Hackaday Prize Entry: Infrared Vein Illumination

Phlebotomy is a fun word, and the fine art of finding veins. While the skill of putting needles in arms is honed by nurses and physicians over the course of decades, there are, of course, technological solutions to finding veins. One of the more impressive medical devices that does this uses near-infrared imaging — basically looking under the skin with almost visible light. These devices cost a fortune.

One project in the Hackaday Prize is looking to change that. It’s a near-infrared vein finder. Instead of the thousands of dollars professional unit costs, this one can be built for under one hundred bucks.

As far as this build goes, veins are illuminated via IR light at about 950nm. The backscatter of this light is captured via a Raspberry Pi NoIR camera, with regular old photography film blocking visible light. From there, it’s just a simple matter of image processing and hitting enhance several times until veins appear on a display.

The team behind this project has already developed a mobile version of the device, complete with 3D printed parts. It’s a handy device and a great entry for the Hackaday Prize.

18 thoughts on “Hackaday Prize Entry: Infrared Vein Illumination

  1. The professional units can also be built for less than a hundred dollars but no manufacturer is going to sale it to you for that much as there are other things that need to be accounted for such as certification and insurance which people don’t seem to take into consideration when looking at the market cost. Nevermind the ongoing cost of maintaining support for years after product introduction.

    1. Sounds like BS in this case, certification and insurgence for a passive device that if it doesn’t work has no result since the physician won’t use it, and which can’t harm a patient.

      1. You are being narrow-minded. For example, a glucometer isn’t typically used by a physician and won’t harm the patient on its own yet it influences medical decision making. If the patient is hypoglycemic and the glucometer reads the blood sample as having high glucose then giving additional insulin can kill the patient. In this instance if the device shows possible veins which aren’t there in the first place then the patient will have to go through the pain of multiple useless attempts of having an IV line started with the additional hazards of causing a hematoma, infection, and damage to other structures such as nerves. You also waste the nurse’s time which could have been better spent on something else such as caring for another patient.

        1. Bad comparison, if a vein indicator doesn’t work then it’s not used, so a comparison with a glucose meter would be that if it goes offline and there is nothing on the display then the doctor won’t use it and will use another method and there is no direct fault in treatment from the device. That’s what I was trying to say, it either works or not, there isn’t really an wrong indication when it’s just a camera showing real world input.
          Now if you add a projector then you could theoretically get a misalignment, but if you add some indicators you can make it visually clear when there is such a misalignment and again you can’t really go wrong can you?

          1. Device failure isn’t just an all or nothing thing. Devices can partially fail and give you wrong information. You also underestimate people being vigilant enough to notice something is wrong and call the results of a device in question. It happens all the time in an environment where it is easy to become complacent and fatigue is high.

          2. I get that with other deices, that they can fail partly, but I’m not sure with this vein detector concept it can fail partly.
            But of course, it’s true that it doesn’t matter to the sue-happy, if they think they can latch on they will.

  2. there was an interesting one i was looking at recently which projected the image of the veins back on to the patients skin.

    I think the main reason for building medical devices is that for some places the cost of the equipment is phobitive and the risks associated with a DIY product are less than the gains realised. its all about risk mitigation

  3. ” The backscatter of this light is captured via a Raspberry Pi NoIR camera, with regular old photography film blocking visible light.”

    This was the genesis of a lot of “see through clothing” photography in the late 1990s as well. I use it to sight in IR lasers in the lab with an el-cheapo webcam with the IR filter flake removed and a bit of film over the lens.

    1. Nope, the only ever real issue with ‘see-through’ was when some Japanese weirdos found out Sony camcorders that had nightvision could see slightly through some bathing suits (making people look ugly mind you), which was disabled by Sony.
      The rest is just BS they try to freak out idiots with.

  4. I can see this being used to find splinters as well. At least someone tried, just because it might be hampered by regulations here does not mean someone else say in darkest Africa can’t use it to save many lives. +100 to the developers!

    (also has a NoIR here and a few random LED screens, may be possible to build something approximating this with a simple used laser barcode scanner and some clever software!)

  5. This is interesting. Veins have built-in valves:

    https://www.ncbi.nlm.nih.gov/pubmedhealth/PMHT0028152/

    Subcutaneous (below the skin) vein valve action is an indicator of blood flow/pressure. As an introduction, see here:

    https://www.ncbi.nlm.nih.gov/pubmedhealth/PMHT0028152/

    For example, if you exercise vigorously you may be able to see the veins in your palm-up wrist pulsing. The valves in the wrist veins are the reason for this (please correct me if I’m wrong, I’m an EE not an MD). This sort of infrared imaging may be useful for obtaining data on relative blood flow and/or pressure. But it’s probably not useful for absolute measurements (difficult if not impossible calibrate).

    Hmmm… How about a wrist-band with a flexible OLED IR emitter array that also act as sensors when multiplexed?

    1. The valve in the vein acts like a diode so that blood only flows in one direction. This is especially important when veins have to act as a conduit to transport blood up against gravity. Failure to stop backflow results in venous distension manifesting as varicose veins.in the lower extremities. The valves cannot pulsate on their own as pusatile flow is solely dependent upon the heart.

      Imaging of blood flow through veins constrained by a single area is not very helpful as there are many more veins in the general vicinity which can carry more or less blood than the ones of interest. An analogy to this would be trying to determine the power consumption of an electronic device by looking at the current flowing through a single arbitrary branch of a circuit – doing that will definitely not help you in getting what you want. In addition, there are veins that are bigger and deeper in the arm that cannot be reached just by shining an IR light onto the surface of the skin as they are behind much more dense tissue (fat and muscle).

    1. Add eight optical mouse image sensors aound a shrouded cree led and then feed that to a projector and let a computer project what was scanned back into the skin.
      Seemingly amazingly I saw pound land has optical mice. Quite tempting to give it a go.

    2. Not for someone who has really deep, small, hard to hit veins. My wife has that problem. It always takes the nurses several tries to stick her. Those are the people you need such a tool for anyway. A bright torch held to her skin shows no veins. The same torch shows tons of veins on me but… I have plenty that can be seen and used with just the naked eye so what’s the point?

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

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