GB2458134A - Pupillometers - Google Patents

Abstract

An integrated binocular pupillometer for assessing relative afferent pupillary defect. Stimulating visible light pulses are applied to each eye in turn by separated illuminable screens. The screens each has a visible fixation point associated therewith and each screen being viewable through an object lens such that the observed field of view is at least 9{.

Description

Pupilometers This invention relates to pupillometers, i.e. to apparatus for determining pupil dimensions in respect of the eyes of a subject, generally (although not necessarily exclusively) a human subject. More particularly the invention concerns a binocular pupillometer for assessing relative afferent pupillary defect in a subject.

Relative afferent pupillary defect is a condition wherein nerve signals from the eye to the brain are unilaterally compromised, for example by retinal lesions or by damage to the optic nerve. In general the right and left pupils react bilaterally to a light stimulus, so that both pupils will contract when such a stimulus is applied to one eye only. In healthy subjects the amounts of pupillar contraction will be the same whichever eye is stimulated; where relative afferent pupillary defect is present, however, both pupils will contract in response to a unilateral stimulus, but this bilateral contraction is smaller when the stimulus is applied to the affected eye.

The traditional clinical procedure for diagnosis of relative afferent pupillary defect is the swinging flashlight test. This employs a small handheld flashlight which is shone into each eye alternately for one or two seconds, the clinician judging whether there is any inequality in the magnitude of the pupillary reactions.

A set of small neutral density filters thay then be used to estimate the magnitude of any observed defect, the flashlight being occluded with such a filter when shone into the healthy eye. The density of the filter is adjusted until filtered illumination of the healthy eye and unfiltered illumination of the affected eye produce equal pupillar responses, this density providing an index of the magnitude of the afferent pupillary defect.

Whilst quick and simple, this procedure is inevitably subjective arid therefore liable to intraoperative variation. Moreover, since the procedure requires attenuation of stimulus to the eye with the larger response, the final measurement will necessarily be made at a relatively low response level, resulting in a loss of resolution and an escalation of measurement errors.

Kalaboukhova et a!., Neuro-Ophthalmology 30, pp. 7-15 (2006), describe an objective method for measuring relative afferent pupillary defect using a custom-built pupillometer consisting of two digital high-resolution video cameras for detection of pupillary response (one for each eye), two white diode lamps allowing alternating stimulating illumination of the eyes, and a background infrared illumination device. Two pasteboard screens fitting on each side of the nose separate the eyes and prevent stray light entering an eye from the contralateral side.

The apparatus is set up in a darkroom and the subject is required to look at a distant dark red fixation light in order to avoid accomodative miosis. Alternating light stimulation of a 0.5 second pulse followed by a I second pause was found to be the best stimulus pattern for relative afferent pupillary defect detection, and it was reported that this form of pupillometry could distinguish eyes with glaucoma from normal eyes with good sensitivity and specificity.

It will be appreciated that the substantial apparatus requirements, including the necessary electronic circuitry as well as the need to provide a dedicated darkroom, limit the applicability of this procedure and render it considerably expensive.

Portable binocular pupillometers in which the stimulating means and the means for detecting pupillary response are commonly housed to form an integrated unit are known in the art, for example as described in WO-A-9407406 and WO-A- 200603 2920. The latter document also describes a pupillometric procedure for diagnosis of relative afferent pupillary defect in which light stimulus pulses are alternately presented to each eye at four or more intensity levels, with each stimulus level optionally being repeated one or more times. One or more pupillary response parameters, such as depth of constriction (e.g. as measured by change in pupil diameter or area), maximum or average speed of constriction, latency period between onset of stimulus and onset of constriction, redilation time or time to reach a specific degree (e.g. 25%, 50% or 75%) of constriction and/or redilation are recorded, and the average values of the chosen parameter(s) for each stimulation level are then plotted (e.g. graphically or digitally) against the level intensities to give a pupillary response profile.

In this procedure the response profiles for the eyes of healthy subjects will substantially completely overlap. For subjects with relative afferent pupillary defect, on the other hand, stimulation of one eye will generate a consistently lower response than does stimulation of the other eye. The extent of the disorder may be quantified by determining the shift between the two profiles along the direction of the stimulus intensity axis, since this represents the light intensity scaling factor required to bring the two profiles into coincidence, and so is an equivalent of the neutral filter density magnitude obtained in the swinging flashlight test.

The present invention is based on the finding that even more precise and accurate diagnosis of relative afferent pupillary defect may be performed using a modified pupillometer as described hereinafter. Measurements so obtained may be of particular value in detecting glaucoma.

Thus according to one aspect of the invention there is provided an integrated binocular pupillometer for assessing relative afferent pupillary defect in a subject and comprising eyepieces for positioning against the eyes of said subject, sensing means adapted to generate and record separate pupillary response data in respect of the left and right eyes, and, commonly housed therewith, left and right stimulating means adapted to apply stimulating visible light pulses independently to said eyes in turn, characterised in that said left and right stimulating means are optically separated by dividing means and each comprise an illuminable screen which has a visible fixation point associated therewith and which is viewable through an object lens positioned between said screen and said eyepiece, said screen and lens being configured such that the observable field of view is at least 9°.

The sensing means for generating and recording pupillary response data may, for example, comprise means for infrared illumination of each eye, such as infrared-emitting diodes. Infrared illumination applied in this way will be scattered back from the iris and may be transmitted through a suitable optical train to form an image of at least the internal boundary of the iris with the pupil. Detector means which may be used to permit generation of pupillary response data from such images conveniently comprise two-dimensional arrays of charge-coupled devices, e.g. as in CCD video cameras. Separate detectors may be used for each eye or images may be transmitted to separate parts of a common sensor (e.g. as described in WO-A-9407406) or alternately to the same part of a single sensor (e.g. as described in WO-A-2006032920).

Pupillary responses which may be detected include depth of constriction, speed of constriction, latency period and redilation time. In a preferred embodiment charges in pupil diameter are recorded; processing of the image data may use techniques such as image'enhancement, automatic measurement of pupil area and application of a circle fitting algorithm in case of variation from absolute circularity in pupil shape.

The screen of each of the two stimulating means is preferably illuminable by a light-emitting diode (most preferably a white light-emitting diode) positioned outside the field of view. Such use of indirect light stimulation, as opposed to light stimulation which is shone directly into the eye, has been found to be particularly beneficial in allowing consistent and reproducible results to be obtained.

The configuration of the screen and object lens such that the observable field of view is at least 90, e.g. 10�0.5°, is also highly ad'antageous in that it ensures that stimulating light reaches a substantial portion of the retina, including the base of the optic nerve.

The visible fixation point associated with each illuminable screen is a further key feature of the pupillometer of the invention, significantly enhancing its versatility and the consistency of measurements obtained. Each such point may conveniently comprise a steadily illuminable coloured light source (e.g. a light-emitting diode, preferably a green light-emitting diode) positioned in contact with or behind an aperture in the respective screen. The light may be relatively dim compared to the stimulating light. Each fixation point and object lens may be configured such that the fixation point essentially appears at infinity, although systems in which the fixation point may be made to appear nearer to suit a particular subject may also be employed.

In the accompanying drawing, which serves to illustrate the invention without in any way limiting the same. Fig. I is a schematic side elevation of a pupillometer according to the invention showing an optical train providing vertical separation of infrared sensing irradiation and visible light stimulation.

In this embodiment, sensing infrared light from infrared-emitting diode IRED is scattered from the iris of the right eye, passes through cold mirror CMI, is reflected by first-surface infrared mirrors IRMI and IRM2 and is focussed by lens L2 onto a CCD imager.

Screen S is illuminable by light-emitting diode LEDI positioned so as to be outside the field of view of the subject. Stimulating light pulses from screen S are transmitted through object lens LI and reflected by visible mirror VMI and cold mirror CMI to the right eye. Green light-emitting diode LED2, positioned behind a small aperture in screen S, provides the visible fixation point, which is positioned at the focal point of object lens Li.

A similar optical train (not shown) is provided for the left eye. It is important that the two trains are optically separated by suitable dividing means in order to avoid cross-contamination of light stimuli to the contralateral.

In operation of the pupillometer according to the invention it is preferred to apply a sequence of stimulating visible light pulses of identical duration, intensity and separation to the left and right eyes alternatively and to measure the amplitude change, e.g. in pupil diameter, caused by these stimuli. Pupillary response may be maximised by selecting a relatively high intensity for the pulses, thereby optimising reproducibility. This is advantageous compared to the above-mentioned procedure from WO-A-2006032920 where response parameters are plotted against signal intensity, since the lower intensity measurements here will introduce greater variability to the results.

The pulse durations are conveniently in the range 0.2-0.5 seconds, with separation times in the range 1-8 seconds, e.g. 1.2-5 seconds. The alternating left and right eye stimuli are preferably symmetrically interleaved.

Accuracy is assisted by recording results for a sequence at 4-8 pulses to each eye, for example 4 pairs of 0.4 or 0.5 second pulses 3 seconds apart (i.e. a sequence 0.4 or 0.5 second left eye stimulus, 3 seconds off, 0.4 or 0.5 second right eye stimulus, 3 seconds off etc.), 7 pairs of 0.2 second pulses 2 seconds apart or, more preferably, 7 pairs of 0.4 second pulses 1.6 seconds apart. This last sequence has been found particularly advantageous in providing optimum precision in measuring pupillary response in a procedure with an overall duration of only about 30 seconds.

It is generally preferred to discard data from the first pair of measurements in such sequences. These tend to be the least accurate since the states of adaptation of the retinae to illumination change substantially following the initial pulse to each eye.

Data from the remaining pairs of the sequence may be averaged to provide traces of pupillary response such as plots of the average pupil diameter change induced by the stimuli. It will be appreciated that these may be separated into direct reflex (e.g. as in lefi pupil response to left stimulus) and consensual reflex (e.g. as in right pupil response to left stimulus). In order to enhance accuracy it is preferred to use only data in respect of direct reflex.

It will be appreciated that such averaging of left and right pupillary responses respectively will smooth out at least some pixel-related noise generated by sensing components such as CCD imagers.

If desired the results may be cleaned of artefactual data caused by blinking.

Thus, for example, if a blink is less than three frames in duration (i.e. <3 x 0.04 seconds at an imaging frame rate of 25 Hz), one may interpolate "good" data from either side of the blink. This may be perfonned manually or by appropriately programmed electronic processing.

The amplitudes A and AL of the pupil diameter changes for the right and left eyes may be normalised by dividing them by the initial (i.e. unstimulated) pupil diameters (Di)R and (Di)L. An afferent pupillary defect is present where the two normalised amplitudes differ, lying in the direction of whichever is the smaller. It may be qualified by the formula APD = 100 (1 -smaller normalised amplitude) %.

larger normalised amplitude In order to optimise the results one may advantageously perform at least two sets of measurements and calculations so as to obtain two or more APD values.

Thus, for example, it is currently preferred to generate three APDs, take the closest two values and reject the third. The final APD is the mean of these two, whilst their standard deviation is a measure of repeatability.

Claims (13)

1. Claims I. An integrated binocular pupillometer for assessing relative afferent pupillary defect in a subject and comprising eyepieces for positioning against the eyes of said subject, sensing means adapted to generate and record separate pupillary response data in respect of the left and right eyes, and, commonly housed therewith, left and right stimulating means adapted to apply stimulating visible light pulses independently to said eyes in turn, characterised in that said left and right stimulating means are optically separated by dividing means and each comprise an illuminable screen which has a visible fixation point associated therewith and which is viewable through an object lens positioned between said screen and said eyepiece, said screen and lens being configured such that the observable field of view is at least 9°.
2. 2. A pupillometer as claimed in claim 1 wherein the sensing means comprise means for infrared illumination of each eye to generate images of the pupil and detector means permitting generation of pupillary response data from images so generated.
3. 3. A pupillometer as claimed in claim I or claim 2 which the screen of each stimulating means is illuminable by a light-emitting diode positioned outside thefield of view.
4. 4. A pupillometer as claimed in claim 3 wherein each said light-emitting diode generates white light..
5. 5. A pupillometer as claimed in any preceding claim wherein each stimulating means is adapted to apply a sequence of visible light pulse of identical duration, intensity and separation.
6. 6. A pupillometer as claimed in claim 5 wherein the left and right stimulating means are adapted to apply equally spaced alternating visible light pulses to the eyes in turn.
7. 7. A pupillometer as claimed in claim 6 wherein the duration of said pulses is in the range 0.2-0.5 seconds and the separation between each pulse is in the range 1-8 seconds.
8. 8. A pupillometer as claimed in claim 7 adapted to apply a total of 4-8 pulses to each eye.
9. 9. A pupillometer as claimed in claim 8 adapted to apply a total of 7 pulses to each eye, the duration of each pulse being 0.4 seconds and the separation between each alternating pulse being 1.6 seconds.
10. 10. A pupillometer as claim in any preceding claim wherein each visible fixation point comprises a steadily illuminable coloured light source positioned in contact with or behind an aperture in the corresponding screen.
11. 11. A pupillometer as claimed in claim 10 wherein said light is coloured green.
12. 12. A pupillometer as claimed in any proceeding claim wherein each fixation point and lens are configured such that the fixation point essentially appears at infinity.
13. 13. A pupillometer as claimed in any preceding claim wherein each screen and lens are configured such that the observable field of view is I 0�0.5°.