Rethinking the Process of Vision

I had discussed this paper previously but recently have had a chance to revisit it at the prodding of my son. If the idea that a “mosaic of three types of cones” forms the foveal region of the retina were true then my concept of light interaction with the retina would have no merit. BUT, IN A CURIOUS WAY INDEED, I BELIEVE THAT THE RESULTS PRESENTED IN THIS PAPER BESIDES REPRESENTING A FACTUAL DESCRIPTION OF LIGHT INTERACTION WITH THE FOVEA SIMULTANEOUSLY AND EXACTLY SUPPORT MY CONCEPT! I will explain and suggest an extension of the RW experiment that might be done to clear this issue.

Parenthetically, the authors seem to assume that the retina forms the image plane of the optical system of the eye..and therefore that its light detection elements must form some sort of analogue of color photographic film or the pixel array of a silicon imaging device. I will discuss this in greater detail later but this view (in my opinion a wrongful view!) permeates the literature of vision science and forms the basis for the results presented in this paper (I demonstrate that the retina forms a diffractometric Fourier transform plane and not an intensity-only sensitive image plane of the eye),

The assumptions of the Roorda Williams letter to Nature:

1.) RW undertake heroic measures to optically visualize individual color detection sites of the retina – that they assume are solely cones. Note that such entities, whatever they are, are small indeed on the retinal surface having dimensions approaching the wavelength of visible light – and thus, the limits of light microscopy. RW employ multiple measurements and complex image registration methods to finally define the final precise imagery. This results in some “blur” admitted by the authors. The question is how much blur? And what is the spatial resolution of the final images? The light detection elements certainly appear as large single diameter cones……but are they?

2.) The methodology of the paper uses wavelength selective bleaching to render the different light detection sites visible. Bleaching is nature’s way of preventing overload and resultant damage to the light detection centers of the retina. It involves in my view a reorientation of the retinal chromophores (whose role I discussed in a recent comment) away from the direction of incident light (this direction, in accord with the measured dichroic orientation of these chromophores, is transverse to the direction of incident light). This method of visualizing individual light detection sites, however, seems appropriate.

3.) It is stated that the images are of the cone mosaic as it appears at one degree of retinal angle (i.e., the parafoveal region). If this is accurate, at this angle, according to George Wald (J. Opt. Soc. Am , Vol.57, No.11, November 1967), quoting Osterberg’s data in his Figure 9 on p.1295, indicates that at one degree cone and rod densities are nearly equal. Precisely, cone density is ~45,000 / mm2 and rod density is ~35,000 / mm2. This is the region of the retina where rods are starting to intrude upon the orderly hexagonal array of cones. One begins to find an increasing number of isolated cone-rod appositions (and even more isolated rod-rod appositions to be discussed below). One degree of retinal angle is a crucial region as cone and rod densities are falling and rising precipitously at this angle. Rod density continues to increase until at 7-8 degrees there are sufficient rods to completely surround each cone (which they do) and define the geometric midband sensitivity point of the retina.

So what are the wavelength detection centers that Roorda and Williams visualize? I would guess that the spatial resolution that they are able to obtain is not sufficient to visualize rod-cone appositions, or, alternatively, blurs them to the extent that they appear similar in size to the “red” interaction sites. The authors make a point about the randomness of the green sites and this is exactly as it should be at this retinal angle following the logic that I pursue.

And what of the infamous blue detection sites shown in Figure 3, that I claim do not correspond at all to “blue sensitive cones”? I would propose that these actually are blue detection sites ….but.. that these sites arise from the few rod-rod appositions that appear as rods continue to crowd into the cone array. The low density of these sites is in exact consonance with this logic. Additionaly is support of this argument, blue sites are found adjacent to, or associated with (except for two outlyers), the green sensitive centers. These would be associated with the few statistically present rod-rod appositions.

I must add that Curcio (“Human Receptor Topography”, Journal of Comparative Neurology, 292:497-523, 1990) seems to disagree with Wald/Osterberg stating in the Abstract “In the fovea, the average horizontal diameter of the rod-free zone is 0.350 mm (1.25 degrees)”. This number seems to have been arrived at using the method described on p.503 of the paper titled “Conversion to Visual Degrees” which is more complex than I choose to manage..so I will leave this point for others to sort out.

Just another thought in regard to the Curcio paper..I have been taken to task for “not understanding” data from this paper wherein it is reported that the size of cones is not constant but rather varies between individuals. I would certainly believe that this is so..genetically mediated. I have proposed rather clearly that the effect of varying the size of cones would be..making the crucial assumption that the refractive properties of the eye are not changed … to offset the geometrically determined midband point (7-8 degrees). Assuming that midband 550 nm wavelengths arrive at the same location, any slight offset would lead to color variant vision or “color blindness” (see Land’s work), Also, the long wavlength limit of the visual response would be extended for individuals having larger cone receptors. Is this measurable?

IN SUMMARY..I WOULD MAKE A PREDICTION AND A REQUEST. IF THE ROORDA GROUP WOULD EXTEND THIS TYPE OF MEASUREMENT TO LARGER RETINAL ANGLES I BELIEVE THAT THEY WOULD FIND THAT THE DENSITY OF GREEN SENSITIVE CENTERS WOULD CONTINUALLY INCREASE UNTIL AT SEVEN TO EIGHT DEGREES REPONSE WOULD BE TOTALLY GREEN. IMAGE RESOLUTION COULD BE RELAXED A BIT AS I BELIEVE THAT IT WOULD BECOME QUICKLY APPARENT THAT THE BEHAVIOR THAT I PREDICT WOULD BE PRESENT. I WISH THAT THIS MEASUREMENT COULD BE UNDERTAKEN.

GCH

This entry was posted on Monday, January 23rd, 2006 at 9:26 am and is filed under Running Commentary . You can follow any responses to this entry through the RSS 2.0 feed. You can leave a comment, or trackback from your own site.