FR2719988A1 - Eye movement control device. - Google Patents

Abstract

A support (12) for a living being, controllably movable in three directions, defines an area (29-I) for the eyes (30-I). A frame (1) comprises two light sources (14), supplying, at the level of the area (29-I) for the eyes (30-I), two infrared light beams (37-I), two cameras (13-I), each for capturing an image derived from the reflection of said light beam (37-I) on an eye (30-I) and processing means for digitalizing the eye images and processing them in real time to calculate the center coordinates of the corneal point (42-I) with reference to the pupil center (41-I). The processing means further control the movement of the support (12) to position the eyes (30-I) precisely in the axis of the cameras (13-I), according to the coordinates of the pupil contour. A mass memory in the microcomputer (2) stores the data relating to a plurality of basic visual tests comprising a selected sequence of visual stimulation images. The response of the eyes to an image provides a gaze direction which is analyzed by processing means.

Description

Device for controlling these eye movements
The invention relates to a device for controlling the eye movements of a living being.

Such devices have been described in particular in
Patents or Patent Applications - rtz 85 12911, FR 06532, US 3 450 466, US 3 869 694,
US 1,579,752, US 4,613,219, WO 91/17705 Ai, and EP 0456166 Ai.

They are based on the work of MMHAITH ("Infrared television recording and measurement of ocular behavior in te human infant", American Psychologist, 1969, 24, 279-283) concerning the illumination of the eye in infrared light, and those by J.MERCHANT and R.MORISSETTE ("Remonte measurement of eye direction allowing subject motion over one cubic foot of space", IEEE Transactions on Biomedical Engineering, 1974, 2, 79-95) concerning the so-called "bright pupil" principle.

Their respective work demonstrated on the one hand that the interior of the eye reflected a part with the infrared illuminating beam, and on the other hand, that if the shooting of said reflected beam was carried out in the axis of the illuminating beam, the image of the pupil appeared clear regardless of the color of the iris.

Finally, MASSÉ demonstrated in 1976 to a tarty of the beam reflected by the cornea orbiting a very bright point, called "corneal point", whose position relative to the image of the pupil varied according to the direction of gaze fixation.

The man in the business has therefore developed a device operating on the principle set out above, in order to follow the eye movements with one eye, sweaters and analyze them.

In these devices, the eye of a living being, who is seated on an active support, is illuminated by a parallel infrared light beam, and observes a stimulation image displayed by a video monitor connected to management means such as 'a microcomputer. The image displayed allows you to test part of the eye activity of the eye, for the specific problems that we want to study.

The beam of infrared light reflected on the eye then contains information concerning the direction of fixation of the eye. I1 is detected by a camera whose optical axis coincides with the axis of the reflected infrared parallel light beam, which camera delivers images of the eye which are digitized by an electronic video card, then stored in a mass memory management resources.

The processing of the images is carried out subsequently by other processing means, such as a microcomputer.

For each image of the eye stored, calculation processing means calculate the coordinates of the center of the corneal point with reference to the coordinates of the center of the pupil.

This leads to data making it possible to know a series of fixing directions, the location and duration of which are determined, and each associated with a stimulation image.

These data are then analyzed, in deferred time, by a specialist who makes a manual report and decides, according to said report, if the subject examined must undergo additional visual examinations.

However, known automated devices do not allow the sequence of visual examinations, since they are generally monotask and monocular. Consequently, it is sometimes necessary to contact several specialists, each capable of carrying out a given visual examination, often manual.

In addition, these automated devices are, for the most part, located in research centers. Clinics and non-specialized circles are almost or totally lacking, which prohibits early and systematic screening in infants under six months, age beyond which the disorders of the visual system are difficult to treat, without risk of disturbing psychomotor development.

Finally, only a few visual examinations, which do not involve the subject examined, are automated or semi-automated. They are generally used by researchers, for ad hoc and non-routine studies.

The object of the present invention is to improve the situation, by providing solutions to these problems.

A first object of the invention is to allow the automated and rapid positioning of the eyes of the living being examined, in the examination position, without external assistance.

A second object of the invention is to authorize the study of both eyes simultaneously.

A third object of the invention is to make available, in the same device, and in fully automated form, all of the known visual examinations.

A fourth object of the invention is to allow the processing and analysis of the data of a visual examination, in real time and simultaneously for both eyes, and to consequently supply an examination report authorizing the chain. - other visual examinations, if necessary.

Finally, a fifth object of the invention is to authorize the optimization of the examination parameters as a function in particular of the age of the subject, and this regardless of his age.

The invention starts from known devices, previously described.

According to a first aspect of the invention, the calculation processing means are capable of calculating, upon receipt of a digitized image, the positions of the center of the pupil and its outline with reference to a template. They also control the step-by-step movement of the support, in order to optimally position at least one eye in the axis of a camera, as a function of a comparison between the positions of the contour of the pupil and those of the template.

Said processing means are capable of interrupting the movement of the support, on which the living being examined is located, when the outline of the pupil is entirely contained in the template.

According to a second aspect of the invention, the mass memory is capable of storing data relating to a plurality of basic visual examinations of variable parameters, and of tests for adjusting the positioning of the eye, each consisting of a selected suite of visual stimulation images.

According to a third aspect of the invention, after positioning the support, the processing means carry out an adjustment test, called "calibration", to calculate a correction factor, specific to each eye, on a correspondence of the corneal point / direction fixation carried out for each image of the eye acquired by the camera. They then correct each response of the eye to a stimulation image according to the correction factor.

According to a fourth aspect of the invention, the processing means further comprise means for recording and analyzing the voice of the living being, as well as a time delay for measuring, if necessary, the time taken by said be alive to respond vocally to an image of stimulation.

According to a fifth aspect of the invention, the support also comprises means for connecting a helmet provided with electrodes, to the processing means, which electrodes slowly collect electrical signals emitted by retinal cells of the living being in the direction of the lobes occipitals of his brain, following the observation, by at least one eye, of a stimulation image. The electrical signals are then correlated to the stimulation images by the processing means.

In a particularly advantageous embodiment of the invention, the device comprises two calibrated infrared light sources and two charge transfer cameras.

Each infrared light source is capable of delivering a reference light beam intended to be reflected on an eye of the living being, then detected by the camera whose axis is coincident with the axis of the reference light beam, which camera provides images of the eye.

In this embodiment, the processing means are able to process, in parallel, and in real time, the images of the eyes provided by the two cameras, and to associate a direction of fixation, for each eye, with a stimulation image. .

Other characteristics and advantages of the invention will appear on examining the detailed description below, and the appended drawings, in which - Figure 1 describes the relationships between the main elements of the device; - Figure 2 illustrates, in detail, the elements connected to the power regulation unit; - Figure 3 illustrates, in three-quarter view, the elements constituting the support; - Figure 4 schematically illustrates the elements of the device allowing the acquisition of images of the eye, as well as the optical paths of the different light beams; - Figure 5 shows schematically the coding, in the form of a logic voltage threshold, of a digitized eye image; FIG. 6 shows the image of the eye displayed by the master video monitor after processing, as well as an example of coding on 16 bits of information relating to this image of the eye; - Figure 7 shows the grid used during the calibration procedure to make a correction in real time of the position of the corneal point; - Figure 8 shows the grid used during the calibration procedure, to make a delayed-time correction of the position of the corneal point; - Figure 9 shows schematically the arrangement of the main elements constituting the device; - Figure 10 is a screenshot of the parameter sheet of a given examination; - Figure 11 shows three copies of screens corresponding to stimulation images displayed on the slave video monitor, and each belonging to a given examination; - Figure 12 is a double graph of the evolution of the coordinates of the corneal point over time; - Figure 13 shows schematically the processing, in real time, by the device, of the data of an examination; and FIG. 14 diagrams the processing, in deferred time, by the device, of the data of an examination.

The attached drawings are, for the most part, certain. Consequently, they form an integral part of the description and can not only serve to supplement it, but also contribute to the definition of the invention, if necessary.

We first refer to Figures 1 and 2.

The purpose of the device is to automatically examine the eye movements of a living being, which requires precise positioning. For this purpose, the device consists of two separate but connected parts.

The first part comprises a support 12 intended to receive, in a preferably seated position, a living being. This part will be described later.

The second part comprises, housed in a frame 1, elements intended either to ensure the management of the device and the visual examinations which it offers, or to process the data delivered by these visual examinations.

Management is carried out by a microcomputer 2 provided with a mass memory 3, capable of storing data relating to all the basic visual examinations offered by the device, as well as adjustment tests. This microcomputer is further connected, on the one hand, to a master video monitor 4, authorizing the display of images and information concerning the management of examinations and, on the other hand, to a microprocessor 5, by a first bus 6 preferably chosen according to ISA or CENTRONIX standards.

The microprocessor 5 is able to process the data delivered by an examination. It is also connected to a first electronic input / output control card 7, thanks to a second input / output bus 8.

This control card 7 allows the microprocessor 5 to interact with all of the elements constituting the device according to the invention.

This control card 7 is also connected in parallel to a second electronic video image digitization card 9, a slave video monitor 10, an electrical power regulation unit 11, and the support 12 for being alive.

Each visual examination consists of a succession of so-called stimulation images intended to be displayed on the screen of the slave video monitor 10.

The digitization card 9 is also connected to two infrared charge transfer cameras (CCD for Coupled
Charge Device) 13-1 and 13-2. Its function is to digitize the images of the eyes of the living being examined, acquired by the two cameras 13-1 and 13-2.

Furthermore, the control unit 11 is further connected to three light sources, including two infrared 14-1 and 14-2 each delivering a reference light beam at 950 nanometers, each intended to illuminate an eye 30-i of the living being examined, and a cold source 15 of neon type intended to illuminate the whole of the room in which the device is installed. This source 15 is adjustable in intensity by the microcomputer 2. The box 11 is also connected to the support 12.

Finally, the control card 7 is also connected to a photodiode 16 to measure the ambient infrared radiation, a rangefinder 17, the diode 18 of which emits at 500 nanometers, to measure the distance between the living being placed on the support, the device, as well as three sensors 18-1 to 18-3 for controlling the position of the support 12 in the three directions (X, Y, Z) defining the space, and a monitoring unit 19 for monitoring the all of the elements described above.

 Ti is now referred to in FIG. 3, which describes in detail the first part of the device, and in particular the support 12. This is the only element of the device not integrated into the frame 1. It comprises on the one hand a folder 20 in which is housed a connection box 21 intended to connect a headset 22 provided with electrodes 23, to the electronic input / output control card 7. On the other hand, it comprises a tray 24 connected to the backrest 20, and resting on a first displacement unit 25-1 allowing its translation parallel to the ground (axis X). This first control box 25-1 is fixed to the end of a leg 26, which is also fixed to a second displacement box 25-2 allowing its translation along a vertical support axis 27 (perpendicular Z axis to the X axis). This axis 27 is fixed on a third displacement unit 25-3 allowing the translation of the support on the ground 28, in a direction Y perpendicular to the directions X and Z.

Furthermore, the connection of the support 12 with the power regulation unit 11 allows the latter to manage the positioning speed of said support 12. Two speeds are provided, a fast and a slow.

Each displacement unit 25 consists of a stepping motor, and an endless screw provided with stops. They are also connected respectively to one of the position sensors 18-1 to 18-3, which supply the microprocessor 5 with data relating to the position of the support 12 relative to the stops of the displacement boxes 25.

Equipped in this way, the support 12 is able to be positioned very precisely in the three directions of space, so that the eyes of the living being seated on the support 12 are optimally placed in the examination position in a eye area 29.

The procedure for positioning the support 12 will be explained later in the description of a complete examination procedure.

Referring now to Figure 4, which describes the device for routing the light beams to the eye area 29, which defines two sub-areas 29-1 and 29-2 intended to materialize the respective location of the left eye 30-1 and that of the right eye 30-2.

To illuminate each eye 30-i, two infrared light sources 14-1 and 14-2 are provided, each consisting of eight light-emitting diodes, emitting a light beam at 950 nanometers.

Each beam is treated on its own channel, the two channels being made up of identical elements.

A source S delivers a beam of infrared light towards a first lens 31-1 which focuses it on a diffuser 32-1, which gives it a diameter of about 10 centimeters. The beam thus diffused is diaphragmed 33-1 then transformed into a parallel beam by a second lens 34-1. The parallel beam can then be filtered by filters 35-1 and 36-1 in order to eliminate the spectral components respectively less than 900 nanometers and greater than 1200 nanometers, which can be dangerous for the eye.

The beam thus treated 37-1 is reflected by a first semi-transparent mirror 38-1, which is oriented so that said beam 37-1 arrives in the sub-area for the eye 29-1 corresponding to it.

Before reflecting on the eye 30-1, it passes through a second semi-transparent mirror 39, common to the two channels, which is oriented so that the slave video monitor 10 can transmit, exactly in the axis of the beam processed from reference 37-1 and of the camera 13-1, the stimulation image which it displays on its screen at the level of the zone for eyes 29. The treated beam 37-1 is then reflected on the stimulated eye 29-1 by the stimulation image, and is responsible for information concerning said stimulated eye 29-1. The beam containing the information 47-1 sets out again in the direction strictly opposite to the treated beam of reference 37-1, crosses the two semi-transparent mirrors 39 and 38-1 then arrives on the camera 13-1 which is located exactly in the beam axis 40-1, which allows it to acquire at least one image of the stimulated eye 29-1.

The same goes for the second treated reference beam 37-2, intended to illuminate the other eye 29-2 (right eye), and detected by the second camera 13-2.

The acquired images are then digitized by the digitization card 9 and then processed by the microprocessor 5.

The latter then transmits to the master video monitor 4 the images of the two stimulated eyes, with reference to templates 41-1 and 41-2 stored in the mass memory 3.

Each image of the eye acquired by a camera 13-i consists of an image of the pupil 41-i and a corneal point 42-i, according to the now well-known principle of corneal reflection.

The images delivered by the 13-i camera include all the gray levels between white and black. After their digitization by the digitization card 9, the images thus digitized are filtered by the microprocessor 5, so as to keep only three levels (at least) unambiguously interpretable, and characterized by their own voltage, as shown diagrammatically in FIG. 5: the white bright (1 volt) for the corneal point 42-i, light gray (0.5 volt) for the pupil 41-i and black (0 volt) for the rest of the field.

The microprocessor 5 can then determine the positions of the contour of the pupil 41-i, which it compares to those of the template 40-i stored in the mass memory 3. When the contour of the pupil 41-i is entirely contained in the template 40-i, the eye 30-i is considered by the microprocessor 5 as being in the examination position, in the sub-area for eyes 29-i. This template 40-i is chosen so that the image of the eye remains included therein despite head movements of +15. Said microprocessor 5 can then calculate the positions of the centers of the pupil 41-i and of the corneal point 42-i, then reference the center of the corneal point 42-i relative to the center of the pupil 41-i previously calculated.

These position calculations are performed by recognizing changes in intensity of logic thresholds in voltage, at each line of analysis of the digitized image. Each change of threshold is coded by three bits, and the coordinates (X, Y) of the place of change are coded respectively on 7 bits and 6 bits, as is shown diagrammatically in FIG. 6. Tables of coordinates (X, Y) of the beginnings and ends of the pupil 41-i and the beginnings and ends of the corneal point 42-i, eliminating the aberrant lines introduced by the movements of the eyelashes or the eyelids.

The image is then retained for further processing if it meets the following requirements: horizontal and vertical diameter of the pupil 41-i sufficiently large, and corneal point 42-i of correct shape.

From the established coordinate tables, the XP and YP coordinates of the center of the pupil are determined over all the horizontal and vertical lines writable in the pupil, by averaging the X and Y coordinates of the midpoints of the horizontal and vertical lines respectively. . An identical calculation is carried out for the coordinates XS, YS of the center of the corneal point 42-i.

Finally, the coordinates XV, YV of the center of the corneal point 42-i with respect to the center of the pupil 41-i are given by
XV = XS - XP and
YV = YS - YP.

A value is assigned to XV and YV, according to predefined criteria, when XP and YP, or XS and YS cannot be calculated, to preserve the possibility of an analysis of the causes of unsatisfactory results.

Knowing the coordinates of a corneal point 42-i with reference to those of the center of the pupil 41-i must be perfectly accurate if one wishes to know precisely the direction of gaze fixation and correlate it to a stimulation image. However, AMSLATER and J. FINDLAY have shown ("The corneal reflexion technique and the visual preference method sources of error", Journal of experimental Child Psychology, 1975, 20, 240-247) that by systematically reporting the position of the corneal point 42-i to that of the center of the pupil 41-i, whatever the direction of the gaze, an error is made due to the exact non-coincidence of the center of rotation of the eye 30-i and the center of the corneal diopter .

In addition, A.BULLINGER and JLKAUFMANN showed ("Technique of recording and analysis of eye movements", Perception, 1977, 6, 345-353) that the error made was all the more important as the direction of the gaze moved away from the central point of the eye 30-i. However, they demonstrated that this problem could be solved by performing a calibration on a limited number of test points.

This calibration is carried out before starting a basic visual examination, then restarted before any new examination to eliminate the risk of significant movements of the head between examinations. It is performed for both eyes simultaneously. The objective of the calibration is to calculate a correction factor on the preliminary reading of the coordinates of the corneal point 42-i.

The calibration procedure used by the device is carried out by means of a positioning adjustment test stored in the mass memory 3. It consists of presenting successively to the living being, on the slave monitor 10, five stimulation images generally, or only two in the case of a newborn, to lighten the calibration phase in the latter case. On each image is placed a single point of reference, the five points materializing the center and the four corners of a grid limiting a visual field of +15. Each image remains displayed on the screen of the slave video monitor 10 for one second.

Two levels of correction are authorized depending on whether the analysis of the visual examination is carried out in real time or in deferred time.

The real-time correction refers to the diagram in FIG. 7, where the calibration grid is a square materialized by a grid (X, Y) of 5 "pitch. Two vectors VO and V1 are also represented.

VO = (V0x, V0y) is the vector obtained when the living being fixes the central point.

 V1 = (Vlx, Vly) is the vector obtained when the living being fixes any point on the screen of the slave monitor 10.

The corneal point therefore presents a drift whose coefficients are Cx = VOx and Cy = V0y.

Under these conditions, the corrected positions in X and Y with respect to the center are given respectively by
(Vix (Vlx - V0x)
x
= (Vly - VGy)
Cy
The correction in deferred time refers to FIG. 8, which is valid for the right eye 30-2. The objective of this second correction is to provide a very precise correction factor when reprocessing the archived data, either in mass memory 3, or on a magnetic tape or a removable floppy disk, and without the constraint of the time of appearance of an image (of the order of 40 milliseconds).

We use the calibration performed before the exam and used for real-time correction, but the grid used is no longer a square but a trapezoid, for which only the step in X is considered to be constant. The position of the gaze in X is therefore given by the previous case:
(V1X - VOx)
V # 1X =
Cx
On the other hand, the position of the gaze in Y is a function of the two directions X and Y, as indicated below. We consider the line d passing through points A and B with respective coordinates (Ax, Ay) and (Bx, By). This has the equation y = ax + b, where:
(By-Ay) (By + Ay) a
a = and b = + x (Bx + Ax)
(Bx-Ax) 2 2
Let V be the uncorrected corneal vector, with coordinates
Vx = Vlx - VOx,
Vy = Vly - V0y, the ordinate of a point E with coordinates (Ex, Ey) belonging to the line d corresponding to the abscissa of V is given by
Ey = (ax Ex) + b with Ex = Vx.

The corrected corneal vector V 'with coordinates (V'x, V'y) is then obtained by a simple rule of three, since the ordinate found is worth +15. So we have Vx XX = 'yjC RX =
y
In the case of an uncorrected corneal vector of ordinate Vy negative, the same reasoning is used, but with the straight line passing through points C and D.

Furthermore, the calibration of the left eye 30-1 is obtained by using the symmetry of the trapezoid of FIG. 6, relative to the vertical axis AB.

Reference is now made to FIG. 9.

At the start of an examination, the operator installs the living being on the support 12, which has been automatically positioned in a reference position at the end of the previous examination. Then, he settles down on his seat 43 and begins to dialogue with the microcomputer 2 thanks to a keyboard with keys 44. The management of the device is carried out under "windows" environment. This environment was chosen for its user-friendliness and for its very widespread use.

However, under windows, it is not yet possible to work in 32-bit time sharing, or multitasking. However, the device is designed for real-time processing of the images of the eyes 30 acquired via the cameras 13. This requires displaying a stimulation image on the screen of the slave video monitor 10, acquiring the images of the two eyes 30 lit by the two treated infrared reference beams 37-1 and 37-2, to process these images in order to know the direction in which the gaze is fixed, to correlate these images to the stimulation image, while displaying the images of the two eyes 30-1 and 30-2 inside the two templates 40-1 and 40-2 on the screen of the master video monitor 4.

All these operations must be carried out in less than 40 ms, or 25 frames per second, in order to perfectly follow the eye movements.

These operations are made possible thanks to the development by the Applicant of control software which short-circuits the clock specific to the Windows environment, and replaces it with its own clock.

To ensure the management of such a device, the Applicant has chosen a microcomputer 2, PC compatible, provided with an i486DX processor operating at least at 33MHz, and assisted by a microprocessor 5, at least to ensure the calculation part. This microprocessor 5 will for example be that manufactured by the company SIEMENS under the reference No. "166. In addition, to ensure the simultaneous processing and display of the images of the two eyes 30 on the screen of the master video monitor 4, the Applicant has chosen to couple the digitization card 9 to a VESA BUS type memory bus interface.

The device informed of the start of the examination energizes its elements. The two infrared light sources 14-1 and 14-2 then each deliver a reference light beam, which is processed by the optical elements described above, then reflected towards a sub-area for the eye 29-i , by the semi-transparent mirror 38-i associated with its optical path. The screen of the master video monitor 4 delivers 40, the eyes 30 of the living being are in the optimal position and the examination can begin. On the other hand, if this is not the case, the microprocessor 5 calculates by how many steps it will have to move each displacement unit 25-i so that the two want 30 are opposite the eye areas 29. In case of If this fails, this comparison procedure begins again. If necessary, the microcomputer 2 uses the range finder 17 to know precisely the distance between the eyes 30 of the living being and said device.

When the living being is correctly positioned, the microcomputer 2 controls the photodiode 16 to measure the infrared light intensity reflected on the face of the living being. The photodiode 16 is connected to an electronic circuit which can trigger an alarm at the level of the master video monitor 4, if a chosen intensity threshold is exceeded. If this threshold is exceeded, the microcomputer 2 is able to command the control unit 11 to adjust the intensity of the infrared light sources 14, so that this intensity is less than said threshold.

Then, the microcomputer 2 controls the calibration procedure described above, then offers the operator on the screen of the master video monitor 4 a choice of basic visual examinations from which the operator can select an examination adapted to the living being examines.

These examinations constitute a database which can be supplemented later by loading a floppy disk, or a magnetic tape, into the microcomputer 2, which is provided for this purpose with at least three slots on the front allowing accept a 5 inch floppy disk, a 3.5 inch, or 5 inch hard drive, and a 3.5 inch external format magnetic cartridge reader / writer.

The device, equipped with its database, managed in a Windows environment, therefore allows the selection by bar of icons of a particular examination adapted to the living being examined, whatever it is, since each examination is configured. Thus, before launching a given visual examination, the microcomputer 2 displays on the screen of the master video monitor 4 the menu presenting the adjustable parameters of said visual examination. An example of a menu for adjusting the parameters of a given examination is shown in the screenshot of FIG. 10. Depending on the age of the living being, or the visual problem which concerns it, we will choose, by example, consider either a particular eye, or both. Then, we will adjust the size of the objects to be displayed on the screen of the slave video monitor read, as well as their speed and their direction of travel.

Three stimulation images, extracted from basic visual examinations, intended to test the peripheral detection of the eyes, are given by way of example in FIG. 11.

The library of visual examinations, which constitutes the database, offers the following examinations - peripheral detection, - acuity scale 1, - acuity scale 2, - visual preferences, - pursuit, - jerks, - Thibaudet, - Optokinetic nystagmus, - visual evoked potential (EPI).

The proposed examinations can be classified into three groups of methods, which relate to visual activity or oculomotility.

The first group concerns behavioral methods.

Many of them have already been automated, but for the study of a single eye, which prohibits binocular examinations. They mainly include the preferential gaze technique, which is particularly suitable for children under the age of four, and the so-called visual target tracking technique.

These two techniques are obviously proposed by the device in monocular or binocular examination.

The second group concerns the subjective methods, which call for an active collaboration of the living being examined. In children, they are generally usable after 30 months. These methods are not subject to automation, since they generally require either a voice response or a manual indication on a twin image of that displayed on the slave video monitor 10.

To this end, the device is provided with a voice module composed of a voice recorder 45 coupled to an electronic voice synthesis card 46 which can recognize a certain number of keywords with reference to the stimulation images displayed on the screen of the slave video monitor 10. Furthermore, the device is able to measure the reaction time, called "integration time", which elapses between the display of a stimulation image and a voice response emitted by the being alive examined following the observation of said image.

The complete automation of this type of examination therefore has a very important advantage, since it gives access to the visual integration time of an image.

For example, in a MONOYER type examination, the operator asks the living being examined to read aloud the characters displayed on the screen of the slave monitor 10. The voice module will transcribe the words or letters sent by the living being examined, and compare them to the content of the displayed image.

The analysis of the examination is carried out character after character, image of stimulation after image of stimulation, by a true / false comparison. Depending on the examination parameters chosen by the operator and the number of correct answers, the microcomputer 2 then assigns an examination mark.

The third group concerns objective methods. They bring together the techniques of visual evoked potential (EPI) and the electroretinogram.

The visual evoked potential is an electrical response of the cortex to light stimulation. The excitation of the retinal cells causes a discharge of action potentials, which are transmitted through the optical pathways to the occipital cortex of the living being examined. These signals are then collected by the electrodes 23 placed on the scalp, facing the two occipital lobes. In the present invention, the electrodes 23 are fixed inside the adjustable helmet 22, and connected to the connection box 21 housed in the backrest 20 of the support 12. The connection box is also connected to an acquisition circuit and digital processing of complex and low amplitude analog signals, which circuit is coupled to a library of mathematical programs, such as Fourier transforms, stored in mass memory 3.

However, the device will only provide a statement of brain activity induced by stimulation images, not an analysis of brain activity. This, as well as the report, must be carried out at the end of the examination by the operator.

The visual evoked potential is a low voltage signal masked by the electroencephalogram, which constitutes background noise. It is therefore necessary to extract the signal of the visual evoked potential from this background noise.

The electroretinogram represents the overall activity potential of retinal tissues induced by light stimulation. This technique requires prior pupil dilation as well as corneal anesthesia. As with the techniques of visual evoked potential, the living being examined is equipped with a helmet 22 provided with electrodes 23. The stimulation is carried out using a specific display of stimulation images on the screen of the video monitor. slave 10. This provides stimulation in full field which illuminates the retina evenly.

We now refer to Figures 12 and 13.

When the choice of the parameters of the visual examination is made, the microcomputer 2 controls the display, on the screen of the slave video monitor 10, of the first image of stimulation of said examination. The camera or cameras 13-i have the time to acquire several images of the eye or eyes 30-i examined. The procedure for acquiring each image and calculating the coordinates of the associated corneal point 42-i, with reference to the center of the pupil 41-i, have been described previously. It then remains at the microcomputer 2, in the event of real-time processing, to correct the coordinates of the corneal points 42-i with the correction factor calculated during calibration, and to deduce from these corrected coordinates the direction of fixation of the gaze, common to the images of the eye or the eyes, in order to correlate them with the image of stimulation displayed.

The procedure is repeated for all stimulation images of the visual examination carried out. The directions of fixation measured are then grouped together on a graph, like that of FIG. 12. This makes it possible to follow over time the movement of the eyes 30 in the two directions X and
Y defining the slave video monitor screen 10.

Three pieces of information can be extracted from these graphs - the gaze adjustment microsaccades, which can be linked to microdisplacements either of the gaze or of the living being examined, or else to the imprecision in the calculation of the coordinates of the corneal point 42-i; - jerks, which attest to the change in direction of gaze fixation in the visual field; and - the "fixations" which characterize the periods during which the gaze remained in the same direction.

The microprocessor 5 is provided with a filter which allows it to keep, for analysis, only the saccades and the fixings.

Theoretically, as soon as a stimulating image differs in content from the previous image, the direction of gaze fixation must change. Knowing the durations of the fixations and the positions over time of the saccades, the microprocessor 5 is able to correlate this information with the positions of the patterns of the stimulation images successively displayed. The microcomputer 2 is then able to deliver a report of the visual examination carried out, after weighting of the results by a coefficient depending on the examination parameters initially chosen, with reference to a table of correspondence examination parameters / coefficient , stored in mass memory 3. This report is displayed on the screen of the master video monitor 4. It can also be printed by a printer 47 connected to the microcomputer 2, on the operator's order, who can select this option in a menu managed by the windows environment, at the end of the exam.

The data of the examination and the report are also stored in mass memory 3. This data can also be transferred either on diskette or on magnetic tape, after selection of options in a menu managed by the Windows environment. (Registered trademark), at the end of the exam.

Reference is now made to FIG. 14.

In case of deferred processing, the deferred time calibration procedure will be applied directly to the coordinates of the corneal points 42-i, then data processing, identical to that performed in real time, will be done.

The processing of data, in deferred time, is automatic when the visual examination calls for eye movement monitoring coupled with vocal participation by the living being examined. This will be the case, for example, for Monoyer type exams.

For this type of examination, deferred processing is made compulsory because the microcomputer 2 cannot manage at the same time (less than 40 ms), both the voice synthesis, the voice response times, and the determination of the fixation directions, and correlate this information to stimulation images.

Whatever the type of treatment, at the end of a first visual examination, the operator may, depending on the results of this first examination, decide to carry out one or more other visual examinations. If he decides to carry out a second examination, he must then repeat the entire procedure described above, including the calibration, but nevertheless without affecting the positioning of the support 12.

Otherwise, when the operator decides that the living being has undergone enough exams, he selects the "end of exam" option in a menu managed by the windows environment.

The microcomputer 2, informed of this decision, will order the repositioning of the support 12 in the reference position. It will then order the regulation unit 11 to switch off the three light sources 14 and 15.

The device has many advantages compared to current devices - the positioning of the living being examined is carried out completely automatically, in an environment that is not very restrictive and user-friendly; - all known visual examinations necessary for screening for visual pathologies are automated; - the parameterization of the exams makes it possible to adapt them to any person, whatever their age, including the infant; - the two ways of acquiring eye images allow binocular examinations to be carried out; - the analysis and processing of the data is carried out in real time, or immediately after the end of the examination, which authorizes the rapid sequence of different visual examinations; and - the device delivers at the end of the examination a report as a function of the examination parameters chosen by the operator at the start of said examination, in all objectivity.

Reference has been made in the description to a device for controlling the positioning of the support by electrical signal flowing in cables. Any remote order transmission device known to those skilled in the art can be used in place of the device described.

As visual examinations must be carried out in places with controlled lighting, provision may be made to manage their lighting, directly at the level of the microcomputer.

If the current examinations are planned to test men, whatever their age, it is clear that they can be adapted to the study of other living beings, such as for example primates.

The system cannot be limited to the prevention of visual pathologies. It can be used in many situations, such as in functional exploration centers, where in occupational health centers to determine if a given job corresponds to a given person, or even in expert offices for to detect possible post-accident simulations, but also for an eye command intended for example for piloting objects in real time.

Claims (15)

Claims 1. Eye control device of the type comprising - a support (12) for being alive, mobile on command in at least one direction, and defining an area for the eyes (29) of said living being, - a frame (1), in which are installed in a fixed position, at least one light source (14), capable of delivering at the level of the eye area (29) a beam of light having selected optical properties (37-i), as well as at least a camera (13-i) capable of acquiring an image resulting from the reflection of said light beam (37-i) on an eye (30-i) of the living being, which image comprises the outline of the pupil (41-i ) and a "corneal" point (42-i) representing the cornea, and - processing means able to digitize the image or images of the eyes provided by the camera (s) (13-i), and comprising, in connection, management processing means able to store in a mass memory (3) a two-dimensional grid containing a template (40-i), as well as m computational processing means capable of processing at least one digitized image, by calculating the coordinates of the center of the pupil (41-i), and that of the center of the corneal point (42-i) with reference to said center of the pupil (41 -i), characterized in that the calculation processing means are capable of calculating, upon receipt of a digitized image, the coordinates of the center of the pupil (41-i) and its outline with reference to the template (40- i), as well as controlling the movement of the support (12), in order to position exactly at least one eye (30-i) in the axis of a camera (13-i), as a function of a comparison between the coordinates of the outline of the pupil (41-i) and those of the template (40-i). 2. Device according to claim 1, characterized in that the processing means are capable of interrupting said movement of said support (12) when the contour of the pupil (41-i) is fully contained in the template (40-i). 3. Device according to one of claims 1 and 2, characterized in that said mass memory (3) is also suitable for storing data relating to a plurality of basic visual examinations of variable parameters, and adjustment tests eye positioning, each consisting of a selected suite of visual stimulation images. 4. Device according to claim 3, characterized in that said stimulation images are displayed by a "slave" video monitor (10), connected to the processing means, then they are transmitted towards the eyes (30) by a semi mirror transparent (39) whose orientation makes it possible to confuse the axis of the image reflected on said mirror (39), and the axis of the camera (13-i). 5. Device according to claim 1, characterized in that the frame (1) further comprises a cold light source (15) of adjustable intensity, for lighting a room in which said device is installed. 6. Device according to one of claims 1 and 2, characterized in that the calculation processing means are capable of establishing a correspondence between a succession of identical coordinates of the center of the corneal point (42-i), and a direction of fixation of the eye (30-i), as a function of a correspondence table coordinates of the corneal point / fixation direction, stored in the mass memory (3), said fixation direction being associated with a stimulation image, and constituting a response.  Device according to one of the preceding claims, characterized in that the processing means are further capable of carrying out, after positioning the support (12), an adjustment test, called "calibration", to calculate a correction factor, proper to each eye (30-i), on the corresponding coordinates of the corneal point / direction of fixation performed for each image of the eye acquired by the camera (13-i), as well as to correct each response as a function of said factor correction. 8. Device according to one of the preceding claims, characterized in that the processing means weight each response to a stimulation image of an examination of given parameters, by a coefficient as a function of an examination parameters correspondence table / coefficient, and deliver an examination report as a function of the corrected and weighted responses associated with each stimulation image of said examination, which report is stored in the mass memory (3) and / or printed on paper by a printer (47) connected to said processing means. 9. Device according to one of the preceding claims, characterized in that the processing means further comprise a "master" video monitor (4), fixed in the frame (1), to display, as desired, at least the image d 'an eye (30-i) acquired by a camera (13-i), the parameters specific to a visual examination, or the report of a visual examination. 10. Device according to one of the preceding claims, characterized in that the processing means further comprise recording (45) and analysis (46) means of the voice of the living being able to measure, if necessary, a time taken by said living being to respond vocally to a stimulation image. 11. Device according to one of claims 1 to 9, characterized in that the support (12) further comprises means for connecting a helmet (22) provided with electrodes (23) to the processing means. 12. Device according to claim 11, characterized in that the electrodes (23) are capable of collecting electrical signals emitted by retinal cells of the living being, in the direction of the occipital lobes of their brain, following observation, by at least one eye (30-i), of a stimulation image, and in that the processing means are also able to correlate said electrical signals with the stimulation image concerned. 13. Device according to one of the preceding claims, characterized in that the processing means also comprise - a power regulation box (11), capable of regulating, on the one hand, the intensity of the light sources (14 , 15) and on the other hand the speed of movement of said support (12), and - a sensor (16) for measuring, after positioning of the support, the intensity of the reference light beam (47-1) reflected on the face of the living being, which sensor (16) is connected to an electronic circuit capable of triggering an alarm at the level of the master monitor (4), if a chosen intensity threshold is exceeded. 14. Device according to one of the preceding claims, of the type comprising two calibrated infrared light sources (14-1 and 14-2) and two charge coupled cameras (13-1 and 13-2), characterized in that each infrared light source (14-1 and 14-2) is capable of delivering a reference infrared light beam intended to be reflected on an eye (30-i) of said living being, then detected by the camera (13-i ) whose axis coincides with the axis of the reference light beam, and in that the processing means are capable of processing in parallel and in real time the images of the eyes provided by the two cameras (13), and to associate a direction of fixation, for each eye (30-i), with a stimulation image. 15. Device according to one of the preceding claims, characterized in that the management processing means comprise - a microcomputer (2) capable of managing the adjustment of the positioning of the support (12) and visual examinations, - a card input / output management (7) to authorize the dialogue between the elements constituting said device, - a digitization card (9), to digitize the images supplied by the cameras (13), and in that the processing means of calculation include a microprocessor (5) able to - calculate the coordinates of the contour and the center of the pupil (41-i), as well as the coordinates of the center of the corneal point (42-i) with reference to said center of the pupil (41 -i), - control the positioning of the support (12) as a function of the comparison between the coordinates of the contour of the pupil (41-i) and those of one of the templates (40-i) contained in the two-dimensional grid. - deducing from said coordinates of the fixing directions, - correcting and weighting the fixing directions as a function respectively of the calibration and of the examination parameters, and - correlating the vocal responses or the electrical signals to stimulation images. 16. Device according to one of the preceding claims, characterized in that the slave video monitor (10) and the cold light source (15) are housed in the frame (1).