My project is a machine that will assist in finding the vein for special medical treatment (inject medicine). It will use a phototransistor ("blue" veins can be seen in infrared light quite good - that's why the opposite, blue light, is used to force drug addicted people away - they cannot see their veins in this light). If the vein is found, the device adjusts in the middle of it and then punctures the vein. If everything worked out fine a message to the medical staff is send by mail, informing them that the treatment was successful. I plan to build everything with fully developed mechanic and pieces from computer junk. I already took apart a DVD drive to use the stepper motor driving the pickup. The steps will be counted and saved where a minimum or maximum can be found - there the vein should be. As a next step I'll use a servo or another stepper from a DVD drive to puncture the vein.
old DVD drive
for stepper motor extraction
Stepper Motor shield
(cheaper than building it myself with 2H-bridges)
What would a sensor (aka phototransistor) "see" when it looks at out skin? I wanted to know this prior to coding to know what to look for - I knew that the blue veins would absorb light differently, but how exactly?
So I took a modified webcam to take some pictures. I'll use the webcam in another project - it can only "see" IR light now. I removed the IR-blocking filter and replaced it with old analog film (developed) that blocks everything except IR. And I attached some SMD-IR-LEDs as well... somehow the autofocus is not working really good, but you can see the huge difference: the veins look very dark in IR light compared to "normal" light in the visible spectrum.
Now I know what to look for - it should be a local minimum concerning the amount of reflected light detected by the phototransistor.
When I successfully detected the vein, I'll upload the code and some pics - and a video of this device in action (puncturing my vein)
BTW: I took photos of my fathers arm, this is not mine :-)
That was luck! I used the stepper shield and an Arduino to drive the motor (stepper), simulating the movement across the arm in search for the vein. Everything worked fine and I just used a 12V power supply for the (yeah, too little) motor. After a minute I reallised something smells funny - the second motor was overheating. Must try with less voltage next time :-)
Today the video of the first build attempts is ready. I will show tomorrow how veins look in infrared light...
... I found out trying to de-solder the ribbon-connector and attach wires instead. It was more heat than the plastic carrying the little connectors could stand - one small connector just vanished within the molten plastic and I could not connect the motor any more .... FAIL.