Conventional perimetry and frequency-doubling technique

CONVENTIONAL PERIMETRY AND FREQUENCYDOUBLING TECHNIQUE

MICHELE IESTER,1,2 MICHELE ALTIERI,1 FRANCESCA NASCIUTI,1 FABIO DE FEO,1 PAOLO VITTONE,2 CARLO E. TRAVERSO1 and GIOVANNI CALABRIA1

1Department of Neurological Sciences, Ophthalmology, Genetic, Clinica Oculistica, University of Genoa; 2Division of Ophthalmology, G. Gaslini Institute; Genoa, Italy

Abstract

Purpose: To evaluate the correlation between the Humphrey Field Analyzer (HFA) and frequencydoubling technique (FDT). Methods: Seventy-six consecutive glaucoma patients were recruited for this study. A diagnosis of glaucoma was made in each study eye using the European Glaucoma Society terminology: visual field and/or an optic disc damage typical of glaucoma with an intraocular pressure (IOP) on no treatment higher than 21 mmHg, and no other reason for optic neuropathy. Visual fields were assessed by the Humphrey perimeter, program 30-2 and frequency-doubling perimetry (FDP), full threshold C-20. HFA mean defect (MD), corrected pattern standard deviation (CPSD), pattern standard deviation (PSD) and FDT MD and FDT PSD were calculated. Results: HFA MD was -5.4 ± 5.2 dB, while FDT MD was -4.7 ± 6.2 au. FDT MD, and PSD showed a significant correlation with HFA MD (r = 0.60, p < 0.001) and HFA PSD (r = 0.71, p < 0.001) or HFA CPSD (r = 0.60, p < 0.001). Conclusions: The significant quantitative correlation found was probably explained by the FDT capability to simultaneously challenge contrast, spatial frequency and temporal modulation, all of which are likely to be affected by glaucoma.

Introduction

The frequency doubling technique (FDT; Welch-Allyn, Skaneateles, NY, ZeissHumphrey, San Leandro, CA) is a new perimetric technique based on a phenomenon called frequency-doubling illusion occurring when a low spatial frequency (<1 cycle/ degree) grating undergoes high counterphase temporal frequency (>15 Hz) flicker. The grating appears to be twice its actual spatial frequency.1,2 FDT is based on this type of stimulus, which is selective for retinal ganglion cells with larger axons also identified as magnocellular (M cells). In particular, FDT is selective to the My cell subpopulation, representing 20% of M cells and just 3-5% of the total ganglion cell population.2,3 FDT should be able to detect very early visual field defects, even when standard threshold perimetry still appears normal, due to the relatively sparse representation of the M cell pathway.

Address for correspondence: Michele Iester, MD, University Eye Clinic, c/o Viale Teano 71/1, 16147 Genoa, Italy. Email: m_iester@hotmail.com

Perimetry Update 2002/2003, pp. 115–119

Proceedings of the XVth International Perimetric Society Meeting, Stratford-upon-Avon, England, June 26–29, 2002

edited by David B. Henson and Michael Wall

© 2004 Kugler Publications, The Hague, The Netherlands

The aim of this study was to compare FDT indices with those of standard threshold perimeters (STP) in glaucomatous patients, and to verify the capacity of FDT to quantify glaucoma damage in different stages.

Patients and methods

Seventy-six subjects were consecutively recruited from the glaucoma center of the Clinica Oculistica, University of Genoa. No patients were excluded on the basis of gender, age, or race. All patients had had at least three previous standard threshold visual field tests and their refractive errors had to fall within the range of from -7 to 7 diopters.

In the study, the visual fields were assessed by a Humphrey Field Analyzer II 750 (HFA), program 30-2, which tests the central 30 degrees of the visual field and by frequency-doubling technology or perimetry (FDT, program C20 full threshold), which tests the central 20 degrees of the visual field.

Patients were classified as having primary open-angle glaucoma or ocular hypertension using the European Glaucoma Society terminology.4

FDT presents stimuli on a black-and-white video monitor with specialized control circuitry interfaced to a microprocessor. An optical system is used to present the stimulus display at optical infinity, with an eyepiece adjustment provided to correct for spherical refraction errors of up to ±7 diopters. Using program C-20, 16 peripheral points and one central point were tested in the periphery up to 20° eccentricity. Stimuli consisted of a 0.25 cycle/degree sinusoidal grating square pattern (10 x 10°) undergoing 25 Hz (50 times/second) counterphase flicker (contrast reversal of light and dark bars). A round target (5° in diameter) was utilized to test the foveal threshold. The details of this technique have been described elsewhere.5,6

Subjects were instructed to fixate on a small central black spot on the video monitor, to refrain from looking at eccentric stimuli, and to depress a response button each time they detected a stimulus.1-3 The frequency-doubling phenomenon is described by the patient as a quick and vertical flickering of the bars.

Patient fixation is measured six times during the test, using blind spot checks according to the Heijl-Krakau method. Responses to this stimulus are interpreted as fixation losses. False answers are checked by periodically presenting a blank trial (six times) or a maximum (100%) contrast stimulus (three times) for false-positive or false-negative answers.

At the end of the FDT test, the results were printed with a gray scale map and a threshold value map, and mean deviation (FDT-MD) and pattern standard deviation (FDT-PSD) indices were calculated. In our study, an FDT test was considered reliable when there were < 2 false-positive answers and fixation losses and no false-negative answers.5

The Humphrey perimetry mean deviation (HFA MD), pattern standard deviation (HFA PSD), and corrected pattern standard deviation (HFA CPSD) were calculated and used for correlation. The results were analyzed by descriptive analysis, and Pearson’s r coefficient was used to compare and correlate perimetric results when the distribution of the data was normal; Spearman coefficient was used when the distribution of the data was annormal. The data were analyzed further using an alternative

technique described by Bland and Altman for assessing the level of agreement between Octopus standard threshold perimetry and FDT indices.7,8

In order to verify the capacity of FDT for quantifying glaucoma damage in different stages, a correlation between HFA MD and the values obtained from the difference between FDT MD and HFA MD was calculated. Furthermore, the cohort of patients was divided into two subgroups based on HFA mean deviation rank distribution, and two subgroups of 38 patients resulted: the early glaucoma group had an HFA MD > -6 dB, while the moderate and advanced glaucoma group had an HFA MD < -6 dB. In each group, the difference between FDT MD and HFA MD was calculated and then, using the unpaired t test, the values of the difference (‘FDT MD – HFA MD’) were compared between the two subgroups. A p value < 0.05 was considered to be statistically significant.

Results

The mean age of the entire group was 62 ± 8.5 years, and the mean refractive error -0.5 ± 0.3 diopters (Table 1).

In the sample as a whole, a statistically significant (p < 0.001) correlation was

Table 1. Descriptive analysis

n = number of eyes. FDT MD = Frequency-doubling technique mean deviation; FDT PSD = Frequencydoubling technique pattern standard deviation; HFA MD = Humphrey field analyzer mean deviation; HFA PSD = Humphrey field analyzer pattern standard deviation; HFA CPSD = Humphrey field analyzer corrected pattern standard deviation. au = arbitrary unit.

found between FDT MD and HFA MD (0.83), between FDT PSD and HFA PSD (0.82), and between FDT PSD and HFA CPSD (0.73). The agreement between the two techniques was calculated using the Bland and Altman7 method. The mean of difference between FDT MD and HFA MD was 1.29 and the limits of agreement were between -5.8 and 8.4. The mean of the difference between FDT PSD and HFA MD was 1.09 and the limits of agreement -2.88 and 5.06, and the mean of the difference between FDT PSD and CPSD was 1.59 and the limits of agreement -2.68 and 5.87. A significant correlation was found between HFA MD and the difference between FDT MD and HFA MD (‘FDT MD’ – ‘HFA MD’) (r = 0.7, p < 0.001). When the

Fig. 1. Scattergram between HFA MD (x axis) and the difference between FDT MD and HFA MD (y axis). On the left side of the graph, most of the points are above the x axis, suggesting that HFA MD is greater than FDT MD. On the right side of the graph, most of the points are below the x axis, suggesting that FDT MD is greater than HFA MD.

entire group was divided into two subgroups based on the HFA MD, the difference between FDT MD and HFA MD was found to vary significantly between the two subgroups. In the early glaucoma group, the difference was –1.34 ± 0.29, which means that FDT MD was greater than HFA MD, while in the moderate and advanced glaucoma groups, the difference was 3.92 ± 0.46, showing that HFA MD was conversely greater than FDT MD values (Fig. 1).

Discussion

Previous studies have shown that FDT has a high sensitivity to differentiate between normal subjects and glaucoma patients, and is able to quantify glaucoma damage.1,2,9 Sponsel et al. showed a good correlation between FDT and Humphrey 30-2 threshold perimetry results in 42 glaucoma patients and 14 normal subjects.10 Iester et al. showed good correlation between FDT and the Octopus perimeter, program G1, in 39 glaucomatous patients, with good agreement using Bland and Altman’s method.5 Furthermore, they found that in patients with severe visual field damage, FDT measured less damage than the Octopus, suggesting that FDT may perform better as a screening test or in patients with early glaucoma rather than in monitoring patients with advanced glaucoma.5 In contrast, however, Kondo et al. showed no correlation between the threshold values of FDT and Humphrey perimetry in 11 normal-tension glaucoma patients.9

Our results confirm the good correlation between the results of the two techniques. Furthermore, it is quite easy to find a correlation when the same parameter is measured by two different techniques, even if the two perimetric techniques test two different visual functions and different widths of the visual field. The Humphrey program 30-2 tests the 30 central degrees of the visual field and FDT, C-20, tests the 20 central degrees with stimuli of different shape, surface, density, pattern distribution, time exposure, and different structural characteristics. The difference in the width

of the tested visual field does not seem to be influenced, indeed correlation between FDT and 30-degree Octopus visual field mean sensitivity was found, similar to the correlation between FDT and 20-degree Octopus visual field mean sensitivity (Iester et al., unpublished data).

Our data showed that FDT MD values are greater than HFA MD values when the early glaucoma group was considered, and that FDT MD values are less than HFA MD values in the moderate and advanced glaucoma groups (Fig. 1). This finding could be due to FDT being more sensitive in early or moderate glaucoma. Indeed, this technique is able to detect the 3-5% of ganglion cell loss and, in particular, the loss of My cells which are sparsely represented. Furthermore these results could also be due either to a difference between the HFA MD and FDT MD scales, or to a more useful dynamic range of FDT for early defects.

In conclusion, due to the characteristics of the FDT scale, this new technique may best be used both to screen populations1,2 and to follow glaucomatous visual field progression in the early and moderate stages of the disease. At the moment, it is important to remember that this is a new method, some of the limitations of which may not yet have been identified.

However, it is necessary to remember that an FDT MD value of -5 dB could correspond to a better HFA MD value, while an FDT MD value of -15 dB could correspond to a worse HFA MD.

Acknowledgment

None of the authors has a proprietary interest in the development and marketing of any products mentioned in the article.

References

1.Johnson CA, Samuels SJ: Screening for glaucomatous visual field loss with frequency-doubling perimetry. Invest Ophthalmol Vis Sci 38:413-425, 1997

2.Quigley HA: Identification of glaucoma-related visual field abnormality with the screening protocol of frequency doubling technology. Am J Ophthalmol 125:819-829, 1998

3.Johnson CA, Demirel S: The role of spatial and temporal factors in frequency-doubling perimetry. In: Wall M, Heijl A (eds) Perimetry Update 1996/97, Proceedings of the 12th International Perimetric Society Meeting, pp 13-19. Amsterdam/New York: Kugler Publ 1997

4.European Glaucoma Society: 1998 Terminology and Guidelines for Glaucoma, ch 2, pp 64-65. Savona: Dogma 1998

5.Iester M, Mermoud A, Schnyder C: Frequency doubling technique in subjects with ocular hypertension and glaucoma: correlation with Octopus perimeter indices. Ophthalmology 107:288-294, 2000

6.Iester M, Capris P, Pandolfo A, Zingirian M, Traverso CE: Learning effect, short-term fluctuation and long-term fluctuation in frequency doubling technique. Am J Ophthalmol 130:160-164, 2000

7.Bland JM, Altman DG: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1:307-310, 1986

8.Flanagan JG, Wild JM, Trope GE: The visual field indices in primary open angle glaucoma. Invest Ophthalmol Vis Sci 34:2266-2274, 1993

9.Kondo Y, Yamamoto T, Sato Y et al : A frequency-doubling perimetric study in normal tension glaucoma with hemifield defect. J Glaucoma 7:261-265, 1998

10.Sponsel WE, Arango S, Trigo Y, Mensah J: Clinical classification of glaucomatous visual field loss by frequency doubling perimetry. Am J Ophthalmol 125:830-836, 1998