ACTA OPHTHALMOLOGICA VOL. 57 1979 XXlV MEETING OF NORDIC OPHTHALMOLOGISTS
Oslo, Norway 13-16 June 1979
CONTRAST SENSITIVITY IN MACULAR DISEASE USING A SMALL-FIELD AND A LARGE-FIELD TV-SYSTEM
JOHAN SJOSTRAND
T h e spatial contrast threshold for a sinusoidal grating of varying contrast and frequency generated with a small-field or a large-field TV-display was examined in normal subjects and in patients with macular disorders. Impairment of contrast sensitivity (reciprocal of contrast) was observed in the different maculopathies investigated. Attenuation of the high- and middlefrequency ranges was an early finding in macular disease, whereas changes including the low-frequency range was observed in more advanced maculopathies. Comparison of results obtained using the small (1.4") or large-field (6"-24") TV-system demonstrated a field-dependence of the contrast sensitivity attenuation in localized macular disorders. In more wide-spread lesions in the posterior pole, a contrast attenuation over the whole frequency range was found also with the largest (24")stimulation field used. T h e study of the contrast sensitivity function supplements the traditional acuity measurements in quantifying the visual loss for objects larger than the resolution limit. I conclude that contrast threshold measurements are not only useful for describing visual loss, but also for tracking progression or recovery. The findings provide some additional insight into the visual difficulties of daily life of patients with a macular disorder. However, the definite role of contrast sensitivity measurements for diagnosing different visual disorders is still lacking.
Key words: contrast sensitivity - maculopathy acuity.
-
spatial frequency - visual
T h e commonly used acuity tests give information about the spatial resolution of fine targets at high contrast. T h e ability to see larger targets and details at low contrast is of equal importance and is usually not tested in clinical investigations. 832
Con trcrst S~n.titi71ity
From clinical practice we know, however, that patients sometimes complain of deterioration of low-contrast vision, even though their visual acuity is within normal limits. A more general description of the visual system is obtained by studying the contrast sensitivity function, i. e. the reciprocal of threshold contrast for a range of sinusoidal gratings that vary in spatial frequency (for review see Campbell 1974; Sekular 1974). By the contrast sensitivity it is possible to study visual functions not tested in the acuity test. The aim of the study has been to investigate the contrast sensitivity function in various maculopathies, and the interest has been focused on the following question: Is contrast sensitivity measurement a useful additional clinical test in the diagnosis, treatment and follow-up of macular disorders?
Material and Methods Patients
Fifteen normal controls (age 19-63 years) and 22 patients (age 19-78 years) with rnacular disease of different types took part in this study. Smal I-field TV-system Vertical, sinusoidal gratings were generated on a television display as previously described by S,jiistrand & Fris6n (1977). T h e television monitor was masked to subtend 1.4 X 1.4 degrees of angle at the eve, when viewed at a distance of 5 m. A white, illuminated mask, which subtended approximately 2 x 2 degrees, surrounded the TV-screen. Spatial frequency (number of cycles per degree of visual angle) was varied between 1.5 and 38 cyclesidegree. Pattern contrast (defined as (IAm.,x + L,&, where L,,,,,, and I,m,,, are the maximum and minimum luminances of the sinusoidal pattern) was varied in 2 dB steps in the 0.6.5-0.0007 range. T h e space average luminance of the gratings was 13.5 cd/m2. T h e contrast sensitivity (reciprocal of contrast threshold) was determined monocularly with optimal correction and natural pupil. Seventeen frequencies were explored twice in a random manner and the contrast threshold determined by raising the contrast from a subthreshold level at a selected spatial frequency until a pattern wasjust visible by the patient. Large-f ield TV-system T h e 20” TV-monitor used in the large-field system was masked to subtend 6 X 6 degree of angle at the eye when viewed at a distance of 2.5 m. T h e TV-monitor was surrounded by a grey screen and by illumination of two daylight tubes, a uniform background luminance of 13 cd/m* was obtained. T h e space average luminance of the grating was 83 cd/m2.The spatial frequency was varied between 0.5 and 30 cyclesidegree. T h e pattern contrast of the sinusoidal grating was varied in 2 d B steps in the 0.6-0.001 range. By changing the distance to the monitor the size of the TV-screen was increased to 24-24 degrees of angle at a distance of 0.6 m. T h e size of the retinal images, using the field size of 1.4, 6, and 24”, respectively, is illustrated in Fig. 1. T h e test distance was controlled by using a chin support and refraction was optimally corrected for the distances used. T h e test procedure was as described above for the small-field system.
833 lcra ophthal. 57. 5
Fig. 1. A fundus picture of a normal control with a schematic drawing of the retinal images using a TV-screen which subtended 1.4 (0), 6, and 24 degrees of arc.
Results Normal subjects
Srnal1;field TV-system. The contrast sensitivity curves (CSC), i. e. the contrast sensitivity plotted as a function of the spatial frequency of the sinusoidal grating, had a bell-shaped form when tested for subjects with normal vision, no previous eye disease, and ophthalmoscopically healthy eyes (Sjostrand & FrisPn 1977) in agreement with previous studies (Schade 1956; Campbell & Robinson 1968). The curves showed a high and low frequency attenuation with a peak sensitivity between 8 and 15 cyclesidegree (Figs. 2-4). Our CSC curves in normal controls using the small-field TV-system ( 1.4") differ from those obtained using the large-field TV-system (6-24") and from those previously reported in that our subjects' peak sensitivity occurred at a higher spatial frequency. Large-field Tv-system. The contrast sensitivity curves of 10 normal controls obtained using the large-field TV-system had a much broader peak with a peak region varying from 1 to 8, 1-10 and 2-12 cyclesidegree for a field size of 6, 12, and 24", respectively (Figs. 7 and 8). A comparison of the peak frequency obtained for 6 normal controls measured on both the small-field (1.4") and the large-field TV-system masked down to 1.4" (see legend Fig. 7) showed a mean difference of 6.5 cyclesidegree for the two systems used (cf. Figs. 2 and 7). The difference in peak frequency therefore presumably reflects the mask and the surroundings and their luminances (cf. Estkvez 8c Cavonius 1976). 834
1000
400
100
40
c > >
t
m Y
10
m
+ C
0 4
4
1
SPATIAL
10
40
FREQUENCY (cycl. I d e d
Fig. 2. CSC (1.4") in normal controls (dotted area indicates range for ten normal subjects aged 19 and 61 years) and three representative cases of different stages of senile maculur drgrwmtion (solid lines). Case I. A. (A,female, 65-years-old) had moderate macular changes and V.A. 0.7; in case K . L . ( 0 ,male, 46-years-old, V . A . 0.4) changes were more advanced; and in case L. L. (m, male, 65-years-old, V. A. 0.2) the changes were extensive. Data of normal controls from Sjostrand & Frisen (1977).
Macular disorders - small-field TV-system
T h e CSC was measured in different types of macular disease: senile macular degeneration, central serous retinopathy, fundus flavimaculatus, vitelliform macular degeneration, and retinal detachment including the macula. The macula is morphologically and functionally non-uniform and since macular disorders may involve the foveal, perifoveal and perimacular region or the receptor or neuronal cells
835
5
1
SPATIAL
FREQUENCY
10
I
50
(cycles I deg )
F g . 3. CSC (1.4")in central sprous retinopathy. Case S. H. (male 42-years-old) with involvement of the right eye (A,lower curve was obtained during the acute phase, V. A. 0.6, and the upper curve following recovery, V . A . 1.5). The CSC of the contralateral left eye (V.A. 1.5) is indicated by filled circles (0).Dotted area shows the range of seven normal controls aged 37 to 6 3 years.
differently, variation in effects on the contrast sensitivity function may be anticipated and is, in fact, demonstrable in this study. Sen& macular degeneration. In senile macular degeneration, pathological changes occur within the macular area (Sarks 1976) giving rise to scotomata of varying size in central vision. The CSC was impaired in senile macular degeneration, and the decrease was notable at high and intermediate frequencies in cases with reductions in central visual acuity to 0.5-0.7 (Fig. 2, case I.A., V.A. 0.7). In advanced cases of macular degeneration abnormality of CSC across the whole spatial frequency range was observed, as illustrated by case L. L. (V. A. 0.2, in Fig. 2).
836
1OC
4c
10
4
1
*
t 2
-
m
fn w
6 C
5 8
1
10
4
SPATIAL
FREQUENCY
40
(cycles I deg )
Fig. 4 . Case B.N., female, 24-years-old. CSC (1.4")of the left eye (0)of this patient with fundus flnuimuculutuc with macular involvement (Stargardt's disease). The upper curve was obtained in 1976 ( V . A . 0.6) and the lower curve after a period of progressive deterioration of central vision 2 % years later (V.A. 0.1). The dotted area represents the range of five normal subjects aged 19 to 3 1 years (from Frisen & Sjostrand 1978).
In 2 patients with macular oedema at the time of the initial test, a subjective improvement and also an improved CSC was demonstrated at a second determination 2-6 months following the initial test. A rough correlation was seen between the contrast sensitivity level between 5 and 10 cycleddegree and the visual acuity in patients selected to be in a stationary phase of disease with no obvious retinal oedema (results not shown). Central serous retinopathy. In this macular disorder macular oedema develops due to leakage through the pigment epithelium. In most cases tested, a detachment of the
837
pigment epithelium in the macular region (Gass 1967) was observed. During the acute phase the patients complained about impaired discrimination of fine details, metamorphopsies and micropsia (Frisen & Frisen 1979). CSC was followed during the course of the disease and during the acute phase contrast sensitivity was decreased for high and intermediate frequencies. Generally, the CSC improved parallel to the increase of visual acuity. Fig. 3 demonstrates the changes in CSC in one case during acute and recovery phase. It is notable that during resolution when the visual acuity had improved to 1.5, the patient still complained about slight blurring of the vision and the CSC was still markedly abnormal in the high- and mid-frequency region (Fig. 3).
\ 1
10
4
SPATIAL
FREQUENCY
40 (cycles
#
deg
1
Fig. 5 .
Case S.L., male 4.i-years-old, CSC (1.4") of the left eye ( 0 )of this patient with r'itrllzform rntlcular drgenrration with slowly progressive deficit in central vision. The upper curve was obtained in 1976 (V. A. 0.7) and the lower curve (V.A. 0.6) 2 % years later. Dotted area shows the range of seven normal controls aged 37 to 63 years.
838
J
01
04
10
GRATING
40
ACUITY
Fig. 6. T h e relation between the mean contrast sensitivity in the mid-frequency range (7.55-12 cyclesideg.) obtained with the small-field TV-system (1.4") and grating acuity. Grating acuity was measured monocularly by means of sinusoidal gratings displayed o n the TV-screen at a distance of 7.9 m (for details, see Andersson & Sjestrand, 1979). Fifteen cases with healed retinal detachment, involving the macula, were examined. t,normal controls 19 to 42-years-old; 0 ,normal controls, 43 to 70-years-old; 0 , macular detachments, aged 19 to 41 years: 0 , macular detachments aged 42 to 70 years; A,fellow eyes for patients aged 19 to 41 years; W, fellow eyes for patients aged 42 to 70 years.
Miscellaneous maculopathies. In this group various progressive maculopathies such as fundus flavimaculatus (Stargardts disease) and vitelliform macular degeneration were investigated. During the time period of this study (3 years) the visual acuity dropped in several of the patients studied. Fig 4 demonstrates the CSC-curve of a 24-year-old female with a fundus flavimaculatus involving the macular region. 839
1000
1000
lo( I
5
t rn
z rn W
rn I-
2
1'
z
8
i
20 SPATIAL FREQUENCY (cyclesldeg.)
1000-
1000
B.N.
100
100
10
10
>
k E t rn W
rn
$ a
+ 0 0
2.5
25
1.25
12.5
SPATIAL FREQUENCY (cycles/deg.)
Fig. 7 Case A. N . with f i c n d ~flnuimocu1atzL.s (the same patient as described in Fig. 4). CSC of the left eye (solid line, V. A. 0.1) of this patient using the large-field TV-system (6-24"). The dotted area shows the range for 5 (So)and 6 (12 and 24") normal controls. To investigate the CSC for an angle of 1.4" at the eye (the same angle as that used with the small-field TV-system) with the TV-system ordinarily used for large-field, the screen was masked to 1.4" with a grey cardboard mask with the same background luminance as the surroundings. Dotted area for the 1.4" field indicates the range for six normal controls. a) 1.4 and 6";b) 12 and 24".
840
1000
1000
S.L
100
--i>? 100
t W z Lo k
2
+
10
10
z 0
2.5
125
25
12.5
SPATIAL FRQUENCY (cycles/deg.)
F z-q 8. Case S. L. with 7~itelliformmacular degeneration (the same patient as described in Fig. 5). CSC of the left eye (solid line V . A . 0.6) with the large-field TV-system (12 and 24"). The CSC for 6" was even more severely depressed (results not shown). Dotted area shows the range for six normal subjects.
Already 3 years ago when the visual acuity was G.6, the CSC curve was impaired, and, when studied half a year ago, both the visual acuity (0.1) and the CSC had dropped further (Fig. 4). Another patient with vitelliform degeneration of the macula, studied during the same time period, is shown as comparison. During the time period of study visual acuity only decreased from 0.7 to 0.6, whereas the CSC depression was markedly augmented over the whole frequency range (Fig. 5). At the same time, the patient experienced increasingly severe metamorphopsies, and the macular status deteriorated as visualized by the ophthalmoscope.
Retinal detachment including the macula. We have investigated a number of patients with rhegmatogenous retinal detachment including the macula, and followed the retinal function with various tests postoperatively (Anderson & Sjostrand 1979). It is known that age and refraction of the patient as well as duration of macular detachment preoperatively are of great importance for the functional outcome (Gundry 8~Davies 1974; Kreissig 1977). Following a successful operation of macular detachment, an initial rapid increase in visual acuity is found, followed by a second phase of prolonged visual improvement extending over more than a year (Kreissig 1977).The spatial processing in the 84 1
/ o h m Sjiistmnd
reattached macula was followed by measuring the contrast sensitivity function and grating acuity during the recovery phase. In a group of patients examined about a year after the operation, an abnormal CSC could be observed in the two-age groups studied (Anderson & Sjiistrand 1979). T h e high and intermediate frequency ranges were affected predominantly. Within the low frequency range the contrast sensitivity essentially was within the normal range. The reduction of the contrast sensitivity in the mid-frequency range was roughly correlated to the visual acuity and grating acuity in these patients (Fig. 6). Macular disorders - large-field TV-system
T o obtain a better insight into the field-related visual problems of patiens with visual disorders, a few of the patients measured with the small-field TV-system were selected for further study, using a large-field TV-system with a visual angle of 6 to 24", subtended at the eye. T h e functional area for summation to threshold for sinusoida1,gratings is a function of both the length of the bars and the number of bars of the grating (Howell & Hess 1979). By using a field of 24" the length of the bars and the number of cycles are above the critical limit at 0,5 cyclesidegree giving rise to a valid contrast sensitivity function at this spatial frequency and above it. With the smaller fields used an artefactual depression of the low spatial frequency end of the contrast sensitivity function is introduced (Estkvez & Cavonius 1976; Howell & Hess 1979). However, in more localized macular lesions, a small field size is important in order to assess contrast sensitivity changes in fovea-near regions of the macula. An illustrative case is the 24-year-old woman (B.N.; cf. Fig. 4) with fundus flavimaculatus with a severe macular involvement, giving her a visual acuity of 0.1 at the time of test with the large-field TV-system. As demonstrated by Fig. 7, a marked depression of the whole CSC curve is found for a field size of 1.4 and 6", using this TV-system. At an angle of 24" subtended at the eye, the CSC curve is almost normalized at lower spatial frequencies, demonstrating the striking field dependence that is found in some of the maculopathies studied. In the case with vitelliform macular degeneration (case S. L.), however, the CSC depression at lower spatial frequencies was observed also with the largest screen size used (Fig. 8) indicating more extended lesions in the posterior pole of the eye.
Discussion During the last decade psychophysical and evoked response methods have been used to measure the capacity of the human visual system to observe contrast patterns (Campbell & Maffei 1970; Campbell 1974; Sekular 1974). Spatially 842
periodic gratings of variable contrast have been used as targets to study the ability to perceive low-contrast patterns under various experimental and clinical conditions. It has been suggested that there may be in the visual system channels tuned to narrow ranges of spatial frequencies, i.e. neurons exist in the visual system which are selectively sensitive to certain sizes of visual targets (Blakemore & Campbell 1969; Campbell & Maffei 1970; Campbell 1974). Our knowledge of the contrast vision in disease, however, is more limited (for discussion see Arden 1978). T h e results reported here have demonstrated that measurement of contrast is not only useful for describing visual loss for objects larger than the resolution limit, but also of value for tracking progression or recovery in agreement with a preliminary study (Sjostrand & FrisPn 1977). The slope of the line in Fig. 6 relating contrast sensitivity to grating acuity in postoperative cases of macular detachment indicates that the contrast sensitivity measurement is as sensitive as grating or any other acuity. In early stages of maculopathies mainly the contrast sensitivity of the high- and middle-frequency range was attenuated, using the small-firld TL-syctem. In advanced cases of maculopathies the pattern of contrast sensitivity depression was similar to that in acute cases of optic nerve disease (Frisen & Sjostrand 1978) and anisometropic amblyopia (Sjostrand 1979) with impairment over a wide range of spatial frequencies. In spite of the disadvantages of using the small TV-screen, this system assessed the contrast sensitivity function of the central spatial vision without interference of the more peripheral vision. It has been shown that contrast sensitivity along a vertical line through the fixation point falls off steadily from a maximum at the centre (Robson & Graham 1979). However, this decline with eccentricity is most marked at high spatial frequencies and is small at lower spatial frequencies. The cases with an impairment over the whole spatial frequency range using the small-field system presumably has an involvement of the central area subtending 1.4"; selective impairment of high spatial frequencies indicate that a peripheral part of the 1.4" stimulation field is spared. Using the large-field TV-system with a field-size ranging from 6 to 24", it was possible to study the contrast sensitivity when retinal summation was allowed over a larger field. A strong field-dependence of the CSC-deficit was found in many of the maculopathies restricted to the central part of macula. In such cases the contrast sensitivity function at lower spatial frequencies was almost normal, with grating field subtending an angle of 24" at the eye (Fig. 7). In cases with extensive lesions in the posterior pole a subnormal CSC over the whole frequency range was still seen with a large stimulation field. Similar findings have been found in the acute phase of optic neuritis. T h e CSC impairment over the whole frequency range appears to be compatible with a disturbance of a majority of the afferent visual channels that 843
subserve central vision. The differential field-dependence provides some insight into daily visual problems of patients with macular disorders. Hess et al. (1978) have investigated a case with a uniocular central 6" scotoma, and they conclude that the macular region contributes significantly to the visibility of objects as large as 2.5 deg. A marked impairment of the low spatial frequency end of the contrast sensitivity curve, however, indicates involvement of a much larger area (cf. Robson & Graham 1979). T h e present study emphasizes that it is important to test CSC at different discrete regions over the visual field to assess regional differences. Previous studies of contrast sensitivity with stimuli of sinusoidal gratings displayed on a TV- or oscilloscope-screen have described abnormalities in contrast detection in patients with myopia (Fiorentini & Maffei 1976), amblyopia (Freeman & Thibos 1975; Levi & Harwerth 1977; Hess & Howell 1978) and suprageniculate lesions of the visual pathways (Rodis-Wollner 1972, 1974, 1976). In a recent study, using a book of printed gratings, Arden & Jacobson (1978) have demonstrated subnormal contrast sensitivity in glaucomatous patients. T h e majority of these studies give no evidence for a selective vulnerability of various spatial frequency ranges. The pattern of CSC attenuation is generally similar for a number of clinical disorders with mainly high-frequency loss in disease such as corneal oedema (Hess & Garner 1977), strabismic amblyopia (Sjostrand 1969), retinitis pigmentosa (Wolkstein et al. 1977) and early maculopathy (this study) and more general depression of CSC in optic nerve disease (Friskn & Sjostrand 1978), anisometropia (Sjostrand 1979) and advanced maculopathies (this study). In conclusion, our study has shown that measurement of spatial contrast sensitivity in visual disorders supplements traditional visual acuity measurements and is useful for tracking progression or recovery of macular disease. Based on the present findings, a less time-consuming test might be constructed using three selected spatial frequencies at high, intermediate and low frequencies. However, the definite role of CSC for diagnosing or distinguishing different clinical disorders still is lacking. T h e possibilities to use contrast measurements other than at threshold should therefore be explored. In our laboratory we are currently investigating the effect of disease on the supraliminal contrast level.
Acknowledgments Supported by the Swedish Medical Research Council (grant 02226), Riksfiirbundet mot Reumatism and Carmen, and Bertil RegnPr's Foundation. I am grateful to Dr. L. Frisbn for valuable advice and constructive discussion and for the loan of his television equipment.
844
References Andersson C . & Sjostrand J . (1979) T h e function of the macula following retinal detachment involving the macula. Actn ophthnl. (Khh.) 57, (in press). Arden G. R . (1978) T h e importance of measuring contrast sensitivity in cases of visual disturbance. Brit. J. Ophthnl. 62, 198-209. Arden C,. B. & JacobsonJ. J. (1978) A simple grating test for contrast sensitivity: Preliminary results indicate value in screening for glaucoma. I n 7 ~ r ~Ophthni. t. Vis. Sci. 27, 23-32. Rlakemore C. & Campbell F. W. (1969) O n the existence of neurons in the human visual system selectively sensitive to the orientation and size of retinal images. ,I. P h j d . (Lmd.) _703,237-260. Bodis-Wollner I. (1974) O n dissociated cisual functions. , L f o m t S i / / c c / .\Jd. 4 I , 38-45. Bodis-Wollner I. (1976) Vulnerability of spatial frequency channels in cerebral lesions. .Vaturr (Lond.) 261, 309-3 1 1. Campbell F. W. (1974) T h e transmission of spatial information through the visual system. The Nruro.wirncP.7 Third Study Program, pp. 95-103. T h e MIT Press, Cambridge Mass. and London. Campbell F. W. & Maffei I. (1970) Electrophysiological evidence for the existence of orientation and size detectors in the human visual system.]. Physzol. (Land.) 207, 635-652. Campbell F. W. & Robson J. G. (1968) Application of Fourier analysis to the visibility of gratings.J. Phyioi. (Lond.) 197, 551-556. Estevez 0.& Cavonius C. R. (1976) Low-frequency attenuation in the detectation of gratings: sorting out the artefacts. Vis. Res. 16,497-500. Fiorentini A. & Maffei L. (1976) Spatial contrast sensitivity of myopic subjects. Vis. Res. Z6, 437-438. Freeman R. D. & Thibos L. N. (1975) Contrast sensitivity in humans with abnormal visual experience.,]. Physiol. (Lond.) 247, 687-7 10. Frisen L. & Sjiistrand J. (1978) Contrast sensitivity in optic neuritis. Docum. Ophthnl. Proc. Ser. 17, 165- 174, Dr. W . J u n k b.v. Publishers, the Hague. Frisen I
Oslo, Norway 13-16 June 1979
CONTRAST SENSITIVITY IN MACULAR DISEASE USING A SMALL-FIELD AND A LARGE-FIELD TV-SYSTEM
JOHAN SJOSTRAND
T h e spatial contrast threshold for a sinusoidal grating of varying contrast and frequency generated with a small-field or a large-field TV-display was examined in normal subjects and in patients with macular disorders. Impairment of contrast sensitivity (reciprocal of contrast) was observed in the different maculopathies investigated. Attenuation of the high- and middlefrequency ranges was an early finding in macular disease, whereas changes including the low-frequency range was observed in more advanced maculopathies. Comparison of results obtained using the small (1.4") or large-field (6"-24") TV-system demonstrated a field-dependence of the contrast sensitivity attenuation in localized macular disorders. In more wide-spread lesions in the posterior pole, a contrast attenuation over the whole frequency range was found also with the largest (24")stimulation field used. T h e study of the contrast sensitivity function supplements the traditional acuity measurements in quantifying the visual loss for objects larger than the resolution limit. I conclude that contrast threshold measurements are not only useful for describing visual loss, but also for tracking progression or recovery. The findings provide some additional insight into the visual difficulties of daily life of patients with a macular disorder. However, the definite role of contrast sensitivity measurements for diagnosing different visual disorders is still lacking.
Key words: contrast sensitivity - maculopathy acuity.
-
spatial frequency - visual
T h e commonly used acuity tests give information about the spatial resolution of fine targets at high contrast. T h e ability to see larger targets and details at low contrast is of equal importance and is usually not tested in clinical investigations. 832
Con trcrst S~n.titi71ity
From clinical practice we know, however, that patients sometimes complain of deterioration of low-contrast vision, even though their visual acuity is within normal limits. A more general description of the visual system is obtained by studying the contrast sensitivity function, i. e. the reciprocal of threshold contrast for a range of sinusoidal gratings that vary in spatial frequency (for review see Campbell 1974; Sekular 1974). By the contrast sensitivity it is possible to study visual functions not tested in the acuity test. The aim of the study has been to investigate the contrast sensitivity function in various maculopathies, and the interest has been focused on the following question: Is contrast sensitivity measurement a useful additional clinical test in the diagnosis, treatment and follow-up of macular disorders?
Material and Methods Patients
Fifteen normal controls (age 19-63 years) and 22 patients (age 19-78 years) with rnacular disease of different types took part in this study. Smal I-field TV-system Vertical, sinusoidal gratings were generated on a television display as previously described by S,jiistrand & Fris6n (1977). T h e television monitor was masked to subtend 1.4 X 1.4 degrees of angle at the eve, when viewed at a distance of 5 m. A white, illuminated mask, which subtended approximately 2 x 2 degrees, surrounded the TV-screen. Spatial frequency (number of cycles per degree of visual angle) was varied between 1.5 and 38 cyclesidegree. Pattern contrast (defined as (IAm.,x + L,&, where L,,,,,, and I,m,,, are the maximum and minimum luminances of the sinusoidal pattern) was varied in 2 dB steps in the 0.6.5-0.0007 range. T h e space average luminance of the gratings was 13.5 cd/m2. T h e contrast sensitivity (reciprocal of contrast threshold) was determined monocularly with optimal correction and natural pupil. Seventeen frequencies were explored twice in a random manner and the contrast threshold determined by raising the contrast from a subthreshold level at a selected spatial frequency until a pattern wasjust visible by the patient. Large-f ield TV-system T h e 20” TV-monitor used in the large-field system was masked to subtend 6 X 6 degree of angle at the eye when viewed at a distance of 2.5 m. T h e TV-monitor was surrounded by a grey screen and by illumination of two daylight tubes, a uniform background luminance of 13 cd/m* was obtained. T h e space average luminance of the grating was 83 cd/m2.The spatial frequency was varied between 0.5 and 30 cyclesidegree. T h e pattern contrast of the sinusoidal grating was varied in 2 d B steps in the 0.6-0.001 range. By changing the distance to the monitor the size of the TV-screen was increased to 24-24 degrees of angle at a distance of 0.6 m. T h e size of the retinal images, using the field size of 1.4, 6, and 24”, respectively, is illustrated in Fig. 1. T h e test distance was controlled by using a chin support and refraction was optimally corrected for the distances used. T h e test procedure was as described above for the small-field system.
833 lcra ophthal. 57. 5
Fig. 1. A fundus picture of a normal control with a schematic drawing of the retinal images using a TV-screen which subtended 1.4 (0), 6, and 24 degrees of arc.
Results Normal subjects
Srnal1;field TV-system. The contrast sensitivity curves (CSC), i. e. the contrast sensitivity plotted as a function of the spatial frequency of the sinusoidal grating, had a bell-shaped form when tested for subjects with normal vision, no previous eye disease, and ophthalmoscopically healthy eyes (Sjostrand & FrisPn 1977) in agreement with previous studies (Schade 1956; Campbell & Robinson 1968). The curves showed a high and low frequency attenuation with a peak sensitivity between 8 and 15 cyclesidegree (Figs. 2-4). Our CSC curves in normal controls using the small-field TV-system ( 1.4") differ from those obtained using the large-field TV-system (6-24") and from those previously reported in that our subjects' peak sensitivity occurred at a higher spatial frequency. Large-field Tv-system. The contrast sensitivity curves of 10 normal controls obtained using the large-field TV-system had a much broader peak with a peak region varying from 1 to 8, 1-10 and 2-12 cyclesidegree for a field size of 6, 12, and 24", respectively (Figs. 7 and 8). A comparison of the peak frequency obtained for 6 normal controls measured on both the small-field (1.4") and the large-field TV-system masked down to 1.4" (see legend Fig. 7) showed a mean difference of 6.5 cyclesidegree for the two systems used (cf. Figs. 2 and 7). The difference in peak frequency therefore presumably reflects the mask and the surroundings and their luminances (cf. Estkvez 8c Cavonius 1976). 834
1000
400
100
40
c > >
t
m Y
10
m
+ C
0 4
4
1
SPATIAL
10
40
FREQUENCY (cycl. I d e d
Fig. 2. CSC (1.4") in normal controls (dotted area indicates range for ten normal subjects aged 19 and 61 years) and three representative cases of different stages of senile maculur drgrwmtion (solid lines). Case I. A. (A,female, 65-years-old) had moderate macular changes and V.A. 0.7; in case K . L . ( 0 ,male, 46-years-old, V . A . 0.4) changes were more advanced; and in case L. L. (m, male, 65-years-old, V. A. 0.2) the changes were extensive. Data of normal controls from Sjostrand & Frisen (1977).
Macular disorders - small-field TV-system
T h e CSC was measured in different types of macular disease: senile macular degeneration, central serous retinopathy, fundus flavimaculatus, vitelliform macular degeneration, and retinal detachment including the macula. The macula is morphologically and functionally non-uniform and since macular disorders may involve the foveal, perifoveal and perimacular region or the receptor or neuronal cells
835
5
1
SPATIAL
FREQUENCY
10
I
50
(cycles I deg )
F g . 3. CSC (1.4")in central sprous retinopathy. Case S. H. (male 42-years-old) with involvement of the right eye (A,lower curve was obtained during the acute phase, V. A. 0.6, and the upper curve following recovery, V . A . 1.5). The CSC of the contralateral left eye (V.A. 1.5) is indicated by filled circles (0).Dotted area shows the range of seven normal controls aged 37 to 6 3 years.
differently, variation in effects on the contrast sensitivity function may be anticipated and is, in fact, demonstrable in this study. Sen& macular degeneration. In senile macular degeneration, pathological changes occur within the macular area (Sarks 1976) giving rise to scotomata of varying size in central vision. The CSC was impaired in senile macular degeneration, and the decrease was notable at high and intermediate frequencies in cases with reductions in central visual acuity to 0.5-0.7 (Fig. 2, case I.A., V.A. 0.7). In advanced cases of macular degeneration abnormality of CSC across the whole spatial frequency range was observed, as illustrated by case L. L. (V. A. 0.2, in Fig. 2).
836
1OC
4c
10
4
1
*
t 2
-
m
fn w
6 C
5 8
1
10
4
SPATIAL
FREQUENCY
40
(cycles I deg )
Fig. 4 . Case B.N., female, 24-years-old. CSC (1.4")of the left eye (0)of this patient with fundus flnuimuculutuc with macular involvement (Stargardt's disease). The upper curve was obtained in 1976 ( V . A . 0.6) and the lower curve after a period of progressive deterioration of central vision 2 % years later (V.A. 0.1). The dotted area represents the range of five normal subjects aged 19 to 3 1 years (from Frisen & Sjostrand 1978).
In 2 patients with macular oedema at the time of the initial test, a subjective improvement and also an improved CSC was demonstrated at a second determination 2-6 months following the initial test. A rough correlation was seen between the contrast sensitivity level between 5 and 10 cycleddegree and the visual acuity in patients selected to be in a stationary phase of disease with no obvious retinal oedema (results not shown). Central serous retinopathy. In this macular disorder macular oedema develops due to leakage through the pigment epithelium. In most cases tested, a detachment of the
837
pigment epithelium in the macular region (Gass 1967) was observed. During the acute phase the patients complained about impaired discrimination of fine details, metamorphopsies and micropsia (Frisen & Frisen 1979). CSC was followed during the course of the disease and during the acute phase contrast sensitivity was decreased for high and intermediate frequencies. Generally, the CSC improved parallel to the increase of visual acuity. Fig. 3 demonstrates the changes in CSC in one case during acute and recovery phase. It is notable that during resolution when the visual acuity had improved to 1.5, the patient still complained about slight blurring of the vision and the CSC was still markedly abnormal in the high- and mid-frequency region (Fig. 3).
\ 1
10
4
SPATIAL
FREQUENCY
40 (cycles
#
deg
1
Fig. 5 .
Case S.L., male 4.i-years-old, CSC (1.4") of the left eye ( 0 )of this patient with r'itrllzform rntlcular drgenrration with slowly progressive deficit in central vision. The upper curve was obtained in 1976 (V. A. 0.7) and the lower curve (V.A. 0.6) 2 % years later. Dotted area shows the range of seven normal controls aged 37 to 63 years.
838
J
01
04
10
GRATING
40
ACUITY
Fig. 6. T h e relation between the mean contrast sensitivity in the mid-frequency range (7.55-12 cyclesideg.) obtained with the small-field TV-system (1.4") and grating acuity. Grating acuity was measured monocularly by means of sinusoidal gratings displayed o n the TV-screen at a distance of 7.9 m (for details, see Andersson & Sjestrand, 1979). Fifteen cases with healed retinal detachment, involving the macula, were examined. t,normal controls 19 to 42-years-old; 0 ,normal controls, 43 to 70-years-old; 0 , macular detachments, aged 19 to 41 years: 0 , macular detachments aged 42 to 70 years; A,fellow eyes for patients aged 19 to 41 years; W, fellow eyes for patients aged 42 to 70 years.
Miscellaneous maculopathies. In this group various progressive maculopathies such as fundus flavimaculatus (Stargardts disease) and vitelliform macular degeneration were investigated. During the time period of this study (3 years) the visual acuity dropped in several of the patients studied. Fig 4 demonstrates the CSC-curve of a 24-year-old female with a fundus flavimaculatus involving the macular region. 839
1000
1000
lo( I
5
t rn
z rn W
rn I-
2
1'
z
8
i
20 SPATIAL FREQUENCY (cyclesldeg.)
1000-
1000
B.N.
100
100
10
10
>
k E t rn W
rn
$ a
+ 0 0
2.5
25
1.25
12.5
SPATIAL FREQUENCY (cycles/deg.)
Fig. 7 Case A. N . with f i c n d ~flnuimocu1atzL.s (the same patient as described in Fig. 4). CSC of the left eye (solid line, V. A. 0.1) of this patient using the large-field TV-system (6-24"). The dotted area shows the range for 5 (So)and 6 (12 and 24") normal controls. To investigate the CSC for an angle of 1.4" at the eye (the same angle as that used with the small-field TV-system) with the TV-system ordinarily used for large-field, the screen was masked to 1.4" with a grey cardboard mask with the same background luminance as the surroundings. Dotted area for the 1.4" field indicates the range for six normal controls. a) 1.4 and 6";b) 12 and 24".
840
1000
1000
S.L
100
--i>? 100
t W z Lo k
2
+
10
10
z 0
2.5
125
25
12.5
SPATIAL FRQUENCY (cycles/deg.)
F z-q 8. Case S. L. with 7~itelliformmacular degeneration (the same patient as described in Fig. 5). CSC of the left eye (solid line V . A . 0.6) with the large-field TV-system (12 and 24"). The CSC for 6" was even more severely depressed (results not shown). Dotted area shows the range for six normal subjects.
Already 3 years ago when the visual acuity was G.6, the CSC curve was impaired, and, when studied half a year ago, both the visual acuity (0.1) and the CSC had dropped further (Fig. 4). Another patient with vitelliform degeneration of the macula, studied during the same time period, is shown as comparison. During the time period of study visual acuity only decreased from 0.7 to 0.6, whereas the CSC depression was markedly augmented over the whole frequency range (Fig. 5). At the same time, the patient experienced increasingly severe metamorphopsies, and the macular status deteriorated as visualized by the ophthalmoscope.
Retinal detachment including the macula. We have investigated a number of patients with rhegmatogenous retinal detachment including the macula, and followed the retinal function with various tests postoperatively (Anderson & Sjostrand 1979). It is known that age and refraction of the patient as well as duration of macular detachment preoperatively are of great importance for the functional outcome (Gundry 8~Davies 1974; Kreissig 1977). Following a successful operation of macular detachment, an initial rapid increase in visual acuity is found, followed by a second phase of prolonged visual improvement extending over more than a year (Kreissig 1977).The spatial processing in the 84 1
/ o h m Sjiistmnd
reattached macula was followed by measuring the contrast sensitivity function and grating acuity during the recovery phase. In a group of patients examined about a year after the operation, an abnormal CSC could be observed in the two-age groups studied (Anderson & Sjiistrand 1979). T h e high and intermediate frequency ranges were affected predominantly. Within the low frequency range the contrast sensitivity essentially was within the normal range. The reduction of the contrast sensitivity in the mid-frequency range was roughly correlated to the visual acuity and grating acuity in these patients (Fig. 6). Macular disorders - large-field TV-system
T o obtain a better insight into the field-related visual problems of patiens with visual disorders, a few of the patients measured with the small-field TV-system were selected for further study, using a large-field TV-system with a visual angle of 6 to 24", subtended at the eye. T h e functional area for summation to threshold for sinusoida1,gratings is a function of both the length of the bars and the number of bars of the grating (Howell & Hess 1979). By using a field of 24" the length of the bars and the number of cycles are above the critical limit at 0,5 cyclesidegree giving rise to a valid contrast sensitivity function at this spatial frequency and above it. With the smaller fields used an artefactual depression of the low spatial frequency end of the contrast sensitivity function is introduced (Estkvez & Cavonius 1976; Howell & Hess 1979). However, in more localized macular lesions, a small field size is important in order to assess contrast sensitivity changes in fovea-near regions of the macula. An illustrative case is the 24-year-old woman (B.N.; cf. Fig. 4) with fundus flavimaculatus with a severe macular involvement, giving her a visual acuity of 0.1 at the time of test with the large-field TV-system. As demonstrated by Fig. 7, a marked depression of the whole CSC curve is found for a field size of 1.4 and 6", using this TV-system. At an angle of 24" subtended at the eye, the CSC curve is almost normalized at lower spatial frequencies, demonstrating the striking field dependence that is found in some of the maculopathies studied. In the case with vitelliform macular degeneration (case S. L.), however, the CSC depression at lower spatial frequencies was observed also with the largest screen size used (Fig. 8) indicating more extended lesions in the posterior pole of the eye.
Discussion During the last decade psychophysical and evoked response methods have been used to measure the capacity of the human visual system to observe contrast patterns (Campbell & Maffei 1970; Campbell 1974; Sekular 1974). Spatially 842
periodic gratings of variable contrast have been used as targets to study the ability to perceive low-contrast patterns under various experimental and clinical conditions. It has been suggested that there may be in the visual system channels tuned to narrow ranges of spatial frequencies, i.e. neurons exist in the visual system which are selectively sensitive to certain sizes of visual targets (Blakemore & Campbell 1969; Campbell & Maffei 1970; Campbell 1974). Our knowledge of the contrast vision in disease, however, is more limited (for discussion see Arden 1978). T h e results reported here have demonstrated that measurement of contrast is not only useful for describing visual loss for objects larger than the resolution limit, but also of value for tracking progression or recovery in agreement with a preliminary study (Sjostrand & FrisPn 1977). The slope of the line in Fig. 6 relating contrast sensitivity to grating acuity in postoperative cases of macular detachment indicates that the contrast sensitivity measurement is as sensitive as grating or any other acuity. In early stages of maculopathies mainly the contrast sensitivity of the high- and middle-frequency range was attenuated, using the small-firld TL-syctem. In advanced cases of maculopathies the pattern of contrast sensitivity depression was similar to that in acute cases of optic nerve disease (Frisen & Sjostrand 1978) and anisometropic amblyopia (Sjostrand 1979) with impairment over a wide range of spatial frequencies. In spite of the disadvantages of using the small TV-screen, this system assessed the contrast sensitivity function of the central spatial vision without interference of the more peripheral vision. It has been shown that contrast sensitivity along a vertical line through the fixation point falls off steadily from a maximum at the centre (Robson & Graham 1979). However, this decline with eccentricity is most marked at high spatial frequencies and is small at lower spatial frequencies. The cases with an impairment over the whole spatial frequency range using the small-field system presumably has an involvement of the central area subtending 1.4"; selective impairment of high spatial frequencies indicate that a peripheral part of the 1.4" stimulation field is spared. Using the large-field TV-system with a field-size ranging from 6 to 24", it was possible to study the contrast sensitivity when retinal summation was allowed over a larger field. A strong field-dependence of the CSC-deficit was found in many of the maculopathies restricted to the central part of macula. In such cases the contrast sensitivity function at lower spatial frequencies was almost normal, with grating field subtending an angle of 24" at the eye (Fig. 7). In cases with extensive lesions in the posterior pole a subnormal CSC over the whole frequency range was still seen with a large stimulation field. Similar findings have been found in the acute phase of optic neuritis. T h e CSC impairment over the whole frequency range appears to be compatible with a disturbance of a majority of the afferent visual channels that 843
subserve central vision. The differential field-dependence provides some insight into daily visual problems of patients with macular disorders. Hess et al. (1978) have investigated a case with a uniocular central 6" scotoma, and they conclude that the macular region contributes significantly to the visibility of objects as large as 2.5 deg. A marked impairment of the low spatial frequency end of the contrast sensitivity curve, however, indicates involvement of a much larger area (cf. Robson & Graham 1979). T h e present study emphasizes that it is important to test CSC at different discrete regions over the visual field to assess regional differences. Previous studies of contrast sensitivity with stimuli of sinusoidal gratings displayed on a TV- or oscilloscope-screen have described abnormalities in contrast detection in patients with myopia (Fiorentini & Maffei 1976), amblyopia (Freeman & Thibos 1975; Levi & Harwerth 1977; Hess & Howell 1978) and suprageniculate lesions of the visual pathways (Rodis-Wollner 1972, 1974, 1976). In a recent study, using a book of printed gratings, Arden & Jacobson (1978) have demonstrated subnormal contrast sensitivity in glaucomatous patients. T h e majority of these studies give no evidence for a selective vulnerability of various spatial frequency ranges. The pattern of CSC attenuation is generally similar for a number of clinical disorders with mainly high-frequency loss in disease such as corneal oedema (Hess & Garner 1977), strabismic amblyopia (Sjostrand 1969), retinitis pigmentosa (Wolkstein et al. 1977) and early maculopathy (this study) and more general depression of CSC in optic nerve disease (Friskn & Sjostrand 1978), anisometropia (Sjostrand 1979) and advanced maculopathies (this study). In conclusion, our study has shown that measurement of spatial contrast sensitivity in visual disorders supplements traditional visual acuity measurements and is useful for tracking progression or recovery of macular disease. Based on the present findings, a less time-consuming test might be constructed using three selected spatial frequencies at high, intermediate and low frequencies. However, the definite role of CSC for diagnosing or distinguishing different clinical disorders still is lacking. T h e possibilities to use contrast measurements other than at threshold should therefore be explored. In our laboratory we are currently investigating the effect of disease on the supraliminal contrast level.
Acknowledgments Supported by the Swedish Medical Research Council (grant 02226), Riksfiirbundet mot Reumatism and Carmen, and Bertil RegnPr's Foundation. I am grateful to Dr. L. Frisbn for valuable advice and constructive discussion and for the loan of his television equipment.
844
References Andersson C . & Sjostrand J . (1979) T h e function of the macula following retinal detachment involving the macula. Actn ophthnl. (Khh.) 57, (in press). Arden G. R . (1978) T h e importance of measuring contrast sensitivity in cases of visual disturbance. Brit. J. Ophthnl. 62, 198-209. Arden C,. B. & JacobsonJ. J. (1978) A simple grating test for contrast sensitivity: Preliminary results indicate value in screening for glaucoma. I n 7 ~ r ~Ophthni. t. Vis. Sci. 27, 23-32. Rlakemore C. & Campbell F. W. (1969) O n the existence of neurons in the human visual system selectively sensitive to the orientation and size of retinal images. ,I. P h j d . (Lmd.) _703,237-260. Bodis-Wollner I. (1974) O n dissociated cisual functions. , L f o m t S i / / c c / .\Jd. 4 I , 38-45. Bodis-Wollner I. (1976) Vulnerability of spatial frequency channels in cerebral lesions. .Vaturr (Lond.) 261, 309-3 1 1. Campbell F. W. (1974) T h e transmission of spatial information through the visual system. The Nruro.wirncP.7 Third Study Program, pp. 95-103. T h e MIT Press, Cambridge Mass. and London. Campbell F. W. & Maffei I. (1970) Electrophysiological evidence for the existence of orientation and size detectors in the human visual system.]. Physzol. (Land.) 207, 635-652. Campbell F. W. & Robson J. G. (1968) Application of Fourier analysis to the visibility of gratings.J. Phyioi. (Lond.) 197, 551-556. Estevez 0.& Cavonius C. R. (1976) Low-frequency attenuation in the detectation of gratings: sorting out the artefacts. Vis. Res. 16,497-500. Fiorentini A. & Maffei L. (1976) Spatial contrast sensitivity of myopic subjects. Vis. Res. Z6, 437-438. Freeman R. D. & Thibos L. N. (1975) Contrast sensitivity in humans with abnormal visual experience.,]. Physiol. (Lond.) 247, 687-7 10. Frisen L. & Sjiistrand J. (1978) Contrast sensitivity in optic neuritis. Docum. Ophthnl. Proc. Ser. 17, 165- 174, Dr. W . J u n k b.v. Publishers, the Hague. Frisen I
