Do I have good vision?

A certain level of good vision is required for many daily activities, some of which may have a legal requirement, such as driving. However, ‘good vision’ is a variable measure and depends on the set standard. Perhaps it is more valuable to have a concept of the limits of our visual capabilities. Vision can be considered in two ways: the optical requirements to achieve an image (i.e. refraction of light by the eye to focus the image on the retina, also known as physiological optics) and the neural processing of visual stimuli by the retina and the brain. The visual process is initiated by the detection of a light signal by photoreceptor cells in the outer retina. Photoreceptor cells convert light energy to an electric stimulus, which is then transmitted to the bipolar cells and onwards to the ganglion cells in the retina (see Ch. 1 , p. 46 ). The information is further transmitted in the axons of these cells (the optic nerves which, after 50% crossover in the optic chiasma, become the optic tracts) to the visual thalamic organ, the lateral geniculate nucleus (LGN). Synaptic contact with neurones in the LGN that project to the cerebral cortex permits onward transmission of the signals via optic radiation to the visual or striate cortex (V1), where they interact with many other neuronal connections from visual cortical cells in the prestriate cortex (V3–V5), and where parcelling out and processing of the signals takes place to build up the final perceived visual image. Input is also received by the visual cortex from many other areas, particularly those controlling general motor function and eye movement, cerebellar and spatial sense, memory and many other functions located in prefrontal cortex. This produces some of what is known as ‘top down’ modulation of visual responses whereby signals received and interpreted in the visual cortex can be influenced by input to the final image (the perceived image) from other areas, such as visual area 4 (V4) which combines elements of object recognition with visual attention (see eFigs 5-1 and 5-8 ).

The brain is constantly shifting its position to meet the requirements of changes in behaviour. It continuously adapts its processing machinery to behavioural demands. Information is therefore transformed, modulated and rechannelled through different neural cortical wiring circuits which have been revealed in series of novel experiments in various animal models which can assess small changes in activity occurring at low frequency. Harris and Thiele suggest that processes involved in selective attention are similar to those involved in state changes; these are summarized in eFig. 5-1 ( ). They include increased activity of cortical neuromodulatory afferents (red (cholinergic), blue (serotonergic) and green (noradrenergic) arrows) which causes a general desynchronization and reduction in spontaneous fluctuation, but may lack the spatial selectivity to desynchronize the patch of cortex representing the attended stimulus. Focused glutamatergic inputs arising from feedback connections could provide this specificity (yellow arrows), causing enhanced desynchronization and sensory responses in the regions of cortex activated by the object of attention. The yellow circle in the visual display indicates the focus of attention, which affects processing in thalamic and cortical areas at specific locations (indicated by the yellow patches). The distorted replication of the visual world in the different areas illustrates (approximately) the known retinotopic organization of these different areas.

Processes involved in attention are thought to be similar to changes occurring in state changes in cortical activity for specific functions. Neuromodulatory effects are identified by the red, blue and green arrows and are explained in the text.

(From .)

In biophysical terms a photoreceptor is capable of detecting a single photon of light (see Ch. 4 , p. 261 ), but in practice what are the limits of detection of a visual stimulus? This depends on the nature of the stimulus and the nature of the ambient conditions in which it is presented. Sensing light is a function of all regions of the retina but the foveal region is specialized for high spatial resolution (visual acuity) and colour detection, served by the small ‘midget’, slow-transmitting ganglion cells (the parvocellular or P system). In contrast, luminance and motion detection are served by the large, fast-transmitting ganglion cells (the magnocellular or M system) that dominate the remaining retina and thus incorporate the entire visual field. According to Barlow’s single neurone doctrine, it should take only one neurone to detect a visual stimulus ( ); however, the question is whether the signal received from the one rod in a thousand stimulated in the dark is sufficient to activate this neurone. Psychophysical studies have shown that both the luminance and colour thresholds for vision are different by orders of magnitude for P neurones between monkeys and humans, suggesting that more than one neurone is involved in detecting a light stimulus. In fact, continuous pooling of information occurs both in the excitatory and inhibitory neuronal activity that is present at all times and that, after a visual stimulus, the changes in the response rate of many neurones are ‘sampled’ by the brain until they reach a certain threshold level, at which point they register and the stimulus is ‘recognized’ (Hurlbert and Derrington, 1993). This perhaps explains how we can sometimes look at an object and yet not ‘see’ it; furthermore, these psychophysical considerations are highly relevant to methods for testing vision, for instance with regard to setting luminance thresholds for studies of visual fields using small transient targets.

Whether or not we have good vision at any moment in time depends on our level of awareness, consciousness and attention to visual stimuli which have many properties such as depth, shape, form, colour texture and more besides, each with its own rate of detection/discrimination.

Sensing colour

Cone photoreceptors are built to sense colour, which they do through cone opsin proteins. There are three types in humans: long (L, red), medium (M, green) and short (S, blue) wavelength cones, each with its own specific opsin (see p. 302 ). Early colour matching experiments in which the colour of a test stimulus is matched by adding together stimuli composed of the three primary colours verified that three colours were all that was required to detect the full spectrum of white light (see below, Colorimetry, p. 302 ). This is known as the trichromatic theory of colour vision. Each colour has properties such as hue and chromaticity.

There are many hues but only three primary colours

Hue is an idealized term for the colour produced by light of a single wavelength. In spite of having only three cone photoreceptors, we are able to distinguish many hues of colour, e.g. lilac and violet. It is therefore clear that any single colour is recognized by a mixture of the three primaries and that there must be overlap in the spectral sensitivity for each primary colour (see below). Theoretically it should be possible to produce light of a single wavelength using a narrow slit on a device such as a monochrometer, but photoreceptor sensitivity is also subject to the intensity of the light, and narrow wavebands of this degree of selectivity are not sufficiently intense to produce a stimulus.

The hue-discrimination curve ( Fig. 5-2 ) describes the physiological limits at which a shift in wavelength can be discriminated as a change in colour. It was derived empirically by using fixed amplitude selective wavelength stimuli (1.5 cycles/300 nm), which found that there were two peaks of discrimination, one in the yellow/orange and a second in the blue/violet range. Monochromatic light therefore is not a practical reality; most colours are in fact tints, i.e. they are unsaturated hues, the degree of unsaturation being determined by the amount of additional white light they require to match them to a hue.

Hue-discrimination curve comparing wavelength discrimination ( y -axis) with changing wavelength ( x -axis). Discrimination of hues varies for any given wavelength, being best at 455 and 535 nm.

Chromaticity is semiquantified ‘colouredness’

Chromaticity refers to ‘colouredness’ and depends on hue, saturation and intensity of light (luminosity). Indeed, hue itself is not independent of the luminosity of the stimulus and chromatic shifts occur as the intensity increases until all hues appear yellow–white (the Bezold–Brucke phenomenon) or as the intensity decreases, when all hues appear achromatic (the Purkinje shift; see below). Any colour can thus be matched by a mixture of the three primary colours plus or minus a proportion of white light to account for unsaturation; these are formally described in the chromaticity chart and can be determined at different levels of lower or higher colour metrics (discrimination) ( Fig. 5-3 ). The International Commission on Illumination (CIE) has developed standard colour ‘observers’ (colorimeters) which have proved valuable in defining ‘true’ colours but do not fully take into account the minimally perceptible differences in colour discrimination (just noticeable differences, JND) based on aspects of photoreception affected by ‘background noise’.

Lower versus higher colour metrics. The chromaticity diagram is long established in the field of colour discrimination. It represents the laws of colour mixing in terms of (x,y). This is the domain of lower colour metrics. The straight lines through x = 0.305, y = 0.323 indicate that this white colour can be obtained, for instance, by mixing, in suitable quantities, light of wavelengths 570 and 465 nm or of wavelengths 600 and 489 nm. The ellipses, drawn here at 10 times their true size, are contours of just noticeably different colours from their central colour. The description of the differences in size, shape and orientation of these ‘JND’ ellipses is the domain of higher colour metrics.

(From .)

However, the chromaticity chart has true practical value, for instance in colour-mixing techniques used routinely in computer programs for producing different colours digitally for image creation and other purposes, and have been developed into a computerized colour vision test. Clinically, colour vision can be tested using hue-discrimination techniques (e.g. the Farnsworth–Munsell 100 hue test) and normal values vary with age (affected by JND effects), peak ability occurring around the age of 19 years. Some effect by rods on cone vision has also been shown by rod function studies (‘background noise’).

Shape, form and depth perception (and more) help to ‘shape’ vision

The discrimination of shape and form is highly developed in primate vision and the cortical localizations which define these functions are now well established. The appreciation of the complexity and sophistication of this aspect of visual perception has in part developed from the realization that the brain recognizes and can categorize objects according to shape irrespective of the angle or distance from which they are viewed, the ambient lighting conditions, or other factors.

Shape processing is achieved by specialized orientation-sensitive cells in the visual cortex, but the extra dimension of form recognition, as in recognition of facial features, requires additional processing. Studies of patients with specific visual defects such as prosopagnosia (inability to recognize familiar faces), however, are powerful indicators of the localization of visual functional sites. Even this apparently specific defect can be subdivided into a perceptual form and an associative form, the latter arising when the patient can perceive the image but cannot draw on visual memory sufficient to ‘remember’ the face. These separate functions have been ascribed to different regions of the brain, although associative face recognition may also involve other senses such as voice recognition ( Fig. 5-4 ).

Processing information for features (of the face) involves a specific neural cortical network, the occipital face area (OFA), the face fusiform area (FFA) and the anterior temporal lobe (ATL), which have a right hemisphere dominance. This region underlies neurological perceptual defects such as prosopagnosia (lesion in the posterior region) as well as person recognition disorders (lesions in the anterior temporal lobe).

(From .)

Much is also known about depth perception. For instance it has been shown that there are specific cortical cells responsive only to disparate but simultaneous orientations of an object, presented to non-corresponding regions of the retina. In addition, clues on the nature of 3D structures can be obtained from motion detection (structure-from-motion). However, although no specific ‘depth appreciation’ cortical region has been identified, the lateral occipital cortex is a favoured area showing much activity. The appreciation of depth is more than simply stereopsis and is built up from many other cues (see below). An understanding of how these modify perception can only come after a description and appreciation of the different types of visual stimuli that separately induce discrete responses in the brain.

What are the limits of detectable light?

As in all biological systems, there is no precise answer to the above question. Light energy comes in quanta (small packets) and it has been estimated that between 50 and 150 quanta of light are required to strike the cornea for a discrete signal to be detected. Of these quanta, only about 10% actually reach the photoreceptors. The detection of this stimulus is not simply a function of photoreceptor stimulation but is subject to ‘dark light’ (effectively background noise in rhodopsin photoisomerization) and is also dependent on higher neural function, and the concept of a visual threshold is more or less a statistical function dependent on how large the stimulus has to be to reach a level of recognition. This is well recognized by anyone who performs a visual field test using an automated visual field analyser.

Dark adaptation curve and retinal sensitivity

The minimum visual stimulus varies depending on ambient light conditions, i.e. whether the stimulus is viewed in the dark or under normal/bright light conditions. In the dark, the eye becomes progressively more sensitive to light stimulation until the light threshold reaches a minimum after about 30 minutes. This is demonstrated in the dark adaptation curve ( Box 5-1 ), which has two components: an early one resulting from increases in the cone sensitivity and a second produced by increases in rod sensitivity. There is also a light adaptation curve for cones in which the sensitivity to light varies as the luminosity increases or decreases within a wide range of high ambient illumination (see below). Thus the shape of the curve can be varied by altering the conditions.

Box 5-1

Dark Adaptation

The normal dark adaptation curve (a) varies if the conditions are varied: with a very small central white target, rods fail to become stimulated at all and the curve flattens out (b). If the cones are first light adapted by weakly stimulating them to maximum sensitivity or by adapting subjects to red light before placing them in the dark, the cone component can be ‘lost’ (c); subjects without cone vision also have no cone component (rod monochromats).

Figure outlining dark/light adaptation responses: (a) mixed rod and cone response of physiological dark adaptation, (b) pure cone response, (c) pure rod response.

(Figure courtesy of H. Dawson.)

Light and dark adaptation are the psychophysical correlates of visual pigment bleaching and regeneration, and can be measured by reflection densitometry. This technique is based on the assumption that light reflected from the unbleached retina will contain lower amounts of 500 nm (peak sensitivity for rods) light than that reflected from the bleached retina, since there will be considerable absorption of 500 nm light by the dark-adapted retina. Reflection densitometry studies permit an evaluation of the photosensitivity of the retina, i.e. the rate at which bleaching takes place for a given intensity of illumination. It has been estimated that the normal retina absorbs 50% of the quanta of light striking the retina but, as discussed above, this is not necessarily associated with a perceived visual stimulus because the absorption by a single rod of a photon of light can have at least three outcomes.

Regeneration of rhodopsin after dark adaptation is slow, taking 30 minutes for completion, with a half-time in humans of 5 minutes. This varies significantly between species. Clearly, the sensitivity of the retina in any individual will depend on the total amount of rhodopsin, and this relationship has been delineated in the Dowling–Rushton equation:

These effects are important in the determination of photosensitivity of discrete regions of the retina where it has been shown that reduced sensitivity can be detected in regions of the retina not exposed to point sources of light. Although this has been attributed to light scatter, there are probably other mechanisms operative here, particularly related to convergence of neural input (see below).

What does adaptation mean at a molecular level? There is considerable evidence to show that dark adaptation and regeneration of rhodopsin are dependent on the local concentration of 11- cis retinal, and the limiting factor for recovery after a large bleach is the rate at which 11- cis retinal is delivered to opsin in the bleached photoreceptors ( Fig. 5-5 ) despite some more recent evidence that some of the retinoid conversion steps occur in the Müller cells also (see Ch. 4 , Fig. 4-69 ). Thus, because a healthy retinal pigment epithelium is central to this process, age-related decline in dark adaptation can be explained on this basis. It is also dependent on termination of the photon-induced signal brought about by efficient phosphorylation of the enzyme rhodopsin ( Ch. 4 , p. 262 ) via rhodopsin kinase, with the subsequent docking of arrestin to the complex, so that free opsin can be made available to bind more 11- cis retinal and respond to a new photon. Absence of rhodopsin kinase (also known as G protein-coupled receptor kinase 1, Grk1) or of arrestin, underlies the pathology of stationary night blindness (Oguchi’s disease) while absence of the equivalent cone opsin kinase (Grk7) causes enhanced S cone syndrome.

Schematic of the MLP rate-limited model. Removal of photoproduct and regeneration of visual pigment is rate-limited by the delivery of 11- cis retinal ( cis RAL) from the retinal pigmented epithelium (RPE) to opsin in the outer segment (OS). IPM, interphotoreceptor matrix; Rh, rhodopsin; M, metarhodopsin.

(From Lamb and Pugh, 2004, with permission from Elsevier.)

Psychophysical evaluation of rhodopsin bleaching has been experimentally tested in humans by comparing a range of dark adaptation curves to different background light levels with the amplitude of the a wave of the electroretinogram, also known as the rod current (see p. 288 ) since both are desensitized to varying degrees by the amount of ambient light. It appears that rhodopsin regeneration and a-wave recovery rates match well.

Cones also regulate their sensitivity in photopic conditions, but it is much more difficult to saturate this response, i.e. cones still adapt at high intensities of steady illumination and the recovery time is very short (100 ms compared to 20–30 minutes for rods). This is probably furnished by Müller cell-derived 11- cis retinal, also under the control of RPE 65 in the Müller cell (see Ch. 4 , Fig. 4-69 ).

In summary, adaptation is exactly what it means: that the retina rapidly adapts to changes in background illumination such that it can respond to increasingly strong or weak stimuli. However, the dynamic range of responses (normally a range from zero to a few hundred impulses per second) over which it functions at any specific level of illumination remains the same and the intensity of the response when it makes one is also the same. Put simply, the retina adapts rapidly to new lighting conditions when there is plenty of light about but slowly when light levels are low.

Melatonin and circadian rhythms

The circadian clock is a process whereby genes regulating various functions such as the sleep–wake cycle, body temperature, immune cell function and behaviour are expressed in a rhythmical manner. At least 11 core clock genes have been discovered, including a set of period genes ( PER 1,2,3) and clock genes ( CLOCK ). These genes have multiple downstream effects on other important regulatory transcription factors such as POPα/β/γ important in immune cell function and genes involved in the synthesis of melatonin in the synthesis of melatonin. The light/dark (sleep/wake) cycle, generated by pacemaker cells in the suprachiasmatic nuclei, drives the production of the pineal gland secretory product, melatonin. Melatonin may also be produced at other sites, including the retina and bone marrow. It is synthesized from tryptophan via serotonin in two steps involving the enzymes serotonin- N -acetyltransferase (NAT) and hydroxyindole- O -methyltransferase (HIOMT) (see Ch. 4 , p. 248 ). Melatonin provides information to the organisms to permit organization of various physiological functions and, because it can adapt to night length, it can promote a seasonal (photoperiod) as well as a diurnal rhythmicity ( Fig. 5-6 ). Apart from its obvious physiological functions, such as sleep–wake patterns, melatonin influences immune diurnal variations in innate immune defence functions such as antioxidation, glucose regulation, blood coagulation enzyme systems and ocular functions such as control of aqueous secretion.

Melatonin acts as an endogenous synchronizer.

(From Claustrat et al., 2005, with permission from Elsevier.)

Melatonin is a methoxyindole, synthesized and secreted principally by the pineal gland at night under normal environmental conditions and binds to two receptors (M1 and 2). The endogenous rhythm of secretion is generated by the suprachiasmatic nuclei and entrained to the light/dark cycle. Light is able to either suppress or synchronize melatonin production according to the light schedule. The nyctohemeral rhythm of this hormone can be determined by repeated measurement of plasma or saliva melatonin or urine sulphatoxymelatonin, the main hepatic metabolite.

The primary physiological function of melatonin, whose secretion adjusts to night length, is to convey information concerning the daily cycle of light and darkness to body physiology. This information is used for the organization of functions, which respond to changes in the photoperiod (seasons). There is still, however, only limited evidence for seasonal rhythmicity of physiological functions in humans related to possible alteration of the melatonin message in temperate areas under field conditions, although there is a reported link between seasonal affective disorder, clinical depression and its control with novel antidepressant drugs based on melatonin receptor stimulation, such as agomelatine. Major clinical depressive illness has also been linked to markedly reduced function of the core clock genes in the brain in a recently reported post-mortem microarray analysis.

The daily melatonin secretion, which is a very robust biochemical signal of night, can be used for the organization of circadian rhythms. Although functions of this hormone in humans are mainly based on correlative observations, there is some evidence that melatonin stabilizes and strengthens the coupling of circadian rhythms, especially of core temperature and sleep–wake rhythms. As the regulating system of melatonin secretion is complex, following central and autonomic pathways, there are many pathophysiological situations where the melatonin secretion can be disturbed. The resulting alteration could increase predisposition to disease, add to the severity of symptoms or modify the course and outcome of the disorder.

Melatonin is also produced by photoreceptors where it can act on melatonin receptors (MRs) in an autocrine manner, as well as on MR+ ganglion cells and other retinal neural cells in a paracrine manner. Thus it regulates the activity of photoreceptors, it acts on horizontal cells stimulated by cones to reduce their responsiveness, but heightens ON-bipolar cells and ganglion cells in some species. In this way melatonin is thought to fine-tune visual function, especially in cone cells under varying ambient light conditions (see eFig. 5-2 ).

This occurs via activation of specific receptors on amacrine, horizontal and photoreceptor cells. Wiechman has proposed a working hypothesis for melatonin paracrine signalling in the retina ( eFig. 5-2 ). Melatonin is normally produced by photoreceptors at night, and diffuses to target cells within the retina that have specific receptors on cells such as GABA-ergic and/or dopaminergic amacrine cells which work in a reciprocal manner to some degree since GABA inhibits dopamine release from amacrine cells. A lower rate of dopamine release from amacrine cells results in lower stimulation of D1 receptors on horizontal cells, which in turn leads to increased coupling of horizontal cells. This would result in an increase in receptive field size and increased sensitivity to light. Lower levels of binding of dopamine to D2 receptors on photoreceptor cells induces an increase in melatonin synthesis. Meanwhile melatonin may bind to horizontal cells to directly inhibit the cellular response to D1 receptor binding. Melatonin may also bind to receptors located on the photoreceptor membrane, which could directly increase rod sensitivity to light, and/or regulate synthesis of melatonin.

Diagram outlining mechanism of how melatonin fine-tunes visual function (see text for details).

(From .)

Interestingly MRs are present on many other ocular tissue cells such as ciliary epithelium, RPE cells, lens cells, corneal endothelium and keratocytes, and stromal cells in the sclera and choroid.

Are two small stimuli equivalent to one large one (summation)?

The threshold for light detection can be measured arbitrarily by setting certain conditions of stimulus size, brightness, pupil size and level of background illumination, and recording how often a subject detects the stimulus. An empirically set level of ‘hits’ or positive detection responses (e.g. 55%) can then be set and expressed in trolands ( Box 5-2 ). Experimentally it has been estimated that at the limit of light detection in the fully dark-adapted eye, the retina is illuminated to a level of 4.4 × 10 ⁻⁵ trolands, which is equivalent to the stimulation of only 1/5000 rods per second. However, if the light is concentrated on one area it will more readily elicit a response and it therefore becomes less practical to think of light energy in terms of area of retinal illumination; instead the minimum flux in light energy required to induce a detectable response is commonly accepted as the threshold and is around 120 quanta per second or, if the stimulus is instantaneous, between 5 and 15 quanta of light.

From this it is clear that stimulation of a single rod is insufficient to produce a visual sensation (even though an electrical response may occur in terms of a change in hyperpolarization of the cell membrane). Approximately 10–15 rods must be stimulated and the summed response must be collected either at the bipolar cell level or within the ganglion cells to induce a visual sensation. These determinations are approximate as fluctuations occur at all levels from the stimulus itself to the responses in each of the different cell types, and the final analysis is based on probabilities of a response taking place.

Temporal summation

When the retina is stimulated in rapid succession by a target, the level of response is the same as when the target is presented continuously for the same total period of time. This is known as temporal summation and is formulated by Bloch’s law, which states: the intensity of the threshold stimulus is inversely proportional to the duration of the stimulus. To a degree this is a difficult psychophysical measurement since at very short intervals it is difficult to distinguish different contrast and duration. Bloch’s law holds true only for a defined period of time as, if the interval between the stimuli were long, the effect would be rapidly lost. Bloch and others found that, in fact, a ‘plateau’ effect was observed. However, Broca and Sulzer found that there was a peak in perception and then there was a decay before a plateau effect (see eFig. 5-3 ).

Investigation of potential different effects can be modelled experimentally. For instance the following experiments reported recently by Rieiro and colleagues demonstrate the actual effects of the different stimuli as varying potential outcomes ( ). In eFig. 5-3 , ( A ) represents two competing models of temporal vision. Bloch’s law postulates a monotonic increase in perceived contrast with increased duration, whereas the Broca–Sulzer effect postulates a peak in perceived contrast with increased duration. In ( B ) subjects fixated on a central cross, and two Gabor patches flashed in succession on opposite sides of the screen. Following stimulus presentation, subjects reported which Gabor had higher contrast. Part ( C ) shows the physical contrasts and stimulus durations used for the comparator and standard stimuli. In the unblocked experiment, all possible combinations were randomized. In the blocked experiment, the different conditions were grouped into four sequential sets of trials or blocks, each with a constant comparator contrast and standard duration and an internally randomized trial sequence. Finally, ( D ) shows the psychometric curve models of the two possible experimental outcomes, colour-coded for different comparator durations. If contrast perception has a peak, as in the Broca–Sulzer effect, the curves will first shift right and then left as stimulus duration increases. If contrast perception follows Bloch’s law, the curves will shift monotonically to the right.

Studies of the potential different effects which may occur in temporal summation (see text for details).

(From .)

In practice the peak described by Broca and Sulzer occurred at about 50–100 ms, and beyond this time there is still some degree of summation, known as partial summation, which decays exponentially.

Recently, the question of temporal summation has been revisited in the context of artificial light (which accounts for about 20% of our energy consumption). Using an experimental design in which intrinsic bias from previous learned experience was omitted, Broca’s peak was detected and was attributed to a differentiation in duration versus contrast which is eliminated probably by higher neural mechanisms ensuring that the same rapid stimulus is identified with short flashes of light. Bloch’s law therefore represents a smoothed-out perception in which the peak of detection/contrast is eliminated by prior learned experience at a subconscious level. If artificial lighting systems were optimally tuned to these temporal summation effects in human vision, for instance by using DC light-emitting diodes, a 20% saving in energy consumption has been estimated, which is not insubstantial.

Visual acuity is not simply a function of cone activity

Vision varies for each individual because of refractive errors and visual physiologists have therefore restricted discussion of normal visual physiology to the emmetropic idealized eye (see eFig. 5-4 ).

The eye has the power to refract (bend) light waves and, as for any lens, this is measured in dioptres. A dioptre (D) is a unit of measurement that describes the strength, or ‘power’, of a lens to bend (refract) light a set amount (degree); the optical power of a lens or curved mirror is equal to the reciprocal of the focal length measured in metres (that is, 1/metres). It is thus a unit of reciprocal length. Thus, a three-dioptre lens brings parallel rays of light to focus at 1/3 metre. The overall refractive power of the eye is around 60D for the normal healthy emmetropic eye and much of its refractive power is attributed to the lens-like (focusing) properties of the cornea (which amount to about 40D), with the remaining 20D due to the ocular lens ( eFig. 5-4 ). The effect of this refractive property of the ocular media (i.e. the tissues through which the light passes) is to focus light rays on the retina, and specifically the fovea, when visual acuity is being measured. Many refractive errors occur in the healthy population, including myopia (short-sightedness), hyperopia (long-sightedness) and astigmatism (non-spherical aberrations of the eye’s refractive power).

Standard dimensions of the eye as related to the model eye.

(From .)

The scientific discipline dealing with the optics of the eye is known as optometry. Many textbooks are available which deal with physiological optics as well as optical devices such as spectacle correction, contact lenses and the optics of intraocular lenses, including large treatises dealing with many aspects of physiological optics as well as shorter text books summarizing refraction and refractive errors for those not intending a career in optometry.

The reader is referred to the following texts as examples of (1) a short comprehensive text and (2) a three-volume in-depth treatise:

• (1)
  

  Hunter DG, West MD. Last-minute optics: a concise review of optics, refraction, and contact lenses. 2010.

  
  
• (2)
  

  von Helmholtz H. Treatise on physiological optics, vols I, II and III. Dover Phoenix Editions; 2005

In addition, several texts are available which deal with the optics of the pseudophakic eye (i.e. the eye with a prosthetic intraocular lens), including official publications of professional bodies such as the American Academy of Optometry and the British College of Optometry.

Visual acuity is a measure of the ability to discriminate two stimuli separated in space. Clinically, this is determined by discriminating letters on a chart, but this task also requires recognition of the form and shape of the letters, processes that involve higher centres of visual perception. Discrimination at a retinal level may therefore be determined by less complex stimuli such as contrast sensitivity gratings. The visual processes that allow discrimination between letters and gratings are fundamentally the same, with finer resolution contrast discrimination at lower luminance levels being provided by some newer test charts, such as the Mars contrast sensitivity charts which are graded in log 0.04 units. Charts that have been customized to test visual acuity in different groups of people, such as the SKILL test (Smith–Kettlewell Institute Low Luminance test), may be a good predictor of eventual development of macular degeneration in older people.

Theoretically, the resolving power of the eye can be derived from an estimate of the angle subtended by a single photoreceptor (about 1.5 µ or 20 minutes (20′) of arc in the case of cones), as this represents the smallest unit distance separating two individually stimulated photoreceptors. This corresponds to about a pixel on a computer screen when viewed at half a metre. However, it is well recognized that the resolving power of the eye can be as great as 0.5′ of arc, for instance when looking for the gap in a Landholt C target, or 4″ of arc when viewing a thin line on an illuminated background. This hyperacuity, or Vernier acuity, is achieved by the complexities of retinal neuronal synaptic organization and is 5–10× greater than ‘standard’ visual acuity, but the limits of acuity are still determined to some extent by the retinal photoreceptor mosaic or ‘grain’.

The highest discriminatory capacity is subserved by cones, although a certain degree of resolution can be achieved by rods. The level of acuity, however, falls off rapidly the greater the distance from the fovea, such that at 5° from the central fovea visual acuity is only one-quarter of foveal acuity. As rod and cone longitudinal dimensions are not sufficiently different to explain the marked difference in acuity, and as the resolving power of the eye is greater than the theoretical limits based on cell size, other mechanisms must underpin acuity. Visual acuity is affected by the luminance of the test object and the degree of adaptation of the observer; dark adaptation increases both rod and cone acuity and therefore is not affected by the sensitivity of cones per se. In contrast, light adaptation increases sensitivity of cones but not rods (see p. 274 ).

Vernier acuity is used in everyday life, for instance in measuring distance with a ruler or detecting the time on a mechanical clock. Vernier acuity is not present in infancy but reaches its highest level of function around the age of 14. It is absent in strabismic amblyopia but may be present in patients with anisometropic amblyopia. Vernier acuity is different from the recently recognized state of supervision, which has been revealed by the use of adaptive optics. Adaptive optics were developed for use in astronomy to minimize optical aberrations and correct higher-order dynamic aberrations caused by such aspects as angle of viewing and accommodation, as compared with correction of static aberrations such as astigmatism and defocus (see below). In essence, by using a wave form sensor, adaptive optics measures phase aberrations in reflected light produced by the imaging light source. When applied to the eye, for instance in the use of wavefront aberrometers and wavefront-guided vision correction in refractive corneal surgery, adaptive optics can theoretically increase acuity to ‘supervision’ levels.

Limits of and limitations on acuity

The letters on reading charts such as the Snellen’s test type and the ETDRS chart ( Box 5-3 ) have been constructed on the assumption that the average person can resolve two points separated by 1′ of arc. If the limit on acuity is in part determined by the single photoreceptor theory (above) then a one-to-one relationship between the photoreceptor and the nerve cell must exist if there is to be no downstream loss of acuity. For foveal cones such a relationship exists between cone cells, midget bipolar cells and midget ganglion cells (see section on retinal connections , pp. 291–296 ), but even midget cells have some interconnections with diffuse bipolar and ganglion cells. In spite of these connections, summation of information should not occur for cells subserving the highest levels of acuity and, indeed, is absent from foveal cone cells but is characteristic of rod cells; furthermore, it has been suggested that the improved visual acuity that occurs under conditions of light adaptation is the result of inhibition of these subsidiary connections.

Box 5-3

Assessment of Visual Acuity Using Standard Letter Charts

Visual acuity in clinical practice is determined as an empirical value based on the assumption that the cone photoreceptor has the ability to discriminate two objects in space subtended by an angle of 1 minute of arc at the nodal point of the eye ( A ). This is measured using a set of charts (optotypes) and standard normal visual acuity equates to the vision of 6/6 or 20/20 (i.e. 1.0 or 100%) when viewing a predetermined standard target (size of optotype letter) at 6 m (UK) or 20 feet (USA). Test conditions describing the ambient illumination, and the illumination of the letters on the chart to provide contrast are also arbitrarily set. The Snellen chart is based on the concept that the smallest spatial target that can be resolved subtends 1 minute of arc at the nodal point of the eye (see above) and although theoretically inaccurate, it serves as a useful parameter. The Snellen chart has rows of letters of decreasing size and is arbitrarily set to produce the standard test at 6 m, although other charts with proportionally smaller letters can be used at shorter distances. The LOGMAR (LOGarithm of the Minimal Angle of Resolution ( B ) is more precisely designed with definitive sizing and spacing of the letters and can provide a more quantitative evaluation of visual acuity. It therefore tends to be the standard for use in clinical trials. As indicated in the text, spatial acuity better than 100% can be achieved, for instance when discriminating the ‘offset’ of a line or edge ( C ). This is termed Vernier acuity. In addition, visual acuity is modified by such factors as glare and contrast, and indeed can be measured as in the contrast sensitivity test using a sine wave grating as shown in ( D ), where diffraction and aberrations have degraded the contrast of a sinusoidal grating pattern.

(From Schweigerling, 2000, with permission from Elsevier.)

Visual acuity is also limited by the physical behaviour of light, such as diffraction and chromatic/spheric aberration. A single point of light small enough to stimulate a single cone will produce diffraction rings in its traverse through the pupil sufficient to stimulate more than one cone. Similarly, the prismatic separation of white light into its constituent wavelengths will lead to the stimulation of several cones of different types. It is clear therefore that resolution of images must be achieved at a post-receptor level and is in fact a function of the receptive field of each ganglion cell unit. Where there is minimal convergence of information from each receptor, i.e. where the one-to-one relationship between receptor and bipolar cell is maintained, then resolution is at its highest and this occurs at the fovea. However, where there is increasing convergence of information, such as with several parafoveal cones synapsing with one bipolar cell, resolution obviously decreases.

The one-to-one relationship, however, does not adequately explain hyperacuity or Vernier acuity. Diffraction and spheric/chromatic aberrations have ruled out the concept of single unstimulated cones occurring between neighbouring stimulated cells; however, it is likely that discrimination is more a matter of degree than absolute responses, i.e. that resolution is achieved by certain receptors being less stimulated than their neighbouring receptors on either side. This is likely to occur with diffraction, where alternating light and dark rings emanate from a point source of light, and with chromatic aberration, where different wavelengths of light are likely to stimulate their respective neighbouring cones to different degrees ( Box 5-4 ). In fact, this fits with the nature of the hyperpolarization response being a graded one in retinal neural cells: as indicated in Chapter 4 (see p. 261 ), only ganglion cells fire ‘classic’ depolarizing action potentials, while other retinal cells have a graded analogue-type ‘tunable’ electrical response. This differential stimulation from cones is registered with the respective bipolar and ganglion cells and, if combined with a minimal degree of receptor convergence, can explain high levels of discrimination. In this way diffraction could explain, at least partly, the ability to resolve a break in a line subtending an angle of less than 10′ of arc, since partial diffraction lines deriving from the edge of the break would ensure differential stimulation of cone receptors over a very small area. Stimulation of any particular cone is also likely to induce local inhibition (via receptive field mechanisms; see below) in neighbouring cones, thus enhancing resolution and ‘sharpening the image’ further.

Box 5-4

Differential Activation of Neighbouring Cones Determines the Limits of Visual Acuity

A beam of light interrupted by a small target will produce diffraction rings at its edges – the central rays in each ring will stimulate one cone (2) more than the weaker peripheral rings will stimulate its neighbours (1) and (3). Our ability to detect the break in the light beam is determined by comparing the differential responses in all the cones in the illuminated region and finding two that produce a similar response to correspond to the edges of the target. Visual acuity is therefore a measure of the retina’s ability to produce different graded responses and not absolute responses.

These concepts are embedded in the canon of knowledge going back to the time of Rayleigh in the late nineteenth century. However, Rayleigh’s law does not fully explain the real condition where diffraction spreading from a double line is slightly greater than that from a single line for the same total amount of light energy. Information theory, adaptive optics and modern electro-optical devices for generating discrete stimuli might help to explain this anomaly.

These considerations also have a number of implications. In particular, the resolving power of the eye is limited by the distance between two images such that a single cone or set of cones is appreciably less stimulated than the rest; the limit of resolution is therefore not an absolute determinant, but depends on conditions such as light and dark adaptation, background illumination and other factors. Most importantly, it depends on the degree of dendritic connections that occur between the affected cones and the neural cells.

The resolving power of the eye therefore depends on:

• •
  

  the distance between two objects

  
  
• •
  

  the degree of light and dark adaptation

  
  
• •
  

  the background illumination

  
  
• •
  

  the extent of the dendritic connections between the cone and neurones.

Contrast sensitivity

Visual acuity is also affected by contrast (sharpness). The finest limits of resolution have been determined by the ability to discriminate a thin white line against a uniform background illumination (0.5′ of arc). The effects of diffraction are such that detection of this line depends on the liminal brightness increment (l.b.i). This increment represents the endpoint at which the differential in brightness between the individual dark/bright oval diffraction rings produced at the edge of the line can be detected; if they are not sufficiently different from the background luminosity, then the line will not be detected. The l.b.i. is determined by the contrast between the light and dark lines, and can be measured quantitatively with a sinusoidal grating (see Box 5-3 ): a spatial pattern where the average luminance remains the same but the contrast between the light and shaded areas can differ.

The degree of contrast (C) is relative to the background luminance ( L ) and is described in terms of the maximum ( L _max ) and the minimum ( L _min ) as follows, also known as the Michelson contrast:

Contrast sensitivity curve showing peak response at midspatial frequencies.

Contrast sensitivity is therefore set by the limits of the grating frequency and is affected by both the optics of the system and the direction of the grating lines, being most sensitive in vertical or horizontal directions. Remarkably, threshold contrast for many targets sits around 1% independently of target size or brightness, which as Peli observes remains unexplained since originally described by Fechner in 1860 (Peli, 2013).

Contrast sensitivity above threshold is, as for any measure of acuity, affected by luminance. In addition, bar width, length and grating motion all affect sensitivity. In the latter there is likely to be significant cortical processing at this level, as there is for ‘line modulation sensitivity’, a technique whereby grating bars are composed of wavy lines and the subject is asked to determine whether the line is straight or not. This technique can provide highly sensitive measures of acuity.

Contrast sensitivity measurements, while an excellent measure of acuity, are sensitive to phase shifts and grating orientation and for absolute measures of object detection, object contrast is critical. In this context, mesopic low-contrast letter acuity is the most sensitive method for revealing small differences in retinal image ‘quality’, which influences ‘recognition’ as opposed to ‘detection’ of an object. The contrast sensitivity function (CSF) has been arbitrarily measured using a set of five spatial frequencies and has been found to be relatively robust. Age and decreased luminance cause a shift to larger frequencies in the CSF. Glare, which is often a side-effect of refractive surgery, affects the CSF at low rather than high spatial frequencies. For clinical purposes, measurement of the CSF (as opposed to measuring the threshold as in most contrast sensitivity charts) is very time consuming and impractical, but recent developments using customized and selective spatial frequencies and contrasts are allowing tailored CSF tests to be applied to specific conditions, e.g. macular degeneration.

Wavelength also affects contrast sensitivity such that at high spatial frequencies the gratings appear to be of the same colour, whereas at low frequencies (i.e. with coarse gratings), colour differences can be detected. Discrimination is poorest with red–green, however, suggesting that for low frequencies rod–cone interactions are important in achieving best visual acuity. Interestingly, contrast sensitivity appears to induce more electrical signal responses in M cells, generally thought to subserve rod function, than in P cells, which are linked to cone function (see below).

Does the retinotopic arrangement of fibres in the cortex have a bearing on acuity?

The representation of retinal ganglion cells in the LGN and cortex is disproportionately larger for foveal midget cells than for ganglion cells elsewhere in the retina. This produces a ‘cortical magnification factor’ for foveal cones over other cones. However, the magnification factor is not related solely to the reduced convergence of foveal cones on ganglion cells (see below), but also to a disproportionate LGN and cortical representation of neurones served by foveal cones. Current use of fMRI has revealed the retinotopic map of the human visual cortex and demonstrated this magnification factor in technicolour ( Fig. 5-8 ).

Dipolar map and isopolar angle maps of human visual areas. These images were prepared using fMRI technology and computer modelling to ‘flatten out’ the visual cortex ( C and F ). The top row shows a map of the occipital cortex indicating the retinotopic location of the stimulus as its eccentricity increases from the fovea; the dipole is coded by colour (brown (fovea) -> orange -> blue -> cyan (periphery) ) displayed on the original cortical surface ( A ), the unfolded cortical surface ( B ), and the cut and flattened cortical surface ( C ). The bottom row shows the polar angle of the stimulus (blue (upper vertical meridian) -> green (horizontal meridian) -> red (lower vertical meridian) ) plotted on the same three surfaces ( D–F ).

(From .)

Best-corrected visual acuity: effects of external factors

Visual acuity is, of course, affected by factors that do not relate directly to the retinal stimulus. These include pupil size, eye movements and binocular viewing.

Pupil size and visual acuity testing in infants

The size of the pupil affects the level of visual acuity in that a reduction in pupil size reduces aberrations but increases the effects of diffraction. Below 3 mm, these effects tend to cancel each other out, and visual acuity is independent of pupil size, although wavefront aberrometry reveals that a pupil size of 2.5–3.0 mm produces the best image quality.

The level of visual acuity attained may also have the reverse effect on the size of the pupil. Luminance affects the level of acuity and the size of the pupil is affected by the light level via well-characterized pupil reflexes ( Box 5-5 ). The size of the pupil also indirectly affects the visual acuity by reducing the amount of light entering the eye when the light stimulus is intense, and conversely increasing light capture under dim lighting conditions. This three-way relationship has been used to develop an objective measure of visual acuity, which may be useful in assessing vision in infants and others who are not able to cooperate in standard visual acuity testing. The test uses a high-resolution infrared pupillometry device to show changes in the amplitude of constriction in response to sine-wave gratings presented on a uniformly illuminated test background. As for contrast sensitivity, there is a peak response in the middle range of frequencies and the threshold for response correlates well with contrast sensitivity estimates of acuity. This pupil response is governed by higher visual pathways, being altered in patients with hemianopia but normal pupil light reflexes; indeed, the phenomenon is well recognized by clinical neuro-ophthalmologists. Infrared pupillometry has been shown to be valuable in studies of delayed visual maturation and to be significantly more reliable than the Rosenbaum card method in which subjective comparisons of pupil size are made.

Box 5-5

Pupillary Light Reflexes

• •
  

  The afferent response commences in photoreceptors, is transmitted to retinal ganglion cells, enters the optic nerve, decussates at the chiasm, traverses the optic tract and terminates in the pretectal nucleus (bypassing the lateral geniculate nucleus).

  
  
• •
  

  Both crossed (via posterior commissure) and uncrossed fibres pass from pretectal nucleus to Edinger–Westphal nucleus (parasympathetic).

  
  
• •
  

  Parasympathetic fibres pass to the III nerve nucleus and leave the brainstem via the III nerve. Fibres synapse in the ciliary ganglion before supplying sphincter pupillae of iris (constriction) via short ciliary nerves.

  
  
• •
  

  Uniocular light stimulus therefore gives rises to bilateral and symmetrical pupillary constriction.

  
  
• •
  

  Melanopsin signals through ipRGCs (see text) to the suprachiasmatic nucleus (circadian response) and the pretectal nucleus (irradiance response).

The pupillary response also receives input from the intrinsically photosensitive retinal ganglion cells (ipRGCs) through melanopsin (see Ch. 4 , p. 249 ) and it is possible using chromatic pupillometry (stimulation of pupil responses at different wavelengths) to separate the contributions from rods, cones and ipRGCs. There are diverse types of ipRGCs with several different functions. For instance the sleep/wake circadian rhythm responders connect with the suprachiasmatic nucleus, while the pupillary responsive neurones synapse in the pretectal (olivary) nucleus. In terms of response to light, ipRGCs are considered to register irradiance or radiating light. They may also be responsible for photoallodynia (the photophobia/light aversion response to very bright light).

The electrical response is initiated by phototransduction

As we have seen (see Ch. 4 , when a photon of light strikes the photoreceptor outer segment, conversion of rhodopsin to the activated molecule induces a series of molecular events culminating in an electrical response. Cells, and particularly neurones, normally exist in a ‘charged’ state in that the inside of the cell is ‘negative’ with respect to the extracellular environment, creating an electrical potential difference across the cell membrane. This condition is maintained by differential distribution of Na ⁺ and K ⁺ ions on either side of the cell membrane. When a neurone is stimulated, there is an initial period of gradually increasing positivity (the generator potential), which culminates in a spike discharge characterized by a rapid depolarization response of the cell. This is achieved by the rapid influx of Na ⁺ through ion channels that are ‘opened’.

In the photoreceptor the reverse situation occurs. Under resting conditions in the dark, the outer segment is maintained in a depolarized state through open (‘leaky’) Na ⁺ channels, which permit the influx of sodium ions from the extracellular space. When light stimulates the outer segment, the sodium channels are abruptly closed, stopping the influx of sodium and thereby leading to a reduced level of depolarization, i.e. a relative hyperpolarization ( Box 5-6 ). This is a direct result of rhodopsin isomerization and is mediated by amplification mechanisms involving cyclic guanosine monophosphate (cGMP) (see Ch. 4 , p. 262 ).

Box 5-6

Dark Currents

Dark currents occur in the resting state (dark adapted eye) owing to ‘Na ⁺ -leaking’ outer segments.

The conversion of light energy to an electric response is dependent on specialized ion channels that tightly control the permeability of the cell membrane to Na ⁺ and Ca ²⁺ . Light stimulation reverses the dark current by closing Na ⁺ channels in outer segments and releasing Ca ²⁺ (and glutamate) at synapses. The cGMP-gated Ca ²⁺ channel and the Na ⁺ /Ca ²⁺ , K ⁺ exchanger are located in the plasma membrane of the photoreceptor, not in the disk stack, but are complexed together with peripherin/rds-rom-1, an integral protein of the disk rim.

The hyperpolarization response is transmitted by a flux in calcium ions along the length of the photoreceptor to the synapse with the bipolar cell (the Ca ²⁺ wave), which is then induced to release its transmitter (glutamate). Bipolar cells may then adapt to one of two responses to glutamate, depending on which type of receptor is induced: an ON response, which is a hyperpolarized state, and an OFF response, which is a depolarization response (see below). Indeed, the hyperpolarized state conferred on the bipolar cell is also transmitted to the horizontal cells in the same region. However, the hyperpolarization response of the bipolar cell is not as steep as that of the photoreceptor in the excited state.

It will be obvious, therefore, that not only is there a resting potential difference across the photoreceptor cell membrane but there is also a potential difference along the length of the photoreceptor in the dark between the relatively depolarized outer segment tip and the hyperpolarized synaptic region of the cell at its interaction with the bipolar cell. This generates the ‘dark currents’ in the eye, which are reversed by the photic current on light stimulation when the photoreceptor tip becomes hyperpolarized (see Box 5-6 ).

Electrophysiology of single retinal cells

Early studies in this field concentrated on the large single neurones that could be obtained from invertebrate eyes and showed that typical action potentials could be obtained, usually preceded by a generator potential ( Box 5-7 ). Surrounding neurones were usually inhibited when action potentials occurred in a single nerve.

Box 5-7

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Embryology and early development of the eye and adnexa
2. General and ocular pharmacology
3. Microbial infections of the eye
4. Biochemistry and cell biology

Stay updated, free articles. Join our Telegram channel

Join