• Ultraviolet Photography
  •  

Veins in NIR

Filters Infrared
27 replies to this topic

#1 Andy Perrin

    Member

  • Members+G
  • 4,093 posts
  • Location: United States

Posted 24 December 2020 - 03:22

Inspired and horrified by the paper, "The use of near-infrared light for safe and effective visualization of subsurface blood vessels to facilitate blood withdrawal in children", I decided I needed to verify that you could see veins in terrifying detail if you use transmitted infrared between 850-900nm.

Attached Image: Screen Shot 2020-12-23 at 10.17.41 PM.png

It turns out that diffuse illumination from a desk lamp makes it hard to capture and you are much better off using a flashlight with a tungsten filament.

For filtration, I used an 830BP40 stacked with the dual-band DB-850 from MidOpt (which passes 820-880nm according to the included chart in the box). So the light here is roughly between 820nm and 870nm.

F/11 13" ISO6400
Attached Image: _DSC8185 UVP.jpg


F/11 1/8" ISO3200
Attached Image: _DSC8176 UVP.jpg

I have no doubt Eka will put this to good use!

Edited by Andy Perrin, 24 December 2020 - 03:28.


#2 Stefano

    Member

  • Members(+)
  • 1,800 posts
  • Location: Italy

Posted 24 December 2020 - 07:41

I tried this with a 850 nm LED and it worked quite well: https://www.ultravio...attach_id=20382

When I will mount my other IR LEDs I can try with each wavelength I have between 730 and 1050 nm, to see if there are differences.

I noticed orange-yellow light is the limit. Green light has a too short wavelength, and is absorbed quickly enough to start noticeably heat you.

#3 Andrea B.

    Desert Dancer

  • Owner-Administrator
  • 8,780 posts
  • Location: UVP Western Division, Santa Fe, New Mexico

Posted 24 December 2020 - 08:27

Or you could use an AccuVein. https://www.accuvein...viewing-system/

I was trying to find what wavelength of IR the AccuVein gadget uses, but I'm getting too tired. It's been a long day.
Andrea G. Blum
Often found hanging out with flowers & bees.

#4 Bernard Foot

    Bernard Foot

  • Members+G
  • 718 posts
  • Location: UK

Posted 24 December 2020 - 08:54

Neat photos, Andy. Looks like that's something else for me to try out over the dark winter months.
Bernard Foot

#5 dabateman

    Da Bateman

  • Members+G
  • 2,771 posts
  • Location: Maryland

Posted 24 December 2020 - 12:59

Yes in the hospital they will use a near IR light to shine on your leg or arm area to find the vein for a connection.
Sadly seem this why too many times.

#6 Andy Perrin

    Member

  • Members+G
  • 4,093 posts
  • Location: United States

Posted 24 December 2020 - 17:10

View PostStefano, on 24 December 2020 - 07:41, said:

I tried this with a 850 nm LED and it worked quite well: https://www.ultravio...attach_id=20382

When I will mount my other IR LEDs I can try with each wavelength I have between 730 and 1050 nm, to see if there are differences.

I noticed orange-yellow light is the limit. Green light has a too short wavelength, and is absorbed quickly enough to start noticeably heat you.
Not much point in trying with the other wavelengths. Look at the graph I posted. The absorption starts to go up rapidly from the water so you lose contrast past 900nm or so.

If I can find a bandpass filter at 875nm at the peak, that would be worth a try. Or an 875nm laser if those exist.

Edited by Andy Perrin, 24 December 2020 - 17:12.


#7 Stefano

    Member

  • Members(+)
  • 1,800 posts
  • Location: Italy

Posted 24 December 2020 - 17:48

...not an exact match, but here is an 880 nm LED: https://www.thorlabs...rtnumber=M880D2 (they are available also on eBay and online in general, even though they are less common than other IR wavelengths).

As far as lasers by Thorlabs, you have 880 nm again and also 976 nm which is spot-on the water peak.

But the peak in the graph is quite broad, so there's no need to be too tight on the wavelength.

Edited by Stefano, 24 December 2020 - 17:48.


#8 Andy Perrin

    Member

  • Members+G
  • 4,093 posts
  • Location: United States

Posted 24 December 2020 - 19:12

Stefano, again, the water peak is not the optimum. There are three important parameters involved, reflection at the surface, absorption, and scattering. Water absorption is definitely a contributor to absorption but the others matter a lot, particularly scattering and absorption by collagen. So 976 peak is irrelevant.

This is exactly why I posted that graph, it shows the true optimum contrast, which is the relevant variable for us.

I feel like people are missing the main approach I take. I usually don’t discover anything new that other people don’t know about. Cadmium posted about dark water ages before I started playing with it. You did the veins before now. Etc.

Instead, my procedure is to research and optimize these phenomena. (By good luck you hit very near the peak though!)

Edited by Andy Perrin, 24 December 2020 - 19:17.


#9 Stefano

    Member

  • Members(+)
  • 1,800 posts
  • Location: Italy

Posted 24 December 2020 - 19:22

Yes, the 976 nm laser was a bit out of context, it is irrelevant here.

I remember trying at 940 nm but my camera had a poor sensitivity and the contrast was not great, and 850 nm worked better. Makes sense reading the graph.

#10 Andy Perrin

    Member

  • Members+G
  • 4,093 posts
  • Location: United States

Posted 24 December 2020 - 19:45

You should try again with better focus! I think you have good hands for this. Mine are too thick.

#11 Stefano

    Member

  • Members(+)
  • 1,800 posts
  • Location: Italy

Posted 24 December 2020 - 19:48

I will for sure when I will have a camera again.

I don't think my hands are thin, yes a white flashlight does shoot through them a tiny bit, but you can find better hands.

#12 Stefano

    Member

  • Members(+)
  • 1,800 posts
  • Location: Italy

Posted 24 December 2020 - 19:55

I also wanted to share this video I remember seeing a couple of years ago: https://youtu.be/8TFHkcfQYvM

Some videos from this channel played a big role in inspiring me to explore invisible light. I don't know if he is a member here, I don't think so.

#13 Bernard Foot

    Bernard Foot

  • Members+G
  • 718 posts
  • Location: UK

Posted 26 December 2020 - 10:41

Couldn't resist trying this. I used a flashgun as the light source, which allows deeper penetration. Filter was a Midwest Optical BN 850.

First, at the knuckle-end of the hand, similar to your original shot:

Attached Image: DSC00034 LoRes.jpg

Now the palm nearer to the wrist - you can se the glint off the wrist watch at upper-left.

Attached Image: DSC00031-1 LoRes.jpg

And finally the wrist itself, where the watch would normally be. Light leakage around the side of the wrist - I may try this again with a mask to prevent this.

Attached Image: DSC00036. LoRes.jpg
Bernard Foot

#14 Stefano

    Member

  • Members(+)
  • 1,800 posts
  • Location: Italy

Posted 26 December 2020 - 10:56

Wow Bernard, those are very nice results! I don't know about my hands, but yours look like to be better than Andy's. The wrist is quite thick, I don't know if I can do that with mines.

#15 Andy Perrin

    Member

  • Members+G
  • 4,093 posts
  • Location: United States

Posted 26 December 2020 - 22:00

Nice! I think the more columnated the light is, the better. You could try with sunshine if you ever have any! But I was thinking one day I'd try with an IR laser.

#16 Stefano

    Member

  • Members(+)
  • 1,800 posts
  • Location: Italy

Posted 26 December 2020 - 22:34

880 nm should work well, as I have suggested above. Your tissues will diffuse the light. I don’t think there is a difference between collimated and un-collimated, but I think having all light going inside you with no spill improves contrast.

#17 Andy Perrin

    Member

  • Members+G
  • 4,093 posts
  • Location: United States

Posted 26 December 2020 - 23:34

Stefano, having the light moving in one direction should definitely help if the idea is to get sharp shadows. Consider the difference between your shadow on a cloudy day vs a sunny day.

#18 Stefano

    Member

  • Members(+)
  • 1,800 posts
  • Location: Italy

Posted 27 December 2020 - 08:48

Yes, I get it, but your skin, muscles and maybe even bones will act as a powerful diffuser. If you want to image 1 mm of skin collimated light will stay more or less collimated, but with several centimeters light will come from all directions at the end. That's also why you don't see bones, even if part of the reason could be that bones seem to be partially transparent to red/IR: https://youtu.be/awADEuv5vWY (go at 4:55 to see a demonstration).

#19 Andy Perrin

    Member

  • Members+G
  • 4,093 posts
  • Location: United States

Posted 27 December 2020 - 09:13

I don't think it's such a great diffuser as you're saying. My book "Biomedical Optics: Principles and Imaging" by Lihong Wang and Hsin-I Wu, in the chapter Monte Carlo Modeling of Photon Transport In Biological Tissue, says the anisotropy factor g, which has a value between -1 and 1 where 0 is isotropic (photon shoots off in all directions equally), is near 0.9:
Attached Image: post-94-0-87344900-1609060188.jpg
So most of the light goes straight ahead!

It's true that this is on a per-collision basis, but even so, I would expect to see SOME advantage to using a columnated source.

Edited by Andy Perrin, 12 January 2021 - 03:28.


#20 Bernard Foot

    Bernard Foot

  • Members+G
  • 718 posts
  • Location: UK

Posted 30 December 2020 - 09:16

Looks like this is not just for IR:

https://www.ebay.co....P4AAOSwmo5Zm61~
Bernard Foot