relevant to you.

    If your normal temperature runs a bit lower, say 97 degrees, then should take that
    into account when figuring out how much your fever is above your normal temperature.
    (If going to make it smart – how about have it remember what each user’s
    normal reading is.)

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.