Tag Archives: retinex theory

Why are our veins blue?

I was asked this, and it’s pretty good question. As I have an MD/PhD myself I figure that I’m fairly reasonably equipped to answer this. Although, I do feel that I need to mention that while having gone to medical school may provide essential knowledge to debunking the popular misconceptions, it won’t actually provide the answer we’re looking for. I’ll explain a bit later where we actually find this elusive little answer. First, the misconceptions.

So why are our veins blue? The first and most common answer that people give is “well, it’s deoxygenation. When blood isn’t exposed to oxygen it’s blue.” But despite its popularity, this one is actually quite wrong. The only thing that I know of with blue blood is something like a horseshoe crab whose blood is bound to copper – it clots significantly faster, so it’s often used for testing purposes.

Deoxygenated venous blood is actually a dark maroon, while oxygenated arterial blood is bright red. You can debunk this little myth yourself if you look at a thin membrane that has blood vessels. Some examples include: the eyes, eyelids, underside of the tongue, and the scrotum (if you’re a boy or happen to have one available). You’ll clearly notice that these are indeed red. So that one’s out.

Okay so if our blood isn’t blue… then maybe our veins are! This is the next most common misconception. Anyone that has ever dissected anything, or even viewed pictures thereof can tell you that veins are most definitely not blue, nor green, nor even purple. Though if you’ve seen a medical diagram it’s a reasonable assumption considering they’re almost always colored blue. But they’re not. If you google image “open surgery” you can see for yourself, though I warn you… these pictures are not for the squeamish. Veins are red. Though, I do know that veins in the brain can appear blue.

Okay so then if our blood isn’t blue and our veins aren’t blue, then why the heck do our veins look blue?!

Believe it or not, this actually has less to do with your internal biology and more to do with our perception of light. Though, it does also have to do with the light absorbing properties of our blood and our skin. Specifically, how light is absorbed by our blood and how it is reflected by our skin. And this is where we get into our answer. For this, I go to the expertise of Dr. Alwin Kienle and associates who specialize in the study of optics and photonics.

What it essentially comes down to is that skin generally reflects light rather than absorbs it while blood absorbs all wavelengths of light, albeit less of the red spectrum.

And well, the jist of it is that the blue wavelength isn’t as good at penetrating skin as the red spectrum. If you have a very thin membrane like mentioned above, then blue light will be absorbed and the vessel will appear red.

But if we go deeper then not nearly as much light is absorbed and much less of the blue/green side of the spectrum than the red. And thus, blue veins are born. It essentially just comes down to optical properties. If you’re more interested in the technical details of, you can grab a pdf on Dr. Kienle and associates research here. I’ll quote below for you the summary of their findings:

To summarize, the reason for the bluish color of a vein is not greater remission of blue light compared with red light; rather, it is the greater decrease in the red remission above the vessel compared to its surroundings than the corresponding effect in the blue. At first it seems astonishing that red light is more attenuated above the vessel than blue light, since, as Table 1 shows, the absorption of blood is much less in the red than in the blue. This is the result of the spectral characteristics of light propagation in tissue. Blue light does not penetrate as deeply into tissue as red light. Therefore, if the vessel is sufficiently deep, the reflectance in the blue will be affected to a lesser extent. Deoxygenated venous blood has a greater absorption coefficient than oxygenated arterial blood in the red spectral region, and this difference of two, rather small, values is amplified because of the long path length of red light in scattering tissue. As a result, veins are more likely to look blue than arteries at the same diameter and depth. Often arteries are not seen at all because they are generally smaller than veins and have thicker vessel walls. It has been shown that a small vessel will look red when close to the surface. However, if a superficial vessel is large it can still look bluish, particularly in the case of the vein. On the other hand, if the depth of a vessel is large, even red remitted light will not be influenced by the vessel, and it will not be seen. We note that, for the calculations here, we assumed an oxygen saturation of 50% for venous blood. This is somewhat arbitrary, but other possible realistic values do not change the conclusions.

As shown in Fig. 8, even above the vein more red than blue light is remitted. Thus, for a complete explanation of the perceived color of the vessel one needs the retinex theory. With the retinex theory the color can be determined by the relative intensities at the three wave bands from a particular scene compared to the surrounding area. The intensities of these wave bands are weighted by spectral functions that represent the human spectral vision. In this study we used single wavelengths that are representative of these spectral functions, and, therefore, a retinex color three space based on selected wavelengths should be applied. However, we chose typical wavelengths for the long-wave, middle-wave, and short-wave regions and made qualitative estimates of the colors. Therefore, we believe that we are justified in applying the usual retinex three-color space. In one example, we calculated the remission at several wavelengths and used the spectral sensitivity of the cones to show that the approach of using only three single wavelengths is justified.

It is interesting to speculate whether retinex theory is necessary for other color perception issues in medicine 1e.g., the color of port-wine stains, vitiligo lesions, blue nevi, age spots, eyes, hair2, or whether the perceived color can be simply related to the absorption spectra of hromophores, possibly modified by the presence of light scattering and measured with a reflectometer. If the problem of vessel color is any guide, it seems that retinex theory may provide an essential step in the description of color perception.

What I don’t quite understand however is how hypoxia is causing the cyanosis of skin (acute arterial thrombosis for instance).

My only guess was perhaps the onset of necrosis such as seen in ischemia, or perhaps the change to venous blood is enough to alter the optical properties of the entire effected region to a blue hue, as it has a greater optical effect than arterial blood. Or perhaps some combination of the two.

Having had a chance to talk to Dr. Kienle, he acknowledged as much albeit mentioning that for a thorough investigation one would have to look more closely to the haemodynamics (and geometrical changes caused by) of cyanosis. Although, he did laugh in saying that he was Physicist, not a Dermatologist. 😛

Tagged , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , ,