Spectroscopic measurements in scleritis: Bluish-red ...

Downloaded from bjo.bmj.com on May 19, 2014 - Published by group.bmj.com

BJO Online First, published on May 2, 2014 as 10.1136/bjophthalmol-2013-304086 Innovations

Spectroscopic measurements in scleritis: Bluish-red or deep red? N P Bannister,1 M J Wakefield,2 A Tatham,3 S L Bugby,1 P M Molyneux,1 J I Prydal2 1

Department of Physics and Astronomy, University of Leicester, Leicester, UK 2 Department of Ophthalmology, Leicester Royal Infirmary, Leicester, UK 3 Princess Alexandra Eye Pavilion, Edinburgh, UK Correspondence to Dr Nigel P Bannister, Department of Physics and Astronomy, University of Leicester, University Road, Leicester LE1 7RH, UK; [email protected] Received 8 August 2013 Revised 12 February 2014 Accepted 29 March 2014

ABSTRACT Purpose To design a slit-lamp mountable spectrometer for the assessment of ophthalmic patients and to illustrate a potential clinical application by measuring the spectral characteristics of inflamed eyes of differing aetiologies. Methods A slit lamp mountable instrument was designed and built, and methods for data analysis developed. Reflectance spectra were recorded from two patients with scleritis, three with non-scleritic red eyes and from two controls with non-inflamed eyes. Results Measurements were reproducible and demonstrated statistically significant differences in the spectral characteristics between the three groups. Spectra from scleritic eyes revealed a relative increase in intensity of long wavelength red light, 650–740 nm, compared with non-scleritic red eyes. These longer wavelengths will be appreciated as dark red. There was no increase in relative intensity in the blue part of the spectrum in scleritic eyes. Conclusions Reproducible measurements of the eye were made using an innovative, slit-lamp mountable spectrometer and its functionality demonstrated by differentiating the spectra from eyes with differing pathologies. While intending only to illustrate one potential application; for the cases examined, our results indicate that inflamed scleritic eyes exhibit a longer wavelength red light with no increase in shorter wavelength blue light. Thus our measurements would seem to confirm that the perceived redness of scleritis differs from other red eyes. However, it is a deeper darker red and not a bluish one as traditionally described.

INTRODUCTION

To cite: Bannister NP, Wakefield MJ, Tatham A, et al. Br J Ophthalmol Published Online First: [please include Day Month Year] doi:10.1136/ bjophthalmol-2013-304086

It has long been observed that the vascular engorgement in scleritis differs from that of conjunctivitis and episcleritis. Duke Elder in 1964 commented that the ‘eye is deeply injected with dark violet vessels’1 and a recent comprehensive review of scleritis refers to the distinctive blueviolet hue.2 These subtle colour differences can be difficult to detect, and are more apparent in natural daylight.2 Spectroscopic methods have been used to examine the human eye,3–5 but there are no studies to date that aim to use spectroscopy to differentiate and describe individual disease processes. The aim of the presented research was to design a slit-lamp mountable spectrometer for the assessment of ophthalmic patients, and to further determine if the subjectively appreciated blue hue of an inflamed eye in scleritis can be confirmed and measured objectively by spectroscopic methods.

Bannister NP, et al. Article Br J Ophthalmol 2014;0:1–4. doi:10.1136/bjophthalmol-2013-304086 Copyright author (or their employer) 2014. Produced

METHOD In order to perform this study a slit lampmountable instrument was built and is described in figures 1 and 2. The converging lens constrains the illuminated area of the subject’s eye and minimises ambient light in the reflected signal. The spectrometer was driven and measurements read out using Ocean Optics’ OOIbase32 software. Laboratory wavelength calibration was achieved using HeNe gas laser sources and features in the solar spectrum, and was stable over time. The effects of changes in ambient lighting (eg, turning room lights and monitor screens on/off) were assessed. Readings were only taken in conditions of less than 0.5Lux ambient. The use of a condensing lens also reduces noise from ambient lighting, and postprocessing reduced any stray light signals to negligible levels using calibration exposures applied after the clinical measurements were taken. Patients with active scleritis were recruited from appropriate clinics and non-scleritic red eyes from the Eye Casualty at Leicester Royal Infirmary between August 2011 and August 2012. Reflectance spectra from an affected region of the eye were acquired in consulting rooms, with the lights off and available blinds closed leaving a low but variable level of ambient light. Several measurements were taken from the inflamed area to allow for small variations in eye position. Measurements of white eyes were taken from all four quadrants of the anterior segment. Data were normalised to 580 nm (intensity at 580 nm=1.0). This was a prominent spectral feature in all plots, but otherwise an arbitrarily selected point. The spectra were then divided by the spectrum from a reflectance standard to correct for instrument response characteristics. This study was approved by the Leicestershire, Northamptonshire and Rutland Research Ethics Committee (2) (Unique Ref 09/H0402/1) and supported by the University Hospitals Leicester Research and Development Department.

RESULTS Two patients with unilateral scleritis were recruited during the study period. Two white eyes and three non-scleritic red eyes were also recruited for comparison. Characteristics of the scleritic and nonscleritic red eyes are summarised in table 1. Clinical photos of the inflamed scleritic eyes are presented in figure 3. The results of the analysis are presented in figure 4. Figure 4 indicates scleritis can be distinguished from control eyes and non-scleritic red eyes by the height and width of the peak at 580–800 nm. In scleritis there is increased signal from 580 nm to

by BMJ Publishing Group Ltd under licence.

1


Innovations

Figure 1 An Ocean Optics LS-1 tungsten halogen lamp generating light in the Visible—Near Infrared (360–2500 nm) feeds a fibre waveguide consisting of six illuminators (I1–I6) and one central reflected signal fibre (R1) terminating in a reflectance probe. A converging lens of focal length 12 mm was mounted 20 mm in front of the reflectance probe and the components held in their required positions using a custom-designed clamp, which could be mounted on the tonometer table of a Haag-Streit slit lamp. The reflected signal is carried in R1 to an Ocean Optics S2000 spectrograph, covering the 200–1100 nm (UV—Visible—Near Infrared) band. The spectrograph is controlled by a computer (PC).

630 nm (yellow to short wavelength red), but more significantly the increase signal continues to longer wavelengths, up to ∼740 nm. A difference between scleritis and control spectra over 500– 580 nm (green) is also seen in the non-scleritic red eyes. Figure 4 also shows increased signal at short wavelengths (below 460 nm, blue to purple) for non-scleritic red eyes compared with scleritic eyes and controls. When comparing conjunctivitis and control spectra, there are statistically significant, though much smaller, differences particularly in the near-infrared beyond 720 nm (non-visible).

Control eye spectral characteristics remained consistent regardless of where the reading was taken from on the sclera. Furthermore, the individual conjunctivitis traces could not be distinguished from one another based on aetiology. Figure 5 shows spectra for one patient with scleritis at presentation and a subsequent visit after treatment had commenced. After treatment, the spectra revert toward a shape similar to that from the control eye. Figure 5 also demonstrates the statistical significance of the measurements in general, with error bars plotted at 15 nm intervals showing significant separation between the traces in the wavelength range of interest.

DISCUSSION

Figure 2 Slit lamp mounted instrument. 2

The data indicate a clear separation in spectral characteristics between inflamed scleritic eyes, non-scleritic red eyes and white eyes. The most striking difference between scleritis and nonscleritic red eyes is the increased relative intensity of the longer wavelength red light (650–740 nm). These longer wavelengths will be appreciated as a darker red. There is no relative increase in the intensity of the reflectance of light in the blue part of the visible spectrum for scleritic eyes. Thus scleritic eyes appear a deep red, not a bluish red as previously described. Two types of scleritis were included in this study. The results may not be generalisable to other forms of scleritis and eyes with recurrent or chronic scleral inflammation. Spectral characteristics of non-scleritic red eyes and white eyes are not as clearly distinguished, though there is evidence of significant separation in the non-visible light spectrum. This indicates that the increase in perceived redness in conjunctivitis Bannister NP, et al. Br J Ophthalmol 2014;0:1–4. doi:10.1136/bjophthalmol-2013-304086


Innovations Table 1 A summary to the characteristics of the scleritic and non-scleritic red eyes included in the study Study eye

Diagnosis

Systemic associations

Relevant investigations

Treatment

Scleritis 1

Nodular scleritis

Wegener’s granulomatosis

PRC-ANCA positive, lung biopsy positive

Scleritis 2

Diffuse scleritis

Rheumatoid arthritis

Rheumatoid factor positive

Non-scleritic 1 Non-scleritic 2

Viral Conjunctivitis Acute bacterial conjunctivitis Acute allergic conjunctivitis

None Contact lens wearer

Adenovirus PCR positive Staphylococcus epidermidis culture positive Culture negative

Systemic prednisolone, mycophenolate mofetil, cyclophosphamide, infliximab Systemic prednisolone, mycophenolate mofetil, cyclophosphamide, infliximab Topical lubricants Topical antibiotic

Non-scleritic 3

Atopic patient previous vernal keratoconjunctivitis

Topical steroids

PRC-ANCA, proteinase-3 anti-neutrophil cytoplasmic antibody.

is due to an increase in density and/or calibre of blood vessels, but not a change in spectral characteristics of the bulk tissue. This is in contrast to scleritic eyes, where it is possible that inflammation of the sclera changes its transparency with resultant contribution to spectra from deeper structures. Prevalence of scleritis is estimated to be 6 cases per 100 000 people.6 This may account for the small number of patients in our study despite yearlong recruitment. However, it was not our aim to collect statistically significant data, but specifically to demonstrate that spectroscopy can be used to identify the nature of the difference in perceived colour between scleritic eyes and other red eyes. Figure 5 demonstrates that spectroscopy may also have a role in disease monitoring and in the diagnosis of other conditions. Clinicians apply specialist knowledge together with visual inspection and a wide range of instrumentation in their diagnosis of conditions. While spectroscopy is not a substitute for this experience and breadth of approaches, it can be a powerful aid. Where the human eye perceives colour and shading, the spectrometer can quantify differences in intensity over narrow ranges of wavelength that are too small to be detected by the eye. This ‘colour decomposition and measurement’ approach may have important roles in diagnosing and assessing a variety of conditions, ophthalmic and otherwise and is the subject of our ongoing work.

Critical to its effective use is reliable calibration. Our use of wavelength standards, such as gas lasers and features in the solar spectrum, shows that the wavelength calibration of our instrument is robust and stable over time. Also important is flux calibration—ensuring that sensitivity as a function of wavelength is accounted for, and that factors such as changes in illumination angle and ambient lighting conditions do not affect the usefulness of the data. Reference measurements are currently performed manually, but can be automated using an additional, parallel channel in the waveguide, so that they are taken simultaneously with the clinical measurement and applied before the data are viewed by the clinician. This will enable a more userfriendly control, capture and analysis using bespoke software and a ‘one shot capture’ approach that performs the analysis described in this paper automatically, within seconds of exposure, and allows the ophthalmologist to concentrate on the patient, not the instrumentation. Implementation of this approach will be the subject of further work which is expected to include hardware revision followed by further clinical trials for validation purposes. In summary, we have presented an innovative, slit-lamp mountable spectrometer that can be used to measure the spectral characteristics of healthy and diseased eyes in a clinical setting. Our results confirm the accepted belief that the redness of inflamed scleritic eyes is of a different colour. However, this

Figure 3 Clinical photos of patients with scleritis (A) Scleritis 1, (B) Scleritis 2 (as described in table 1).

Figure 4 Comparison of different measurement groups: (A) scleritis (red) and conjunctivitis (blue); (B) scleritis (red) and controls (green); (C) conjunctivitis (blue) and controls (green). Different shades of each individual colour represent different patients’ eyes. The colour bar under the wavelength axes indicates the colour of light at that wavelength. Error bars omitted for clarity. Bannister NP, et al. Br J Ophthalmol 2014;0:1–4. doi:10.1136/bjophthalmol-2013-304086

3


Innovations authors have contributed to editing and approving the final version. JIP, NPB and AT were instrumental in the conception, ethical approval and funding applications for the project. MJW, SLB and PMM were responsible for the instrumentation, patient recruitment, data collection, troubleshooting, data analysis and data presentation. Funding This work was supported by an EMDA/ERDF Innovation Fellowship grant number HIRF-493 and an STFC Mini IPS grant number ST/K003054/1. Competing interests None. Patient consent Obtained. Ethics approval Leicestershire, Northamptonshire & Rutland Research Ethics Committee (2) (Unique Ref 09/H0402/1). Provenance and peer review Not commissioned; externally peer reviewed. Data sharing statement Raw spectral data is available from the corresponding author.

REFERENCES

Figure 5 Comparison of a scleritic eye at diagnosis (red), after a period of treatment (blue) and the non-scleritic control eye (green). Colour bar indicates the appearance of light of a given wavelength to the human eye. Error bars are based on Poissonian statistics, where the uncertainty in a measurement of N events is N1/2. Plots of the same colour indicate repeat measurements in the same eye.

1 2 3

4

is not a blue-violet hue, as previously described, but a deeper red due to the contribution of longer wavelength visible light.

5

Contributors All listed authors have been involved in the development of the published article, and have either contributed to the text, figures or both. All listed

6

4

Duke-Elder S. Diseases of the outer eye, part 2: cornea and sclera. In: Duke Elder S, Leigh AG, eds. System of ophthalmology, Vol. VIII. London: Kimpton, 1964, p. 1015. Okhravi N, Odufuwa B, McCluskey, et al. Scleritis. Surv Ophthalmol 2005;50:351–63. Sorbara L, Simpson T, Duench S, et al. Comparison of an objective method of measuring bulbar redness to the use of traditional grading scales. Cont Lens Anterior Eye 2007;30:53–9. Duench S, Simpson T, Jones LW, et al. Assessment of variation in bulbar conjunctival redness, temperature and blood flow. Optom Vis Sci 2007;84:511–16. Jay GD, Racht J, McMurdy J, et al. Pont-of-care noninvasive haemoglobin determination using fiber optic reflectance spectroscopy. Conf Proc IEEE Eng Med Biol Soc 2007;2007:2932–5. Galor A, Thorne JE. Scleritis and Peripheral Ulcerative Keratitis. Rheum Dis Clin North Am 2007;33:835–54.

Bannister NP, et al. Br J Ophthalmol 2014;0:1–4. doi:10.1136/bjophthalmol-2013-304086


Spectroscopic measurements in scleritis: Bluish-red or deep red? N P Bannister, M J Wakefield, A Tatham, et al. Br J Ophthalmol published online May 2, 2014

doi: 10.1136/bjophthalmol-2013-304086

Updated information and services can be found at: http://bjo.bmj.com/content/early/2014/05/02/bjophthalmol-2013-304086.full.html

These include:

References

This article cites 5 articles http://bjo.bmj.com/content/early/2014/05/02/bjophthalmol-2013-304086.full.html#ref-list-1

P