August 2013
Volume 54, Issue 8
Free
Multidisciplinary Ophthalmic Imaging  |   August 2013
A Prototype Hyperspectral System With a Tunable Laser Source for Retinal Vessel Imaging
Author Affiliations & Notes
  • Sunni R. Patel
    Department of Ophthalmology and Vision Sciences, University Health Network, Toronto Western Hospital, Toronto, Canada
  • John G. Flanagan
    Department of Ophthalmology and Vision Sciences, University Health Network, Toronto Western Hospital, Toronto, Canada
    School of Optometry, University of Waterloo, Waterloo, Ontario, Canada
  • Ayda M. Shahidi
    Department of Ophthalmology and Vision Sciences, University Health Network, Toronto Western Hospital, Toronto, Canada
  • Jean-Philippe Sylvestre
    Photon etc., Montréal, Québec, Canada
  • Chris Hudson
    Department of Ophthalmology and Vision Sciences, University Health Network, Toronto Western Hospital, Toronto, Canada
    School of Optometry, University of Waterloo, Waterloo, Ontario, Canada
  • Correspondence: Sunni R. Patel, FP 6-206, Toronto Western Hospital, Department of Ophthalmology and Vision Sciences, VSRP/UHN, 399 Bathurst Street, Toronto, Canada, M5T 2S8; sunni_patel@hotmail.com
Investigative Ophthalmology & Visual Science August 2013, Vol.54, 5163-5168. doi:10.1167/iovs.13-12124
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Sunni R. Patel, John G. Flanagan, Ayda M. Shahidi, Jean-Philippe Sylvestre, Chris Hudson; A Prototype Hyperspectral System With a Tunable Laser Source for Retinal Vessel Imaging. Invest. Ophthalmol. Vis. Sci. 2013;54(8):5163-5168. doi: 10.1167/iovs.13-12124.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose.: To describe the technology and determine the within-session repeatability of manual retinal reflectance measurements of arterioles and venules using a prototype hyperspectral retinal camera.

Methods.: Six healthy young volunteers (three males, average age 26 ± 4 years) had five repeated sets of retinal images captured between 500 and 600 nm at 5-nm intervals using a newly developed hyperspectral retinal camera. Optical densities were manually extracted for first-degree arterioles and venules and the repeatability of retinal reflectance was compared sequentially. The SDs of the differences between sequential mean values were used as an indication of the variance, while the coefficient of repeatability (COR) and intraclass correlation coefficient (ICC) were used to assess repeatability.

Results.: The mean difference between each sequential measure was calculated using 21 images from each of the five spectral cubes. The SDs of these values ranged from 0.01 to 0.06 OD units and from 0.01 to 0.07 OD units for first-degree arterioles and venules, respectively. The COR ranged from 0.02 to 0.11 OD units (relative to a mean OD of 0.15 [0.06–0.23] OD units) for arterioles and 0.03 to 0.14 OD units (relative to a mean OD of 0.25 [0.17–0.31] OD units) for venules. Good reliability (P < 0.001) was found for arterioles (ICC: 78.8%–94.4% with a Cronbach's α of 89.6%–97.6%) and for venules (ICC: 63.7%–92.1% with a Cronbach's α of 86.2%–98.1%).

Conclusions.: Manual optical density determination with this novel hyperspectral camera showed very good intrasession (and intraobserver) repeatability with a small degree of variance that should form the basis of reliable retinal oxygen saturation values in future imaging research studies. Future automation of retinal vessel reflectance image analyses will likely further improve this repeatability.

Introduction
Physiologically it is understood that oxygen saturation levels in retinal arterial and venous blood vessels vary substantially, with differences averaging at approximately 40%. Retinal oximetry can be used to help understand the mechanisms behind retinal health and pathophysiological disease processes impacting the retina by investigating such differences in retinal arterial and venous blood vessel oxygen saturation and, ultimately, this technology may potentially be adopted in clinical practice. 
Hyperspectral imaging was originally developed for remote sensing and astronomy applications. More recently, the use of this imaging technique alongside fluorescence microscopy has also been used in clinical research. 14 Although the application of hyperspectral technology in retinal imaging is relatively new, the concept is clinically appealing. 5,6  
Hyperspectral imaging determines the spectral absorption characteristics for multiple narrow spectral bandwidths within the UV to infrared spectrum. Using two-dimensional images at each specific wavelength, a “spectral cube” can be created whereby the third-dimension aspect of the image cube represents tissue absorbance values as a function of the wavelength. Thus, the molecular content within each image can be determined using the known absorption spectra of given molecules. 
Oxygenated (HbO2) and deoxygenated (Hb) hemoglobin are well-established molecules in terms of their absorption criteria, and, as previously reported by other retinal oximetry studies, the quantification of these molecules in the retina has clear clinical relevance to the pathophysiology of retinal vascular disease. 
Here, we introduce a prototype system that enables spectrophotometric imaging of retinal vessels using a tunable laser source (TLS) coupled to a custom-built fundus camera. In this article, preliminary retinal spectral images and optical density (OD) results are presented from six healthy volunteers using a nonflash modified hyperspectral retinal camera (HRC) with rapid sequential wavelength presentation from 500 to 600 nm at 5-nm intervals. 
The purpose of this article was to describe the technology and to also determine the reliability of the within-session repeated measures in healthy volunteers using this prototype hyperspectral retinal camera. 
Methods
The hyperspectral imaging system is based on a custom-built mydriatic fundus camera (Photon etc., Montreal, QC, Canada) incorporating a TLS as the light source (Fig. 1). 
Figure 1
 
Schematic representation of the HRC incorporating the CCD within the custom-built retinal camera and the TLS as the light source.
Figure 1
 
Schematic representation of the HRC incorporating the CCD within the custom-built retinal camera and the TLS as the light source.
The TLS is able to transmit wavelengths within a spectral range of 420 to 1000 nm (visible to near-infrared) with a bandwidth of 2 nm, allowing rapid wavelength selection from the stable and powerful super-continuum light source (Leukos-SM-30-OEM; Leukos Innovative Optical Systems, Limoges, France), which delivers approximately 100 mW over 420 nm to 2400 nm. The resulting tunable light source provides safe and comfortable illumination of the retina and eliminates the use of a conventional Xenon flash lamp. The TLS is electronically tunable and the system incorporates an automatic spectral calibration system to achieve precise and accurate (<1 nm) wavelength selection. During image acquisition, a small fraction of the illumination light is directed onto a power meter to permit correction for temporal light power fluctuations. 
The instrument includes a sensitive 1.3-MPixel (1392 × 1040 pixel) 14-bit charge-coupled device (CCD) camera (Pixelfly USB; PCO AG, Kelheim, Germany) for high-definition imaging. Approximately 37.4° diagonal field of view is maintained by the opto-mechanics of the system and the instrument is capable of focusing up to ±9 diopters. 
The imaging system is controlled using PHySpec (Photon etc.), a custom-designed software that allows for the acquisition of hyperspectral images and post hoc analysis to extract reflected light intensity of the retinal vessels. 
Participants
Images were acquired from six healthy volunteers (three males, mean age 26 years; age range, 20–32) recruited at the Retinal Research Laboratory within the Department of Ophthalmology at the Toronto Western Hospital. Written informed consent was obtained from all the volunteers and the study protocol adhered to the tenets of the Declaration of Helsinki. The study was approved by University Health Network and University of Waterloo Ethic Boards. 
In all volunteers, best-corrected visual acuity was 0.00, or better, using the logarithm of the minimum angle of resolution (LogMAR) acuity chart and there was no evidence of ocular disease. IOPs (using Goldmann tonometry) were taken before pupil dilatation with Tropicamide 1% (Mydriacyl; Alcon Canada, Inc., Mississauga, Canada) and a full ocular examination (slit-lamp fundus biomicroscopy examination) was performed before all measurements were taken. Additionally, all volunteers acclimatized to a room temperature of 20 to 22°C for 20 minutes before baseline blood pressure and pulse oximetry measurements were taken simultaneously (CardioCap/5; Datex-Ohmeda, Louisville, KY). 
Imaging Procedure
The volunteer's head was stabilized using the chin rest and forehead support and the volunteer was then asked to maintain eye fixation on a red-fixation light in front of the contralateral eye (Fig. 2). 
For each volunteer, 21 retinal images were acquired by two trained operators (SRP or AMS) using wavelengths between 500 and 600 nm at 5-nm intervals to generate one spectral cube using an exposure time of 80 ms for each wavelength with an average acquisition time of 10 seconds. 
Five repeated sequences (i.e., five spectral cubes) were taken for all volunteers. Live retinal images at a wavelength of 570 nm were initially brought into focus. The 570-nm wavelength allowed good contrast of the blood vessel and facilitated the focusing procedure. This wavelength is also close to the midrange of the spectral band of interest (500–600 nm) in this study, which ensured that a reasonably good focus was achieved on the full spectral band of interest. Indeed, a previous study using an eye model (OEMI-7; Ocular Instruments, Bellevue, WA) showed that the focus remained reasonably good at ±150 nm around the wavelength where focus adjustment was performed. Once a sharp live image was achieved, the participants were instructed to avoid blinking while fixating on the external target until the automatic acquisition procedure was finished. 
Image Processing.
Following spectral image acquisition, the final spectral cubes were normalized and registered independently by a single researcher (SRP). 
The optics used by the system creates parasitic reflections responsible for a central “hot spot” and peripheral darkening that need to be primarily eliminated. A normalization procedure was implemented on all acquired spectral cubes to correct for image ghosting and spectral responses of the retinal camera (Fig. 3). 
Figure 3
 
Left: Fundus image at 570 nm (right eye [RE] −0.50 diopter cylindrical [DC]) taken before normalization. Right: Same image after normalization procedure.
Figure 3
 
Left: Fundus image at 570 nm (right eye [RE] −0.50 diopter cylindrical [DC]) taken before normalization. Right: Same image after normalization procedure.
More precisely, the raw retinal data cube (Ceye ) was corrected/normalized using the following equation (equation 1) and two reference cubes, the baseline cube (Cbaseline ) and the white cube (Cwhite ), with their respective associated powers (P) for each image, as well as dark images (Idark ):  The baseline cube is obtained during an acquisition where a light trap (beam dump) is placed at the position of the eye to image the parasitic reflections caused by the retinal camera. The white cube is obtained by imaging the surface of a material with high diffuse reflectance (Spectralon; Labsphere, Inc., North Sutton, NH) through a triplet lens mimicking the optical properties of the eye. A Gaussian filter is applied to the white cube to smooth any surface imperfection. The dark images are obtained when no light is reaching the CCD detector (plastic cap put in front of the objective lens of the retinal camera and light source turned off).  
The resulting cube (Ceye,norm ) is theoretically normalized/corrected for the ghosts (with the baseline), the spectral and spatial intensity response of the HRC (with the white), and the temporal incident light fluctuations (with the power values and the dark image). 
A registration procedure was used on the normalized spectral cubes to correct for wavelength-dependent optical deformations (scaling) and fine eye movement (translation). The scaling correction is constant in time and was determined during the instrument set-up and qualification (Photon etc.). PHySpec includes a registration procedure allowing translation correction between images in the cube. To adequately translate the images, the optic disc and other retinal features within an individual's images (such as large vessels or distinguishing areas) that are similar in appearance in all the sequential spectral images were chosen at the reference wavelength of 570 nm. For alignment purposes exclusively, the registration procedure was enhanced using a customized Gaussian filter (filter size equal to 101). 
Image Qualification.
The normalized and registered spectral cubes consisting of 21 spectral images (Fig. 4) were then assessed namely for image quality, focus, and registration. The spectral images were then also assessed for consistency against the already known absorption characteristics of HbO2 and Hb (i.e., it was expected that at wavelengths greater than 600 nm that the arterioles would appear less optically dense than venules). 
Figure 4
 
Typical images taken by the HRC to create a hyperspectral cube of retinal images between 500 and 600 nm at 5-nm intervals with an exposure time of 80 ms.
Figure 4
 
Typical images taken by the HRC to create a hyperspectral cube of retinal images between 500 and 600 nm at 5-nm intervals with an exposure time of 80 ms.
Optical Density Determination.
The system is capable of automatic intensity spectra determination at manually specified locations. Using the custom software (PHySpec; Photon etc.), a linear profile of a fixed length (20–30 pixels long, 2 pixels wide [meaning that the extracted spectra are averages of at least 40 pixels], within two-disc diameter of the center of the optic nerve head) was manually selected within and on either side of the vessel wall. The intensity values for both arterioles and venules were then extracted (within and outside the vessel) along the length of each profile for each wavelength. The ODs of the superior temporal vessels were measured by calculating the log ratio of the measured reflected light intensity of the adjacent retinal area to the selected vessel area (IR ) over to the measured intensity of within the selected vessel (IV ) :    
Statistical Analysis
All data were tested for normal distribution using the Kolmogorov-Smirnov test before applying any statistical tests. All data were found to be normally distributed (P > 0.05), so parametric tests were applied using Statistica (Statsoft, Tulsa, OK). The arteriolar and venular retinal reflectance values from each acquired image (i.e., five images at each of the 21 wavelengths) were used (as opposed to average values) and compared sequentially using Bland-Altman, coefficient of repeatability (COR), intraclass correlation coefficient (ICC), and SDs of the differences between sequential mean analysis. A P value less than 0.01 was adopted to minimize the influence of Type II errors induced by multiple comparisons (i.e., which results in a loss of power to detect real differences). 
Results
The baseline characteristics for the volunteers are presented in Table 1
Table 1
 
Average Baseline Characteristics of the Six Healthy Volunteers in This Study
Table 1
 
Average Baseline Characteristics of the Six Healthy Volunteers in This Study
IOP, mm Hg ± SD 13 ± 1
Systolic blood pressure, mm Hg ± SD 108 ± 5
Diastolic blood pressure, mm Hg ± SD 66 ± 4
Heart rate, mm Hg ± SD 81 ± 3
Pulse oximeter oxygen saturation, % ± SD 98.3 ± 0.8
The average (averaged for all five repeated measures) OD values for the arterioles and venules are shown graphically in Figure 5 for all the volunteers. 
Figure 5
 
The average arteriolar (left) and venular (right) OD values at each given (5-nm) imaged wavelength from 500 to 600 nm for all of the volunteers.
Figure 5
 
The average arteriolar (left) and venular (right) OD values at each given (5-nm) imaged wavelength from 500 to 600 nm for all of the volunteers.
Figure 6 shows the typical presentation (example of one participant) of repeated measures for arterioles and venules when plotting the OD values for each wavelength derived from each acquired hyperspectral cube. These values were used individually for repeated measure analysis, as shown in Tables 2 and 3. The values given for intraclass correlation, Cronbach's α, and correlations all reached statistical significance (P < 0.001). 
Figure 6
 
Typical example from a volunteer showing OD values for arterioles (left) and venules (right) for each of the five acquired hyperspectral cubes.
Figure 6
 
Typical example from a volunteer showing OD values for arterioles (left) and venules (right) for each of the five acquired hyperspectral cubes.
Table 2
 
Results Showing Ranges of Values for ODs, SDs, Mean of Differences Between Sequential Measures, and CORs for Repeated Measures in Arterioles and Venules
Table 2
 
Results Showing Ranges of Values for ODs, SDs, Mean of Differences Between Sequential Measures, and CORs for Repeated Measures in Arterioles and Venules
Mean OD (Range) SD of Difference Between Means Mean of Differences COR
Arteriole 0.15 (0.06–0.23) 0.01–0.06 −0.08–0.09 0.02–0.11
Venule 0.25 (0.17–0.31) 0.01–0.07 −0.06–0.06 0.03–0.14
Table 3
 
Results Showing Limits of Agreement, ICC, Cronbach's α, and Correlation Values for Repeated Sequential Measures of Arterioles and Venules
Table 3
 
Results Showing Limits of Agreement, ICC, Cronbach's α, and Correlation Values for Repeated Sequential Measures of Arterioles and Venules
Limits of Agreement ICC Cronbach's α Correlation Between Sequential Measures, r 2
Arteriole −0.06–0.06 78.8%–94.4%* 89.6%–97.6%* 0.800–0.912*
Venule −0.04–0.07 63.7%–92.1%* 86.2%–98.1%* 0.671–0.945*
Discussion
The application of modified camera systems that image the retinal vessel at two or more wavelengths has been reported in recent literature. Early in vivo measurements of retinal vessel oxygen saturation have been reported using a dual-wavelength measurement approach. 79 However, the use of multispectral imaging is now gaining pace, with recent work showing success in image acquisition and oxygen saturation acquisition. 6,1013 Indeed, models based on only two or three wavelengths may be too simplistic to reliably and accurately provide retinal oximetry measurements, as the light reflectance measurements of the blood vessels are complex and influenced by a myriad of factors (including the diameter of the vessel, retinal pigmentation, lens yellowing, light scattering and autofluorescence occurring in the vitreous and cornea, and illumination uniformity). Models based on spectral-rich datasets are expected to be much more powerful in that respect. 14,15 It was suggested that increasing the number of discrete wavelengths at which the ocular fundus reflectance is measured could increase, by the square root of the considered wavelength number, the precision of the blood oximetry evaluation. 16 To our knowledge, however, no study compared directly the experimental results obtained with a two- or three-wavelength method with a method based on hyperspectral datasets. 
The study demonstrated that hyperspectral imaging can be achieved using a modified nonflash camera equipped with a high-definition CCD camera and TLS. The data reveal promising within-session repeatability of the current HRC system for repeated hyperspectral imaging within session (COR 0.02–0.11 OD units for arterioles and 0.03–0.14 OD units for venules). This compares well to current commercially available two-wavelength instruments that have quoted ICC values between 0.91 and 0.94 for retinal branch arteries and between 0.84 and 0.88 for retinal branch veins. 17  
Other reproducibility data report an SD of 3.7% in arteries and 5.3% in veins, 18 whereas recent data for the same system showed an SD for repeated measurements of 1.0% and 1.4% in retinal arteries and veins respectively. 19 The former values fall very similar to those reported in our study and are as expected with a new prototype system. 
Also consistent with others, 17 venous values exhibit more deviation as compared with arterioles, but whether this is due to physiological variations in oxygen levels, vessel properties of the veins, or the imaging capabilities of the system is yet unknown. 
As this instrument is at its very early stage of development, a number of technical limitations are potential sources of error. First, eye movements are likely to occur during the 10 seconds that are required to obtain the 21 images sequentially. These eye movements may affect the illumination conditions for a given retinal region from one frame to another. Although the general form of the illumination intensity as a function of the position is corrected for by the normalization process, more subtle effects, such as shadowing, and specular and parasitic reflections, may still be present. Furthermore, variations in local oxygen content throughout the cardiac cycle may also occur during the acquisition period. Future optimization of the instrument should result in shorter acquisitions (possibly 1–3 seconds), as frequently observed in optical coherence tomography imaging and limit these effects. Also, it is important to note that the automation of procedures has not yet been implemented within the system. Thus, the determination of OD values is performed manually, which inevitably introduces variability in the process (for example, the line selections performed in the five spectral cubes of a given subject were most probably not from exactly the same location, although effort was made to repeat the selections as closely as practically possible). Furthermore, the image registration algorithms implemented in this preliminary version of the analysis software could not correct for the rotation often observed in the images due to eye movements during data acquisition. Moreover, the perfectible optics design of the retinal camera is responsible for a number of image artifacts that could not be completely corrected in the image-processing step and this may contribute to the data variability observed. However, as image analysis (J-PS) and reflectance extraction and statistical analysis (SRP) was performed by only one user, this did reduce the amount of error that could possibly affect the results. 
The next step will be to extract retinal oximetry values from the hyperspectral data. Different approaches have been proposed for this purpose, including curve fitting 5,14 and curve integration. 10,20 To determine the best possible approach, a study is currently under way where hyperpspectral datasets are collected during gas provocation (i.e., that the systemic oximetry is precisely controlled). To identify the best algorithm for retinal oximetry, we will aim to maximize the correlation between the arteriolar retinal oximetry and systemic oximetry values. 
Finally, it is pertinent to compare the technology of the HRC based on a TLS used in this study to other hyperspectral instruments described previously. Systems based on push-broom 10,20 and pinhole 14 spectrograph collect spectral-rich information from a line or a point, respectively, of the retina. Reconstitution of a full two-dimensional image is impractical in humans, as the process requires complete immobilization of the eye for the duration of the acquisition, lasting several seconds. In comparison, the system used in our study obtains two-dimensional images for an array of wavelengths. The common landmark features (e.g., optic disk, vessels, macula) can therefore be used to realign the images with theoretically subpixel precision. Furthermore, in contrast to snapshot retinal imaging systems, 11,21,22 where all wavelengths are acquired simultaneously, meaning that the total area of the camera's sensor needs to be separated in the same amount of wavelengths imaged, a bigger field-of-view and/or higher spatial resolution can be achieved with the HRC, as the full sensor surface is used for an image. Last, instruments based on a discrete number of light-emitting diodes 23 and optical filters 12 offer less flexibility and considerably limit the accessible spectral information compared with a tunable laser source. 
In summary, this article has described a novel hyperspectral prototype for spectral imaging of the retina that can potentially be used in the future to acquire retinal vessel blood oxygen saturation values. By considering the limitations of ocular imaging encountered by other retinal oximetry studies, namely longer acquisition and exposure times, flash exposure, and limited wavelength intervals, this new instrument may be promising in acquiring more refined and faster measurements of nonflash exposure retinal oximetry measurements in vivo that can potentially be applied to human retinal vascular disease. 
Acknowledgments
The authors thank Mark Dunne and Richard Armstrong, Aston University, Birmingham, UK, for statistical advice. 
Disclosure: S.R. Patel, None; J.G. Flanagan, Photon etc. (F), Heidelberg Engineering (C), Carl Zeiss Meditec (C), Optometric Glaucoma Society (S); A.M. Shahidi, None; J.-P. Sylvestre, Photon etc. (E); C. Hudson, Photon etc. (F) 
References
Martin ME Wabuyele MB Chen K Development of an advanced hyperspectral imaging (HSI) system with applications for cancer detection. Ann Biomed Eng . 2006; 34: 1061–1068. [CrossRef] [PubMed]
Sorg BS Moeller BJ Donovan O Cao Y Dewhirst MW. Hyperspectral imaging of hemoglobin saturation in tumor microvasculature and tumor hypoxia development. J Biomed Opt . 2005; 10: 44004. [CrossRef] [PubMed]
Yudovsky D Nouvong A Schomacker K Pilon L. Assessing diabetic foot ulcer development risk with hyperspectral tissue oximetry. J Biomed Opt . 2011; 16: 026009. [CrossRef] [PubMed]
Holzer MS Best SL Jackson N Assessment of renal oxygenation during partial nephrectomy using hyperspectral imaging. J Urol . 2011; 186: 400–404. [CrossRef] [PubMed]
Mordant DJ Al-Abboud I Muyo G Spectral imaging of the retina. Eye (Lond) . 2011; 25: 309–320. [CrossRef] [PubMed]
Nourrit V Denniss J Muqit MM High-resolution hyperspectral imaging of the retina with a modified fundus camera. J Fr Ophtalmol . 2010; 33: 686–692. [CrossRef] [PubMed]
Hickam JB Frayser R Ross JC. A study of retinal venous blood oxygen saturation in human subjects by photographic means. Circulation . 1963; 27: 375–385. [CrossRef] [PubMed]
Beach JM Schwenzer KJ Srinivas S Kim D Tiedeman JS. Oximetry of retinal vessels by dual-wavelength imaging: calibration and influence of pigmentation. J Appl Physiol . 1999; 86: 748–758. [PubMed]
Hardarson SH. Retinal oximetry. Acta Ophthalmol . 2013; 91: 1–47. [CrossRef] [PubMed]
Khoobehi B Beach JM Kawano H. Hyperspectral imaging for measurement of oxygen saturation in the optic nerve head. Invest Ophthalmol Vis Sci . 2004; 45: 1464–1472. [CrossRef] [PubMed]
Kashani AH Kirkman E Martin G Humayun MS. Hyperspectral computed tomographic imaging spectroscopy of vascular oxygen gradients in the rabbit retina in vivo. PLoS One . 2011; 6: e24482. [CrossRef] [PubMed]
Calcagni A Gibson JM Styles IB Claridge E Orihuela-Espina F. Multispectral retinal image analysis: a novel non-invasive tool for retinal imaging. Eye (Lond) . 2011; 25: 1562–1569. [CrossRef] [PubMed]
Bedard N Hagen N Gao L Tkaczyk TS. Image mapping spectrometry: calibration and characterization. Opt Eng . 2012; 51: 111711. [CrossRef] [PubMed]
Vucea V Bernard PJ Sauvageau P Diaconu V. Blood oxygenation measurements by multichannel reflectometry on the venous and arterial structures of the retina. Appl Opt . 2011; 50: 5185–5191. [CrossRef] [PubMed]
Mordant DJ Al-Abboud I Muyo G Validation of human whole blood oximetry, using a hyperspectral fundus camera with a model eye. Invest Ophthalmol Vis Sci . 2011; 52: 2851–2859. [CrossRef] [PubMed]
Schweitzer D. Calibration-free measurement of the oxygen saturation in human retinal vessels. Proc SPIE . 1995; 2393: 210–218.
Lasta M Palkovits S Boltz A Reproducibility of retinal vessel oxygen saturation measurements in healthy young subjects. Acta Ophthalmol . 2012; 90: e616–e620. [CrossRef] [PubMed]
Hardarson SH Harris A Karlsson RA Automatic retinal oximetry. Invest Ophthalmol Vis Sci . 2006; 47: 5011–5016. [CrossRef] [PubMed]
Palsson O Geirsdottir A Hardarson SH Retinal oximetry images must be standardized: a methodological analysis. Invest Ophthalmol Vis Sci . 2012; 53: 1729–1733. [CrossRef] [PubMed]
Beach J Ning J Khoobehi B. Oxygen saturation in optic nerve head structures by hyperspectral image analysis. Curr Eye Res . 2007; 32: 161–170. [CrossRef] [PubMed]
Johnson WR Wilson DW Fink W Humayun M Bearman G. Snapshot hyperspectral imaging in ophthalmology. J Biomed Opt . 2007; 12: 014036. [CrossRef] [PubMed]
Ramella-Roman JC Mathews SA Kandimalla H Measurement of oxygen saturation in the retina with a spectroscopic sensitive multi aperture camera. Opt Express . 2008; 16: 6170–6182. [CrossRef] [PubMed]
Everdell NL Styles IB Calcagni A Multispectral imaging of the ocular fundus using light emitting diode illumination. Rev Sci Instrum . 2010; 81: 093706. [CrossRef] [PubMed]
Figure 1
 
Schematic representation of the HRC incorporating the CCD within the custom-built retinal camera and the TLS as the light source.
Figure 1
 
Schematic representation of the HRC incorporating the CCD within the custom-built retinal camera and the TLS as the light source.
Figure 3
 
Left: Fundus image at 570 nm (right eye [RE] −0.50 diopter cylindrical [DC]) taken before normalization. Right: Same image after normalization procedure.
Figure 3
 
Left: Fundus image at 570 nm (right eye [RE] −0.50 diopter cylindrical [DC]) taken before normalization. Right: Same image after normalization procedure.
Figure 4
 
Typical images taken by the HRC to create a hyperspectral cube of retinal images between 500 and 600 nm at 5-nm intervals with an exposure time of 80 ms.
Figure 4
 
Typical images taken by the HRC to create a hyperspectral cube of retinal images between 500 and 600 nm at 5-nm intervals with an exposure time of 80 ms.
Figure 5
 
The average arteriolar (left) and venular (right) OD values at each given (5-nm) imaged wavelength from 500 to 600 nm for all of the volunteers.
Figure 5
 
The average arteriolar (left) and venular (right) OD values at each given (5-nm) imaged wavelength from 500 to 600 nm for all of the volunteers.
Figure 6
 
Typical example from a volunteer showing OD values for arterioles (left) and venules (right) for each of the five acquired hyperspectral cubes.
Figure 6
 
Typical example from a volunteer showing OD values for arterioles (left) and venules (right) for each of the five acquired hyperspectral cubes.
Table 1
 
Average Baseline Characteristics of the Six Healthy Volunteers in This Study
Table 1
 
Average Baseline Characteristics of the Six Healthy Volunteers in This Study
IOP, mm Hg ± SD 13 ± 1
Systolic blood pressure, mm Hg ± SD 108 ± 5
Diastolic blood pressure, mm Hg ± SD 66 ± 4
Heart rate, mm Hg ± SD 81 ± 3
Pulse oximeter oxygen saturation, % ± SD 98.3 ± 0.8
Table 2
 
Results Showing Ranges of Values for ODs, SDs, Mean of Differences Between Sequential Measures, and CORs for Repeated Measures in Arterioles and Venules
Table 2
 
Results Showing Ranges of Values for ODs, SDs, Mean of Differences Between Sequential Measures, and CORs for Repeated Measures in Arterioles and Venules
Mean OD (Range) SD of Difference Between Means Mean of Differences COR
Arteriole 0.15 (0.06–0.23) 0.01–0.06 −0.08–0.09 0.02–0.11
Venule 0.25 (0.17–0.31) 0.01–0.07 −0.06–0.06 0.03–0.14
Table 3
 
Results Showing Limits of Agreement, ICC, Cronbach's α, and Correlation Values for Repeated Sequential Measures of Arterioles and Venules
Table 3
 
Results Showing Limits of Agreement, ICC, Cronbach's α, and Correlation Values for Repeated Sequential Measures of Arterioles and Venules
Limits of Agreement ICC Cronbach's α Correlation Between Sequential Measures, r 2
Arteriole −0.06–0.06 78.8%–94.4%* 89.6%–97.6%* 0.800–0.912*
Venule −0.04–0.07 63.7%–92.1%* 86.2%–98.1%* 0.671–0.945*
×
×

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.

×