July 2015
Volume 56, Issue 8
Free
Retina  |   July 2015
Impact of Macular Pigment on Fundus Autofluorescence Lifetimes
Author Affiliations & Notes
  • Lydia Sauer
    Department of Ophthalmology University Hospital Jena, Jena, Germany
  • Dietrich Schweitzer
    Department of Ophthalmology University Hospital Jena, Jena, Germany
  • Lisa Ramm
    Department of Ophthalmology University Hospital Jena, Jena, Germany
  • Regine Augsten
    Department of Ophthalmology University Hospital Jena, Jena, Germany
  • Martin Hammer
    Department of Ophthalmology University Hospital Jena, Jena, Germany
    University of Jena, Center for Medical Optics and Photonics, Jena, Germany
  • Sven Peters
    Department of Ophthalmology University Hospital Jena, Jena, Germany
Investigative Ophthalmology & Visual Science July 2015, Vol.56, 4668-4679. doi:10.1167/iovs.14-15335
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Lydia Sauer, Dietrich Schweitzer, Lisa Ramm, Regine Augsten, Martin Hammer, Sven Peters; Impact of Macular Pigment on Fundus Autofluorescence Lifetimes. Invest. Ophthalmol. Vis. Sci. 2015;56(8):4668-4679. doi: 10.1167/iovs.14-15335.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: To characterize the macular region and to investigate the influence of the macular pigment (MP) on fundus autofluorescence (FAF) lifetimes in vivo.

Methods: Forty-eight healthy subjects with a mean age of 24.1 ± 3.6 years (range, 20–37 years) were included. A 30° retinal field was investigated using the fluorescence lifetime imaging ophthalmoscope (FLIO), based on a Heidelberg Engineering Spectralis system, detecting FAF decays in a short (498–560 nm; ch1)- and a long (560–720 nm; ch2)-wavelength channel. The mean fluorescence lifetime τm was calculated from a 3-exponential approximation of the FAF decays. Macular pigment optical density (MPOD) was measured by one-wavelength reflectometry, and macular optical coherence tomogram (OCT) scans were recorded. Correlations between τm and MPOD were analyzed.

Results: The τm showed shortest values at the macular region with a mean of 82 ps (ch1) and 126 ps (ch2). We found a strong correlation of τm to the MPOD (ch1: r = −0.760; ch2: r = −0.663; P < 0.001), as well as a topologic agreement of shortest τm with highest MPOD.

Conclusions: Macular pigment, which is known to have very short fluorescence decays, considerably contributes to the macular autofluorescence (AF). This study gives indirect evidence for a strong impact of MP on macular τm, although no direct measurement of MP autofluorescence lifetimes in vivo is possible at this point. Potentially, imaging the FAF lifetimes could lead to a novel methodology for the detection of macular pigment properties and pathology-induced changes in the living human retina.

In vivo fluorescence lifetime imaging (FLI) of the human fundus, a new method developed by Schweitzer et al.,1,2 provides metabolic information of the retina. The clinically used fundus autofluorescence (FAF) intensity imaging shows only the total autofluorescence (AF) that with age is increasingly dominated by lipofuscin.36 In contrast, FLI provides not only spectral but additional fluorescence lifetime information, thus allowing a better discrimination of fluorophores. Each fluorophore exhibits a specific fluorescence lifetime based on its molecular structure. Furthermore, the lifetime of a fluorophore can be influenced by its microenvironment; but regardless of possible changes, the fluorescence lifetime of a fluorophore is basically independent of its concentration (except for self-absorption/self-quenching at high concentrations).79 Fluorescence lifetime imaging offers several distinct advantages over FAF imaging: (1) Fluorescent biomolecules with overlapping fluorescence emission spectra but different lifetimes can be distinguished1013; (2) changes in the biological microenvironment may be detected through an altered fluorescence quenching7,10,12,14; (3) fluorescence lifetimes are independent of the fluorophore's concentration and therefore more robust than fluorescence intensities7,13; (4) the inherent principle of FLI allows the assignment of individual fluorophores/lifetimes to anatomical structures, for example, within the retina2; and (5) the metabolic state of cells can be assessed by different lifetimes for free and protein-bound endogenous fluorophores.7,10,14,15 
Based on two-photon excited AF imaging of ex vivo porcine retinal cells,16 we previously assigned the shortest lifetimes found in vivo to the melanin within retinal pigment epithelium (RPE) cells. Intermediate as well as long lifetimes were assigned to the neuronal layers. In vivo studies conducted by Schweitzer et al.2 reported a connection of the longest lifetimes with collagen as well as with the crystalline lens. Furthermore, lifetime changes in early stages of retinal diseases such as age-related macular degeneration (AMD), diabetic retinopathy, and Alzheimer's disease were investigated.1719 According to Klemm et al.,9 metabolic retinal changes are often partially or completely reversible, whereas morphological changes are not. In order to establish the fluorescence lifetime imaging ophthalmoscope (FLIO) for clinical diagnostic usage, it is of utmost importance to unravel the underlying molecular mechanisms and assign lifetimes to certain fluorophores. 
A connection of short macular lifetimes with xanthophyll has recently been speculated.20 Xanthophyll, mainly consisting of the carotenoids lutein and zeaxanthin,21 is located in the plexiform layers and photoreceptor axons (Henle's layer) in the macula lutea and thus is called macular pigment (MP).2224 A third component, mesozeaxanthin, was found to be formed inside the retina.25 Macular pigment is known to attenuate the ocular FAF by absorbing the short-wavelength excitation light, causing a hypofluorescent central macula in FAF images.26 Additionally, a direct fluorescence from macular carotenoids in vivo has been described.27 Different MP distribution patterns were found by Sharifzadeh et al.,28 and the MP amount is known to correlate with the macular thickness (MT).29 In vitro, the long carotenoids lutein and zeaxanthin (10 and 11 conjugated double bonds, respectively) show fluorescence maxima at wavelengths between 500 nm and 550 nm (S2 → S0 transition) and a very short femtosecond fluorescence lifetime of approximately 200 to 250 fs.28,30 In this study, the hypothesis that MP has an influence on the FAF lifetime was investigated. Further correlations with possible implications, such as age and MT, were investigated in an exploratory manner. 
Methods
Subjects
Forty-eight volunteers between 20 and 37 years of age (mean age 24.1 years, SD = 3.6 years) were examined between September 2013 and March 2014 at the University Hospital of Jena, Germany. Their medical histories were recorded; any systemic disease such as diabetes mellitus or hypertension led to exclusion. Volunteers with previous or general eye diseases except for refractive errors up to ±6 diopters were excluded. Another exclusion criterion was a history of eye surgery. This study adhered to the tenets of the Declaration of Helsinki and was approved by a local ethics committee. Prior to all investigations, written informed consent was obtained from the subjects. 
Study Protocol
The intraocular pressure was measured by noncontact tonometry. Only one eye was examined in each participant, and all pupils were dilated with 0.5% Tropicamid (Mydriaticum Stulln; Pharma Stulln GmbH, Stulln, Germany). Fluorescence lifetime images were acquired 30 minutes after pupil dilation in a dark room; each measurement took approximately 2 to 5 minutes. In all cases, FLI was the first measurement performed. After a defined interval of 5 minutes, it was followed by the measurement of the macular pigment's optical density (MPOD), which was also carried out in a dark room. After measuring the MPOD, fundus photography was performed for documentation. Finally, a frequency-domain optical coherence tomogram (OCT, Cirrus; Carl Zeiss Meditec AG, Jena, Germany) was recorded and an eye examination was conducted to exclude subjects with any retinal abnormality or disease. The slit-lamp exam was carried out last to avoid possible bleaching effects. The same investigator conducted all examinations. 
FLIO Setup and Image Acquisition
To record FAF lifetimes, the FLIO by Heidelberg Engineering (Heidelberg, Germany) was used. FLIO is based on a Heidelberg Engineering Spectralis system and relies on the principle of time-correlated single-photon counting (TCSPC).31 It has been characterized recently by Dysli et al.20 FLIO records fluorescence decays and is therefore equipped with a pulsed diode laser, emitting pulses of 89 ps full width at half maximum (FWHM) with a frequency of 80 MHz at a wavelength of approximately 473 nm. With a nominal average laser power P0 = 200 μW, the energy E per pulse would be 2.5 pJ, resulting in a single pulse peak power P = 2.5 × 10−12 J/89 × 10−12 s = 28 mW. With a laser spot size of approximately 75 μm2 on the retina, this would result in an intensity of 28 mW/75 μm2 = 37 kW/cm2. Most likely, such a laser power density, when applied with a continuous wave laser on a stationary retina spot, would damage the retina. However, firstly, the duty cycle for the extreme high-laser power density is very low, that is, 89 ps/12.5 ns = 7*10−3. Therefore, the mean power (critical for thermal damage) is more than 2 orders of magnitude lower, that is, 196 μW/75 μm2 = 261 W/cm2. Secondly, due to the fast scanning procedure, only a few single laser pulses (eight pulses with total energy of 20 pJ) sum up at one spot until proceeding to the adjacent pixel (pixel clock 10 MHz corresponds to 100 ns). According to the IEC 60825-1:2007 standard, the International Electrotechnical Commission standard that adopts regulations of the American National Standards Institute (ANSI) Z136.1-200032 standard, the accessible exposure limit (AEL) for class 1 laser products is deduced for a single 89-ps laser pulse at 470 nm as follows: 
Accessible single pulse exposure limit:    
Accessible single pulse exposure for FLIO:    
Therefore, the FLIO fulfills, with respect to single pulse exposure, the laser class 1 criteria by far; the pulse energy is almost a factor 104 smaller than the AEL for single pulses, defined by the standard. Additionally, according to Delori et al.,33 the maximum permissible radiant power through the pupil for a pulsed exposure MPΦB,PLS for FLIO is approximately 2.8 mW. Therefore, FLIO is using an average laser power well below the limits recommended by the American National Standards Institute (ANSI Z136.1-2000). 
While scanning, the laser power is typically distributed over an area of 30° × 30°, that is, approximately 9 × 9 mm2, with a rate of 9 Hz. The individually recorded fluorescence photons are detected by two hybrid photon-counting detectors (HPM-100-40; Becker & Hickl GmbH, Berlin, Germany) in two separate spectral channels (Fig. 1A). The channels were chosen by Heidelberg Engineering according to previous studies conducted by our group.2 Channel 1 (ch1) detects photons within the range of 498 to 560 nm, a range where several retinal fluorophores emit, for example, reduced nicotinamide adenine dinucleotide (NADH), flavin adenine dinucleotide (FAD), advanced glycation end-products, collagen/elastin. Channel 2 (ch2) has a detection range of 560 to 720 nm and is assumed to be predominantly influenced by the emission of lipofuscin.2 The detected photons are recorded into 1024 time channels for each pixel, resulting in a photon arrival histogram representing the fluorescence decay (Figs. 1B–D). This finally results in AF intensity and lifetime images for both spectral channels with a pixel array size of 256 × 256 pixels each. Additionally, FLIO includes a high-contrast confocal infrared reflectance (IR) image for eye tracking in order to register each recorded frame properly and analyze each fluorescence photon at its correct spatial location. At least 1000 photons were recorded as a minimal signal threshold in ch1 at the fovea. 
Figure 1
 
FLIO setup and images. Overview of the general FLIO setup as provided by Heidelberg Engineering (A). Three representative photon arrival histograms (blue dots) of characteristic areas obtained from one FLIO scan of a healthy fundus, (B) macula, (C) temporal periphery, and (D) optic nerve head; fitted curve (red) and instrument response function (green curve). Typical false color-coded mean FAF lifetime (τm) images of the spectral channel 1 (E) and channel 2 (F) and a fluorescence intensity picture (G) overlaid with the standard ETDRS grid.
Figure 1
 
FLIO setup and images. Overview of the general FLIO setup as provided by Heidelberg Engineering (A). Three representative photon arrival histograms (blue dots) of characteristic areas obtained from one FLIO scan of a healthy fundus, (B) macula, (C) temporal periphery, and (D) optic nerve head; fitted curve (red) and instrument response function (green curve). Typical false color-coded mean FAF lifetime (τm) images of the spectral channel 1 (E) and channel 2 (F) and a fluorescence intensity picture (G) overlaid with the standard ETDRS grid.
Given an instrument response function (IRF) FWHM of 172 ps (ch1) and 153 ps (ch2), FLIO has a practical time resolution of approximately 30 ps (theoretically approximately 15 ps). The IRF was measured using a 25 μM Eosin Y solution (Sigma-Aldrich Chemie Gmbh, Munich, Germany), which additionally contains 5 M potassium iodide. The Eosin Y fluorescence can be excited from 350 to 500 nm and ranges between 450 and 680 nm with a sufficient intensity. Based on the reported Rose Bengal fluorescence lifetime of approximately 16 ps when dissolved in 5 M potassium iodide,34 it is reasonable to assume that Eosin Y, another fluorescein derivative, shows similar characteristics, which we can confirm. Additionally, we compared IRFs based on Eosin Y fluorescence with IRFs measured using scattered excitation laser light and found no differences regarding the shape as well as the width. Furthermore, the FLIO recordings of the IRF showed no sign of variation across the images. The acquisition time was set to 2 to 3 minutes, which is comparable to a typical fundus image recording. 
Analysis of FLIO Data
The photon arrival histogram was determined for each of the 65,536 recorded pixels. We analyzed the data using the Software SPCImage 4.4.2 (Becker & Hickl GmbH). The FAF decays were approximated using a 3-exponential fit according to    
I (t) is the fluorescence intensity at time t and I0 the initial, maximal intensity; α is the amplitude of each lifetime (τ) component, and ⊗ indicates the convolution integral with the IRF. 
Thereby we obtain the three different lifetimes: τ1, τ2, and τ3. Their amplitude-weighted mean, τm, is calculated according to    
In order to reduce noise, a sliding average using a 5 × 5-pixel squared kernel was employed. χr2 was used as a goodness-of-fit criterion. Due to the fact that the FAF does not completely decay within the used time window of 12.5 ns, all fits were carried out in the incomplete multiexponential decay mode. A tail fit was used starting at the fluorescence peak. The τm was the main parameter for the statistical analysis and comparison. 
To quantify the macular lifetimes in healthy subjects, we averaged the fitted data over a circular area centered at the foveola, using the software FLIO-reader (ARTORG Center for Biomedical Engineering Research, University of Bern, Bern, Switzerland). This software averages lifetimes over defined areas. We used two different circular areas, one 1 mm in diameter (C) and a smaller one 0.35 mm in diameter (CS) (Figs. 2B, 2C). C is the same size as the central area in the standard Early Treatment Diabetic Retinopathy Study (ETDRS) grid (Fig. 1G), as well as the assumed distribution area of MP. Therefore, area C was used for analytical comparison. The second grid was utilized to further examine τm in the central foveolar region, reducing possible influences of surrounding inner retinal layers or, in general, areas without MP. To illustrate the FAF lifetime images, the software FLIMX (Institute of Biomedical Engineering and Informatics, Technische Universität Ilmenau, Ilmenau, Germany) was used. 
Figure 2
 
Macular τm. Distribution of short macular mean AF lifetime (τm) within our study group for both spectral channels and both utilized grid sizes (A). The grid sizes C (B) and CS (C) are depicted within a false color-coded τm image.
Figure 2
 
Macular τm. Distribution of short macular mean AF lifetime (τm) within our study group for both spectral channels and both utilized grid sizes (A). The grid sizes C (B) and CS (C) are depicted within a false color-coded τm image.
Macular Pigment Density
The optical density of MP was measured with the Visucam 500 (Zeiss Meditec AG, Jena, Germany). According to a standard protocol, the device provides the integral of MPOD over the macula, the so-called MPOD volume, which is equal to the product of the distributional area size (AS) and the mean optical density (mean OD) of MP. Fundus photography as well as pigment measurements was carried out. According to the user's manual, the MPOD volume has a standard deviation of 9% when examinations are carried out by different investigators and on different cameras. We assume our SD to be lower than that because the same investigator used one Visucam for all measurements. 
As we are aware of the fact that juvenile reflexes corrupt the MPOD volume calculations, we applied a correction based on a mask that includes all reflexes and a threshold cutoff (ImageJ35). More specifically, the area temporal of the macula without reflexes served as a reference to define the threshold. All gray values above the threshold and within the mask were removed, eliminating all juvenile reflexes to a sufficient extent. Those corrected images were analyzed again through the Visucam 500 software. 
Given the results by Sharifzadeh et al.,28 we assumed the MP to be mainly distributed within a circle 1 mm in diameter centered at the fovea. Therefore, to compare the MPOD volume with fluorescence lifetime means, the area C with 1-mm diameter was analyzed and compared in particular detail. The total amount of MP was evaluated by the Visucam parameter “volume” and correlated with mean AF decay times across the same area. 
Optical Coherence Tomography
A frequency-domain OCT was performed. The 512- by 128-pixel macular cube scan was obtained, as well as a high definition 5-line scan of the macular region, providing information about retinal anatomy and thickness. As thickness parameter, the central subfield thickness (CST) was utilized. 
Statistical Analysis
For all statistical analyses, SPSS 21 (SPSS, Inc., Chicago, IL, USA) was employed. To test for significant lifetime differences between areas C and CS, a t-test for paired samples was used. We correlated the lifetimes of area C, the MPOD volume, its contributing parameters AS and mean OD, the CST, and the age with each other. All conducted correlation tests are based on the Pearson correlation coefficient. To verify the linearity of the considered correlation and check for nonlinear correlations, a regression analysis was carried out and the coefficient of determination (R2) was analyzed. We applied a Holm-Bonferroni correction for all multiple testing to all correlations. P values of 0.05 and below indicated significance. Selected partial correlations were carried out for Holm-Bonferroni corrected significant correlations to further analyze the direct or indirect influence of individual parameters. The τm of area CS was correlated only to the MPOD volume; therefore, no additional correction was applied. To show if the independent variables age and CST significantly affect the correlation of τm with MPOD volume, a multivariate regression analysis was performed. All plots were generated with Origin 9.1 (OriginLab Corp., Northampton, MA, USA). 
Results
Autofluorescence Lifetime of the Healthy Macula
Representative FAF lifetime images for a healthy subject are depicted in Figures 1E and 1F (ch1 and ch2, respectively). Additionally, photon arrival histograms for ch1 are presented for three characteristic areas (Figs. 1B–D). In all subjects, the shortest decay times were found in the macular region. The obtained lifetime distributions within the areas C and CS for both channels are illustrated in Figure 2. Values for τm across C ranged from 47 to 123 ps in the short-wavelength channel with a mean of 82 ± 18 ps. Area C lifetimes in the long-wavelength channel varied from 101 to 155 ps between the subjects, with a mean of 126 ± 13 ps. The short- and long-wavelength channel lifetimes for the more MP-specific, foveolar area CS were 52 ± 12 ps (range, 33–86 ps) and 105 ± 16 ps (range, 77–132 ps), respectively. Area CS mean AF lifetimes were significantly shorter (P < 0.001) compared to area C for both channels. 
Four representative examples of mean FAF lifetimes are summarized in Figures 3A and 3B. Evidently, the depicted area of shortest AF decays (reddish color) varies in size between different subjects. By taking a closer look, we found a trend across all 48 volunteers: A shorter τm averaged over the macular region area C (τmC) is associated with a larger reddish area, thus indicating a variable distribution of a fast-decaying fluorophore. 
Figure 3
 
Topographic distribution of short macular τm. Typical mean FAF lifetime (τm) images (false color-coded) of the spectral channels 1 (A) and 2 (B) and their corresponding OCT images (C) of four healthy subjects (ad). Their respective central subfield thickness (obtained from OCT) and MPOD volume are plotted below (D).
Figure 3
 
Topographic distribution of short macular τm. Typical mean FAF lifetime (τm) images (false color-coded) of the spectral channels 1 (A) and 2 (B) and their corresponding OCT images (C) of four healthy subjects (ad). Their respective central subfield thickness (obtained from OCT) and MPOD volume are plotted below (D).
In order to potentially identify the origin of short macular AF lifetimes, we investigated the influences of MP on macular τm. Exploratory, other possible influences were investigated. The significant correlations and P values are summarized in Table 1
Table 1
 
Correlations of τm, MP, MT, and Age
Table 1
 
Correlations of τm, MP, MT, and Age
The Influence of Macular Pigment on Macular τm
The MPOD volume in our study population varied within a range of 1423 to 8772 arbitrary units (a.u.; mean: 5575 ± 1753 a.u.). Figure 3D depicts our finding that the amount of MPOD volume increases with a decreasing τmC. This is statistically supported by a negative correlation of the macular τm with the amount of MP (Fig. 4). Precisely, τm of area C strongly correlates with the MPOD volume across the same area (ch1: r = −0.76, P < 0.001; ch2: r = −0.66, P < 0.001). No significant interaction terms of age (P value ch1: 0.918; ch2: 0.690) and CST (P value ch1: 0.691; ch2: 0.912) on the τm–MPOD volume relationship were found. The results of the multivariate regression analysis are summarized in Table 2. On average, a difference of the MPOD volume by 1000 a.u. shortens the AF lifetime of area C by 7 ps (ch1) and 5 ps (ch2). 
Figure 4
 
Correlations of mean FAF lifetimes (area C) and MPOD volume. A scatterplot of linear correlations for mean FAF lifetimes (τm) of area C to the MPOD volume (ch1: r = −0.76; ch2: r = −0.66; P < 0.001).
Figure 4
 
Correlations of mean FAF lifetimes (area C) and MPOD volume. A scatterplot of linear correlations for mean FAF lifetimes (τm) of area C to the MPOD volume (ch1: r = −0.76; ch2: r = −0.66; P < 0.001).
Table 2
 
Multivariate Regression Analysis
Table 2
 
Multivariate Regression Analysis
We found different MP distribution patterns, which we refer to as cone-shaped with high central and low parafovolar MP levels, in 7 eyes (14.6%), broad-shaped with high central as well as high parafovolar MP levels in 35 eyes (72.9%), and ring-shaped with lower central MP concentration and a ring-like high concentration in parafovolar areas in 6 eyes (12.5%). 
Figure 5 visualizes the statistical correlation of short τm to high MPOD. Fundus autofluorescence lifetime images and three-dimensional (3D) plots (ch1), as well as 3D MPOD volume plots of three subjects with different MPOD distributions (Fig. 5A: broad-shaped distribution, Fig. 5B: ring-shaped distribution, Fig. 5C: cone-shaped distribution), are illustrated. The color-coded lifetime images are shown in Figures 5A1, 5B1, and 5C1, where the color cutoff is set at 200 ps to emphasize the macular region. In addition to the en face images, the macular τm (approximately 12° field) are plotted in 3D projections (Figs. 5A2, 5B2, 5C2). Their corresponding 3D MPOD volume plots are depicted below (Figs. 5A3, 5B3, 5C3); these include foveal reflexes to a different extent. Although these reflexes are still visible in the MPOD volume images, the algorithm for the MPOD volume calculation is corrected. However, these artifacts are absent in the FLIO measurement. Generally, the shortest mean AF lifetimes as well as the highest amounts of MP are colocated at the foveal region, and the ring-shaped MPOD distributions show a ring-shaped FAF lifetime pattern as well. Furthermore, the proportions of the short macular τm distributions have a striking similarity to those of MP. Consequently, the 3D lifetime plots are inversely reflected in the 3D MPOD volume plots. In all subjects, peaks of MPOD volume can be assigned to areas with the shortest lifetimes across the macula. 
Figure 5
 
Topographic similarity of τm and MPOD volume. Macular mean AF lifetime (τm) and MPOD volume for three representative, healthy subjects with different spatial distributions of MP ([A] female, 23 years, broad-shaped MP; [B] female, 22 years, ring-shaped MP; [C] male, 27 years, cone-shaped MP); channel 1 τm images (A1, B1, C1); corresponding 3D projection of macular τm (A2, B2, C2) and 3D MPOD volume plots (A3, B3, C3).
Figure 5
 
Topographic similarity of τm and MPOD volume. Macular mean AF lifetime (τm) and MPOD volume for three representative, healthy subjects with different spatial distributions of MP ([A] female, 23 years, broad-shaped MP; [B] female, 22 years, ring-shaped MP; [C] male, 27 years, cone-shaped MP); channel 1 τm images (A1, B1, C1); corresponding 3D projection of macular τm (A2, B2, C2) and 3D MPOD volume plots (A3, B3, C3).
As both the AS and the mean OD of measured MP contribute to the parameter MPOD volume, we investigated their influences separately. Based on partial correlations, we found that the variability of the MPOD volume can be almost equally accounted for by the AS (r = 0.93, P < 0.001) and the mean OD (r = 0.90, P < 0.001) of MP. Therefore, those parameters influence the macular τm indirectly through the MPOD volume. 
As long as there are multiple fluorophores at the human fundus, the obtained mean FAF lifetime sums up as a mixture of multiple contributions. By evaluating area C, we are looking at differences in the fractional contribution of MP to the total AF. We found that a higher fraction of xanthophyll leads to shorter mean AF lifetime. The interindividually different distribution of MP over area C results in different fractions of xanthophyll fluorescence and, finally, in the correlations of τm with MPOD volume found here. 
In the central area CS, on the other hand, the amount of MP is highest (except for a ring-shaped MP distribution). Assuming that MP substantially contributes to the AF of the macular region, this leads to the following consequences: (1) Within area CS, the fraction of MP AF is less variable than in area C; (2) the fluorescence intensity of MP is highest; (3) MP's possible influence as an optical filter is most effective (4) due to the anatomical shape of the foveola, a possible influence of inner retinal layers is mostly reduced; and (5) the fractional contribution of retinal fluorophores such as lipofuscin is lowest. 
Given that MP is highest in area CS and its AF lifetime is independent of its concentration, we would expect the correlation of τm with MPOD volume to be lower in area CS than in C. And this is exactly what we found: The correlation decreased from r = −0.76 to −0.53 in ch1 (ch2: r = −0.66 to −0.57). Therefore, we are convinced to approach, within the resolution limit of FLIO, the AF lifetime of MP in vivo by investigating area CS
Influence of Macular Thickness
A corresponding macular OCT image is shown for each subject in Figure 3C: A wider foveola was found in eyes with a larger area of short macular τm. It is known that the central foveal thickness correlates with the foveal pit diameter.36 Therefore, we analyzed the CST to quantify its influence. The CST ranged from 218 to 312 μm (mean: 260 ± 21 μm). Our study population tended to show higher amounts of MP with thinner maculae. Accordingly, CST correlates inversely with the MPOD volume (r = −0.53, P < 0.001). 
We investigated the impact of CST on MPOD volume, AS, and mean OD separately by using partial correlations. Thereby, significant influences of CST were found only on the AS of measured MP (r = −0.42). A weak and nonsignificant correlation of the CST to the mean OD was found when the AS was taken as a control parameter. 
The larger AS of MP and thereby the higher amount of MPOD volume in thinner maculae indirectly cause an influence of CST on macular τm (Fig. 3D). Therefore, correlating CST with τm (area C) revealed that eyes with a smaller CST exhibit a shorter mean AF lifetime in the macular region (ch1: r = 0.5, P < 0.001; ch2: r = 0.44, P < 0.01). However, when conducting a partial correlation with the MPOD volume as a control parameter, this correlation disappears (ch1: r = 0.174, P = 0.25; ch2: r = 0.119, P = 0.44). 
Influence of the Subject's Age
The volunteer's age shows a medium to weak correlation with τm of area C. After applying the Holm-Bonferroni correction, possible age influences were significant only for ch2 (P < 0.05), not for ch1. No significant influence was found for all additionally considered parameters (CST, MPOD volume, AS, and OD of MP), which correlate weakly (or less) with the age. 
Discussion
The mean FAF lifetime distribution within a 30° field, which we found for all 48 subjects, agrees with previous observations and comprises typical anatomical patterns: The longest τm can be assigned to the area of the optic disc; intermediate τm are found across the whole retina; and short τm are found in the macular region. These local differences in a typical FAF lifetime image were first reported by Schweitzer et al.1 in 2004. The short macular lifetimes have been discussed as originating from within the RPE.2 However, a connection of short FAF lifetimes with lutein and zeaxanthin has recently been speculated.20 Additionally, direct fluorescence from the macular carotenoids has been reported in the context of Resonance Raman–based MP measurements.27 To our knowledge, this study is the first that describes a very likely contribution of the MP to the AF lifetime measured at the posterior pole in vivo. 
Fluorescence Lifetime Imaging of the Human Retina
The reproducibility of the FLIO has recently been demonstrated by Dysli et al.20 as well as by Klemm et al.9 The software FLIO reader supports this by offering comparable means of fluorescence lifetimes in different areas. Autofluorescence lifetimes of tissues in vivo may depend on different influences such as redox state, pH, and metabolic state as well as the maturation or senescence of cells.8,37,38 Detecting retinal fluorophores involved in metabolic processes, such as NADH,2,39 FLIO may provide new insights on alterations in early stages of retinal diseases. Therefore, it offers great potential for noninvasive and nondestructive early-stage diagnostics.40 
It is of great importance to assign measured lifetimes to specific fluorophores in order to decipher and fully understand changes in early pathological stages. We previously evaluated the lifetime contributions within the retina by using biexponential approximations: The shortest lifetimes were found to originate in RPE cells (τm: 260 ps), intermediate lifetimes in the neuronal retina (τm: 460 ps), and long lifetimes in the crystalline lens (τm: 1320 ps).41 Consequently, the short component in FLIO was assigned to the RPE, which was thought to have a stronger impact in the macular region due to its smaller retinal thickness and thereby reduced influence of inner retinal layers. An assumed influence of those layers is based on the observation of a substantial AF emission from within the neuronal retina, as has been shown by two-photon microscopy of porcine retinae ex vivo.16 However, the Henle fiber layer, containing MP, has not been in the focus of any previous FAF lifetime study. 
Influence of Xanthophyll on Macular FLI
Our results suggest a striking influence of MP fluorescence on the macular AF lifetime. We demonstrated that macular τm show a strong correlation to the amount of MP, independent of MT and subject's age. Due to the emission spectra of xanthophylls with a peak around 500 to 550 nm, the short-wavelength channel is more strongly influenced by this fluorescence than the long-wavelength channel. Lutein and zeaxanthin with 10 and 11 double bonds, respectively, emit fluorescence by relaxing directly from the S2 (instead of S1) to the S0 state.4245 The autofluorescence lifetThe AF lifetime of xanthophylls in vitro has been demonstrated to be very short, within the range of approximately 200 to 250 fs.28,30 We found longer macular AF lifetimes in vivo, where the macular fluorescence decayed in the lower picosecond range. The shortest average τm that we found within the central foveal area CS were 52 ± 12 ps (ch2: 105 ± 16 ps). These lifetime values clearly differ from those of xanthophylls in solution, which is probably due to the following: (1) The retina contains a composition of multiple fluorophores rather than a single component, and therefore its τm is very likely a superposition. The same applies to area CS, which presumably contains a large fraction of MP but to a certain extent also a fraction of longer-decaying fluorophores, such as lipofuscin or melanin. This is supported by the nonzero correlation of τm across area CS to the MPOD volume; (2) assuming that area CS would contain only MP, the resulting τm would still be longer compared to the in vitro decay time of a few hundred femtoseconds. This is due to the fact that FLIO's time resolution is limited to approximately 30 ps; (3) additionally, Billsten et al.46 found that the lifetime of the nonradiative S1 state of zeaxanthin is prolonged either when bound to the xanthophyll binding protein or in self-assembled aggregates.47 This may also affect the lifetime of the radiative S2 state, which, however, has not been demonstrated yet; and (4) the individual AF of the human lens may also have an additional but small impact on the fluorescence lifetimes measured in vivo. 
By characterizing the FAF, Delori et al.4,48 described a major (lipofuscin) and a minor retinal fluorophore. The minor fluorophore was especially detectable inside the fovea, with no age dependency and a fluorescence spectrum of 520 to 580 nm with a peak at 520 to 540 nm. Delori et al.49 suggested assigning the minor fluorophore to FAD, which has a weak fluorescence emission around 520 nm. Xanthophylls in vitro show a fluorescence emission of approximately 500 to 580 nm and a peak at 520 to 540 nm.28 This emission spectrum is very similar to the one described by Delori et al.4 The fluorescence lifetime of FAD is known to be around 2.3 to 3.0 ns (free) and around 150 ps (protein bound).2,5052 In cells, FAD is almost exclusively bound to proteins, resulting in a very weak fluorescence. Based on that, FAD cannot be excluded from contributing to the macular AF. However, flavoproteins can be found in mitochondria across the entire retina, not only within photoreceptor inner segments, but also to a large extent in Müller cells. Given this presumably homogenous distribution of FAD across the retina, there is no reason why FAD should strongly correlate with the MP distribution within the macula. We therefore suggest that the foveal AF is not solely based on the minor fluorophore FAD, as concluded by Delori et al.,49 but is rather a superposition of at least the xanthophylls and FAD. This suggestion is supported by the strong correlation of MPOD volume with τm; we therefore assume a major contribution of the xanthophylls to the foveal AF. We also assume that this correlation would be even higher if the FLIO did not have a resolution limit of approximately 30 ps. 
The MP distribution varies between individuals. The fractional contribution of MP is reflected in the FAF lifetimes; a higher fractional contribution of the xanthophylls leads to a shorter τm. This is especially visible in area C, where the different distribution patterns of MP28 show the greatest variance. 
With respect to the possible lipofuscin AF at the macular region, it is safe to argue that quenching processes by MP are negligible, if present at all. An average length of at least 20 to 35 μm for the photoreceptor outer segments in which no MP is present has been reported.2224,53,54 This excludes all short (Dexter)- and long (Foerster)-range quenching processes known. 
At the macular region, the AF is mainly a superposition of lipofuscin and MP fluorescence. Here, a mixture of lipofuscin components, located within the RPE, is spatially separated from an anterior layer containing MP, which acts as an excitation as well as emission filter. Therefore, it is reasonable to argue that the dependence of τm with MP is a filter effect, differently affecting single lipofuscin components, rather than a fluorescence contribution of the MP itself. However, this can be excluded because (1) FLIO uses a narrow excitation at 473 nm, and thus MP uniformly reduces the excitation of the lipofuscin AF; (2) lipofuscin and all their individual and identified components have similar, unstructured and broad emission spectra with a maximum centered around 600 nm4; (3) the small spectral overlap of MP absorption up to 520 nm28,46,55,56 and lipofuscin fluorescence emission are affecting only ch1; and (4) there is no absorption of the fluorescence emission by MP in ch2 but a correlation of τm with MPOD volume. 
Taking the reported MP fluorescence spectra (e.g., zeaxanthin: 550 ± 90 nm)28,55 into account, this provides a more consistent explanation. The fraction of MP fluorescence is larger in ch1 as compared to ch2. Likewise, τm and MPOD volume correlate more strongly for ch1. 
Finally, based on our findings and the current evidence available, we are strongly convinced that MP is not only acting as an excitation and emission filter on the lipofuscin AF, attenuating its contribution at the macula to a large extent, but also mainly contributing to the macular AF and thereby leading to the shortest AF lifetimes observed in the human retina. 
Additionally, we cannot completely exclude an influence of lipofuscin. To a small extent, it very likely contributes as background to the shortest AF lifetimes in the macula, since we measured an average τm (CS in ch1) of approximately 50 ps instead of 30 ps (resolution limit). 
In vitro, all lipofuscin components show longer τm (1352 ps)2 than the shortest mean AF that we found in the macular region in vivo. For the component A2E, however, different AF lifetimes have been reported (16 and 189 ps).2,57 In any case, a decline of A2E distribution at the macular region has been found.5861 Therefore, neither lipofuscin nor A2E can explain the short macular τm
Additionally, we investigated full-thickness idiopathic macular holes, full-layer defects of the retina located at the fovea centralis (Sauer L, Peters S, Schmidt J, Schweitzer D, Augsten R, Hammer M, unpublished data, 2015). The lifetimes inside the macular holes, without the influence of the neuronal retinal layers containing MP, are significantly longer compared to intact macular regions. 
Anatomical Influences
The foveal anatomy very likely influences the MP distribution, which is calculated as MPOD volume from the distributional AS and the mean OD of MP. The anatomical parameter measured in this study was the CST. In our study population, we found a medium negative correlation for CST with the MPOD volume. More specifically, we were able to show that the MT mainly influences the area across which the MP is distributed (AS), whereas the mean OD is less influenced. In conclusion, eyes with a smaller CST tend to contain a higher amount of MP at a given mean OD. An influence of MT on the MPOD has also been demonstrated previously.29 Additionally, thinner maculae seem to be connected with wider foveolae.36 Therefore, we speculate that the protective MP may show a wider distribution in those eyes to cover the entire foveal region. 
Autofluorescence lifetimes of area C showed a strong correlation to the central thickness of the same area. However, a partial correlation with the MPOD volume as a control parameter revealed a very weak, nonsignificant correlation. Thus, we assume that the foveal thickness has no direct but rather an indirect influence on the macular τm via AS. 
Age Influences
A slow decline of MP density with increasing age has been reported28; the MPOD may also depend on age.62,63 However, a different study found no such age influence.64 Our study showed only a very weak and nonsignificant age influence. Since we investigated only young subjects up to the age of 37, we cannot exclude an age dependency for an older subject group. 
The only age dependency we found was for τm, which was medium to weak. But this is an independent influence on the lifetimes, which may be through an accumulation of lipofuscin with age and/or through an increase of lens fluorescence. The hypothesis of increasing lipofuscin fluorescence with age is supported by the fact that the correlation of τm with age is higher for ch2, in which the lipofuscin AF has a greater impact. 
Evaluation of FLIO Data Analysis
Our findings of τm for area C (ch1: 82 ± 18 ps; ch2: 126 ± 13 ps) differ from previously published results (ch1: 124–390 ps; ch2: 189–355 ps).20 This is due to several reasons: Primarily, the fit approaches differ. By using two exponentials and a binning of 3 × 3 pixels, we found that the approximation process, carried out by the software, fails for some pixels. This also applies for the macular region. The obtained fit curves for those pixels do not describe the entire AF decay: They tend to cut off the fluorescence maximum, thereby leaving out the short fluorescence component. This results in apparently longer mean AF lifetimes, which can obviously be identified, for example, as blue pixels (= long τm) in the reddish macular region of the color-coded lifetime image. Evaluating all of those pixels within area C may therefore lead to a longer τm. To avoid incorrect fits, which in this case result from a small number of photons, either the number of binned pixels or the photon collection time needs to be increased. Therefore, we decided to use a binning of 5 × 5 pixels instead of 3 × 3 pixels, yielding a higher photon count per pixel. Additionally, we found a 3-exponential fit to more accurately represent our recorded FAF decays, which was based on evaluating χr2 values and the homogeneity of residual plots. Although the longest lifetime (τ3) contributed to τm only with a small amplitude, the goodness-of-fit criterion χr2 was consistently favorable, especially for the short component (τ1). With this approach, the incorrectly fitted pixels mentioned were absent in all fitted FAF lifetime images and lower χr2 values were obtained. A fit using two components and a binning of 3 × 3 pixels serves the purpose of a better spatial resolution, but seems unfavorable for the aforementioned reasons. 
In addition to a different fitting approach, the range of τm given in the previous study20 refers not only to area C, but also to the whole ETDRS grid (Fig. 1). Lifetimes are shortest only in area C, whereas the surrounding areas show longer decay times. 
Furthermore, due to older participants in the mentioned study group (age range, 22–61 years),20 the influence of longer FAF lifetimes with increasing age has most probably a stronger impact on τm, as compared to our study group (age range, 20–37 years). 
Evaluation of MPOD
In the current study, reflectometry at one wavelength was used for the MPOD measurement. We previously demonstrated a good correlation of this method to the two-wavelength AF method.65 An adequacy of 95% for the single-wavelength method with a 488-nm laser exposure has been shown.28 Recently, comparing this method to other MPOD measurements, weak correlations were found.66,67 However, Creuzot-Garcher et al.67 analyzed subjects between 23 and 30 years of age, disregarding juvenile reflexes. Those reflexes, caused by a smooth juvenile retina, mainly occur in patients up to the age of 35 to 40 years and corrupt the MPOD evaluation: Bright reflexes intensify the contrast in the single-wavelength reflectance method, thereby overestimating the amount of MP. Therefore, the reflectometry measurement of MPOD using the Zeiss fundus camera Visucam 500 is recommended by the manufacturer for patients above an age of 40 years. 
Since our subjects were exclusively younger than 40 years, we corrected the pigment density images manually by removing juvenile reflexes. This manual correction may have introduced a small systematic error for the absolute MPOD volume, but reinstalled the comparability and reproducibility for a younger subject group and eliminated an artificial age dependency for MPOD. 
Light scattering in the lens may also affect the one-wavelength reflectance measurement.68 By investigating only young subjects with clear lenses, we assume the impact of different lens scattering to be negligible. The reflectance-based measurement of MPOD is corrupted neither by the AF of a crystalline lens nor by that of the MP itself. This is a clear advantage over measuring MPOD by the depletion of the FAF. 
Conclusions
This study demonstrates that xanthophylls, which are known to have very short fluorescence decays, considerably contribute to the macular AF and especially to the AF lifetime. A direct measurement of MP autofluorescence lifetimes in vivo is not possible at this point; however, an indirect evidence of its influence is undeniable. We demonstrated that the foveal AF cannot be accounted for by the sole contribution of lipofuscin and FAD, as was previously assumed, but rather by a substantial contribution of xanthophyll fluorescence. Different distribution patterns of MP are reflected in FAF lifetime images and can be distinguished. Although lipofuscin is reported to show a greater fluorescence efficiency (10−2)57 than MP (10−4),69 its fluorescence excitation as well as emission is strongly reduced by MP in the retinal context, resulting in a still weak, but more prominent, MP fluorescence. The impact of this fluorescence is especially important to be considered when analyzing the lifetimes. Thus, FLI is providing additional information to the FAF imaging and potentially, imaging the FAF lifetimes could lead to a novel methodology for the detection of MP properties and pathology-induced changes in the living human retina. 
Nevertheless, in the context of a controversial discussion regarding the methods of MPOD evaluation, a comparative study employing an alternative technique for the measurement of MP is required to validate our findings. Additionally, further studies are necessary to evaluate the exact impact of both xanthophylls, lutein and zeaxanthin. 
Even though, as known, the foveal autofluorescence is weak, FLIO can estimate a contribution of MP to the macular τm. Therefore, the fluorescence of xanthophylls should be taken into account when investigating the macular region. 
Acknowledgments
The authors thank Heidelberg Engineering for providing the FLIO as well as for their technical assistance. The authors also thank Matthias Klemm for providing his FLIMX software to illustrate the FLIO images and the ARTORG Center for Biomedical Engineering Research from University of Bern for providing the software FLIO reader. 
Supported by the German Research Council, Grant No. HA 4430/3-1. 
Disclosure: L. Sauer, None; D. Schweitzer, None; L. Ramm, None; R. Augsten, None; M. Hammer, P; S. Peters, None 
References
Schweitzer D, Hammer M, Schweitzer F, et al. In vivo measurement of time-resolved autofluorescence at the human fundus. J Biomed Opt. 2004; 9: 1214–1222.
Schweitzer D, Schenke S, Hammer M, et al. Towards metabolic mapping of the human retina. Microsc Res Tech. 2007; 70: 410–419.
Eldred GE, Lasky MR. Retinal age pigments generated by self-assembling lysomorphotropic detergents. Nature. 1993; 361: 724–726.
Delori FC, Dorey CK, Staurenghi G, Arend O, Goger DG, Weiter JJ. In vivo fluorescence of the ocular fundus exhibits retinal pigment epithelium lipofuscin characteristics. Invest Ophthalmol Vis Sci. 1995; 36: 718–729.
Sparrow JR, Boulton M. RPE lipofuscin and its role in retinal pathobiology. Exp Eye Res. 2005; 80: 595–606.
Holz FG, Spaide RF. Medical Retina: Focus on Retinal Imaging. Berlin: Springer; 2010: 20–21.
Lakowicz JR. Principles of Fluorescence Spectroscopy. Baltimore: Springer US; 2007: 741.
Becker W. Fluorescence lifetime imaging—techniques and applications. J Microscopy. 2012; 247: 119–136.
Klemm M, Dietzel A, Haueisen J, Nagel E, Hammer M, Schweitzer D. Repeatability of autofluorescence lifetime imaging at the human fundus in healthy volunteers. Curr Eye Res. 2013; 38: 793–801.
Wagnieres GA, Star WM, Wilson BC. In vivo fluorescence spectroscopy and imaging for oncological applications. Photochem Photobiol. 1998; 68: 603–632.
Urayama PK, Mycek MA. Fluorescence lifetime imaging microscopy of endogenous biological fluorescence. In: Mycek M-A, Pogue BW, eds. Handbook of Biomedical Fluorescence. New York: Marcel-Dekker, Inc.; 2003: 211–236.
Elson D, Galletly N, Talbot C, et al. Multidimensional fluorescence imaging applied to biological tissue. In: Geddes C, Lakowicz J. eds. Reviews in Fluorescence 2006. New York: Springer; 2006: 477–524.
Chorvat D, Chorvatova A. Multi-wavelength fluorescence lifetime spectroscopy: a new approach to the study of endogenous fluorescence in living cells and tissues. Laser Phys Lett. 2009; 6: 175–193.
Marcu L. Fluorescence lifetime techniques in medical applications. Ann Biomed Eng. 2012; 40: 304–331.
Schweitzer D, Gaillard ER, Dillon J, et al. Time-resolved autofluorescence imaging of human donor retina tissue from donors with significant extramacular drusen. Invest Ophthalmol Vis Sci. 2012; 53: 3376–3386.
Peters S, Hammer M, Schweitzer D. Two-photon excited fluorescence microscopy application for ex vivo investigation of ocular fundus samples. In: So P, Beaurepaire E, eds. European Conference on Biomedical Optics (ECBO) Munich, Germany. Optical Society of America; 2011: 808605.
Schweitzer D, Klemm M, Quick S, et al. Detection of early metabolic alterations in the ocular fundus of diabetic patients by time-resolved autofluorescence of endogenous fluorophores. Proc SPIE Int Soc Opt Eng. 2011; 80871G.
Jentsch S, Schweitzer D, Schmidtke KU, et al. Retinal fluorescence lifetime imaging ophthalmoscopy measures depend on the severity of Alzheimer's disease. Acta Ophthalmol. 2015; 93: e241–e247.
Schweitzer D, Deutsch L, Klemm M, et al. Fluorescence lifetime imaging ophthalmoscopy in type 2 diabetic patients who have no signs of diabetic retinopathy. J Biomed Opt. 2015; 20: 61106.
Dysli C, Quellec G, Abegg M, et al. Quantitative analysis of fluorescence lifetime measurements of the macula using the fluorescence lifetime imaging ophthalmoscope in healthy subjects. Invest Ophthalmol Vis Sci. 2014; 55: 2106–2113.
Bone RA, Landrum JT, Tarsis SL. Preliminary identification of the human macular pigment. Vision Res. 1985; 25: 1531–1535.
Snodderly DM, Auran JD, Delori FC. The macular pigment. II. Spatial distribution in primate retinas. Invest Ophthalmol Vis Sci. 1984; 25: 674–685.
Trieschmann M, van Kuijk FJ, Alexander R, et al. Macular pigment in the human retina: histological evaluation of localization and distribution. Eye. 2008; 22: 132–137.
Krinsky NI. Possible biologic mechanisms for a protective role of xanthophylls. J Nutr. 2002; 132: 540S–542S.
Landrum JT, Bone RA. Lutein, zeaxanthin, and the macular pigment. Arch Biochem Biophys. 2001; 385: 28–40.
Delori FC, Goger DG, Hammond BR, Snodderly DM, Burns SA. Macular pigment density measured by autofluorescence spectrometry: comparison with reflectometry and heterochromatic flicker photometry. J Opt Soc Am A Opt Image Sci Vis. 2001; 18: 1212–1230.
Ermakov IV, Ermakova MR, Gellermann W. Simple Raman instrument for in vivo detection of macular pigments. Appl Spectrosc. 2005; 59: 861–867.
Sharifzadeh M, Bernstein PS, Gellermann W. Nonmydriatic fluorescence-based quantitative imaging of human macular pigment distributions. J Opt Soc Am A Opt Image Sci Vis. 2006; 23: 2373–2387.
Meyer zu Westrup V Dietzel M, Pauleikhoff D, Hense HW. The association of retinal structure and macular pigment distribution. Invest Ophthalmol Vis Sci. 2014; 55: 1169–1175.
Kandori H, Sasabe H, Mimuro M. Direct determination of a lifetime of the S2 state of beta-carotene by femtosecond time-resolved fluorescence spectroscopy. J Am Chem Soc. 1994; 116: 2671–2672.
Becker W. The bh TCSPC Handbook. 6th ed. Berlin: Becker & Hickl GmbH; 2014.
ANSI. American National Standard for the Safe Use of Lasers. ANSI Z 136.1-2000. Orlando FL: Laser Institute of America; 2000.
Delori FC, Webb RH, Sliney DH. American National Standards I. Maximum permissible exposures for ocular safety (ANSI 2000), with emphasis on ophthalmic devices. J Opt Soc Am A Opt Image Sci Vis. 2007; 24: 1250–1265.
Luchowski R, Szabelski M, Sarkar P, et al. Fluorescence instrument response standards in two-photon time-resolved spectroscopy. Appl Spectrosc. 2010; 64: 918–922.
Schneider CA, Rasband WS, Eliceiri KW. NIH, Image to ImageJ: 25 years of image analysis. Nat Methods. 2012; 9: 671–675.
Tick S, Rossant F, Ghorbel I, et al. Foveal shape and structure in a normal population. Invest Ophthalmol Vis Sci. 2011; 52: 5105–5110.
Tanaka F, Tamai N, Yamazaki I, Nakashima N, Yoshihara K. Temperature-induced changes in the coenzyme environment of D-amino acid oxidase revealed by the multiple decays of FAD fluorescence. Biophys J. 1989; 56: 901–909.
Lakowicz JR, Szmacinski H, Nowaczyk K, Johnson ML. Fluorescence lifetime imaging of free and protein-bound NADH. Proc Natl Acad Sci U S A. 1992; 89: 1271–1275.
Schweitzer D, Quick S, Schenke S, et al. Comparison of parameters of time-resolved autofluorescence between healthy subjects and patients suffering from early AMD [in German]. Ophthalmologe. 2009; 106: 714–722.
Miura Y, Huettmann G, Orzekowsky-Schroeder R, et al. Two-photon microscopy and fluorescence lifetime imaging of retinal pigment epithelial cells under oxidative stress. Invest Ophthalmol Vis Sci. 2013; 54: 3366–3377.
Schweitzer D. Metabolic mapping. In: Holz F, Spaide R, eds. Medical Retina. Berlin, Heidelberg: Springer; 2010: 107–123.
Snyder R, Arvidson E, Foote C, Harrigan L, Christensen RL. Electronic-energy levels in long polyenes: S2 .fwdarw. S0 emission in all-trans-13,5,7,9,11,13-tetradecaheptaene. J Am Chem Soc. 1985; 107: 4117–4122.
Cosgrove SA, Guite MA, Burnell TB, Christensen RL. Electronic relaxation in long polyenes. J Phys Chem. 1990; 94: 8118–8124.
Decoster B, Christensen RL, Gebhard R, Lugtenburg J, Farhoosh R, Frank HA. Low-lying electronic states of carotenoids. Biochim Biophys Acta. 1992; 1102: 107–114.
Andersson PO, Gillbro T. Photophysics and dynamics of the lowest excited singlet-state in long substituted polyenes with implications to the very long-chain limit. J Chem Phys. 1995; 103: 2509–2519.
Billsten HH, Bhosale P, Yemelyanov A, Bernstein PS, Polivka T. Photophysical properties of xanthophylls in carotenoproteins from human retinas. Photochem Photobiol. 2003; 78: 138–145.
Billsten HH, Sundstrom V, Polivka T. Self-assembled aggregates of the carotenoid zeaxanthin: time-resolved study of excited states. J Phys Chem A. 2005; 109: 1521–1529.
Delori FC. Spectrophotometer for noninvasive measurement of intrinsic fluorescence and reflectance of the ocular fundus. Appl Opt. 1994; 33: 7439–7452.
Benson RC, Meyer RA, Zaruba ME, Mckhann GM. autofluorescence—is it due to flavins? J Histochem Cytochem. 1979; 27: 44–48.
Nakashima N, Yoshihara K, Tanaka F, Yagi K. Picosecond fluorescence lifetime of the coenzyme of D-amino acid oxidase. J Biol Chem. 1980; 255: 5261–5263.
Koziol B, Markowicz M, Kruk J, Plytycz B. Riboflavin as a source of autofluorescence in Eisenia fetida coelomocytes. Photochem Photobiol. 2006; 82: 570–573.
Skala MC, Riching KM, Bird DK, et al. In vivo multiphoton fluorescence lifetime imaging of protein-bound and free nicotinamide adenine dinucleotide in normal and precancerous epithelia. J Biomed Opt. 2007; 12: 024014.
Greeff R, Graefe AK. Die mikroskopische anatomie des sehnerven und der netzhaut. Leipzig; 1900.
Rodieck RW. The First Steps in Seeing. Sunderland MA: Sinauer Associates; 1998: 38–55.
Frank HA, Bautista JA, Josue JS, Young AJ. Mechanism of nonphotochemical quenching in green plants: energies of the lowest excited singlet states of violaxanthin and zeaxanthin. Biochemistry. 2000; 39: 2831–2837.
Krinsky NI. Possible biologic mechanisms for a protective role of xanthophylls. J Nutr. 2002; 132: 540S–542S.
Lamb LE, Ye T, Haralampus-Grynaviski NM, et al. Primary photophysical properties of A2E in solution. J Phys Chem B. 2001; 105: 11507–11512.
Bhosale P, Serban B, Bernstein PS. Retinal carotenoids can attenuate formation of A2E in the retinal pigment epithelium. Arch Biochem Biophys. 2009; 483: 175–181.
Ablonczy Z, Gutierrez DB, Grey AC, Schey KL, Crouch RK. Molecule-specific imaging and quantitation of A2E in the RPE. Adv Exp Med Biol. 2012; 723: 75–81.
Ablonczy Z, Higbee D, Anderson DM, et al. Lack of correlation between the spatial distribution of A2E and lipofuscin fluorescence in the human retinal pigment epithelium. Invest Ophthalmol Vis Sci. 2013; 54: 5535–5542.
Ablonczy Z, Higbee D, Grey AC, Koutalos Y, Schey KL, Crouch RK. Similar molecules spatially correlate with lipofuscin and N-retinylidene-N-retinylethanolamine in the mouse but not in the human retinal pigment epithelium. Arch Biochem Biophys. 2013; 539: 196–202.
Beatty S, Murray IJ, Henson DB, Carden D, Koh H, Boulton ME. Macular pigment and risk for age-related macular degeneration in subjects from a Northern European population. Invest Ophthalmol Vis Sci. 2001; 42: 439–446.
Obana A, Gohto Y, Tanito M, et al. Effect of age and other factors on macular pigment optical density measured with resonance Raman spectroscopy. Graefes Arch Clin Exp Ophthalmol. 2014; 252: 1221–1228.
Werner JS, Donnelly SK, Kliegl R. Aging and human macular pigment density. Appended with translations from the work of Max Schultze and Ewald Hering. Vision Res. 1987; 27: 257–268.
Schweitzer D, Jentsch S, Dawczynski J, Hammer M, Wolf-Schnurrbusch UE, Simple Wolf S. and objective method for routine detection of the macular pigment xanthophyll. J Biomed Opt. 2010; 15: 061714.
Dennison JL, Stack J, Beatty S, Nolan JM. Concordance of macular pigment measurements obtained using customized heterochromatic flicker photometry, dual-wavelength autofluorescence, and single-wavelength reflectance. Exp Eye Res. 2013; 116: 190–198.
Creuzot-Garcher C, Koehrer P, Picot C, Aho S, Bron AM. Comparison of two methods to measure macular pigment optical density in healthy subjects. Invest Ophthalmol Vis Sci. 2014; 55: 2941–2946.
Komar B, Rauscher FG, Wiedemann R, Dawczynski J. Macular pigment optical density measurements by one-wavelength reflection photometry—influence of cataract surgery on the measurement results. Graefes Arch Clin Exp Ophthalmol. 2014; 252: 1717–1727.
Bhosale P, Li B, Sharifzadeh M, et al. Purification and partial characterization of a lutein-binding protein from human retina. Biochemistry. 2009; 48: 4798–4807.
Figure 1
 
FLIO setup and images. Overview of the general FLIO setup as provided by Heidelberg Engineering (A). Three representative photon arrival histograms (blue dots) of characteristic areas obtained from one FLIO scan of a healthy fundus, (B) macula, (C) temporal periphery, and (D) optic nerve head; fitted curve (red) and instrument response function (green curve). Typical false color-coded mean FAF lifetime (τm) images of the spectral channel 1 (E) and channel 2 (F) and a fluorescence intensity picture (G) overlaid with the standard ETDRS grid.
Figure 1
 
FLIO setup and images. Overview of the general FLIO setup as provided by Heidelberg Engineering (A). Three representative photon arrival histograms (blue dots) of characteristic areas obtained from one FLIO scan of a healthy fundus, (B) macula, (C) temporal periphery, and (D) optic nerve head; fitted curve (red) and instrument response function (green curve). Typical false color-coded mean FAF lifetime (τm) images of the spectral channel 1 (E) and channel 2 (F) and a fluorescence intensity picture (G) overlaid with the standard ETDRS grid.
Figure 2
 
Macular τm. Distribution of short macular mean AF lifetime (τm) within our study group for both spectral channels and both utilized grid sizes (A). The grid sizes C (B) and CS (C) are depicted within a false color-coded τm image.
Figure 2
 
Macular τm. Distribution of short macular mean AF lifetime (τm) within our study group for both spectral channels and both utilized grid sizes (A). The grid sizes C (B) and CS (C) are depicted within a false color-coded τm image.
Figure 3
 
Topographic distribution of short macular τm. Typical mean FAF lifetime (τm) images (false color-coded) of the spectral channels 1 (A) and 2 (B) and their corresponding OCT images (C) of four healthy subjects (ad). Their respective central subfield thickness (obtained from OCT) and MPOD volume are plotted below (D).
Figure 3
 
Topographic distribution of short macular τm. Typical mean FAF lifetime (τm) images (false color-coded) of the spectral channels 1 (A) and 2 (B) and their corresponding OCT images (C) of four healthy subjects (ad). Their respective central subfield thickness (obtained from OCT) and MPOD volume are plotted below (D).
Figure 4
 
Correlations of mean FAF lifetimes (area C) and MPOD volume. A scatterplot of linear correlations for mean FAF lifetimes (τm) of area C to the MPOD volume (ch1: r = −0.76; ch2: r = −0.66; P < 0.001).
Figure 4
 
Correlations of mean FAF lifetimes (area C) and MPOD volume. A scatterplot of linear correlations for mean FAF lifetimes (τm) of area C to the MPOD volume (ch1: r = −0.76; ch2: r = −0.66; P < 0.001).
Figure 5
 
Topographic similarity of τm and MPOD volume. Macular mean AF lifetime (τm) and MPOD volume for three representative, healthy subjects with different spatial distributions of MP ([A] female, 23 years, broad-shaped MP; [B] female, 22 years, ring-shaped MP; [C] male, 27 years, cone-shaped MP); channel 1 τm images (A1, B1, C1); corresponding 3D projection of macular τm (A2, B2, C2) and 3D MPOD volume plots (A3, B3, C3).
Figure 5
 
Topographic similarity of τm and MPOD volume. Macular mean AF lifetime (τm) and MPOD volume for three representative, healthy subjects with different spatial distributions of MP ([A] female, 23 years, broad-shaped MP; [B] female, 22 years, ring-shaped MP; [C] male, 27 years, cone-shaped MP); channel 1 τm images (A1, B1, C1); corresponding 3D projection of macular τm (A2, B2, C2) and 3D MPOD volume plots (A3, B3, C3).
Table 1
 
Correlations of τm, MP, MT, and Age
Table 1
 
Correlations of τm, MP, MT, and Age
Table 2
 
Multivariate Regression Analysis
Table 2
 
Multivariate Regression Analysis
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×