Circadian Field Photometry

80
Circadian field photometry December 1, 2006 PETTERI TEIKARI [email protected] PROJECT WORK OF MEASUREMENT SCIENCE AND TECHNOLOGY FOR THE COURSE “S-108.3120 PROJECT WORKCourse credits: ECTS Points Grade (1-5): Supervisor’s signature: M.Sc. Tuomas Hieta

Transcript of Circadian Field Photometry

  • Circadian field photometry December 1, 2006

    PETTERI TEIKARI [email protected]

    PROJECT WORK OF MEASUREMENT SCIENCE AND TECHNOLOGY FOR THE COURSE S-108.3120 PROJECT WORK

    Course credits:

    ECTS Points

    Grade (1-5):

    Supervisors signature:

    M.Sc. Tuomas Hieta

  • Symbols and abbreviations................................................................................................. 3 1. Introduction ............................................................................................................ 4 2. Circadian photobiology ......................................................................................... 6

    2.1 Circadian rhythms .........................................................................................................7 2.2 Circadian clock mechanism ..........................................................................................8 2.3 Physiology of the eye ....................................................................................................9

    2.3.1 Ophthalmological optics ..................................................................................................... 12 2.3.2 Pupil pathways .................................................................................................................... 15 2.3.3 Eye movements ................................................................................................................... 17

    2.4 Light characteristics ....................................................................................................19 2.4.1 Spectrum ............................................................................................................................. 19 2.4.2 Spatial distribution .............................................................................................................. 22 2.4.3 Intensity............................................................................................................................... 23 2.4.4 Timing................................................................................................................................. 24 2.4.5 Duration .............................................................................................................................. 25 2.4.6 Photic history ...................................................................................................................... 26 2.4.7 Polarization ......................................................................................................................... 26

    3. Eye and photometric measurements .................................................................. 27 3.1 Electrical activity ........................................................................................................27

    3.1.1 Electroretinogram (ERG) .................................................................................................... 27 3.1.2 Electooculogram (EOG)...................................................................................................... 29

    3.2 Eye tracking ................................................................................................................30 3.2.1 Pelz et al. (2000, 2004) ....................................................................................................... 32 3.2.2 Li et al. (2006): openEyes ................................................................................................... 33

    3.3 Pupil size .....................................................................................................................37 3.3.1 Video-driven infrared pupillography .................................................................................. 38 3.3.2 Photorefractometry.............................................................................................................. 40 3.3.3 Digital photography ............................................................................................................ 41

    3.4 Digital-imaging circadian photometry ........................................................................43 3.4.1 Circadian-weighed luminancephotometers (Gall et al., 2004) ............................................ 43 3.4.2 Digital photography (Hollan et al., 2004) ........................................................................... 45

    3.5 Dosimeters ..................................................................................................................47 3.5.1 LichtBlick (Hubalek et al., 2004)........................................................................................ 48 3.5.2 Daysimeter (Bierman et al., 2005) ...................................................................................... 49

    4. Dosimeter design and simulation........................................................................ 54 4.1 Eyetracker and/or pupil size measurement..................................................................54 4.2 Dosimeter ....................................................................................................................56

    4.2.1 Photodiode-based dosimeter ............................................................................................... 56 4.2.2 Spectroradiometer-based..................................................................................................... 61

    5. Conclusions ........................................................................................................... 63 6. References ............................................................................................................. 65

  • Symbols and abbreviations 3

    SYMBOLS AND ABBREVIATIONS wavelength max peak wavelength W/cm2 microwatt per square centimeter A/Hz amperes per root hertz A/W amperes per watt acv circadian action factor Ap pupil area Ar area of the image at the retina As source area B noise bandwidth [Hz] B() action spectra for blue light hazard (ICNIRP) B() biological/circadian action spectra b-lx blue-lx, unit for blue-colored illuminance BY hypothetical luminous efficiency function for

    circadian responses. c() circadian action function cd/m2 candelas per square meter, unit for luminance dB decibel DC direct current dp pupil diameter dr diameter of the image at the retina ds diameter of the source E Energy Ec corneal irradiance en noise voltage density Er retinal irradiance f frequency f focal length g gram h Plancks constant Hg high-pressure mercury Hz Hertz, unit for frequency Id dark current If feedback (gain) current Ijn Johnson noise current In noise current in noise current density In,e noise current from en in,e noise current density from en Ip photocurrent Itot total noise current J/cm2 Joules per square centimeter K Kelvin, unit for (color) temperature kB Boltzmann constant kbauds/s kilobauds per second lm/w lumens per watt, unit for luminous efficacy Ls source radiance lx lx, unit for illuminance mA milliampere mAh milliampere hour MB megabyte Mbps megabits per second MHz megahertz mm millimeter M melanopsin-containing retinal ganglion cell

    spectral efficiency nm nanometer, normal unit for wavelength nV nanovolt pA picoampere Pr retinal power P spectral irradiance at the eye [W/m2/nm] Rf feedback (gain) resistance Rsh shunt resistance R photosensitivity S S cone spectral efficiency T temperature [K] V()

    spectral sensitivity curve for photopic vision V/Hz volts per root hertz V rod spectral efficiency function V10 photopic spectral sensitivity for centrally fixated

    large target

    VDC volts, direct current W watt Xec circadian radiation quantity Xv photometric radiation quantity angular subtense of the source lens transmittance luminous flx [lm] e, Spectral radiance [Wm-2] ohm, unit for resistance s solid angle [sr]

  • Symbols and abbreviations 4

    +/+ wild type mice ADC analog-to-digital converter AgCl silver chloride CBT core body temperature CBTmin core body temperature minimum CCD charge-coupled device CCT correlated color temperature CD compact disc CIE International Commission on Illumination CMOS complementary metal oxide semiconductor CP constant posture CR constant routine CRH corticotropin-releasing hormone CT circadian time DLMO dim light melatonin onset DLMOff dim ligh melatonin offset DLMOn dim light melatonin onset DMH dorsomedial hypothalamic nucleus dmSCN dorsomedial suprachiasmatic nucleus dSPZ dorsal subparaventricular zone EB eye blink ECG electrocardiogram EEG electroencephalogram EOG electrooculogram ERG electroretinogram ERP early-receptor potential EW Edinger-Westphal nucleus fMRI functional magnetic resonance imaging FOV field of view FWHM full width at half maximum GNU GNU's Not Unix GPL General Public License GTP ganosine triphosphate hbw half bandwidth hbw half bandwidth IC integrated circuit IEEE-1394 Institute of Electrical and Electronics Engineers,

    standard 1394 (FireWire, i.LINK) INL inner nuclear layer ipRGC intrinsically photosensitive retinal ganglion cell IR infrared IRC intensity response curve IRED infrared LED KRG potassiumretinogram LASIK laser-assisted in situ keratomileusis LCD liquid crystal display LED light emitting diode LGN lateral geniculate nucleus LRP late-receptor potential MPO medial preoptic region mRGC melanopsin-containing retinal ganglion cells M-RGC magno-retinal ganglion cells NIF non-image forming ONL outer nuclear level PF prefrontal PLR pupillary light reflex PPRF paramedine pontine reticular formation PRC phase response curve P-RGC parvo-retinal ganglion cells PRK photorefractive keratectomy PU Pupillary Unrest PUI Pupillary Unrest Index RANSAC Random Sample Consensus REM rapid eye movement sleep RGC retinal ganglion cell RHT retino-hypothalamic tract RI retinal illuminace RMS root-mean-square RS-232 a standard for serial binary data interconnection SCN suprachiasmatic nucleus SD standard deviation SEM slow eye movement SEM standard error of the mean SPD spectral power distribution USB Universial Serial Bus

    UV ultraviolet VDT video display terminals VDU video display unit VLPO ventrolateral preoptic nucleus vlSCN ventrolateral suprachiasmatic nucleus vSPZ ventral subparaventricular zone

  • Introduction 5

    1. INTRODUCTION

    Over 150 years since the discovery of the retinal rod and cone photoreceptors in 1834, it has been believed that, both visual and biological effects induced by light would be dependent on those two traditional photoreceptors. However in 2002, through the discovery of a novel photoreceptor in the eye by David Berson et al. [1] views have changed on how human vision system works. The novel photoreceptor, intrinsically photosensitive retinal ganglion cell (ipRGC) is one of the known ~20 ganglion cells in human retina. It has been estimated that of all retinal ganglion cells (RGC), 0.25% are photosensitive ipRGCs [2]. It was found [3] that the novel photoreceptor is responsible mainly for regulating light-induced human biological rhythms (circadian rhythms) by synchronizing body to environmental light/dark-cycle. It has also been proposed [4] to mediate light-induced increase in alertness, pupillary responses, and as a possible target for seasonal depression treatment. This novel photoreceptor may have many consequences for practical applications both in general lighting and lighting for special groups (e.g. elderly, shift workers, patients suffering from seasonal depression) so it has become a great interest of research in the lighting community [5]. However, the characteristics of non-image forming (NIF) visual system differ from the conventional visual system based on cones and rods. The NIF visual system has a higher threshold for activation, requires longer exposures for activation, depends on the location of light source in the visual field, and most importantly has different spectral characteristics with the peak wavelength (max=480 nm) shifted towards the blue part of visible spectrum. Given that the spectral characteristics are different with the NIF visual system; conventional photometric illuminance can not be used to quantify the NIF responses in humans. As the temporal characteristics (duration and timing of light exposure) differ from visual system, simple measurement of task illuminance is not sufficient to determine the NIF effective light response for example during normal office work. Measurement situations with NIF effective lighting can be divided roughly into two categories: measurement in strictly controlled laboratory studies and to field studies and workplace measurements. This work addresses the problems measuring the NIF effective light exposure by means of a literature review in field conditions using a portable head-mounted light dosimeter. However given that the biological effects of given light exposure are not completely unknown, the simple measurement of light exposure is not sufficient to produce new knowledge on the NIF visual system. Both in field and laboratory studies physiological measurements are needed to study the causality between biological effects and light. Within the scope of this work the physiological measures are only briefly reviewed. In reality when designing a dosimeter, the simultaneous measurement of measures such as ERG, EOG, EEG and ECG with light exposure should be taken into account (e.g. electromagnetic compatibility and space/weight restrictions). For a review of the physiology related to light exposure and NIF responses, the Masters thesis of the author is suggested [6]. In this work chapter 2 reviews briefly the circadian photobiology that is needed to understand the needs of the new measurement device needed to quantify the NIF responses. Chapter 3 reviews the literature on existing technologies for measuring NIF effective light exposure and the related measurements such as pupil size, eye tracking, electrical activity of eye, and NIF effective ambient luminosity distribution. In Chapter 4 the possible improvements and cost-cutting methods for the measurement are reviewed with some basic signal-to-noise ratio comparisons.

  • Circadian photobiology 6

    2. CIRCADIAN PHOTOBIOLOGY

    The whole concept of circadian rhythm is the most essential point of this work so its essential to clarify what it really means. Circadian (circa, meaning approximately and dies, meaning day) clock exists practically in every organism starting from cyanobacteria to humans [7-10]. Genuine circadian rhythms are generated totally endogenously without external cues (zeitgebers=time givers) like light. Light is thought to be the strongest zeitgeber, but in reality some weaker zeitgebers (nonphotic entrainment) can have even greater effect on human circadian rhythms in certain special cases. These weaker zeitgebers include social interaction, sleep/wake schedules, food, drugs, auditory and olfactory stimuli, temperature, and exercise [11]. In general circadian rhythms are advantageous to organisms to anticipate changes in the environment, such as the rising and setting of the sun. Other rhythms that affect our bodies include ultradian, which are cycles that are shorter than day, for example, the milliseconds it takes for a neuron to fire, or a 90-minute sleep cycle. Theres also infradian, referring to cycles longer than 24 hours like monthly menstruation for example. There are also seasonal rhythms (photoperiod) like those in hibernation, and reproduction [12]. People have noted the existence of daily rhythms throughout the history. In 1729, Jean Jacques DOrtous de Mairan, a French astronomer, had a sharp insight into how to test whether a daily rhythm is internal or completely dictated by external stimulus [13]. He studied this by keeping a heliotrope (plant) in the dark and noticed that its leaves kept opening and closing despite the lack of light. De Mairan however didnt conduct any further experiments on this subject. In the 1850s Karl von Frisch, Gustav Kramer and Colin Pittendrigh each independently discovered compelling evidence for internal clocks in animals [14]. These investigations were the beginning of the modern field of circadian photobiology. In the 1960s Jrgen Aschoff, a German researcher conducted a human study on circadian rhythms [15]. He built a bunker in which volunteers lived for several weeks. Most of the time they were totally isolated from real world without any sunlight, and occasionally the door to bunker was opened revealing the real world. Aschoff noticed in his study that even though volunteers were isolated, their sleep-wake cycles persisted. But he also found that their biological clock got desynchronized with the real world because their endogenous cycle wasnt exactly 24 hours, average periodicity being 24.5 hours. Aschoffs final conclusion was that when the door was reopened, the subjects readjusted their clocks meaning that it was possible to reset their circadian clock by external cues. Still Aschoff didnt know how this circadian clock works, and whether it could be located in a particular part of the body. Normally animals have been divided into nocturnal (active during night) and diurnal (active during day), but there exists also some adaptability to environmental situation at circadian level, which is the case with Finnish bats which are nocturnal mainly during the warm months of the summer. However, in the spring and fall, when environmental factors no longer favor night flight (fewer insects to eat during the colder nights, and fewer birds to compete with and to prey on bats during the day), their activity cycle shifts to the daylight hours [16,17]. In the 1970s scientists demonstrated that the mammalian clock is located in a part of the brain called hypothalamus, specifically in a set of neurons on each side of the

    Figure 1. Rhythm of Finnish bats. The gray region indicates the night period and the black bars show the actual activity time for the bats on each day [17].

  • Circadian photobiology 7

    brain called the suprachiasmatic nucleus (SCN). It was noticed that when removing the SCN from rats and hamsters, the animals lost their normal biological rhythms. However when transplanting the SCN back to into hamsters, the normal rhythms were restored. Recently experiments have shown that the SCN can be completely isolated from the animal and still measure chemical and physiological signals from it in vitro conditions. Later SCN and the other hypothalamic nuclei involved in regulation of circadian rhythms will be reviewed.

    2.1 CIRCADIAN RHYTHMS Cyclical fluctuations around 37C in core body temperature (CBT) are perhaps the best documented circadian rhythm. Gierse [18] had already shown in 1842 that his own oral temperature revealed a maximum temperature in the early evening and minimum in the early morning. Aschoff et al. [19,20] showed that circadian rhythm is caused both by changes in heat production and changes in heat loss, and concluded that heat production undergoes a circadian rhythm which is phase advanced by 1.2 h with respect to the circadian rhythm of heat loss, and this delay is caused by bodys inertia and because transport of heat takes time. This individual regulation of the heat production and heat loss results in much finer tuning of the CBT rhythm than if only one of these components were regulated [21]. It has been proposed [22] that body temperature represents the underlying mechanism regulating performance. The speed of thinking and performance depends on the level of metabolic processes in neurons in the cerebral cortex. However, the interrelationship between thermoregulatory and sleepiness/performance regulatory mechanisms is rather complex and not fully understood [21]. CBT can be easily measured continuously by using a rectal thermistor (e.g. Harvard Apparatus YSI 400 Series [23]). Another common circadian rhythm measured in chronobiological studies, is the circadian rhythm of melatonin hormone. Melatonin (C13H16N2O2; molecule weight232,278 g/mol), 5-methoxy-N-acetyltryptamine, is a hormone produced primarily by pinealocytes in the pineal gland (located in the brain) [24]. It can be considered to be a reliable marker of the circadian phase as it is secreted in very strict circadian manner peaking during the night. It is synthesized and secreted at night in both day-active and night-active species [25], thereby acting as a signal for the length of day and night. Despite its robust circadian behavior many mechanisms of melatonin are still unclear [26]. Abnormal melatonin levels caused by lighting at wrong biological time in night-shift workers have been connected to increased risk of breast cancer in women [27-30]. The typical circadian variations of plasma melatonin and core body temperature are seen in Figure 2 [31]. Melatonin levels are usually used as a

    Figure 3. Relative (%) circadian phase markers using melatonin. DLMO, dim-light melatonin onset, DLMOFF dim-light melatonin offset [33].

    Figure 2. Plots of (A) endogenous plasma melatonin, and (B) core body temperature with data folded at endogenous circadian period as determined by core body temperature for each subject. Abscissa refers to biological time which corresponds different clock time in every individual. In average minimum in CBT (CBTmin) occurs around 04:00 hours [31].

  • Circadian photobiology 8

    marker for the phase shifts in circadian rhythms. This phase-shift in practice means that light exposure can delay or advance the onset of nocturnal melatonin rhythm [32]. Typical methods to assess the time of nocturnal melatonin surge can be seen in Figure 3 [33]. Other circadian rhythms include the diurnal rhythms of cortisol [34], thyrotropin (TSH) [35], prolactin [36], vasopressin [37], and growth hormone (GH) [38] among many others.

    2.2 CIRCADIAN CLOCK MECHANISM In this chapter different brain regions and the hormones involved in the regulation of circadian rhythms are briefly reviewed. The information presented here should be taken with caution as all the presented areas require further research. Hypothalamus is a structure in the brain located below the thalamus and it regulates various metabolic and autonomic processed [39]. Given its central position in the brain and its proximity to the pituitary (Figure 4) it is involved as an integrator of both sensory and contextual information. Hypothalamus consists of various nuclei (Figure 5). A lot about the hypothalamus is still unknown, but some actions are at least partially understood and can be described at basic level. Suprachiasmatic nuclei (plural form of nucleus) are nuclei in the hypothalamus situated immediately above the optic chiasm (Figure 5) on either side of the third ventricle in anterior hypothalamus. The SCN is one of

    four nuclei that receive nerve signals directly from the retina through retinohypothalamic tract (RHT, Figure 4); the others are lateral geniculate nucleus (LGN), the superior colliculus and the pretectum. In the 1970s the biological clock was located in SCN [40,41], and it was shown that SCN contain genetically driven clock mechanism that ensures a nearly 24 hour cycle [42]. Precise estimation of the periods of the endogenous circadian rhythms of melatonin, core body temperature, and cortisol in healthy individuals living in carefully controlled lighting conditions indicates that the intrinsic period of the human circadian pacemaker averages 24.18 hours with a tight distribution that is consistent with other species [43]. Circadian rhythmicity is abolished by SCN lesions [41] and restored by SCN transplants [44]. Traditionally SCN has been subdivided into a dorsomedial shell (dmSCN) and a ventrolateral core (vlSCN) based on retinal innervation and phenotypically distinct cell

    Figure 5. The hypothalamus, showing the location of the suprachiasmatic nucleus (SCN), which in mammals is the primary biological clock. [39]

    Figure 4. Schematic summary of targets influenced by photosensitive retinal ganglion cells. Projections to the SCN from the retinohypothalamic tract (RHT) [39].

  • Circadian photobiology 9

    types [45,46], while this subdivision has also been criticized for simplifying the SCN structure [47]. Intrinsically rhythmic cells are largely confined to the SCN shell [48], receive little retinal innervation [46], and displays delayed clock gene expression following phase-shifting light exposure [49]. Cells in the SCN core receive direct retinal innervation [50] and express c-fos, Per1 and Per2 in response to phase-shifting light pulses [51-53]. Cells in the SCN core oscillate in response to light stimulus. Light exposure always increases firing rates in SCN neurons [54], although light induces clock gene expression in the SCN only during the night [55]. The simplified assumption that SCN is responsible solely for circadian rhythms is inadequate for in-depth understanding of the human circadian rhythms. Currently human circadian rhythms are thought to be controlled via multioscillator organization hypothalamic nuclei [56-58]. SCN provides three major output pathways. One pathway runs into the medial preoptic region (MPO) and then up into paraventricular nucleus of thalamus. A second pathway runs to the retrochiasmatic area and the capsule of the ventromedial nucleus. The third pathway, which contains the largest portion of the SCN efferent (going away, opposite is afferent) flow, runs mainly to vSPZ and dSPZ with smaller proportion terminating to the DMH. Also small numbers of SCN axons innervate directly the areas that are involved in feeding, wake-sleep cycles and secretion of hormones such as melatonin (presumably through dorsal parvicellular portion of the paraventricular nucleus [59]) and corticotrophin-releasing hormone (CRH) [60]. The further examination of circadian clock mechanism is beyond the scope of this work.

    2.3 PHYSIOLOGY OF THE EYE The simplified anatomy of an eye is shown n Figure 6 [61]. The pupil allows light to enter the eye. It appears dark because of the absorbing pigments in the retina. The pupil is surrounded by beautifully pigmented iris, which is a circular muscle controlling the amount of light entering the eye. Both pupil and the iris are covered by a transparent external surface called the cornea. This is the first and most powerful lens of the optical system of the eye and allows, together with the crystalline lens the production of a sharp image at the retinal photoreceptor level. The purpose of the lens is to focus light onto the back of the eye. The lens is encased in a capsular-like bag and suspended within the eye by tiny guy

    wires called zonules. The cornea is continuous with the sclera, the white of the eye, which forms part of the supporting wall of the eyeball. Furthermore this external covering of the eye is in continuity with the dura of the central nervous system. The sclera and the cornea form the external layer of eye.

    Figure 6. a) Vertical, and b) horizontal sagittal section of the adult human eye [61].

    a) b)

  • Circadian photobiology 10

    Retina is the sensory part of eye and part of the central nervous system. The central point for image focus (the visual axis) in the human retina is the fovea. The optic axis is the longest sagittal distance between the front or vertex of the cornea and the furthest posterior part of the eyeball. It is about the optic axis that the eye is rotated by the eye muscles. In the center of the retina is the optic nerve, a circular oval white area. From the center of the optic nerve radiate the major blood vessels of the retina. Approximately 17 degrees (4.5-5 mm), or two and half optic disc diameters to the left of the optic disc (or optic nerve head is the point in the eye where the optic nerve fibers leave the retina), can be seen the slightly oval-shaped, blood vessel-free reddish spot, the fovea, which is at the center of the area known as the macula. It is a small and highly sensitive part of the retina responsible for detailed central vision. A circular field of approximately 6 mm around the fovea is considered the central retina while beyond this is peripheral retina stretching to the ora serrata. The optic nerve contains the ganglion cell axons running to the brain and, additionally, incoming blood vessels that open into the retina to vascularize the retinal layers and neurons. A radial section of a portion of the retina reveals that the ganglion cells (the output neurons of the retina) lie innermost in the retina closest to the lens and front of the eye, and the classical photosensors (the rods and cones) lie outermost in the retina against the pigment epithelium and choroid (Figure 7A [62]). All vertebrate retinas are composed of three layers of nerve cell bodies and two layers of synapses (Figure 7B). The outer nuclear layer (ONL) contains cell bodies of the rods and cones, the inner nuclear layer (INL) contains cell bodies of the bipolar, horizontal and amacrine cells and the ganglion cell layer contains cell bodies of ganglion cells and displaced amacrine cells. Between these layers are areas called neuropils where synaptic contacts occur.

    Traditionally cones and rods have been thought to be the only photoreceptors in mammalian retina, but after the discovery of the novel photoreceptor ipRGCs the exact roles of all three photoreceptors are not fully understood. The rod system is specialized for vision at very low light levels, but with the expense of poor spatial resolution. When only rods are activated the perception is called scotopic vision. With only rods active it is impossible to neither sense color differences or to make exact visual discriminations. The cone system has a very high spatial resolution, with color sensing abilities in the expense of poor light sensitivity. At about the level of starlight the cones begin to contribute to vision and they become more and more dominant as light level increases. At very high light levels such as in sunlight, only cones are active and rods are totally saturated [63]. This condition is called photopic vision. The area between scotopic and photopic vision is called mesopic

    (A)

    Figure 7. (A) Simple diagram of the organization of the retina. (B) 3-D block of a portion of human retina. [62]

    (B)

  • Circadian photobiology 11

    vision, which is characterized by contribution of both rods and cones. The estimated upper luminance limit for mesopic vision is 3-10 cd/m2 [64]. Spectral sensitivities for photopic, mesopic and scotopic vision can be seen in Figure 8 [65]. The retina contains about 20 different retinal ganglion cells (RGCs) [66], which basically are responsible for the output of visual data to the brain. At the basic level ganglion cells can be divided in two ways, either by their receptive field with the division to magno (M-) and parvo (P-) cells, or by their polarization response to light (ON and OFF cells). M-RGCs terminate in the magnocellular layer of the lateral geniculate nucleus (LGN) of the thalamus, and P-RGCs terminate in the parvocellular layer of the LGN. The conventional view was that ganglion cells got their commands from rods and cones, and ganglion cells did not have any light-sensitive properties themselves [1]. In the beginning of 1980s, however, behavioral studies especially those of Foster and colleagues, began to challenge this model [67]. Photic entrainment exhibited high thresholds, low-pass temporal filtering and long-term temporal integration that seemed difficult to explain with the conventional model of cones and rods. This was backed up by studies made with blind mice [68-70] with severe degeneration of classical photoreceptors as well as studies done with certain blind humans [71]. However, it was not clear at all that the receptor for circadian phase would be found from the eye. In non-mammalian animals light penetrating directly to brains acts as circadian pacemaker. In mammals, however, many studies were made and no impact on circadian phase could be shown after eye removal [72-75]. Interestingly, one study reported a bright light behind the knee phase-shifting circadian rhythm [76], but the results could not be replicated [77,78] making this explanation a bit unlikely at the moment. The discovery of circadian photoreceptor was at last made by David M. Berson et al. [1]. The novel photoreceptor is abbreviated as ipRGC (intrinsically photosensitive retinal ganglion cells), or as mRGC (melanopsin-containing retinal ganglion cell, mRGC) due to the photopigment responsible for the noticed non-image forming (NIF) effects. Melanopsin was first discovered by Ignacio Provencio and his colleagues [79,80], and is named by the cells in which it first was isolated: the dermal melanophores of frog skin. The two main differences of ipRGCs compared to cones and rods, are that light depolarizes ipRGC while the opposite happens with rods and cones; and ipRGCs are far more sluggish compared to rod and cones, response latencies being as long as a minute. The results are not consistent about the peak wavelength of melanopsin-pigment. Qiu et al. [81], and Panda et al. [82] show that melanopsin max is very close to 480 nm, but Melyan et al. [83] and Newman et al. [84] suggest that melanopsin has max closer to 420-430nm. The most likely explanation for this kind of large difference was that Newman et al. [84] were the only ones who determined the direct absorption spectrum of melanopsin in vitro conditions whereas all the other studies were done in vivo conditions [85]. Peak absorption spectrum of 420-430 nm might well be the intrinsic peak wavelength for melanopsin, but it would not be the actual peak wavelength responsible of the wide range of the biological effects mediated by ipRGCs. There is also some preliminary evidence that some cones contain also melanopsin and are involved in circadian phototransduction [86].

    Figure 8. Spectral sensitivity functions of the eye. In photopic vision, when cones are active, the sensitivity follows the function V() with a peak wavelength of 555nm. At very low light levels only rods are active, and spectral sensitivity follows V()-function with a peak wavelength of 505nm. The Vmes() is one example of the possible mesopic spectral sensitivity as no consensus exists on it yet. The V10() is is the photopic spectral sensitivity for centrally fixated large target [65].

  • Circadian photobiology 12

    2.3.1 Ophthalmological optics

    Figure 9 shows the human visual fields, which are divided first to monocular (one eye) and binocular visual fields (two eyes), and then further into superior/inferior and nasal/temporal visual fields. Figure 9B shows how the image is inversed onto the surface of the retina, and how the different quadrants of monocular visual fields are related to the binocular vision. It is important to notice from Figure 9C that light reaching nasal (inner) part of the retina is coming from peripheral visual field and vice versa, and the same thing happens with superior/inferior visual fields, where the object (e.g. sky) in the superior visual field is projected to the inferior part of the retina [39]. Binocular visual field is larger than either of the monocular visual fields. Forehead, nose and cheeks limit visual field so that it is larger horizontally than vertically. Binocular visual field is horizontally about 190, and below the horizontal level about 70-80 and above 50-60 [87]. It should be noted that the human visual field is much larger than normal 35mm lens used in cameras. However, visual processing is not uniform across the visual field: 25% of cortex is devoted to the central five degrees of the field of view [88].

    Figure 9. Projection of the visual fields onto the left and right retinas. (A) Projection of an image onto the surface of the retina. The passage of light rays through the optical elements of the eye results in images that are inverted and left-right reversed on the retinal surface. (B) Retinal quadrants and their relation to the organization of monocular and binocular visual fields, as viewed from the back surface of the eyes. (C) Projection of the binocular field of view onto the two retinas and its relation to the crossing of fibers in the optic chiasm. Points in the binocular portion of the left visual field (B) fall on the nasal retina of the left eye and the temporal retina of the right eye. Points in the binocular portion of the right visual field (C) fall on the nasal retina of the right eye and the temporal retina of the left eye. Points that lie in the monocular portions of the left and right visual fields (A and D) fall on the left and right nasal retinas, respectively. [39].

  • Circadian photobiology 13

    Human eye as an optical instrument is briefly reviewed here as the actual retinal illuminance or irradiance depends on the optical characteristics of the eye. Figure 10B [89] shows the simplified version of the human eye as an optical system. The size of the pupil determines (pupil diameter dp) the light entering the eye. Figure 10A [89] shows the wavelength dependent transmittance (from cornea to retina) and retinal absorption. Transmittance is important for the actual retinal irradiance whereas retinal absorption affects the amount of retinal damage from light exposure.

    Between the corneal irradiance Ec, the retinal irradiance Er, and the radiance of the source Ls, the following relation exists [89]:

    p

    rr

    r

    r

    2s

    2

    2p

    2s

    2p

    2r

    2s

    sssc AAE

    AP

    r

    Afd4

    r

    Ad

    f4Er

    ALLE

    =

    pi=

    pi=== (1)

    r

    pcs2

    pr A

    AEL

    f4d

    E

    =

    pi= (2)

    Where, Ec = corneal irradiance Ls = source radiance s = solid angle [sr] As = source area r = distance between the source and lens Pr = retinal power = lens transmittance dp = pupil diameter f = focal length, can be estimated to be 1,7 cm [89,90] Er = retinal irradiance Ar = area of the image at the retina Ap = pupil area

    The image size (diameter) of the source at the retina can be calculated quite simply [89]:

    fr

    fdd sr == (3)

    Where, dr = size of image at the retina ds = size of the source f = focal length, can be estimated to be 1,7 cm [89,90] r = distance between the source and lens = angular subtense of the source

    Transmittance

    Absorption

    %

    Figure 10. (A) Transmittance of optical radiation from cornea to retina and the absorption at the retina [89]. (B) Eye as an optical system [89].

    (A) (B)

  • Circadian photobiology 14

    The size of the image at the retina depends on distance between the source and lens. In practice the retinal irradiance can de different while the corneal irradiance is the same. Larger angular subtense produces smaller retinal irradiance and smaller then consequently produces larger retinal irradiance [91]. This is why the corneal irradiance (or illuminance) should always be controlled when conducting experiments by neutral density filter rather than moving light sources further from eye as circadian responses ultimately depend on the retinal irradiance. Figure 12 shows the wavelength dependent average transmittance of human lens [92-95]. It can be seen that in the visible part of spectrum newborn lens does not have significant wavelength dependence. In the age group of 20-29 years the transmittance of the blue part of the visible spectrum is slightly attenuated, and in the age group 60-69 years the attenuation is really significant due to yellowing of the lens. In visual responses the human brain can compensate the attenuation of the blue light in a manner that the world does not appear to be less blue for the older people [96,97]. The transmittance of the human lens in different age groups accompanied with the spectral transmittance of intraocular lens (used after cataract removal surgery) [98], the proposed melatonin suppression curve, and the cornea is shown in Figure 11 [99]. Corneal spectral transmittance remains relatively constant in aging as supported by the study by Beems et al. [100] that the corneal transmission for donors younger than 45 yr (n = 3, 2243 yr) did not differ significantly from that of donors older than 45 yr (n = 5, 6787 yr) at any wavelength.

    Figure 12. The average transmittance of human lenses for three different age groups as a function of wavelength [92,95].

    WAVELENGTH

    %

    Figure 11. Transmittance data for lens: 14 years (); 49 years (+); 92 years () (after Weale, 1985 [93]); mean lens data (X) (after Stockman and Sharpe, 2000 [63]); intraocular lens (*) (after Mainster, 1986 [98]); cornea () (after Beems and van Best, 1990 [100]). The heavy continuous curve shows the relative sensitivity of the presumed photopigment (after Thapan et al., 2001 [116]. Graph from Charman [99].

  • Circadian photobiology 15

    2.3.2 Pupil pathways

    It was assumed in earlier days that pupil light response (PLR) was driven by a single subcortical pathway and this was because of persisted pupillary light reflex in cortically blind people [101]. Recently however this hypothesis has been replaced with a theory of bilateral signaling involving two different pathways [102]. Pupillary reflexes have been divided into steady-state pupil size depending on the ambient light level, and brisk and transient constriction of pupil size depending on rapid changes in light flux, which is also described as dynamic PLR response. Shining light in the eye thus leads to an increase in the activity of pretectal neurons, which stimulates the Edinger-Westphal neurons and the ciliary ganglion neurons they innervate, thus constricting the pupil. Pupil-related pathway is shown in Figure 13.

    Pupil size is determined by iris movement, which is controlled by two antagonistic muscles, the sphincter and the dilator. Activation of the sphincter of the iris causes the pupil to constrict (i.e., miosis), this being largely under parasympathetic control and involving ciliary ganglion. Dilator is under sympathetic control, and causes the pupil to dilate (mydriasis) through superior cervical ganglion controlled by EW nucleus. Sympathetic system is associated with fight-or-flight responses with epinephrine and norepinephrine stimulation. Parasympathetic system is the opposite and it is sometimes called the rest and digest system for its ability to relax and slow down the functions of organs (slowing heart beat and increasing its constricting power). In practice general arousal through increased sympathetic activity will cause pupil dilation independent of the ambient light level, and vice versa. Like with all lens systems, the size of the pupil

    Figure 13. The circuitry responsible for the pupillary light reflex. This pathway includes bilateral projections from the retina to the pretectum and projections from the pretectum to the Edinger-Westphal nucleus. Neurons in the Edinger-Westphal (EW) nucleus terminate in the ciliary ganglion, and neurons in the ciliary ganglion innervate the pupillary constrictor muscles. Notice that the afferent axons activate both Edinger-Westphal nuclei via the neurons in the pretectum [39].

    Figure 14. Example of dynamic pupil light reflex responses to flashes of increasing luminance contrast, i.e., L/Lb = 0.3, 0.6, 0.9, 1.2, 1.5 & 2.15 [102].

  • Circadian photobiology 16

    (aperture) determines the amount of light entering to retina (retinal illuminance, RI). Pupil size also controls the aberrations and depth of the field of the eye in the same manner as in camera. A smaller aperture (larger f-value) will enhance depth of field and reduce aberrations [103]. PLR response can be divided into steady-state and transient (dynamic) component. The steady-state component is determined by ambient light level and is characterized by neural mechanisms that response to overall light flux changes, large dynamic range and exhibit large spatial summation. Dynamic PLR response is very rapid to rapid light flux change as seen in Figure 14 [102]. Observed transient constriction would need from neurons the following properties: limited spatial summation, band-pass temporal response characteristic, and high contrast gain. A light stimulus always depends on both components, but the relative contribution each component makes to constriction will depend on size of the stimulus, its luminance contrast, onset temporal characteristics and location in the visual field. As observed in Figure 14 pupil constriction is greater with higher luminance contrast, when pupil response is more dominated by steady-state component. Participation of steady-state component could be even further increased with larger stimulus size. Despite involvement of both rod and cone photoreceptors in determining pupil size [104-106], there is an increasing amount of evidence pointing out that ipRGCs play some role in pupillary controls as functional pupillary light reflex (PLR) has been shown to be retained in rodent models of retinal generation (impaired cone/rod function) [107-111]. The results of the spectral sensitivity for the pupillary reflex obtained by Alpern and Campbell [105] can be seen in Figure 15. It can be seen that photopic pupil response is close to the photopic spectral efficiency curve V() and the scotopic pupil response curve is close to the scotopic spectral efficiency curve V(). However, pupil size has been noticed to be smaller under light with higher CCT [136] (8000 K) compared to light sources with CCT=4100 K slightly in contrast with the curves presented in Figure 15. Pupil responses have also been noticed to be larger on exposure of the nasal part of the retina (temporal visual field) [112], having similar spatial characteristics as melatonin suppression (as later noticed) [150].

    Figure 15. (A) Mean spectral sensitivity curve for the photopic pupil response (max 550 nm) of two subjects. Differential threshold measurements () are plotted for 2 sec flashes of a 2 test patch centrally fixated and seen against a continuous blue background. Interrupted line, CIE photopic luminosity curve; solid line, mean results of psychophysical measurements of photopic luminosity (flicker photometry) on the same two subjects with the same apparatus. (B) Solid line-Deviations of the pupil results under scotopic conditions (max 500 nm) from the CIE spectral sensitivity data (25 V criterion) and corrected for the absorption in the eye media (double passage); , the b-wave of the ERG of the dark-adapted eye. Mean results from two observers [105].

    (A) (B)

  • Circadian photobiology 17

    2.3.3 Eye movements

    There are four basic types of eye movements: saccades, smooth pursuit movements, vergence movements, and vestibule-ocular movements. They all have their own controlling neural circuitry. Eye movements are very important as high visual acuity is restricted only to fovea, and eye is always trying to direct fovea to new objects of interest (foveation). Russian physiologist Alfred Yarbus demonstrated in his experiments in the 1960s the pattern of eye movements while examining an object [113]. Yarbus used contact lenses with small mirrors attached to them to track eye movements. Results can be seen in Figure 17 revealing subjects gaze while viewing a bust of Queen Nefertiti. Thin lines represent the quick, ballistic movements (saccades) and the denser spots represent points of fixation where the observer paused to take in visual information (only a few tens of milliseconds). First types of eye movements, saccades, are rapid, ballistic movements that abruptly change the point of fixation. The amplitude of saccades can range from small correction movements (with reading) to larger (gazing around a room) movements. The rapid eye movements during REM-sleep are also saccades. Time behaviors of saccades are illustrated in Figure 16, which shows that there is about 200ms delay if an already fixated target starts to move. This delay is used to compute an appropriate correction and if the target keeps moving, another computation is needed. This is the main problem with saccades as both the amplitude (how far) and the direction of the movement should be computed as accurate as possibly. The amplitude of the movement is controlled by firing duration of lower motor neurons of the oculomotor nuclei. Figure 18 shows the control of horizontal movement using lateral and medial muscles. The direction of the movement is determined by which eye muscles are activated. In principle any given direction could be controlled just summing different eye muscle activity, but in reality the complexity of such mechanism would be great. Basically the control has been divided into two gaze centers: paramedine pontine reticular formation (PPRF) or a horizontal gaze center; and rostral interstitial nucleus or vertical gaze center. Centers can be separately activated and the rotational movements are determined by relative contribution of each center.

    Figure 17. The eye movements of a subject viewing a picture of Queen Nefertiti. The bust on the left is what the subject saw; the diagram on the right shows the subject's eye movements over a 2-minute viewing period [39].

    Figure 16. The metrics of a saccadic eye movement. The red line indicates the position of a fixation target and the blue line the position of the fovea. When the target moves suddenly to the right, there is a delay of about 200 ms before the eye begins to move to the new target position [39].

    Figure 18. Motor neuron activity in relation to saccadic eye movements. The experimental setup is shown on the right. In this example, an abducens lower motor neuron fires a burst of activity (upper trace) that precedes and extends throughout the movement (solid line). An increase in the tonic level of firing is associated with more lateral displacement of the eye [39].

  • Circadian photobiology 18

    Computation of the movements does not take place at gaze centers as they get their input from the superior colliculus of the midbrain and region called frontal eye field (Brodmanns area 8) as seen in Figure 19. Both areas respond to visual stimuli and have specific visual and motor maps equivalent to retinotopic mapping. The responses of superior colliculus are better known than frontal eye field. The simplified relation between superior colliculus and frontal eye field is the following: the frontal eye field projects to the superior colliculus and the superior colliculus projects to the PPRF on the contralateral side (Figure 19), as it does also to vertical gaze center which is excluded from the picture for the sake of clarity. The frontal eye field then controls the eye movements by activating selected populations of superior colliculus neurons. It can also project directly to PPRF and control eye movements independently of the superior colliculus. Frontal eye field is also responsible for systematic scanning of visual field to locate an object of interest from background noise. It was thought in early 1970s when the collicular maps were found that saccadic movements could be easily estimated using visual/motor map matching. However later it has been found that saccade movements dont necessarily even need visual stimuli. As seen in Figure 20, nonvisual stimuli like auditory or somatic stimuli can activate motor neurons and produce saccade movements. Also it has been discovered that animals can be trained not to make saccades when an object appeared to visual field, which led to a development of more complex models as seen in Figure 20. We can see that theres a direct connection between motor and visual neurons, which probably provide the substance for the very short latency (~100ms) reflex-like express saccades, which have been notices even after the destruction of the frontal eye fields. The second type of eye movement, smooth pursuit movements are much slower tracking movements designed to keep a moving stimulus on the fovea. Smooth pursuing movements are under voluntary control as person can decide whether to follow some object or not. However only highly trained individuals can make smooth pursuing movement without actual moving target to follow, in reality most people just end up making a saccade. Traditionally these movements were tested placing a subject inside a rotating cylinder with vertical stripes, but nowadays same test can be done using a screen with series of horizontally moving vertical stripes. The eyes follow the stripe end of their excursion followed by a quick saccade to opposite direction for a pursuit of new stripe. This kind of mixed fast and slow movement of the eyes is called optokinetic nystagmus. This is illustrated in Figure 22, where after a quick saccade eyes are able to follow the moving target smoothly. Vergence movements align the fovea of each eye with targets located at different distances from the

    Figure 19. The relationship of the frontal eye field in the right cerebral hemisphere (Brodmann's area 8) to the superior colliculus and the horizontal gaze center (PPRF) [39].

    Figure 20. The superior colliculus receives visual input from the retina and sends a command signal to the gaze centers to initiate a saccade. The terminals of the visual neuron are located in the same region as the dendrites of the motor neuron [39].

  • Circadian photobiology 19

    observer. But unlike other eye movements, vergence movements are disconjugate (or disjunctive) meaning that eyes move to opposite directions, converging for close objects and diverging for far objects. Convergence caused by near-field stimuli (or near reflex triad) involves also pupillary constriction to increase depth of field. Vergence movements are the slowest speed eye movement although latency being less than with saccades. They are also very small in amplitude, typically a few degrees. The last types of eye movements, vestibulo-ocular movements, mean the compensation of eyes to movement of head. When tilting your head you notice that your fixating point remains more or

    less at same point of your retina. The name vestibulo comes from vestibular system, which main element is vestibular nuclei that is situated in inner ear acting as accelerometer and spatial position guide. The system extends through a large part of the brainstem; simple clinical tests like the vestibulo-ocular response can be used to determine brainstem involvement and possible damages, even on comatose patients. The vestibular system detects brief, transient changes in head position and produces rapid corrective eye movements. However it is relatively insensitive for slow changes. For example if the vestibulo-ocular reflex is tested with continuous rotation and without visual cues about the movement of the image (i.e., eyes closed), the compensatory eye movements cease after only about 30 seconds. A person with vestibular damage finds it difficult or impossible to fixate on visual targets while the head is moving, a condition called oscillopsia (bouncing vision).

    2.4 LIGHT CHARACTERISTICS In this chapter the light characteristics linked to the novel photoreceptor are reviewed in regard to human circadian rhythms. The basic understanding of this chapter is essential in designing the measurement equipment for light exposure.

    2.4.1 Spectrum

    The peak wavelength of circadian responses is shifted towards the blue end of the spectrum compared to the traditional visual spectral sensitivities for photopic (V(), max=555nm), mesopic (max between photopic and scotopic peak wavelengths) and scotopic (V(), max=508nm) vision. According to current knowledge, the peak wavelength seems to be around 480 nm [2] for ipRGCs. A series of action spectra are presented in Table 1 concurrent with the discovery of melanopsin (Provencio et al. [114], 1998) and ipRGCs (Berson et al. [1], 2002).

    Figure 22. The metrics of smooth pursuit eye movements. These traces show eye movements (blue lines) tracking a stimulus moving at three different velocities (red lines). After a quick saccade to capture the target, the eye movement attains a velocity that matches the velocity of the target [39].

    Figure 21. Vestibulo-ocular eye movement (slow) resulting from head rotation. This slow component is also called physiological nystagmus, Fast eye movement are saccades that reset the eye position [39].

  • Circadian photobiology 20

    Table 1. Analytic action spectra for circadian, ipRGC, and ocular responses (modified from Brainard, 2006 [85]).

    Species Biological responses Stimuli tested Peak [nm]

    First Author Year

    Human (Homo sapiens) Plasma melatonin suppression

    8 fluence-response curves (hbw 10-15nm)

    Est. max=464 (446-477)

    Brainard [115] 2001

    Human (Homo sapiens) Plasma melatonin suppression

    6 fluence-response curves (hbw 5-13nm)

    Est. max=480 (457-462)

    Thapan [116] 2001

    Mouse (Mus musculus) Pupillary light reflexes 6 fluence-response curves (+/+) (hbw 10nm) (rd/rd cl)

    Est. max=480 or 508

    Lucas [117] 2001

    Human (Homo sapiens) Cone cell ERG-wave 7 fluence-response curves (hbw 10nm) Est. max=479

    Hankins [118] 2002

    Rat (Rattus norvegicus) ipRGC cellular depolarization

    6/10 fluence-response curves (hbw 10nm) Est. max=484

    Berson [1] 2002

    Mouse (Mus musculus) Circadian phase shift 7 fluence-response curves (rd/rd cl) (hbw 10nm) Est. max=481

    Hattar [119] 2003

    Mouse (Mus musculus), purified mouse melanopsin in vitro

    Melanopsin-catalyzed GTP--35S uptake

    Single irradiances of 4 restricted bandwidths (hbw 10-30nm)

    Est. max=424 (420-44)

    Newman [84] 2003

    Monkey (Macaque nemestrina)

    ipRGC cellular depolarization

    10 fluence-response curves (hbw 15-20nm) Est. max=482

    Dacey [2] 2005

    Wild-type and retinally degenerate strains are indicated by (+/+) and (rd/rd cl). Est. = Estimated max from fitting data to spectral sensitivity curves or to visual photopigment nomograms. hbw =half-bandwidth (hbw smaller than 10nm is considered monochromatic).

    The recent range for max of circadian responses has been from 459 to 484 nm with the clear exception of 420 nm by Newman et al. [84], which study was done in vitro conditions and does not necessarily represent the in vivo behavior of melanopsin as already reviewed with melanopsin. Neither the study by Lucas et al. [117] does not identify a max in the blue part of the spectrum in pupillary responses of wild-type (+/+) mice, which was also found in earlier studies [120-122] for phase shifting locomotor activity. It could be that the intact rodent retina combines input from ipRGCs and classic visual photoreceptors (cones more likely) for phase shifting and pupillary responses. In contrast when mice do not have functional cones or rods, their retinal sensitivity appear to shift towards shorter wavelengths [119,122,123]. In photometry, Abney's Law for additivity [124] has been used as a hypothesis for the linear behavior of luminance perception. Additivity means that the total luminance of a non-monochromatic light is the sum of the weighted spectral radiations of the component wavelengths. However, additivity does not hold for all lighting conditions. Additivity failures occur both in photopic and mesopic vision [125], also referred as the Abney Effect [126], that recognizes the failure of the basic law. To make things even more complex this Abneys Effect is known to be in error also [127]. In photopic vision, additivity failure called sub-additivity occurs when the perceived brightness is less than the sum of the component perceived brightnesses. This phenomenon is apparently due mainly to non-linear cone-cone interactions and is also called the Helmholtz-Kohlrausch effect [128]. For example mixing monochromatic red light with monochromatic green light of equal brightness can be seen less bright than either of the two lights alone [129]. In mesopic vision only the magnocellular channel appears to obey Abney's law of additivity [130]. The first research by Figueiro et al. [131] studying the circadian spectral opponency in humans compared the melatonin suppression effects of blue light emitting diodes (LEDs) and clear mercury (Hg) vapor lamps. Blue LEDs produced an illuminance of 18 lx (29

  • Circadian photobiology 21

    W/cm2) at subjects eyes when the 175 W Hg lamp produced an illuminance of 450 lx (170 W/cm2) at the eye. The radiant power of polychromatic Hg lamp was set to produce at least equal or higher melatonin suppression than the blue LED if additivity was to exist in circadian response, following the univariance principle [132]. Results revealed a statistically significant difference between the LED and Hg lighting conditions, with the LED condition resulting stronger melatonin suppression in contrast to the theory of additivity. The best-fitting function from the results is shown Figure 23 which are relatively close match to the empirical action spectrum for melatonin suppression by Brainard et al. [225] (r2=0,86) and by Thapan et al. [116] (r2=0,84) both from 2001. In conclusion the larger melatonin suppression by photopically less-powered LED would indicate that spectral opponency exists in human circadian system, and the results from studies done with monochromatic light sources could not be generalized to normal polychromatic sources used in architectural lighting. The first circadian phototransduction model to incorporate the suggested spectral opponency [131] was presented by Rea et al. [133] in 2005. Compared to the previous models [91,134,356], it is much more ambitious while still maintaining relatively simple mathematical format. It is not limited on modeling the ipRGCs or melanopsin, but it incorporates the basic mechanisms of other retinal neurons involved in circadian phototransduction, as cones and rods are also been proposed to be involved in circadian responses [119,135]. However, while the model is based upon a synthesis of a wide range of existing literature in neuroanatomy, electrophysiology, and psychophysics, main emphasis is on the results got on melatonin suppression. Rea et al. [133] admit that this model is highly likely to be changed as it lacks the more advanced features of circadian phototransduction, but it still is a large step towards more realistic models. Figure 24 shows the action spectra of the proposed model. The proposed model [133] was later tested by the same authors [136] with the results showing a relatively good fit to the proposed model when tested with two polychromatic light sources. The conventional views has been after the discovery of ipRGCs that all non-image forming (NIF) functions have the same action spectra as it for example found that short-wavelength light (460 nm) is more effective in alertness-promoting than light at 550nm [137-139], but a recent study by Revell et al. [140] revealed that light at 420 nm was more effective in alertness-promoting than light at 470 nm. This would mean that the action spectrum presented for melatonin suppression [116,133,225] would not be accurate for alertness promotion. This could mean that human melanopsin could be really most sensitive to short wavelengths at 420-430

    Figure 23. Hypothetical opponent action spectrum for melanopsin consistent with the present [131] and previous results [116,225]. Curve from Figuiero et al. [131].

    Figure 24. Predictions of the model to the constant criterion spectral sensitivity data of Brainard et al. [225] and of Thapan et al. [116]. Graph by Real et al. [133].

  • Circadian photobiology 22

    nm as shown in some studies [83,84]. However, this peak wavelength of 420-430 nm is in contrast with the other data from melanopsin action spectrum [2,81,82], cone ERG [118], and circadian phase shifting [141]. Human eye undergoes age-related changes in total and in wavelength-dependent transmittance. It would be natural to assume that these changes would have some kind of impact to circadian phototransduction as well. As in both early studies by Brainard et al. [225] and Thapan et al. [116] the action spectra were corrected for absorption in the lens (as authors wanted to obtain an action spectrum that applies to the irradiance level at the retina), it is reasonable to study what are the real differences in circadian responses due to properties of cornea, aqueous, lens and vitreous, which light has to pass before reaching the retina. As noticed already in Figure 11, corneal transmittance is relatively constant between the range of 400 nm to 600 nm and above all the corneal transmittance does not differ significantly as a function of age at any wavelength [100]. Although absorption and scatter in humours may have minor effects, the most relevant part of the human eye is the crystalline lens as it has been noted to yellow with age thus attenuating short-wavelength light. Another significant age-related change in ophthalmologic optics is the senile miosis [142], where pupil diameter changes with age under both light-adapted (diameter decreases) and dark-adapted (diameter increases) conditions. The relative pupil area has its maximum value at the age of about 15 years and is reduced throughout adult [99], and it this factor alone reduces the retinal illuminance to half in the eye of a 70-year-old. The following equation can be used to calculate the relative efficiency of light for suppression of melatonin with regard to the age according to Charman et al. [99]:

    = d )S(A)(T)(T)(ER LC (4) Where, R = effective irradiance at the retina

    E() = spectral irradiance of the source at the cornea TC() = transmission of the cornea (near-axial path) TL() = transmission of the lens (near-axial path) A = pupil diameter (near-axial path) S() = melatonin action spectrum

    According to this formula, the efficiency of light should decrease with increasing age; however an experimental study failed to verify this assumption [143] and concluded that there is no correlation between efficiency of melatonin suppression and age of the subjects. No gender-related differences in melatonin suppression have been yet discovered [144]. The values of TC(), TL() and A are measured when light enters the eye along a near-axial path assuming that no effects of the Stiles-Crawford type [145,146] will occur, as the ganglion cells lie anterior to the outer segments of the receptors which are responsible for any waveguiding effects.

    2.4.2 Spatial distribution

    Relatively little is known about the spatial distribution of melanopsin-containing ganglion cells (ipRGCs) in human eye. This knowledge is important in knowing where to place the lights in order to produce the maximal biological responses. The results from various studies [147-150] indicate that a significant gradient in density of melanopsin-containing retinal ganglion cells is present both in the horizontal and in the vertical direction. The highest density of melanopsin-containing retinal ganglion cells (ipRGCs) would seem to occur in the inferior nasal area of the retina corresponding to upper (superior) temporal (lateral) visual field. The ratio between temporal retina and nasal retina for melatonin

  • Circadian photobiology 23

    suppression was 0.54 in the study by Visser et al. [147] and 0.59 in the study by Rger et al. [150]. The difference was even larger between upper (superior) and lower (inferior) retina in the study by Glickman et al. [149] where melatonin suppression was ~6,3% for upper retina whereas it was ~29,1% after a 90 minute 200 lx polychromatic light exposure (percentages plotted from the graphs). This would indicate a ratio of 0.22 between upper and lower ratio.

    2.4.3 Intensity

    Despite of the relatively large amount of studies on circadian phototransduction, only a few systematic studies [157,192] have been done on the influence of light intensity on the phase shifting and melatonin suppression. The early studies done with subjects which allowed to self-select their sleep-wake cycle showed that only bright light could affect human circadian rhythms [151,152], one study reporting a threshold of 1500 lx [153] and the others showed a significant phase shift with illuminances as high as 4000 lx [154] and 5000 lx [155]. In the human study by Boivin et al. [156] the phase resetting response was reported to increase with light intensity in a nonlinear manner. In the study by Zeitzer et al. [157] the intensity response curve (IRC) between illuminance and the phase resetting response was also found to be nonlinear. This found nonlinearity is consistent with a cube-root compression of illuminance as a function of the illuminance and (phase resetting) response, reported previously for visual perception [158]. In non-human mammals, the intensity dependence of both phase shifting of the circadian pacemaker and acute suppression of melatonin have been well characterized [159-161]. In general, the results obtained by Zeitzer et al. [157] are the most commonly used as a reference for the light intensity required for melatonin phase shift (Figure 25A) and melatonin suppression (Figure 25B). As little as ~100 lx of (corneal) light could produce half of the maximal phase delay shift found at 10000 lx and that 90% of the asymptotic maximum response could be achieved with 550 lx. This would indicate that human circadian pacemaker is highly sensitive to ordinary room light and that minor changes in room light intensity could have a major impact on entrainment of the human circadian pacemaker. This is not consistent with some previous studies [154,162,163] which failed to

    Figure 25. Illuminanceresponse curve of the human circadian pacemaker. The shift in the phase of the melatonin rhythm (A), as assessed on the day following exposure to a 65 h experimental light stimulus, has been fitted with a four parameter logistic model using a nonlinear least squares analysis. Acute suppression of plasma melatonin (B) during the light exposure also has been fitted with a four parameter logistic model using a nonlinear least squares analysis. The logistic models predict an inflection point of the curve (i.e. the sensitivity of the system) at 120 lx. Saturation of the phaseshift response is predicted to occur with 550 lx and saturation of the melatoninsuppression response is predicted to occur with 200 lx. Individual subjects are represented by , the model by the continuous line, and the 95% confidence intervals by the dotted lines [157].

  • Circadian photobiology 24

    find significant phase resetting with room light but is however supported by several studies [164-166,192] with similar results. However, it should be noticed that there seems to be maximal melatonin suppression rate that is independent of the light intensity used [167,168]. Results of McIntyre et al. show approximately a rate of 1,5% per minute of light exposure until reaching an asymptotic level between 30 and 60 min. Similar observations have been made from animal studies [169-171]. A comparable is found in electronics and known as the slew rate, whereby the output of an amplifier cannot keep up with rapid changes in the input. Maximum nocturnal melatonin suppression would be about 45-50% after 30 minute of bright light. The accurate measurement of retinal illuminance is more difficult than measuring horizontal task illuminance as retinal illuminance depends on the angle of gaze, position of the head, pupil size [172], lens transmission [99], and possible photophobic response such as squinting [173-175]. For example, Sliney [175] estimated that squinting results in a log unit reduction in retinal illuminance compared to the estimated retinal illuminance using photometric and pupillometric measurements. In practice this means that higher corneal illuminance can produce smaller melatonin suppression than lower corneal illuminance even though pupil size is measured continuously as occurred in a study by Figueiro et al. [136]. An example of the relation between photopic illuminance at cornea can be seen in Table 2 from the study by Figueiro et al. [136], where it can be seen that light source with a CCT (correlated color temperature) of 8000K produced larger melatonin suppression with 1000 lx than with 300 lx. It can be also seen that as corneal illuminance increased the pupil size decreased, and the mean pupil are area was smaller with light source with higher CCT.

    Table 2. Corneal irradiance, mean pupil area, retinal illuminance value and mean melatonin suppression (meanS.E.M.) for each lighting condition [136]. Light source

    Photopic illuminance at cornea (lx)

    Irradiance at cornea (W/cm2)

    Mean pupil area (mm2)

    Retinal illuminance (lx mm2)

    Mean melatonin suppression (%)

    30 8.2 19 573 -3% (11%) 100 27 12 1150 10% (4%) 300 82 8.9 2670 38% (7%) 4100 K 1000 270 5.8 5800 38% (6%)

    30 9.7 16 492 10% (8%) 100 32 10 1010 32% (7%) 300 97 8.2 2460 47% (4%) 8000 K 1000 320 5.0 5000 34% (9%)

    2.4.4 Timing

    The amount by which a discrete light pulse can change the timing of the circadian system is phase dependent, and this phase dependency is described by phase response curves (PRC). In general, there are two general PRC morphologies: a low amplitude PRC with maximal phase shifts of a few hours (Type 1), and a high amplitude PRC with phase shifts as large as 12 h (Type 0) [176,177]. In Type 0 resetting, the resetting stimulus affects both the phase and amplitude, and a stimulus of appropriate strength applied at a critical phase can in theory reduce the amplitude of oscillation in zero (singularity) [177-179]. Single bright light elicits phase shifts in humans consistent with Type 1 PRC [180,181] showing typically phase advances of ~2 h and maximum phase delays of ~3 h. In both the Type 1 and Type 0, phase shifts in response to light are observed during the biological night when humans are habitually asleep in the dark. A recent study by Khalsa et al. (2003) [182] was the most comprehensive study on human PRC so far. In this study a 9 day in-laboratory study protocol was used preceded by 2 weeks of regular 8 h sleep schedule based upon subjects (n=43) habitual sleep and wake times. The results [182] supported previous findings [183,184] that there is no dead zone

  • Circadian photobiology 25

    (when no phase shift is elicited by bright light) in the human PRC. Three different PRCs from the study can be seen in Figure 26, which differ by the phase markers used. Figure 26A uses melatonin midpoint, Figure 26B dim light melatonin onset (DLMOn), and Figure 26C dim light melatonin offset (DLMOff) as the phase marker for circadian rhythm (for phase marker details see Figure 3). The transition from delays to advances in the critical region (CT 0) is rapid, while the transition from phase advances to phase delays during the subjective day is more gradual. The phase shifts measured by DLMOff are smaller than those measured by DLMOn [182], which is consistent with the results obtained from rodent studies. It has been hypothesized that there may be two coupled oscillators, an evening or E oscillator associated with melatonin onset, and a morning or M oscillator associated with melatonin offset [185-188].

    2.4.5 Duration

    Traditionally bright light experiments have consisted of 2 to 8 hour continuous exposures [189-192], based on the assumption that bright light exposure is consistent with the Bunsen-Roscoe law that states that the effect is independent (within a certain general time frame) of the duration of exposure as long as the radiant exposure is the same [91,193]. However evidence from animal experiments [194-196] would suggest that the same phase-shifting than with continuous exposure could be achieved with intermittent light exposure with less radiant energy. The response of human circadian system has not been very well quantified even though the exposure to bright light is typically intermittent in everyday life [197-200]. Kronauer et al. [201,202] have proposed a revisal model for the resetting effect of light. The model is partly based on experiments comparing the effects of continuous and intermittent bright light stimuli (~9 500 lx) over a ~5-h period. The results of a study by Rimmer et al. [203], designed to test the model of Kronauer et al. [201,202], suggest that an intermittent bright light stimulus, interrupted by intervals of complete darkness that exceed the light exposure can significantly phase shift the human circadian pacemaker. When bright light occupied only 31% of the total stimulus, 70% of the median resetting response was observed. Furthermone, when bright light occupied 63% of the total stimulus, nearly 90% of the median resetting response was preserved. These findings also indicate that the brief intermittent exposures to bright light that are normally encountered in everyday life (during the night and day) [197,199,200] may have a greater impact on circadian entrainment than was previously recognized [198-200,204,205]. Studies by Boivin and James [206], by Baehr [207], and Gronfier et al. [208] also support this

    Figure 26. Phase advances (positive values) and delays (negative values) are plotted against the timing of the centre of the light exposure relative to the melatonin on the pre-stimulus CR (defined to be 22 h), with the core body temperature minimum assumed to occur 2 h later at 0 h. Using A) melatonin midpoint, B) dim light melatonin onset (DLMOn), and C) dim light melatonin offset (DLMOff) as marker for circadian phase. Data points from circadian phases 618 are double plotted. The filled circles represent data from plasma melatonin, and the open circle represents data from salivary melatonin in one subject. The solid curve is a dual harmonic function fitted through all of the data points. The horizontal dashed line represents the anticipated 0.54 h average delay drift of the pacemaker between the pre- and post-stimulus phase assessments. The fitted peak-to-trough amplitude of the DLMOn PRC (5.41 h) appears slightly larger than that of the DLMOff PRC (4.60 h) [182].

    C B A

  • Circadian photobiology 26

    proposed model [201,202] indicating that sustained periods of intensely bright light are not necessary for resetting the human circadian system.

    2.4.6 Photic history

    It has been shown that resetting response of circadian pacemaker can be attenuated by a preceding nonsaturating stimulus in animals [209], and phase shifting in mammalian has been shown to be maximal after prolonged exposure to complete darkness before a stimulus [210,211]. Also a study by Hebert et al. [212] made in humans showed significant differences (with large inter-individual variability) differences in melatonin suppression after different (dim vs. bright) light history. These studies would suggest that the light-mediated melatonin suppression could also be modulated by prior photic history. Study by Smith et al. [213] revealed a significant difference in melatonin suppression between two lighting history conditions, with a mean suppression of 71,2% (7,1%) in the approximately 200 lx prior light history condition vs. a mean suppression of 85,7% (6,5%) in the approximately 0,5 lx prior light history condition. The results [213] demonstrate that a prior light history alters light-mediated melatonin suppression, while it is impossible to determine whether dim background potentiates or relatively bright background diminishes the strength. It would seem important always to control the prior light history when examining melatonin suppression. Findings suggest that a controlled photic history of 63 hours before a light stimulus is sufficient to change the suppression effect of the subsequent light stimulus. However it is not possible to determine the exact time from this study [213] to be a sufficient control time and further investigation is needed for more quantitative results.

    2.4.7 Polarization

    The differences between nonpolarized and vertically polarized light to melatonin suppression were investigated by Brainard et al. [214] in 2000. Six subjects participated in the study and they were exposed to four different light intensities: 20, 40, 80 and 3200 lx (for saturation response) with their pupils dilated with cyclopentolate HCl. The results of the study [214] revealed hat there is a significant correlation between the light intensity used and melatonin suppression, but no significant differences between nonpolarized and polarized light.

  • Eye and photometric measurements 27

    3. EYE AND PHOTOMETRIC MEASUREMENTS

    Literature review is done on the available methods to measure eye movement, pupil size and light entering the eye for proper quantification of circadian effective light exposure. Also the electrical activity recordings of the eye ERG and/or EOG are reviewed as they can be used for further quantification of physiological responses of light. Even though within the scope of this work they are not integrated to the dosimeter device but the possible recording of EOG/ERG should be taken account when designing the dosimeter.

    3.1 ELECTRICAL ACTIVITY Two the most typical measurements of the electrical activity of the eye are electroretinogram (ERG) and electrooculogram (EOG). ERG is more clinical utility used for the diagnostics of various eye diseases but it can also be used to measure circadian responses as done by Hankins et al. [215]. EOG is commonly used in studies measuring changes in alertness as well as in EEG (electroencephalography) studies for the elimination of the eye blink artifacts in EEG recordings.

    3.1.1 Electroretinogram (ERG) Electroretinogram (ERG) is a device measuring the electrical activity of the eye. Figure 27A [216] shows schematically the basic measurement setting with a special saline filled contact lens with an Ag/AgCl electrode placed on top of cornea As it is shown in Figure 27B [216], light pulse induces a potential change and four common ERG waveforms (the a, b, c, and d waves) are marked to the picture. It should be noted that notations can differ in literature.

    The first (a) component is early-receptor potential (ERP), which appears almost instantaneously after onset of light. The amplitude of the ERP depends directly upon stimulus intensity and the concentration of visual pigment in the outer segments of the photoreceptors. Therefore, the ERP is believed to reflect dipole changes in the visual pigment molecules due to conformational changes that are elicited by photon absorption. The ERP has been used in research to follow non-invasively the concentration of the visual pigment during light adaptation and in the dark following an exposure to bright light that causes substantial pigment bleaching [217]. This is followed by (b) late-receptor potential (LRP), which has a small latency (1-5ms) and is found to be maximal near the synaptic endings of the photoreceptors therefore

    Figure 27. (A) The transparent contact lens contains one electrode, shown here on horizontal section of the right eye. Reference electrode is placed on the right temple. (B) Typical vertebrate ERG waveform in response to a 2 s light flash [216].

  • Eye and photometric measurements 28

    reflecting the outputs of the photoreceptors. ERG and b waves can be used to study the diurnal variation in the cone pathway [218] which is related to the diurnal transition between processes optimized for high (photopic) and low (scotopic) light levels [219-222]. Cone b wave-implicit time appears to be regulated by environmental irradiance as an adaptation to the varying demands of the solar cycle [222]. This regulation seems to be driven by the novel photoreceptor (ipRGC), and by studying the irradiance and wavelength dependent reduction in b wave-implicit time it is possible to study the spectral sensitivity of the novel photoreceptor as done by Hankins et al. [223]. Even though ERG has not been used in further studies examining the non-image forming (NIF) responses in humans, it could be added to some experimental designs to provide supplemental information in addition to typical measures (e.g. melatonin, CBT). For example Hankins et al. [223] controlled retinal illumination by using a custom-built Ganzfeld dome illuminator [224] (similar apparatus as Goldman perimeter [225,226]) with the ERG electrodes attached bilaterally beneath each eyelid with a forehead reference ground [222]. The c-wave is now known to originate in the pigment epithelium after the discovery of potassiumretinogram (KRG) [227]. C-wave is also called 'The standing potential of the eye'. Although the c-wave originates from the pigment epithelium, it depends upon the integrity of the photoreceptors, because light absorption in the photoreceptors triggers the chain of events leading to the decrease in extracellular concentration of potassium ions. Therefore, the ERG c-wave can be used to assess the functional integrity of the photoreceptors, the pigment epithelial cells and the interactions between them. The d-wave is only evident when the ON and OFF phases of the ERG response are separated in time, by using light stimuli with long duration (>100ms). With shorter durations d-wave tends to be combined with the b-wave. ERG measurements can be also used to determine the perception of flickering lights as with 100 Hz fluorescent lamp flicker [228,229]. Even though no visual perception of flicker exist the response to flicker can be seen in ERG, and this flicker perception will most likely cause the problems associated with clearly visible flicker such as headaches and fatigue. Also cone and rod ERG responses can be isolated using different (colored) stimuli as seen in Figure 28A [231]. Rods are also incapable of following long light flicker (not really evident in short timescale of Figure 28A [231]), and it is possible to determine the involvement of rods and cones.

    (A)

    (B)

    Figure 28. (A) Cone and rod ERGs can be isolated using dim flash stimuli into photopic (cone) and scotopic (rod) signals [231]. (B) Using different rates (flicker) of stimulus presentation also allows rod and cone contributions to the ERG to be separated. Even under ideal conditions rods cannot follow a flickering light up to 20 per second whereas cones can easily follow a 30 Hz flicker, which is the rate routinely used to test if a retina has good cone physiology [231].

  • Eye and photometric measurements 29

    3.1.2 Electooculogram (EOG) The electrooculogram measures the potential that exists between the cornea and Bruch's membrane at the back of the eye. The potential produces a dipole field with the cornea approximately 5 millivolts positive compared to the back of the eye, in a normally illuminated room. Although the origin of the EOG is the pigment epithelium of the retina, the light rise of the potential requires both a normal pigment epithelium and normal mid-retinal function. Elwin Marg named the electrooculogram in 1951 and Geoffrey Arden [230] developed the first clinical application [231]. It far less invasive method compared to ERG as seen in Figure 30 [231]. The electrodes of the EOG are normally placed at the outer canthi of each eye, one slightly above the cantomeatal place, the other slightly below [232]. EOG is also the abbreviation for electro-olfactogram, which is used to determine electrical responses of different smells and scents, and thus is totally different device [233]. EOG is frequently the method of choice for recording eye movements in sleep and dream research [234], in recording eye movements from infants and children, and in evaluating reading ability and visual fatigue. And the most important application within this work is its use in slow eye movement (SEM) measurement with EEG to assess alertness. Figure 29 [231] shows a 10-second periods of eye movement back and forth between two red LED lights placed 30 degrees apart inside a Ganzfeld measurement device. After training the patient in the eye movements, the lights are turned off. About every minute a sample of eye movement is taken as the patient is asked to look back and forth between the two lights. As various types of eyelid and eye movement patterns have been shown to respond to sleep loss and to correlate with sleepiness in a variety of protocols [235-237] indicating that EOG could be in theory used to assess sleepiness objectively [238]. However, in practice the EOG recordings have shown too large inter-individual differences making objective alertness assessment still a thing of the future [239-241]. Aserinsky and Kleitman [242] described SEMs during drowsiness preceding sleep onset and during light sleep. Kuhol and Lehmann [243] found that SEMs became larger and more regularly sinusoidal when simultaneous showing of the EEG was noted during sleep onset. Slow (0,25 Hz), pendular, horizontal eye movements wer