System and associated method for determining subjective values ​​of optical characteristics of at least one corrective lens adapted for an eye of a subject

Abstract

The invention relates to a system for determining the value of a measure associated with subjective values ​​of optical characteristics of at least one corrective lens adapted for an eye of a subject by carrying out a subjective test in which the subject is asked to describe his visual performance under a plurality of real optical conditions by selecting a descriptive answer from a set of possible descriptive answers, the system comprising a computer having one or more memories and one or more processors, wherein: an initial model giving the likelihood of each descriptive answer in the set of descriptive answers for each of a plurality of theoretical optical conditions is stored in one of the one or more memories of the computer, the one or more processors of the computer being programmed to: a) collect (500) the results of the subjective test, the subjective test being carried out in the following manner: the subject is asked to describe his visual performance under each of the plurality of real optical conditions by selecting a descriptive answer from the set of possible descriptive answers for each real optical condition, the results comprising each selected descriptive answer and the corresponding real optical condition, b) determine (800, 900) the value of the measure by taking into account the initial model and each collected descriptive answer.

Description

Technical Field

[0001] The present invention relates to a system and a method for determining subjective values ​​of optical characteristics of at least one corrective lens adapted for an eye of a subject. Background Art

[0002] In order to provide a visual device adapted to a subject's visual impairment, it is necessary to estimate this visual impairment.

[0003] A first possibility is to measure objective values ​​of optical characteristics of the subject's eye.

[0004] It is also known to determine the subjective values ​​of the optical characteristics of the corrective lenses desired by a subject by means of subjective tests.

[0005] In practice, during a subjective test protocol, the subject's vision is tested using a phoropter that allows the successive placement in front of the subject's eye of lenses having different values ​​of the optical characteristics.

[0006] This testing protocol can be implemented using a classic phoropter or a phoropter with two variable-power compound lenses.

[0007] In a classic phoropter, different lenses with fixed, predetermined focal powers are placed successively in front of each eye of a subject. For example, lenses with different spherical powers are placed successively in front of one eye of a subject during successive trials. From one trial to the next, the spherical power is increased by a predetermined step size. This step size is typically 0.25 diopters (D) or 0.125 D.

[0008] The subjective test protocol can also use a modified phoropter including variable power lenses. Such phoropter / variable lenses are described, for example, in the following documents: US 20160331226, US 2017027435 or WO 2017 / 021663.

[0009] On each trial of the subjective test, the subject is asked to express a visual assessment corresponding to the current lens placed in front of his eye, indicating his preferred visual state between the two presented to him, or whether he is unable to decide between the two. For example, the two visual states may correspond to his vision of two different images through the current lens, or may correspond to his vision through the previous lens and his vision through the current lens.

[0010] In practice, each trial of the subjective test protocol may correspond to an assessment given by a subject, for example, during a two-color test. During this two-color test, the subject is presented with an image comprising an optotype displayed on a red background on one side and an optotype displayed on a green background on the other side. If the subject has better vision with the current lens for the optotype on the red background, the spherical power should be reduced in the next presented lens; if the subject has better vision with the current lens for the optotype displayed on the green background, the spherical power should be increased in the next presented lens.

[0011] As a variant, on each trial, with the current lens placed in front of his eye, the subject is asked to assess the quality of his vision through the current lens compared to the previous lens: he is asked whether the current lens provides better or worse vision than the previous lens, or whether he is unable to decide between the two. In the latter case, both lenses provide the subject with similar visual quality.

[0012] Therefore, prior art subjective testing relies on a predetermined, standardized protocol. This protocol may use the subject's characteristics as input for the first trial of the protocol, for example to determine the optical characteristics of the test lens initially placed before the subject. Otherwise, no further customization is provided.

[0013] Therefore, the protocol may take a long time to perform or be inaccurate in the case of certain subjects, such as children. Summary of the Invention

[0014] It is therefore an object of the present invention to provide a system for determining a subjective value of a measure associated with a subjective value of an optical characteristic of at least one corrective lens adapted for an eye of a subject, thereby allowing a more efficient testing protocol, reducing the time required to perform the subjective test and improving the reliability of the subjective test.

[0015] Reliable means that the subjective value of the optical characteristic is accurate and repeatable.

[0016] According to the present invention, the above objects are achieved by providing a system for determining a measure value associated with a subjective value of an optical characteristic of at least one corrective lens adapted to an eye of a subject by carrying out a subjective test in which the subject is asked to describe his visual performance under a plurality of real optical conditions by selecting a descriptive answer from a set of possible descriptive answers, the system comprising a computer having one or more memories and one or more processors, wherein:

[0017] an initial model giving a likelihood of each descriptive answer in a set of descriptive answers for each theoretical optical condition in a plurality of theoretical optical conditions is stored in one of the one or more memories of the computer,

[0018] - the one or more processors of the computer are programmed to:

[0019] a) collecting results of the subjective test, the subjective test being performed by asking the subject to describe his visual performance under each of the plurality of real optical conditions by selecting, for each real optical condition, one descriptive answer from the set of possible descriptive answers, the results comprising each selected descriptive answer and the corresponding real optical condition,

[0020] b) Determining the value of the measure by considering the initial model and each of the descriptive responses collected.

[0021] The measure associated with the subjective value of the optical characteristics of at least one corrective lens adapted to the eye of the subject may be any kind of measure and in particular:

[0022] - the value of the optical characteristic itself,

[0023] a statistically determined measure based on a plurality of values ​​of the optical characteristic (such as an average value, a weighted average value, a standard deviation),

[0024] - an indication of a confidence level associated with the measured value of the optical characteristic.

[0025] In step b), the initial model is taken into account to determine the value: this is the case as long as at least one trial of the subjective test is performed using the initial model. For example, the initial model can be used to determine a modified model used in further trials to determine parameter values ​​of the subjective test, such as the step value for increasing / decreasing an optical characteristic of an optical component placed in front of the subject's eye, and / or to directly determine the value of the measure.

[0026] Thanks to the system according to the present invention, the determination of the metric takes into account the initial model, which gives the probability of each descriptive answer in the set of descriptive answers for each of a plurality of theoretical optical conditions. By comparing the actual optical conditions with the theoretical conditions, the most likely answer of the subject under the actual optical conditions can be determined. Furthermore, based on this most likely answer, the subjective testing protocol can be modified to be more efficient, thereby reducing the time required to perform the subjective test and increasing its reliability.

[0027] The subjective test can also be adapted to the subject by determining the initial model based on the subject's personal characteristics.

[0028] Other advantageous, non-limiting features of the system according to the invention are listed below:

[0029] - in step b), the computer is programmed to compare the initial model with at least one descriptive answer collected and to determine the value of the measure based on this comparison; in particular, the computer is programmed to compare the initial model with each descriptive answer collected and to determine the value of the measure based on this comparison;

[0030] - the initial model comprises at least one of the following: an initial model curve, an initial model formula, and an initial model data set;

[0031] - the initial model comprises an initial model curve, and the computer is programmed to perform the following sub-steps in step b):

[0032] - plotting at least one of the collected descriptive responses selected by the subject against the corresponding real optical conditions and superimposing said initial model curve on the thus obtained graph such that the initial model curve fits the graph,

[0033] - determining said value of said measure taking into account the relative positions of said initial model curve and said graph;

[0034] - the initial model comprises an initial data set, and the computer is programmed to perform the following sub-steps in step b):

[0035] - statistically processing the initial data set and the collected descriptive responses, and

[0036] - taking this statistical processing into account to determine said value of said measure;

[0037] - the computer is programmed to perform the following sub-steps in step b):

[0038] - modifying the initial model taking into account at least one descriptive answer of the subject and the corresponding real optical conditions to obtain a modified model, and

[0039] - comparing the modified model with the collected descriptive answers and determining the value of the measure based on this comparison; in particular, modifying the initial model taking into account each descriptive answer of the subject and the corresponding real optical condition to obtain a modified model, and comparing the modified model with each collected descriptive answer and determining the value of the measure based on this comparison;

[0040] - the modified model comprises at least one of: a modified model curve, a modified formula, and a modified model data set;

[0041] - the modified model comprises a modified model curve and the computer is programmed to perform the following sub-steps in step b):

[0042] - plotting at least one, possibly each, collected descriptive answer selected by the subject against the corresponding real optical conditions and superimposing said modified model curve on the thus obtained graph such that the modified model curve fits the graph,

[0043] - determining said value of said measure taking into account the relative positions of said modified model curve and said graph;

[0044] - the modified model comprises an initial data set, and the computer is programmed to perform the following sub-steps in step b):

[0045] - statistically processing the modified data set and the collected descriptive responses, and

[0046] - taking this statistical processing into account to determine said value of said measure;

[0047] - the initial model takes into account one or more personal characteristics of the subject;

[0048] - the personal characteristic comprises at least one value of:

[0049] - Social parameters: such as age, gender, geographical origin, eye care history,

[0050] -Morphological parameters: such as pupil size, pupil distance,

[0051] - Optical parameters: such as type and / or value of refractive error, astigmatism, eye-lens distance,

[0052] - Setting parameters: such as the starting point of the subjective test, the type of stimulus, the distance to test distance or near vision, the distance between the phoropter and the eye,

[0053] - visual parameters: such as acuity, previous answers to subjective tests, monocular dominance, binocular vision, eye movements,

[0054] - behavioral parameters: such as: rapidity of previous answers to subjective tests, sensitivity to changes in at least one optical characteristic of at least one ophthalmic lens;

[0055] - the computer is further programmed to determine at least one initial parameter value of the subjective test based on personal characteristics of the subject;

[0056] - said computer being programmed to determine said at least one initial parameter value of the subjective test taking into account said initial model;

[0057] - said optical characteristics of the corrective lens suitable for the subject include at least one of the following:

[0058] - spherical power in the distance and / or near vision zones,

[0059] - cylindrical power and / or axis position in the distance and / or near vision zones or a combination thereof,

[0060] - prismatic power and / or axis position or a combination thereof in the distance and / or near vision zones,

[0061] - filter transmittance in the distance and / or near vision zones;

[0062] - The descriptive answer set includes the following answers:

[0063] -Better visual performance in first real optical conditions,

[0064] - better visual performance in second real optical conditions,

[0065] - no perceived difference between the first real optical condition and the second real optical condition;

[0066] - said initial model is statistically determined based on a reference data set previously collected when performing said subjective test on a plurality of reference subjects, said reference data set comprising:

[0067] - individual characteristics of each reference subject,

[0068] - Actual optical conditions and corresponding responses of each reference subject;

[0069] - the initial model is determined based on the average value or weighted average value of this reference data set;

[0070] - the initial model is determined based on an average or weighted average of a portion of the reference data set selected based on one or more personal characteristics of the reference subject and / or the initial model is determined using a machine learning algorithm and / or a neural network algorithm;

[0071] - a reference answer graph is obtained by plotting the answer of each reference subject against the corresponding real optical condition, and the initial model is determined based on distribution characteristics of the reference answer graph;

[0072] - the plurality of reference subjects are selected from a group of reference subjects having one or more personal characteristics that are the same as or similar to the corresponding personal characteristics of the subject;

[0073] - said computer being programmed to carry out said subjective tests and to determine, taking into account said initial model, said real optical conditions for subsequently carrying out said subjective tests;

[0074] - said computer being programmed to carry out said subjective tests and to determine, taking into account said modified model, said real optical conditions for subsequently carrying out said subjective tests;

[0075] - said computer being programmed to determine said value of said measure based on the position of the initial or modified model curve relative to the graph when said position is determined after a predetermined number of trials or when this position has not been modified by more than a predetermined amount between two consecutive trials of a subjective test;

[0076] - said computer being programmed to compare each collected descriptive answer corresponding to a real optical condition with said initial or modified model, and when this comparison shows that the difference between a particular collected descriptive answer corresponding to a particular real optical condition and the initial, respectively modified model exceeds a predetermined threshold, this particular collected descriptive answer is identified as inconsistent;

[0077] - When the number of inconsistent responses exceeds a predetermined threshold, modifying the initial or modified model to reduce the number of inconsistent responses.

[0078] The invention also relates to a method for determining a measurement value associated with a subjective value of an optical characteristic of at least one corrective lens adapted to an eye of a subject by means of a subjective test in which the subject is asked to describe his visual performance under a plurality of real optical conditions by selecting a descriptive answer from a set of possible descriptive answers, the method comprising the following steps:

[0079] i) providing an initial model giving a likelihood of each descriptive answer in a set of descriptive answers for each theoretical optical condition in a plurality of theoretical optical conditions,

[0080] j) performing said subjective test by asking the subject to describe his visual performance under each of said plurality of real optical conditions by selecting a descriptive answer from said set of possible descriptive answers for each real optical condition,

[0081] k) collecting each descriptive answer of the subject and the corresponding actual optical condition,

[0082] 1) Determining the value of the measure by considering the initial model and each descriptive answer collected.

[0083] This method is typically implemented by a system according to the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] The following description, with reference to the accompanying drawings, will make clear what the present invention encompasses and how it is accomplished. The present invention is not limited to the embodiment(s) shown in the accompanying drawings. Accordingly, it should be understood that where features mentioned in the claims are followed by reference numerals, the inclusion of such reference numerals is solely for the purpose of enhancing the intelligibility of the claims and is in no way intended to limit the scope of the claims.

[0085] In the attached figure:

[0086] - Figure 1 is a schematic block diagram representation of an embodiment of the method according to the invention using the apparatus according to the invention,

[0087] - Figure 2 is a schematic diagram of the responses (squares) of a plurality of subjects obtained during the respective subjective tests plotted against the theoretical optical conditions associated with the spherical power of the lens placed in front of each subject's eye in each respective trial of the subjective test based on the actual refractive correction of each subject, in comparison with the graphs (solid and dashed lines) of three different initial models determined on the basis of these data (as will be described below, the responses represented by the value "0" correspond to responses in which the subject did not distinguish between the two real optical conditions tested);

[0088] - Figures 3 to 7 is a graphic representation of the subjects' responses (squares) obtained in successive trials of the subjective test corresponding to real optical conditions (here the spherical power of the test lens in diopters), fitted by the initial model curve (full line). DETAILED DESCRIPTION

[0089] The invention belongs to the field of conception and manufacture of visual devices adapted to the optical characteristics of at least one eye of a subject in order to improve the subject's vision by compensating for the visual defects of this eye.

[0090] Figure 1 A schematic representation of the steps of an embodiment of a method according to the invention for determining a measurement value associated with a subjective value of an optical characteristic of at least one corrective lens adapted for an eye of a subject is shown.

[0091] The values ​​of the metrics are determined by subjective testing.

[0092] The measure associated with the subjective value of the optical characteristics of at least one corrective lens adapted to the eye of the subject may in particular be:

[0093] - said subjective value of the optical characteristic itself,

[0094] - a statistically determined measure based on a plurality of subjective values ​​of the optical characteristic (such as mean value, weighted mean value, standard deviation),

[0095] - an indicator of a confidence level associated with said subjective value of the optical characteristic.

[0096] The optical characteristics of the corrective lenses can be any optical characteristics that can be used to compensate for any kind of visual impairment in a subject.

[0097] In particular, the optical characteristics of the corrective lens suitable for the subject include at least one of the following:

[0098] - the spherical power of the lens in the distance and / or near vision zones,

[0099] - Add light from below,

[0100] - cylindrical power and / or axis position or a combination thereof in the distance vision zone and / or near vision zone of the lens,

[0101] - the prismatic power and / or axis position or a combination thereof in the distance and / or near vision zones of the lens,

[0102] - the filter transmission in the distance vision zone and / or the near vision zone of the lens.

[0103] The spherical power corresponds to the refractive power of the lens given by the spherical component of the shape of the front and back of the lens, measured in diopters.

[0104] The cylindrical power corresponds to the optical power of the lens given by the cylindrical component of the shape of the front and back faces of the lens, expressed in diopters.

[0105] The axial position corresponds to the orientation of the cylindrical component of the lens shape.

[0106] Cylinder power and axis position describe the cylindrical component of an ophthalmic lens determined to compensate for the astigmatism of a subject's eye. Instead of being determined in the standard polar coordinate form with a measure and orientation decomposition, a vector decomposition of the cylindrical component as described in the applicant's document PCT / EP2018 / 061207 can be used.

[0107] The vector decomposition of the cylindrical components is done according to two orthogonal directions J0, J45.

[0108] For example, the decomposition along these two J0, J45 directions corresponds to replacing the classic sphero-cylindrical notation (spherical power S, cylindrical power C, axis position) with a triplet of orthogonal values ​​(M, J0, J45), which is defined as the equivalent spherical lens with focal power M=S+C / 2 and two Jackson cross cylindrical lenses, one of which has an axis position of 0° and focal power J0=(-C / 2)*cos(2*axis position), and the other cylindrical lens has an axis position of 45° and focal power J45=(-C / 2)*sin(2*axis position), as the astigmatism decomposition in polar coordinate form (C, axis position).

[0109] To more closely resemble the standard Jackson cross cylinder procedure, the two orthogonal directions can correspond to the initial astigmatism direction and its perpendicular direction.

[0110] Lower add may be defined as the difference in spherical power between distance and near vision, for example between the distance and near vision zones of a progressive lens.

[0111] A prism is a wedge-shaped optical component. When light passes through a prism, it is deflected toward the base, or larger side, of the prism.

[0112] Because corrective lenses typically have uneven thickness, they behave like a prism. A subject viewing an object through the lens will see the object as slightly off because the perceived image appears to originate from the off-center light rays.

[0113] The prismatic power corresponds to the prismatic displacement of the image and is measured in prismatic diopters, where one prismatic diopter is equal to one centimeter of displacement at a distance of one meter.

[0114] Filter transmittance corresponds to the ratio of the intensity of the light emerging from the corrective lens to the intensity of the incident light.

[0115] Furthermore, subjective values ​​relating to binocular balance, astigmatism, lower addition or binocular spherical power can also be additionally determined during the subjective test.

[0116] After determining the spherical and cylindrical components of the correction required for the subject, it is known that the binocular balance of the eyes may need to be adjusted. In order to achieve a comfortable correction for the subject's eyes, it is indeed useful to ensure that the image quality seen through the adapted lenses with the determined adapted spherical power is similar in both eyes.

[0117] Determining the spherical power value for each eye may not be as accurate as determining the value for both eyes. Known methods allow for separating the image seen by each eye, such as using two polarizers oriented at 90°, one of which is placed in front of each eye, so it is useful to compare the image quality seen by each eye with the corresponding adapted lens with the adapted spherical power. If one of the adapted lenses is determined to provide a better image than the other, the spherical power of this lens can be modified to achieve a similar image quality for the subject.

[0118] The corresponding optical characteristic can then be equal to the difference in the spherical power of the two lenses.

[0119] The corresponding optical characteristic may also be equal to the difference between the difference between the spherical powers of the two lenses initially determined before the binocular balance test and the difference between the spherical powers of the two lenses after the binocular balance test.

[0120] After adjusting binocular balance, the spherical power of both lenses can be adjusted. Based on the spherical power values ​​previously determined, for example, after a binocular balance test, the spherical power of both lenses is varied simultaneously by the same increment, and the subject evaluates the quality of binocular vision. The subject is then asked to compare the image quality seen with the two spherical powers varied.

[0121] More precisely, the method according to the present invention comprises the following steps:

[0122] i) providing an initial model giving a likelihood of each descriptive answer in a set of descriptive answers for each theoretical optical condition in a plurality of theoretical optical conditions,

[0123] j) performing said subjective test by asking the subject to describe his visual performance under each of said plurality of real optical conditions by selecting a descriptive answer from said set of possible descriptive answers for each real optical condition,

[0124] k) collecting each descriptive answer of the subject and the corresponding actual optical condition,

[0125] 1) Determining the value of the measure by considering the initial model and each descriptive answer collected.

[0126] This method is implemented by a system according to the invention for determining a measurement value associated with a subjective value of an optical characteristic of at least one corrective lens adapted for an eye of a subject.

[0127] The system includes a computer having one or more memories and one or more processors.

[0128] According to the invention, an initial model giving a likelihood of each descriptive answer in a set of descriptive answers for each theoretical optical condition in a plurality of theoretical optical conditions is stored in one of the one or more memories of the computer.

[0129] The one or more memories of the computer may include permanent or temporary memory.

[0130] Such memory may be local or remote. The memory may be packaged with the processor or may be remotely accessible via a (wired) network cable or via a wireless connection. Thus, the initial model may be stored in the local memory of the computer or retrieved from a server / cloud.

[0131] In a preliminary step of the method according to the invention, the initial model ( Figure 1 300).

[0132] The initial model gives a likelihood of each descriptive answer in the set of descriptive answers for each of a plurality of theoretical optical conditions.

[0133] The initial model may include at least one of: an initial model curve, an initial model formula between the answer and the theoretical optical condition, and an initial model data set.

[0134] In particular, the initial model may be continuous in the case of a curve or a formula, or may be discontinuous in the case of a data set.

[0135] The initial model curve can correspond to a corresponding initial model formula or a graphical representation of an initial model data set that relates the answer to the theoretical optical conditions, or can be determined based on such an initial model data set.

[0136] The initial model curve or formula or data set is, for example, statistically determined based on a reference data set previously collected when performing the subjective tests on a plurality of reference subjects, the reference data set comprising real optical conditions and corresponding answers to each subjective test performed on each reference subject.

[0137] In said preliminary step of the method according to the invention, said data set ( Figure 1 100).

[0138] Preferably, for all tests performed, such as the classical refraction test, the dichromatic test, the Jackson Cross cylinder test, the binocular balance test, every answer for every correction value tested is collected and can be considered to determine the initial model.

[0139] The reference data set also preferably comprises at least one personal characteristic of each reference subject, which at least one personal characteristic is stored relative to other data related to this reference subject.

[0140] The personal characteristics include at least one value from the following:

[0141] - social parameters such as age, gender, geographical origin,

[0142] - Eye care history (past prescriptions, eye diseases / problems...),

[0143] - morphological parameters, such as pupil size and pupil distance,

[0144] - optical parameters such as type and / or value of refractive error, in particular spherical refractive value, cylindrical refractive value, cylindrical axis, astigmatism, eye-lens distance,

[0145] - setting parameters such as starting point of subjective test, type of stimulus, distance to test distance or near vision, distance between phoropter and eye,

[0146] - visual parameters such as acuity, previous answers to subjective tests, monocular dominance, binocular vision or eye movements,

[0147] - behavioral parameters such as: rapidity of previous answers to subjective tests, sensitivity to changes in at least one optical characteristic of at least one ophthalmic lens.

[0148] Further information relating to the reference subject is preferably collected in the reference data set, such as the final refraction or final optical characteristics determined by each test performed.

[0149] The reference data set is collected, for example, by saving the above-mentioned data for each reference subject after performing each subjective test.

[0150] Data may be automatically collected from devices used to perform subjective tests anywhere in the world and programmed to transmit such data to the computer.

[0151] In an embodiment, the initial model is determined based on statistical characteristics of a reference data set.

[0152] In particular, the initial model is determined based on an average value or a weighted average value of the reference data set.

[0153] This means that the initial model curve or formula or data set is determined by fitting a modified data set that includes the average or weighted average of a reference data set.

[0154] For example, three different possible answers to a performed subjective test can be associated with numerical values, and these values ​​can be averaged for fixed real optical conditions.

[0155] Possible answers include:

[0156] - better visual performance in the first real optical condition, such as the code "1", that is, in Figure 2 The top is represented by the ordinate value 1,

[0157] - better visual performance under the second real optical condition, for example coded as "-1", that is, under Figure 2 The top is represented by the vertical coordinate value -1

[0158] - There is no perceived difference between the first real optical condition and the second real optical condition, for example coded as "0", that is, Figure 2 Up is represented by a ordinate value of 0.

[0159] In the case where the initial model curve, formula, or data set is determined by fitting a modified data set including the mean or weighted mean of a reference data set, if 10 responses corresponding to a value of -1, 8 responses corresponding to a value of 0, and 2 responses corresponding to a value of +1 are obtained for a given spherical power value, then the mean of the responses associated with the given spherical power value is -0.4 (measured as a percentage of -40%). The mean responses can be given by the same subject or by different subjects. Finally, the mean of the data set is not a priori equal to -1, 0, or 1, but rather represents the probability of obtaining a 1, 0, or -1 response.

[0160] The initial model may also be determined based on the distribution envelope of the reference data set. A reference answer graph is obtained by plotting the answer of each reference subject against the corresponding real optical conditions, and the initial model is determined based on the distribution characteristics of the reference answer graph.

[0161] For example, a distribution envelope can be defined to include data within a given percentile or at a given distance from the mean. For example, for a reference data set, all data collected for a particular real-world optical condition are distributed around the mean or median. The median corresponds to the 50% distribution: 50% of the values ​​collected for this real-world optical condition are above the median, and 50% are below the median.

[0162] Other statistically interesting values ​​can be defined, for example, by the 10% and 90% values ​​of the distribution: 90% of the values ​​collected for this real optical condition are above the value at the 10% percentile of the distribution, while 10% are below this value; and 10% of the values ​​collected for this real optical condition are above the value at the 90% percentile of the distribution, while 90% are below this value.

[0163] Examples of different initial models obtained by using different parts of the distribution of the reference data set are as follows: Figure 2 shown.

[0164] The solid curve represents the initial model obtained by considering the data of the 50% distribution.

[0165] The two dotted curves represent the initial model obtained by considering the values ​​of 10% (curve M10) and 90% (curve M90) of the data distribution. Figure 2 In the case of , the data shown are corrected as described below.

[0166] Any other distribution of the reference data set, whether corrected or not, may also be used. For example, the value at the 25th percentile (first quartile) or the value at the 75th percentile (third quartile) may be used.

[0167] In an embodiment, the initial model considers the descriptive responses given by all reference subjects, rather than selecting reference subjects based on their personal characteristics.

[0168] In this case, it may be necessary to correct the reference data set to account for differences in the individual characteristics of the reference subjects, and then determine the initial model based on the corrected reference data set.

[0169] For example, the theoretical optical conditions of the initial model are related to the optical components placed in front of the subject when the reference subject gives the answers plotted on the ordinate. These theoretical optical conditions can be the values ​​of the optical characteristics of the lens placed in front of the subject, or can be derived from the values ​​of these optical characteristics, for example by applying an offset or multiplication factor.

[0170] The following is an example of how to determine the initial model taking into account reference data of reference subjects with different actual spherical refractions.

[0171] To achieve this, the theoretical optical conditions of the initial model are defined relative to the actual refraction of each reference subject. For example, the theoretical optical conditions can correspond to the actual optical conditions of the reference data set with an offset. This offset is calculated to center the data around a predetermined refraction value. The initial model is then determined based on the corrected actual optical conditions corresponding to the offsetted actual optical conditions.

[0172] For example, in practice, a subjective test is first performed on a reference subject, which is usually performed according to methods known to those skilled in the art. The actual refractive value of the reference subject is determined by this test.

[0173] The actual refractive value is then subtracted from the optical characteristic values ​​of the lenses used in each trial of the subjective test in order to obtain the reference subject's response under theoretical optical conditions relative to the actual refractive value. The reference data obtained for each reference subject can then be compared, and in particular averaged.

[0174] The embodiments described here can of course be combined and the initial model can be determined based on data from a group of subjects selected for comparison that have been calibrated.

[0175] Figure 2 An example of an initial model obtained using this protocol is shown.

[0176] Figure 2 The answer values ​​of different reference subjects are shown, centered on their actual refraction values ​​and superimposed around -0.5D.

[0177] More precisely, the ordinate shows three different possible answers to the subjective test performed.

[0178] For example, as mentioned above, these possible responses are:

[0179] - First choice or answer: better visual performance under the first real optical condition, coded as "1",

[0180] -Second choice or answer: better visual performance under the second real optical condition, coded as "-1",

[0181] -Third choice or response: No perceived difference between the first real optical condition and the second real optical condition, coded as "0".

[0182] The abscissa shows the theoretical optical conditions.

[0183] exist Figure 2 In the example of FIG, the theoretical optical conditions are determined based on the spherical power value of a lens placed in front of a reference subject, as described below.

[0184] Consider a group of reference subjects with a refraction of -0.5D plus or minus 0.2D.

[0185] In this example, the responses of all reference subjects in this group are centered around -0.5 D. This means that for a reference subject in this group whose refraction was ultimately determined to be -0.53 D, the values ​​of the responses obtained are plotted or saved as values ​​+0.03 D corresponding to the theoretical optical conditions corresponding to the real optical conditions used during the subjective test, so that the results are centered at -0.5 D.

[0186] Thus, a corrected reference data set is generated which includes the answers and the corresponding theoretical optical conditions.

[0187] The corrected reference dataset was obtained by Figure 2 The initial model for the likelihood of each descriptive answer in the descriptive answer set for each of the plurality of theoretical optical conditions is obtained here based on the corrected reference data set by determining a curve or formula that fits the corrected reference data set.

[0188] In the examples described here and shown in the figures, the general formula used for fitting is one of the sigmoid general formulas:

[0189] (1-2 / (1+e (-s*(x-x0)) ), minimized using RMS.

[0190] In the general formula for the S-type mentioned above, S is the maximum slope of the S-type, x0 is the relative position of the S-type, and x is the abscissa of the fitted data (the true optical condition value). RMS is defined as the root mean square of the difference between the ordinate given by the answer from the data and the ordinate given by the formula being tried to fit.

[0191] Other general formulas can be used to determine the initial model by fitting or interpolating the reference data, such as step function, linear, polynomial, or any other type of function.

[0192] In the example initial model shown, the initial model is scaled to vary between 1 (corresponding to 100% likelihood of the first choice or answer) and -1 (corresponding to 100% likelihood of the second choice or answer).

[0193] In general, the slope of the portion of the initial model curve included between the 1 and -1 ordinates and intersecting the abscissa axis is larger for subjects with a higher sensitivity to changes in real optical conditions (e.g., changes in the refractive power of a lens placed in front of their eyes) and smaller for subjects with a lower sensitivity.

[0194] Alternatively, the initial model can also include a model data set which includes an average / weighted average of the responses associated with the corresponding theoretical optical conditions.

[0195] Preferably, at least one personal characteristic of the subject is taken into account for determining the initial model.

[0196] exist Figure 1 In the embodiment of the method according to the invention shown, during said preliminary step at least one personal characteristic of the subject is acquired ( Figure 1 200).

[0197] The personal characteristic of the subject may be any of the personal characteristics listed above.

[0198] Preferably, this personal characteristic is related to the type of visual impairment of the subject.

[0199] This personal characteristic may also be related to the age of the subject.

[0200] The personal characteristics may for example include age, and / or type / value of refractive error, and / or distance between the eye and the lens of the eyewear.The type of subjective test performed may also be considered.

[0201] For example, different initial models can be determined for subjects with myopia, hyperopia, subjects within a predetermined age range, etc.

[0202] Preferably, the initial model is determined taking into account a plurality of personal characteristics of the subject. The initial model is determined, for example, by a series of steps, each step taking into account one personal characteristic.

[0203] An initial model can then be obtained using supervised learning (e.g. decision trees or neural networks), where the learning phase is constructed for all data collected on the reference subject:

[0204] - For input: values ​​of personal characteristics and optical characteristics (spherical power, cylindrical power, axis position, etc.) in the corresponding test of the subjective test

[0205] - Output: The response (or likelihood of certainty in the answer) associated with each trial. For example, select 1 for sure (100% likelihood), select 1 for maybe (60% likelihood), select no (0% likelihood), select 2 for maybe (-60% likelihood), select 2 for sure (-100% likelihood). The output can be a discrete or continuous variable.

[0206] In particular, an initial model associated with each personal characteristic or different sets of initial models can be determined, with the probability of each initial model being associated with them, taking into account the personal characteristics of the subject, in particular the subject's sensitivity to changes in the real optical conditions used, taking into account his age and / or taking into account his expected refractive error (nature and / or value).

[0207] For example, in the case of an initial model obtained by fitting reference data with an S-type general formula, the slope of the portion of the initial model curve that intersects the abscissa axis can be adjusted depending on the subject's sensitivity to changes in real optical conditions (e.g., changes in at least one optical characteristic of at least one ophthalmic lens).

[0208] The optical characteristics here correspond to the real optical conditions of the subjective tests.

[0209] Subjects with a higher sensitivity to changes in the optical characteristics will indeed benefit from a higher slope in the initial model, as it will allow the subjective test to be performed more quickly.

[0210] The slope of the initial model can also be adjusted depending on the age of the subject, his level of comprehension of the test (e.g., whether he speaks the language of the test fluently), or his level of fatigue during the test. Older and younger subjects, or subjects with poor comprehension and subjects who are generally fatigued, will benefit from using an initial model with a lower slope, which will lead to more gradual testing, thus giving more trials to allow checking of the obtained answers.

[0211] Finally, the slope of the initial model can also be adjusted depending on the subject's refractive error. For example, for a subject with myopia and / or astigmatism: the higher the correction value required by the subject, the lower the slope of the initial model should be.

[0212] Furthermore, the evolution of the subject's answers during the subjective test can also be taken into account.

[0213] Some subjects will indeed provide different responses when the actual optical conditions progress in one direction or the other. More precisely, in the case of changing the optical power of a lens, a subject's response when the optical power is reduced may differ from that when the optical power is increased.

[0214] Due to the training effect, the responses at the end of the subjective test can be closer to the expected responses based on the initial model than the responses at the beginning of the subjective test.

[0215] On the contrary, for some subjects, due to subject fatigue, the responses at the end of the subjective test may be further away from the expected responses according to the initial model than the responses at the beginning of the subjective test.

[0216] The subject's behavior with respect to the training effect / fatigue during the subjective test is also a personal characteristic that can be taken into account when determining the initial model (or a further modified model) by giving different weights to the responses collected in the reference data based on the time period of the subjective test when the reference data was collected. Naturally, responses with greater reliability (no fatigue after training) will be given a higher weight.

[0217] This process for determining the initial model is very useful, especially when the individual characteristics under consideration are many.

[0218] Machine learning algorithms and / or neural network methods may also be used to determine the selection among all personal characteristics and the weight of each personal characteristic.

[0219] This procedure can be used on a selected group of reference subjects.

[0220] In an embodiment, the initial model is determined based on a value, an average or a weighted average of a portion of a reference data set.

[0221] The part of the reference data set considered is preferably selected based on one or more reference personal characteristics of a reference subject.

[0222] More precisely, a portion of the reference data set under consideration is selected so as to keep data obtained for reference subjects whose one or more reference personal characteristics match the corresponding personal characteristics of the subject of the current tested eye.

[0223] The reference personal characteristics of the reference subject may be the same as the personal characteristics of the subject, or may be related to the personal characteristics of the subject within a predetermined range.

[0224] In other words, the plurality of reference subjects are selected from a group of reference subjects having one or more personal characteristics that are the same as or similar to the corresponding personal characteristics of the subject.

[0225] For example, for myopic subjects, the initial model is determined by considering a reference data set collected from selected myopic reference subjects. For myopic subjects previously known to have a refraction of -1 D, the initial model is determined by considering a reference data set collected from selected myopic reference subjects having a final refraction determined to be -1 D or a final refraction determined to be between -0.9 D and -1.1 D, or between -0.5 D and -1.5 D, or between 0 D and -2 D. The data for the selected reference subjects can be corrected as described above to be centered around the actual refraction or visual defect of each reference subject.

[0226] The initial model can then be used directly with data collected during subjective testing performed on the subjects to determine the measure.

[0227] However, in a preferred embodiment, the subjective test is adjusted based on the initial model ( Figure 1 400).

[0228] In practice, the computer is programmed to determine at least one initial parameter value of the subjective test based on personal characteristics of the subject.

[0229] Personal characteristics such as age, sensitivity of the subject to changes in the optical characteristics of the lenses, level of understanding and / or fatigue, nature and value of the refractive error may in particular be taken into account.

[0230] When a subject is determined to be unlikely to adapt, such as in the case of older subjects (adults), the initial refractive values ​​of the subject's eyes, e.g. measured objectively with an autorefractor, will be used as starting values ​​for the subjective test without any modification.

[0231] When a subject is determined to be prone to adaptation, such as for a younger subject (child), a hyperopic subject, or a fatigued subject, an additional value is added to the subject's initial refractive index. The higher the determined tendency of the subject to adapt, the higher the additional value determined to the initial refractive index.

[0232] For example, the initial model may be taken into account to determine an initial value of an optical characteristic of an optical component placed in front of the subject's eye or an initial value of a step value for increasing or decreasing this optical characteristic.

[0233] For example, considering the Figure 2 In the case of an initial model of average data, if the first trial uses an optical lens with a refractive power equal to -0.375 D, an answer of 2 is expected based on the initial model. A second trial is performed with an optical lens with a power of -0.75 D, and an answer of 1 is expected. If the answer given by the subject matches the expected answer, the initial model is confirmed; otherwise, it is adjusted, e.g., modified, to take into account the subject's answer.

[0234] Generally speaking, based on the initial model and initial parameter values ​​(e.g. values ​​from a previous prescription), the computer is programmed to perform two first trials with values ​​of the real optical conditions, so that when performing the two consecutive first trials, one answer 1 and one answer 2 are obtained. This allows the total number of required trials to be reduced.

[0235] The computer is programmed to take into account the initial model to determine the at least one initial parameter value of the subjective test.

[0236] A subjective test was then performed on the subjects.

[0237] According to the present invention, the one or more processors of the computer are programmed to:

[0238] a) collecting results of the subjective test, the subjective test being performed by asking the subject to describe his visual performance under each of the plurality of real optical conditions by selecting a descriptive answer from the set of possible descriptive answers for each real optical condition, the results comprising each selected descriptive answer and the corresponding real optical condition ( Figure 1 500),

[0239] b) determining the value of the measure by considering the initial model and each descriptive answer collected ( Figure 1 800, 900).

[0240] Step a)

[0241] During the subjective test, the subject is asked to describe his visual performance under a plurality of real optical conditions by selecting a descriptive answer from the set of possible descriptive answers.

[0242] This subjective test can be performed within the method according to the invention or independently thereof.

[0243] Each of the plurality of real optical conditions typically corresponds to a given value of an optical characteristic (such as spherical power, cylindrical power, axis position or transmission coefficient) of an optical component (such as a test filter or test lens placed in front of the subject's eye).

[0244] Subjective testing is usually performed using a specific device called a "phoropter." This phoropter includes at least one test optic with variable optical characteristics (e.g., a test lens with variable optical power) or a plurality of test optics with different fixed, predetermined optical characteristics (e.g., a plurality of lenses with different fixed, predetermined optical powers), and at least one support adapted to support the test optic or one of the plurality of test optics. The support is adapted to be positioned in front of the subject's head so that the test optic, supported by the support, can be placed in front of the subject's eye to be tested.

[0245] Variable lenses are described, for example, in the following documents: US 20160331226, US 2017027435 or WO 2017 / 021663.

[0246] The device can also include other test optical components with different optical characteristics, such as spherical lenses, cylindrical lenses, prisms, linear or circular polarization filters, color filters, transmission filters, slits, badal systems, reflectors, active lenses, deformable mirrors, SLMs (spatial light modulators), Alvarez lenses, etc.

[0247] In practice, the subject's vision is tested during a subjective testing protocol using the phoropter to successively place test optics, such as lenses having different values ​​of the optical characteristic, in front of at least one eye of the subject.

[0248] In the case of testing lenses, the optical power is increased or decreased by a predetermined step value from one step of the subjective test to the next. This step value is typically 0.25 diopters (D) or 0.125 D. The successive steps of the subjective test are referred to herein as "trials".

[0249] For each test lens placed in front of his eye, the subject is asked to express a visual assessment corresponding to an indication of the preferred visual state of the two visual states presented to him, or whether he is unable to decide between the two. The two visual states may, for example, correspond to his vision through the previous lens and his vision through the current lens, or may correspond to his vision of two different images through the current lens.

[0250] In practice, this part of the test protocol may correspond, for example, to the assessment given by a subject during a two-color test. During this two-color test, the subject is presented with an image comprising an optotype displayed on a red background on one side and an optotype displayed on a green background on the other side. If the subject has better vision with the current lens for the optotype on the red background, the spherical power should be reduced in the next presented lens; if the subject has better vision with the current lens for the optotype displayed on the green background, the spherical power should be increased in the next presented lens.

[0251] As a variant, for each current lens placed in front of his eye, the subject is asked to evaluate the quality of his vision through the current lens compared to the previous lens: he is asked whether the current lens provides better or worse vision than the previous lens, or whether he cannot decide between the two. In the latter case, both lenses provide the subject with similar visual quality.

[0252] The phoropter may be part of the device according to the invention described herein or distinct therefrom. In the latter case, it may comprise a communication device together with the device according to the invention.

[0253] As mentioned above, generally speaking, the descriptive answer set includes, for example, the following answers:

[0254] -Better visual performance in first real optical conditions,

[0255] - better visual performance in second real optical conditions,

[0256] - No difference is perceived between the first real optical condition and the second real optical condition.

[0257] The first real optical condition and the second real optical condition may correspond to observing the same image with different test optical components having different values ​​of the optical characteristic, or to observing different images with the same test optical component, such as an image comprising an optotype displayed on a red background on one side and an optotype displayed on a green background on the other.

[0258] Step b)

[0259] In step b), the computer is for example programmed to compare the initial model with each of the collected descriptive answers and to determine the value of the measure based on this comparison.

[0260] In case the initial model comprises an initial model curve, the computer is programmed to perform the following sub-steps in step b):

[0261] b1) plotting each of the collected descriptive responses selected by the subject against the corresponding real optical conditions and superimposing said initial model curve on the thus obtained graph such that the initial model curve fits the graph,

[0262] b2) determining the value of the measure taking into account the relative positions of the initial model curve and the graph.

[0263] The steps for plotting the collected descriptive responses are not intended to be in any way representative of the graphs and / or other plots that may or may not be displayed to the user and / or subject.

[0264] In an embodiment, said subjective value of an optical characteristic of at least one corrective lens adapted for an eye of a subject is determined directly by comparing answers collected for said subject with said initial model.

[0265] In step b1), the initial model is not modified to fit the graph, but the relative positions of the model curve and the graph are modified to minimize the distance between the model curve and the graph in a global manner, for example, using quadratic optimization to minimize the distance.

[0266] Use as in Figure 2 In the example of , the initial model determined, the placement of the initial model relative to the data collected for the subject may give the actual refraction of the subject at the intersection between the initial model and the abscissa.

[0267] As a variant, the computer is programmed to perform the subjective test. Figure 1 In the embodiment of FIG. 5 , block 500 corresponds to performing subjective testing and collecting corresponding data.

[0268] In this case, the computer is programmed to adjust the parameters of the subjective test taking into account the initial model ( Figure 1 For example, a computer is programmed to determine the real optical conditions used subsequently for performing the subjective tests taking into account the initial model.

[0269] Advantageously, step b1) is performed during each trial of the subjective test, in other words, for each answer given by the subject. For each new answer given by the subject, this answer is added to the graph, and the relative position of the model curve and the graph is modified so as to globally minimize the distance between the model curve and the graph. Then, during each trial, the intermediate value of the measure can be determined, and the parameters of the subjective test can be modified before the next trial is performed. In particular, the step size between two consecutive values ​​of the optical characteristic of the test lens used during the test can be adjusted. Alternatively, the next value of the optical characteristic of the test lens can be determined directly.

[0270] The subjective test is then adjusted to take into account the data that has been collected for the subject ( Figure 1 501).

[0271] Figures 3 to 7 An example of this process is shown in Figures 1 and 2. In these figures, the squares show the positions of the plotted data of the subjects: the answers given under the corresponding real optical conditions (here the spherical power of the test lens in diopters). The initial model is represented by a curve.

[0272] Figure 3 The results of the first trial are shown, where the subject only gave one answer. The initial model was placed so that it passed through this single point. Therefore, the position of the initial model curve is uncertain.

[0273] After the second trial, the second response is plotted and the initial model curve is shifted to fit the graph ( Figure 4 ).

[0274] After the third trial, the third response is plotted and the initial model curve is shifted accordingly.

[0275] Figure 6 and Figure 7 The corresponding next two trials are shown.

[0276] The intersection point PV of the initial model curve and the abscissa gives the predicted value of the eye's spherical refraction ( Figure 7 ).

[0277] The diopter step size of the test lens can therefore be adjusted so that the spherical power of the next test lens is close to the predicted value of the eye's spherical refraction.

[0278] The computer may be programmed to determine the value of the measure based on the position of the initial model curve relative to the graph when the position is determined using a predetermined number of trials or when the position does not modify by more than a predetermined amount between two consecutive trials of a subjective test, for example, when the position is determined after at least 4, 10 or 20 trials, or when the position does not modify by more than 0.25D, 0.12D or 0.06D between two consecutive trials of a subjective test.

[0279] The use of an initial model allows the exact position of the curve to be quickly obtained.

[0280] In case the initial model comprises an initial data set, the computer is programmed to perform the following sub-steps in step b):

[0281] - statistically processing the initial data set and the collected descriptive responses, and

[0282] - said value of said measure is determined taking into account this statistical processing.In practice, a law describing the variation of a reference data set is determined by means of said statistical processing, and this law is then applied to the answers of the subjects during the subjective test.

[0283] According to an advantageous feature of the device and method of the invention, the initial model can be modified during any trial of the subjective test, i.e. each new test optical component ( Figure 1 600).

[0284] In particular, the initial model ( Figure 1 502).

[0285] In an embodiment, an initial model may be determined by considering a first personal characteristic of the subject. In any trial of a subjective test, the initial model may be modified to consider a second personal characteristic of the subject that is different from the first personal characteristic. The second personal characteristic may be considered in addition to or instead of the first personal characteristic.

[0286] For example, the initial model can be modified in real time to take into account the subject's behavior when responding: facial features, facial expressions, voice, hesitations, and the time that elapses between the presentation of each new real optical condition and the subject's answer.

[0287] As for the time that elapses between the presentation of each new real optical condition and the subject's answer, it can be compared with a threshold value: if the time that has elapsed exceeds said threshold value, the answer is not taken into account.

[0288] In an embodiment, the initial model may be modified to take into account the responses collected for the subject.The subject is then included in the reference subjects, whose responses are used to determine the model.

[0289] The modified model thus obtained is then used to determine the measure.

[0290] In this case, the computer is programmed to perform the following sub-steps in step b):

[0291] - modifying the initial model taking into account each descriptive answer of the subject and the corresponding real optical conditions to obtain a modified model, and

[0292] - Compare the modified model to each descriptive response collected ( Figure 1 700 ) and determining the value of the metric based on this comparison.

[0293] The modified model is determined and used just like the initial model described above.The modified model may include at least one of: a modified model curve, a modified model formula between the answer and the theoretical optical condition, and a modified model data set.

[0294] When the modified model comprises a modified model curve, the computer is programmed to perform the following sub-steps in step b):

[0295] - plotting each of the collected descriptive responses selected by the subject against the corresponding real optical conditions and superimposing said modified model curve on the graph thus obtained such that the modified model curve fits the graph,

[0296] - determining said value of said measure taking into account the relative positions of said modified model curve and said graph.

[0297] Alternatively, if the modified model comprises a modified data set, the computer is programmed to perform the following sub-steps in step b):

[0298] - statistically processing the modified data set and the collected descriptive responses, and

[0299] - taking this statistical processing into account to determine said value of said measure.

[0300] In practice, the computer is programmed to determine at least one of the real optical conditions for performing the subjective test by taking into account the modified model. In particular, the computer can be programmed to determine each of the real optical conditions for performing a posteriori trials of the subjective test by taking into account the modified model.

[0301] The computer is programmed to determine the value of the measure based on the position of the modified model curve relative to the graph when the position is determined using a predetermined number of trials or when this position does not modify by more than a predetermined amount between two consecutive trials of a subjective test. Figure 1 800).

[0302] Furthermore, the computer is programmed to compare each collected descriptive answer corresponding to a real optical condition with the initial or modified model, and when this comparison indicates that the difference between a particular collected descriptive answer corresponding to a particular real optical condition and the initial, respectively modified model exceeds a predetermined threshold, this particular collected descriptive answer is identified as inconsistent ( Figure 1 900).

[0303] In particular, a criterion for calculating the distance from an initial or modified model curve can be used to detect incorrect or uncertain answers. When the distance between the answer graph and the curve of the initial or modified model is higher than a predetermined threshold or higher than the "don't know" answer, the answer is identified as inconsistent.

[0304] Thus, an indicator of the confidence level associated with each measured subjective value of the optical characteristic can be determined based on the distance between the answer graph and the curve of the initial or modified model. For example, the indicator indicates that the confidence level of the answer corresponding to the graph that is closer to the model is higher.

[0305] When an inconsistent answer is thus detected, an alert can be issued to the destination of the operator of the subjective test.Inconsistent answers are preferably not taken into account when determining the measure.

[0306] Preferably, the experiment of the subjective test for which the inconsistent response was collected is repeated and another response is collected to replace the inconsistent response that was not considered ( Figure 1 901).

[0307] Preferably, when the number of inconsistent responses detected during the subjective test exceeds a predetermined threshold, the initial or modified model is modified to reduce the number of inconsistent responses ( Figure 1 902).

[0308] The initial or modified model is then modified. For example, a model based on an average reference data set can be changed to a model based on a distribution of the reference data set, or a model based on individual characteristics of the subjects can be changed to a more general model based on all reference data.

[0309] Atypical behavior can be detected. For example, younger subjects tend to answer faster than older subjects. If a younger subject is slower, it could be due to a specific question or a misunderstanding of the subject.

[0310] Generally speaking, during a subjective test, initial values ​​for the actual optical conditions are determined taking into account the individual characteristics of the subject, such as previous prescriptions (if any), and an initial model selected or determined for this subject. The subsequent values ​​for the actual optical conditions are determined by a computer based on the initial model.

[0311] For example, as already mentioned, the computer is programmed to carry out two first trials with the values ​​of the real optical conditions, so that an answer 1 and an answer 2 are obtained when carrying out the two successive first trials.

[0312] The correspondence between the actual responses and the expected responses based on the initial model is monitored. The subject's actual responses for given real optical conditions are compared with the expected responses based on the initial model for the corresponding theoretical optical conditions.

[0313] In the event that the actual answer matches the expected answer, the initial model may be retained or modified for further improvement. When the number of actual answers that differ from the expected answer is below a first difference threshold, the actual answer may be considered to match the expected answer.

[0314] If the actual answer differs from the expected answer, the initial model is stopped and the subjective step is resumed in a traditional manner, or the initial model is modified. When the number of actual answers that differ from the expected answer is higher than a second difference threshold, the actual answer can be considered different from the expected answer.

[0315] According to a first possibility, another predetermined model may be tried.

[0316] According to a second possibility, the data currently collected for the subject may be added to the reference data and a modified model determined based on this updated reference data comprising the data collected for the subject.

[0317] According to a third possibility, a new model can be determined based on data collected for the subject, current and past data. The last solution is particularly useful when past data is available.

[0318] Indeed, the present invention also allows the use of data from past eye examinations, which allows a better understanding of the evolution of behavior. This can be used to prevent and / or detect the evolution of cataracts or to prevent and / or detect other evolving refractive defects (myopia) or pathologies (age-related macular degeneration, etc.). For example, determining that an initial model that was successful for a subject in the past is completely unsuitable for this subject later on may be a sign of a new problem (such as cataracts).

Claims

1. A system for determining a metric value associated with a subjective value of an optical characteristic of at least one corrective lens adapted for an eye of a subject by performing a subjective test in which the subject is asked to describe his visual performance under a plurality of real optical conditions by selecting a descriptive answer from a set of possible descriptive answers, the system comprising a computer having one or more memories and one or more processors, wherein: - an initial model giving a likelihood of each descriptive answer in a set of descriptive answers for each theoretical optical condition in a plurality of theoretical optical conditions is stored in one of the one or more memories of the computer, - the one or more processors of the computer are programmed to: a) collecting (500) results of the subjective test, the subjective test being performed by asking the subject to describe his visual performance under each of the plurality of real optical conditions by selecting, for each real optical condition, one descriptive answer from the set of possible descriptive answers, the results comprising each selected descriptive answer and the corresponding real optical condition, b) determining (800, 900) the value of said measure by considering said initial model and each descriptive answer collected, The initial model includes an initial model curve, and the computer is programmed to perform the following sub-steps in step b): - plotting at least one of the collected descriptive responses selected by the subject against the corresponding real optical conditions and superimposing the initial model curve on the thus obtained graph such that the initial model curve fits the graph, - determining the value of the measure taking into account the relative positions of the initial model curve and the graph.

2. The system according to claim 1, wherein: In step b), the computer is programmed to compare the initial model with the at least one descriptive answer collected and to determine the value of the measure based on this comparison.

3. The system according to claim 1, wherein: The initial model comprises an initial data set, and the computer is programmed to perform the following sub-steps in step b): - statistically processing the initial data set and the collected descriptive responses, and - taking this statistical processing into account to determine said value of said measure.

4. The system according to claim 1, wherein: The computer is programmed to perform the following sub-steps in step b): - modifying the initial model taking into account each of the subject's descriptive responses and the corresponding real optical conditions to obtain a modified model, and - comparing the modified model to each of the collected descriptive responses and determining the value of the measure based on the comparison.

5. The system according to claim 4, wherein: The modified model includes a modified model curve, and the computer is programmed to perform the following sub-steps in step b): - plotting each of the collected descriptive responses selected by the subject against the corresponding real optical conditions and superimposing the modified model curve on the thus obtained graph such that the modified model curve fits the graph, - determining said value of said measure taking into account the relative positions of said modified model curve and said graph.

6. The system according to any one of claims 4 to 5, wherein: The modified model includes the modified data set, and the computer is programmed to perform the following sub-steps in step b): - statistically processing the modified data set and the collected descriptive responses, and - taking this statistical processing into account to determine said value of said measure.

7. The system according to claim 1, wherein: The initial model takes into account one or more personal characteristics of the subject.

8. The system according to claim 1, wherein: The descriptive response group includes the following responses: - Better visual performance in first real optical conditions, - better visual performance in second real optical conditions, - No difference is perceived between the first real optical condition and the second real optical condition.

9. The system according to claim 1, wherein: The initial model is statistically determined based on a reference data set previously collected when performing the subjective test on a plurality of reference subjects, the reference data set comprising: - individual characteristics of each reference subject, - Real optical conditions and corresponding responses for each reference subject.

10. The system according to claim 9, wherein: The initial model is determined based on an average or weighted average of a portion of the reference data set selected based on one or more personal characteristics of the reference subject and / or the initial model is determined using a machine learning algorithm and / or a neural network algorithm.

11. The system according to any one of claims 9 to 10, wherein: The plurality of reference subjects are selected from a group of reference subjects having one or more personal characteristics that are the same as or similar to the corresponding personal characteristics of the subject.

12. The system according to claim 4, wherein: The computer is programmed to determine the value of the measure based on the position of the initial or modified model curve relative to the graph when the position is determined after a predetermined number of trials or when this position has not been modified by more than a predetermined amount between two consecutive trials of the subjective test.

13. The system according to claim 1, wherein: The computer is programmed to compare each of the collected descriptive answers corresponding to the real optical condition with the initial model, and when this comparison shows that the difference between a specific collected descriptive answer corresponding to a specific real optical condition and the initial model exceeds a predetermined threshold, this specific collected descriptive answer is identified as inconsistent.

14. A method for determining a quantitative value associated with a subjective value of an optical characteristic of at least one corrective lens adapted to an eye of a subject by means of a subjective test in which the subject is asked to describe his visual performance under a plurality of real optical conditions by selecting a descriptive answer from a set of possible descriptive answers, the method comprising the following steps: i) providing (300) an initial model giving a likelihood of each descriptive answer in a set of descriptive answers for each theoretical optical condition in a plurality of theoretical optical conditions, j) performing said subjective test by asking said subject to describe his visual performance under each of said plurality of real optical conditions by selecting a descriptive answer from said set of possible descriptive answers for each real optical condition, k) collecting (500) each descriptive answer of the subject and the corresponding real optical condition, l) determining (800, 900) the value of said measure by considering said initial model and each descriptive answer collected, wherein the initial model comprises an initial model curve, and wherein in order to determine the value of the measure, the method comprises: - plotting at least one of the collected descriptive responses selected by the subject against the corresponding real optical conditions and superimposing the initial model curve on the thus obtained graph such that the initial model curve fits the graph, - determining the value of the measure taking into account the relative positions of the initial model curve and the graph.

Images