Rethinking the Process of Vision

I will begin by repeating the quote from Albert Rose’s treatise “Vision Human and Electronic” (Plenum Press) p.29 (emphasis mine):

“In the scheme of evolution, vision has an almost unique role. One can conceive, for example, that further evolutionary developments might lead to a larger brain capacity, a more evolved nervous system, or a variety of enhancements of current functions. It is not conceivable that the sensitivity of the visual process can be significantly enhanced. The visual process is at an absolute terminal point in the evolutionary chain. To the extent that the visual process now succeeds in counting each absorbed photon….”

This ability of the vision process has never been explained and seems never to have elicited interest to the vision field.

A summary:

I have been fascinated with the idea that in looking at an object only a foot in front of me I am “looking into the past”. Light reflected from that object to my eye requires a nanosecond (10^-9 sec.) to arrive at my retina so that I can tell only what happened to that object a nanosecond ago. I can never see it “instantly”. Inherent in this mental exercise is the idea that my eye “passively” receives this image on the retina subsequently transiting it to the brain, my consciousness etc. with time not being considered in the process.

But time is inherent in the process of vision.

I will proceed in the next few days to lay out an entirely new logic for the process of vision that follows from a physical explanation for the eye’s ability to sense single quanta. To achieve this level of sensitivity the vast retinal array, must be considered to be composed of individually acting receptors or “light detection devices” (defined as the receptor appositions that I postulate) each of which can detect and generate a measurable signal from the interaction of single quanta (the “quantum limit”). I will show that receptor appositions form such devices with a reaction time in the picosecond (10^-12 sec) time domain (with reaction time being defined as the time between light “stopping” in the mass of the detection center and the generation of an electrically measurable signal for subsequent use in the image formation process).

The retinal array (comprising 200,000+ independent devices in the fovea alone) of quantal signals thus derived forms a near instantaneous image in this fast time domain on the retinal surface.. This “entire quantal image” continuously “updates” a spatially and temporally coherent image stream (?) being transferred via the optic nerve. to the visual centers of the brain The optic nerve is composed of 1.9 million separate fibers appearing as a coherent (i.e., imaging) fiber bundle Transmission of the image through the optic nerve is ionic and therefore slow (10^-3 sec or millisecond) time. This slow mode of transport combined with the finite length of the optic nerve combine to “slow the visual image to human scale proportions”, i.e., compatible with the time scale in which nerve signals can reach extremities etc. I propose that this time is the genesis of the incorrect idea that the “reaction time of the eye” is from 0.1 to 0.2 seconds.

(It intrigues me that evolution must have defined the length of the optic nerve, i.e., the physical distance from eye to brain to bring the “instantaneous” quantal visual image into a time scale useful to human proportions!)

Therefore, I believe that the visual image is actually detected and processed in two separate time scales : 1. initially )in the picosecond or fast time domain in which the quantal image is latent on the retina and, 2. a secondary, millisecond, time domain introduced by transport through the optic nerve to transfer the intact  image into a slower time useful to the human system.

I propose therefore that the eye, in contrast to what been historically believed, actually views a “progressing ephemeral instant” of time that is subsequently slowed for compatibility with the human system. The former is of interest to physics with the latter slower time domain  of concern  to biology.

It should be obvious that even within the process of vision itself we are “looking back in time”. We are doomed by the design of the human system itself to always view the past! Might this be the underlying reason for the physics conundrum of a direction or “arrow” of time?

Finally perhaps this explanation provides some insight into the “instant” that has been of interest to philosophers over the ages (and the physicist Julian Barbour). And moreover……there are some four orders of magnitude between the picosecond time domain invoked here and the frequency of light. What information might be contained in that span of time?

This is meant to be a quickly written summary written on a Friday afternoon. I will follow in the coming days with supporting data as to the speed of response requirement of retinal detection devices. Etc.

GCH / 9/22/06/revised 9/23/06

I dislike using this word in a scientific context although one frequently sees it used to describe quantum physics implying that no physical understanding of the effects can be reached (an oft quoted statement is that “the mathematics work, don’t bother with any physical interpretation”, i.e., an understanding expressed in words). David Bohm spent a great deal of time in his later years on this subject believing that “underlying variables” existed that might lead to a physical explanation for quantum effects. Bohm concluded, interestingly, that an entirely new language had to be devised to explain these effects. This is how far things have gone. I do not believe that anything is beyond physical understanding. But I find myself thinking of the dreaded word in conjunction with the vision process.

Following from the last posted Comment regarding the quantum limit of vision, it occurred to me that an entirely new paradigm for the process is possible ( or evident?). I suddenly came to visualize the brain (the “mind”, “I”,“ consciousness”) as “ looking out” from the visual cortex of the brain through the coherent bundle of fibers of the optic nerve to a high speed, continuously changing, view of the external world impressed on the retina of the eye. I had termed this view, being in picosecond time, as “instantaneous” but upon reflection it should more properly be termed “near instantaneous”. This view is in contrast to the traditional one (the “other way”) where an image is impressed on the retina of a passive “camera-like” eye that views the external scene and transmits that image information to the brain for “processing”. Might the processing precede - or ultimately be found to be coincident with - the external viewing?

And what might follow here are my preciously expressed comments on vision as an “active” rather than the assumed “passive” process with the eye (at the level of single quanta) actively “interrogating” or “in communication with” external reality? (See my comment of Feb.12,2006 “Previous Thoughts on Consciousness and Time”). In this context I have proposed that the retina possesses the properties of a phase conjugate mirror which would be a necessary component of an “active system”.

And therein I believe lies the crux of the matter - the time domain. My proposal, based on what I believe are sound physical principles, brings the vision process closer to the “instant”, or what might be termed, the “quantum instant”. I would note that even if the vision receptors that I propose are able to react in picosecond (10-12 sec) time there are left some two to three orders of magnitude between this time and the frequency of light. Might even more information be contained in this even faster time domain?

GCH
9/14/06

A quote from Albert Rose’s book “Vision Human and Electronic” (Plenum Press) p.29:

“In the scheme of evolution, vision has an almost unique role. One can conceive, for example, that further evolutionary developments might lead to a larger brain capacity, a more evolved nervous system, or a variety of enhancements of current functions. It is not conceivable that the sensitivity of the visual process can be significantly enhanced. The visual process is at an absolute terminal point in the evolutionary chain. To the extent that the visual process now succeeds in counting each absorbed photon….”

I believe that an understanding of how the eye detects at the quantum limit, i.e., is able to “count” individual photons or quanta of light will lead (finally!) to a fundamental understanding of the vision process. The single credible explanation for the light sensitivity of the eye extending to the ultimate level of sensitivity - the quantum limit - must invoke very fast time of the order of 10>-12 sec in the detection process . This time is, in essence, involves the “stopping time” of light (or the photon) in the absorbing mass constituted by the retinal outer segments and the time-associated reduction of noise that competes with signal.
____________________________________________________

A correspondent provides references to femtosecond chemical measurements that support the above:

“As established experimentally, the quantum of light impinging on a rod and being absorbed by the molecule of a rhodopsin, “launches” the photochemical reaction (occurrence of electric impulse) for 0.2 picoseconds !,( El-Sayed M.A, Tanaka I, Molin Y. (1995). Ultra Fast Processes in Chemistry and Photobiology, Blackwell Science., Kandori1, H, Shichida, Y., Yoshizawa, T., (2001). Photoisomerization in Rhodopsin, Biochemistry 66) Curiously, that is exactly the time required for a quantum of light to reach a photosensitive cell, as the length of a rod is about 60 μm”

__________________________________________________________________

In explaining the quantum limit capability, it must be realized at the outset that the retina of the eye as taught in this work should be thought of as a circular array of generic “quantum wires” spaced according to physics principles to detect three discrete electromagnetic wavelengths. At the point where light interacts (the length of the outer receptor segments) this array forms a huge number (200,00+ in the fovea alone) of high speed, probably picosecond (or 10>-12 sec.) responsive, optical detection devices. These devices are formed between quantum wires as I have proposed (see “Diagram of the Fundamental Light Detection Device…” on the web page). Each of these devices independently is sufficiently sensitive and free of noise to detect single quanta of light. One envisions then a dense, circularly symmetric, light detecting image-forming array with each element (“pixel”) able to detect (“count”) single light quanta and form an electrically useful signal that is transferred forward in the vision process. In electronic light detection technology this would be termed the “first stage” of the detector.

The total image in the above is therefore present in very fast time - probably ranging from the 10>-12 sec to 10>-9 sec (the latter to allow for retinal processing) time domain. This image thus is ‘there” and being continuously “updated” - and “viewed” by the visual centers of the brain - in this time frame. More about this next.

(I have completed a preliminary analysis that I will present shortly showing that, in the visual scene, sufficient photons are incident on the retina in this time frame to produce a statistically proper image. Remember that the retina is at the Fourier or focal point of the optical system of the eye. The pupillary constrictive system of the eye evolved to continually reduce photon intensity on the retina to biologically tolerable values. There is no dearth of photons incident on the retina!)

Then …the “instantaneously updated” visual image is transmitted via the 1.9M fibers of the optic nerve to the visual centers of the brain. This conduction path is centimeters (how many?) long and the conduction is known to be ionic. Ionic mechanisms are slow and might generally be said to lie in the millisecond (10>-3 sec.) range. I would guess that evolution employed the combination of the finite length of the optic nerve (the distance from the eye to the brain centers involved) and the slow ionic conduction mechanism to arrive at a proper “reaction time” consistent with muscle movement etc. of the human system. In a sense the visual image had to “slowed down” to achieve this match. This is the oft-quoted “reaction time” of the eye approximating 0.2 sec. I propose that this time actually contains two separate times as discussed herein and that the measured “reaction time” of the eye has nothing to do with the process of forming the visual image.

We thus now have the picture of an “instantaneously updated image” that is both spatially and temporally coherent being transmitted through the coherently arranged fibers of the optic nerve to the brain - with the image itself appearing at the visual cortex of the brain!

With this realization one is struck with the question: just what is viewing what here? The traditional view of a passive eye/retina transmitting images to the visual cortex of the brain, OR the brain “looking out” through the coherent optic nerve bundle to the retina acting as a “window on the external world” with it’s imagery changing and being continuously updated in very fast time. Questions?

GCH
9/13/06

(I am posting this Comment in very preliminary form..will workmore on it shortly / GCH)

From the Albert Rose reference discussed previously (“VISION HUMAN AND ELECTRONIC”, Plenum Press) the first few lines from Chapter I are worth quoting:

“It would be difficult to find a more cogent confrontation between physics and biology than in the visual process. …”

(Also, a quote that I have come upon (that will certainly be disputed as outdated….but?) by the English physicist Ernest Rutherford who is credited as the discoverer of nuclear energy: “All science is either physics or stamp collecting.”)

____________________________________________________

I have come to believe that answering the never explained question as to how the eye detects light at the quantum limit is the most fundamental question that can be asked of vision science. It’s answer will open up entirely new lines of thought about vision.

Again quoting Rose (p.29 of the above reference):

“In the scheme of evolution, vision has an almost unique role. One can conceive, for example, that further evolutionary development might lead to a larger brain capacity, a more involved nervous system, or a variety of enhancements of current functions. It is not conceivable that the sensitivity of the visual process can be significantly enhanced. The visual process is at an absolute terminal point in the evoutionary chain….”.

I think that the meaning of the term “quantum limit” should be obvious to most. It implies the specific detection and registration in the visual chain of individual particles of light - photons.

(NOTE: I would substitute the term “quantized interaction” for “photon”. As shown in this work the eye actually evolved to detect light as the wave of classical physics adjacent to quantized electrons (the absorbing mass or retinal receptor). My term does not in any way change the meaning of the interaction - it can be variously termed as being “quantized” or “wave defined”. I will continue to use “photon” however and the reader will understand).

To be clear, Rose recounts the many references documenting that the eye is able to sense individual photons renging from two to a few hundreds in number -remembering that each detection is an individual event. I do not believe that there can be any doubt about this.

Rose uses the term to “count” individual photons. Again quoting from p.2:

“The quantum character of light is a hard constraint. Nature could, in a physical sense, do no more and, in a survival sense, do no less than devise a photon counter…..” (emphasis mine).

Now, how does the eye manage to accomplish this? The term “counter” implies segregation of single photon interaction “events” in time, i.e., that they are distinctly separated from each other in the detection process.

It seems appropriate to recycle here some prose that I had written in my Comment of 8/08/06:

“Rose, following from his background in electronics (he was the inventor of the Vidicon tube that had a major impact on the development of television!), led him to propose that an amplification (or electronic ‘gain’) of a million or a factor of 10>6 must somehow follow the original photon interaction signaland be operative in thebiological eye. This is one way to express the system requirement for single photon detection but there are others. Rose could provide no idea as to how the biological eye might accomplish such amplification.

I would note that the limits of contemporary photonic technology allow detection (and counting) of single photons by reducing the temperature of the detector to a few degrees above absolute zero or, at room temperature, by using the electronic gain induced by thousands of volts (as Rose was really proposing) in a vacuum photomultplier tube. The eye performs this single photon detecton function in a biological structure above room temperature - at body temperature!

In photonic technology any signal detection process involves two factors – the inherent (or original) signal level balanced against the random electronic noise that might interfere with (or obscure) that signal level in this time frame in the overall detection system. One must achieve an adequate ‘signal-to-noise ratio’ for unambiguous detection of an event. Accepting the original signal level and reducing whatever noise might be present in that time frame is another strategy for low level signal such as single photon detection.

First, what constitutes the ‘fundamental signal’ when a light wave interacts with an ‘antenna’ absorbing site on the retina as I propose. Such centers (or ‘devices’) are comprised of two adjacent receptors (or more precisely, the quantum- confined electron spaces that the receptors provide) and the wavelength-defining space between them. Light interacts as the wave of classical physics in this central space imparting a different amount of energy to each receptor (with the difference a function of the direction that the light ray entered the device) The energy thus imparted then mechanically/electrically effects an isomeric transition of the retinal molecules contained within each receptor.This then constitutes the initial ‘signal’ that must subsequently processed to provide the visual image.

So what constitutes this most fundamental light detecting devic? I have diagrammed such a deice on my web page (“Important Material”, “Diagram of the Fundamental….). First, the device is is extremely small in consonance with speed of response requirements having a cross section smaller than light wavelength and length the dimension of the receptor outer segments (~50 microns). The time response of any electronic device is dependent upon its dimensions. This is the reason why microcircuitry technology is always seeking smaller ‘feature size’ (i.e., smaller devices) to satisfy the ever present need for increased speed. The charge stopping time (or the time when the signal is ‘there’) I estimate for a device of this dimension to be of the order of 10>-12 seconds (or a picosecond). This is an important number and note that it has nothing to do with subsequent, much slower, “biological signal processing” or transduction time in sub-retinal circuitry or in transit to the brain.

Then, the retina is composed of an array of these very fast devices. I would contend that the initial (and complete) information to form the visual image is present on the retina in some time approximating picosecond ( 10>-12 second) time.

Now back to the subject of ‘signal to noise’ ratio:

Every radiation detection system has an inherent ‘time constant’ that can be considered a ‘time window’ that is ‘left open’ long enough for the signal to be recognized as such and, for example, be electronically transferable to an ‘amplification stage’ for further processing. One property of this ‘time window’ is that electronic noise in the system (i.e., random electron events coming in time) are integrated. The longer the time window is left open the higher the noise that is accumulated within the window. This ultimately constitutes the noise level that obscures the signal. The fundamental rule is therefore that one wants the overall act of detection to be accomplished in as short a time as possible, ideally in a time approaching the ‘signal-is-there’ time

A number of us published a paper on the subject of this optimization in 1976

__________________________________________________________________

Tove,P.A., Cho,Z.H., Huth,G.C., “The Importance of the Time Scale in Radiation Detection Exemplified by Comparing Conventional and Avalanche Semiconductor Detectors”, Physica Scripta, 13, 83-92, (1976).
__________________________________________________________________

I will not go into details of this work here but only to note that the relevant time constant of detection systems of the time was in the microsecond (10>-6 sec.) range. Thus the ‘time window’ was left open for some six orders of magnitude longer than desirable accumulating noise and raising the level of signal that could be detected.

Now to the eye. Validating the 1935 retinal measurements of Osterberg, I have shown that the retina forms the focal or Fourier plane of the converging lens system of the eye. There really can be no doubt about this. It is a property of this plane, as Feynman so cogently explained, that light rays are brought into ‘time convergence’ at this point with light rays that pass through the center of the lens being slowed to allow refracted rays from the thinner outer part of the lens to ‘catch up’. I have written about this elsewhere on this page.

The central fovea of the retina can be used as the prime example with all light rays entering the eye brought to such a time convergence at this point. I have written that such time ‘convergence’ might be defined as ‘zero time’ (or as near as quantum limitations will allow) or what might be thought of as the ‘the absence of time’ or the ‘instant’ of time. Thus the visual ‘signal’ comes as close to the optimum time described above as one might imagine.

The light detection devices of the retina are thus seen to be consistent with the time requirements of Fourier plane imaging so, in fact, the entire content of information necessary to form the visual image is ‘there’ in picosecond time.

I believe that the invocation of the picosecond time domain answers the question as to how the eye detects single quantum events. It is not that the signal is somehow amplified but rather that the contravening noise level is reduced.

With the above insight as to the time regime, it occurs to me that the terminology of ‘detecting single quantum events’ is not strictly correct. This capability might more properly be termed that the ability of the eye to ‘discriminate in time the interaction of single quantum events’. As I have shown the retina is capable of logic function in the picosecond time domain. This opens up many new avenues of thought.

For example, I believe that analysis will show that at normal light levels entering the pupillarily (?) constricted eye the individual devices of the retina will be able to separate individual quantum events. I am working on this.

END FOR NOW