White Noise Field Campimetry
(Aulhorn's snow field campimetry)

Ulrich Schiefer, M.D.

Summary

White noise field campimetry – also called snowfield campimetry – is a method directly visualizing scotomas, thus enabling patients to immediately detect and interactively describe their visual field defects (VFDs). This is realized by a presenting an equal number of randomly distributed small black and white square dots (12´x 12´) with a flicker frequency of 30 Hz on a video display unit. Scotomas are usually described as “clouds”, which are differentiated from the surrounding normal noise field by an apparent change in brightness perception and / or change in noise perception. Subjects respond by delineating the scotoma with their finger tips on a computer (touch) screen. Subsequently, the examiner can ask for further characteristics of the noise field defects in a standardized way. Fixation is monitored via an infrared video camera.

White noise field campimetry has been validated by finding good agreement with conventional perimetry in a series of 368 eyes of 368 patients with visual pathway lesions.

The sensitivity of snowfield campimetry was assessed by comparing its overall scotoma detection rates (80.7; 95%-confidence interval [C.I.] 76.7% - 84.3%) with those of conventional white-on-white perimetry i.e. automated threshold-oriented slightly supraliminal static perimetry (84.2%; 95%-C.I. 80.5% - 87.6 %), performed on the Tuebingen Automated Perimeter (TAP) in the above-mentioned cohort. Sensitivity of the method was highly dependent on location (i.e. site) of the lesion and scotoma depth.

The white noise stimulus was also presented to 198 eyes of 198 ophthalmologically normal subjects in order to assess the specificity of this method, being 82.3% (95-C.I. 76.9 – 87.1%).

A “floor” effect, i.e. missing detection at the low end of the dynamic range, does exist insofar as very subtle relative defects may be not detected by the patients. Furthermore, the method seems to have a reduced or even missing sensitivity towards specific conditions of visual pathway lesions (e.g.: extensive retinal lesions or long-standing post-geniculate lesions).  A “ceiling” effect is found in case of subtotal visual field loss, most probably due to fixation problems and a remnant normal area, which may be too small for acting as a sufficient reference region. Furthermore, advanced visual field loss located outside the video display cannot be detected.

Due to its high spatial resolution within the central visual field, white noise field campimetry can easily detect even the smallest areas of circumscribed visual field loss, for instance caused by Purtscher´s retinopathy, which are often missed by conventional automated static perimetry due to the usually comparatively coarse grid.

Since white noise field defects are immediately perceived by the patients, this method is well suited for screening purposes and for assessing spontaneously occurring short-term visual alterations, such as obscurations or induced visual field changes, e.g. during stress tests.

Introduction / History of development

Approximately 3% to 3.5% of all subjects suffer from a visual field defect, and thus a lesion of the visual pathway. The majority of scotomas are of the negative type, i.e. they cannot be perceived by the patient. Local adaptation, “filling-in” phenomena and slow progression of manifest scotoma may explain why subjects spontaneously contact an ophthalmologist only in very advanced stages of visual field loss. This is well-known in glaucoma patients. Furthermore, it is a difficult task motivating patients for potentially life-long medication or even surgical therapy if they do not really actively experience the functional impact of the already existing visual field defect.

At the suggestions of one of her glaucoma patients, Elfriede Aulhorn in 1987 described a new examination technique, called white noise field campimetry ^1;2. In the initial state, a black and white TV monitor without broadcast reception was used for presenting a white noise stimulus (Fig. 1). More recent versions use a computer-generated noise field ^3;4 and have added a touch-screen device, a fixation control unit and a completely software-generated snowfield stimulus for further enhancing clinical and scientific applicability ^1;5(Fig. 2).

Methods

White noise field campimetry is performed in a darkened room using a high resolution video display unit (VDU) for stimulus presention. The details of this method have been published ^1;3-5.

With the campimeter an area (“radius”) of 35° (horizontal) and 24° (vertical) of the central visual field can be examined, at a distance of 30 cm between the eye and the surface of the screen. An equal number of black and white randomly distributed black and white square dots with a flicker frequency of 30 Hz is used, comparable to the snowfield of a TV screen without broadcast reception. Each dot has a visual angle of 12´x 12´. The luminance of the light dots is 200 cd/m² and that of the dark ones 0.8 cd/m², respectively.  A stimulus in the center of the screen is presented to stabilize central fixation. Direct inspection and video monitoring of the examined eye is used to control fixation. The head of the examined subject is positioned in a combined head and chin rest. Adequate thin rim glasses are used for near correction if necessary. The fellow eye is covered with an opaque patch.

The examiner asks the patient in a standardized way about the perception of the noise field. Most of the disturbances (scotomas) are described as “clouds”. The clouds can be differentiated from the surrounding normal snowfield by the following qualities:

The border of the “cloud” is indicated by the patient with the tip of his finger directly on the surface of the monitor and can be directly documented by a touch screen option (or – in former version – by directly following the path of the patient´s indicating finger with a computer mouse).

For clinical evaluation, white noise field campimetry was compared to results of threshold-related, slightly supraliminal automated static grid perimetry, obtained with the Tuebingen Automated Perimeter (TAP). With this conventional perimetric technique a comparatively high spatial resolution (191 test location with centripetal condensation and polar arrangement within the central 30° visual field) is achieved at the cost of an ultimate scotoma depth resolution.

Test times of white noise field campimetry range from 1 to 5 minutes per eye, including standardized questions about the perception of the noise field and documentation of the result.

Validity 

Extensive studies have been performed in order to test whether white noise field campimetry is capable of detecting visual pathway lesions of various origin (see chapter “Sensitivity and Specificity”).

PET as well as fMRI studies (Fig. 3) were able to show that a (hemifield) random noise stimulus is conducted to the visual cortex and is able to elicit specific activity in the contralateral hemisphere ⁶.

Sensitivity and Specificity

The specificity of white noise field campimetry was estimated by presenting this stimulus to 198 eyes of 198 opthalmologically normal subjects (age range 11 to 74 years)⁷. Subjective statements, which characterized the flicker perception as “completely normal over the entire screen” (n = 84) or described a “somewhat changed” flicker perception at the very edges of the campimeter (n = 21) or a “more distinct” flicker perception in the central snowfield area (n = 19) or finally reported on “transient clouds”, which disappeared immediately after blinking (n = 19), were regarded as normal. This evaluation procedure resulted in a specificity of 82.3% (95-C.I. 76.9 – 87.1%).

Sensitivity for defect detection was investigated with subjects having white noise field campimetry and conventional automated static threshold-oriented, slightly supraliminal perimetry (30° TAP, see chapter “Methods”) on the same day in 368 eyes of 368 patients with visual pathway lesions from various origins. Overall scotoma detection rates were 80.7% (95%-C.I. 76.7% - 84.3%) for white noise field campimetry, compared to 84.2% (95%-C.I. 80.5% - 87.6%) with conventional TAP perimetry. In the vast majority, noise field defects were noticed by combined changes in noise (most frequently reduced or even defective) and brightness perception (with reports on “dark” or “bright” clouds being roughly equally distributed). 

In contrast to conventional static perimetry, sensitivity of white noise field campimetry turned out to highly dependent on e.g. scotoma depth and also location (i.e. site) of the lesion. Relative glaucomatous visual defects in TAP perimetry (, i.e. stage I, according to the Aulhorn classification ⁸) were detected by white noise field campimetry in 65.0% of cases. With absolute glaucomatous scotomas, detection rate was 83.8% without connection to the blind spot (Aulhorn stage II) and increased to 100% for all stages of more advanced glaucomatous visual field loss. A typical finding is shown in Figure 4.

In 162 eyes with homonymous visual field defects due to the post-chiasmal visual pathways lesions, detection rate of conventional TAP perimetry was 98.8% (95%-C.I. 95.7% - 99.9%), whereas sensitivity of white noise field campimetry was only 76.5% (95%-C.I. 70.5% - 82.2%)¹. More recent studies give evidence for higher detection rates of white noise field campimetry for those cases with involvement of the visual cortex ⁹.

Reliability and Test-retest Variability

To the author´s knowledge, no systematic studies addressing reliability and re-test variability of white noise field campimetry have been undertaken so far.

Floor and Ceiling Effects

A “floor” effect, i.e. missing detection at the low end of the dynamic range, does exist insofar as very subtle relative defects may be not detected by the patients (see also chapter “Sensitivity and Specificity”). Furthermore, the method seems to have a reduced or even missing sensitivity towards specific conditions of visual pathway lesions (e.g.: extensive retinal lesions or long-standing post-geniculate lesions; this aspect is also addressed in “Sensitivity and Specificity” chapter). 

A “ceiling” effect is found in case of subtotal visual field loss, most probably due to fixation problems and a remnant normal area, which may be too small for acting as a sufficient reference region. Subjects with inability to fixate (e.g. due to nystagmus) or profound central scotomas may be unable to perceive the noise field stimulus in an adequate manner.

Furthermore, even advanced visual field loss located outside the video display cannot be detected.

Practicality

The test is practical, requiring a black and white TV monitor without a broadcast reception; in the high end version, a calibrated, high-resolution colour monitor with integrated touch screen, a control monitor, and a video-pupillographic device have to be integrated in an appropriate computer system graphics card is used. A custom built lens holder that fits on the top of the monitor along with custom made lenses is necessary.

Subject Acceptance

To the author's knowledge, no systematic studies addressing subject acceptance have been published so far. Our experience is that most of the patients had no difficulty with the test.

Robustness Against Errors 

White noise field campimetry is a subjective method in twofold ways: on the one hand it depends on the intellectual abilities of the examined subject to perceive and adequately describe the observed phenomena, on the other it is influenced by the recognition and interpretation of the examiner. Further errors can be introduced by the fact that patients, who for first time in their life perceive their scotomas, will tend to directly look at these phenomena. This spontaneous gaze movement inevitably shifts the scotoma, which might then be dislocated outside the examination area and thus disappear. Careful, fixation continuous control by the examiner is therefore mandatory in this method.

Subjects with improper correction, inability to fixate or profound central scotomas may not be able to perceive the noise field stimulus in an adequate manner.

As with conventional automated perimetry, tear film, eye brow and eye lash artifact can occur.

Normative Standards 

Specificity data is available on 198 eyes of 198 ophthalmologically normal subjects (age range 11 to 74 years)^{7 1}. Sensitivity data were obtained from 368 eyes of 368 patients with various pathway lesions (see chapter “Sensitivity and Specificity”)⁷.

Further Developments

In a recent version the noise field stimulus is exclusively software generated and presented on a calibrated, high-resolution colour monitor with integrated touch screen; the results are recorded on a separate examination screen ^1;5  (Fig. 2). This enables the patients to delineate their scotomas on the examination monitor without any disturbance by visible borders. Noise field characteristics (i.e. brightness, movement of the dots and colour characterstics) within the demarcated scotoma region(s) or within the entire examination screen can be interactively varied in order to assess further quantitative scotoma features.

In a pilot study white noise field stimulus has also been used for remote screening via a regular TV program ^10;11.

Originally developed by: Elfriede Aulhorn and Gert Köst

Further developments by: Ulrich Schiefer and Guido Stercken-Sorrenti

Commercially available: Please contact OCULUS Inc., Dutenhofen, Germany or further information; for screening purposes, a regular black and white TV screen with no broadcast reception seems to be sufficient.

Figures

Reference List

   1.   Schiefer U. Rauschfeldkampimetrie. Stuttgart: Kohlhammer, 1995.

   2.   Aulhorn E. White noise field campimetry. Lecture at 85. German Ophthalmological Society Meeting. 1987.

   3.   Aulhorn E, Köst G. Rauschfeldkampimetrie. Eine neuartige perimetrische Untersuchungsmethode. Klin Monatsbl Augenheilkd 1988;192:284-8.

   4.   Aulhorn E, Köst G. In: Heijl A, ed. Perimetry Update 1988/89. Proceedings of the VIIIth International Perimetric Society Meeting. Amsterdam: Kugler & Ghedini, 1989: 331-6.

   5.   Schiefer U, Stercken-Sorrenti G. Ein neues Rauschfeldkampimeter. Klin Monatsbl Augenheilkd 1993;202:60-3.

   6.   Schiefer U, Skalej M, Kolb R et al. Cerebral activity during visual stimulation - a positron emission tomography and functional magnetic resonance imaging study. German J Ophthalmol 1996;5:109-17.

   7.   Schiefer U, Pfau U, Selbmann HK et al. Sensitivität und Spezifität der Rauschfeldkampimetrie. Ophthalmologe 1995;92:156-67.

   8.   Aulhorn E, Karmeyer H. Frequency distribution in early glaucomatous visual field defects. Docum Ophthal Proc Series 1977;14:17-83.

   9.   Kolb M, Petersen D, Schiefer U et al. Scotoma perception in white-noise-field campimetry and postchiasmal visual pathway lesions. German J Ophthalmol 1995;4:228-33.

10.   Gisolf AC, Kirsch J, Selbmann HK, Zrenner E, Schiefer U. in: Wall M, Heijl A, eds. Perimetry Update 1996/1997. Amsterdam: Kugler Publications, 1997: 49-50.

11.   Schiefer U, Gisolf AC, Kirsch J et al. Rauschfeld-Screening - Ergebnisse einer Fernseh-feldstudie zur Detektion von Gesichtsfelddefekten. Ophthalmologe 1996;93:604-16.