Method, system and computer program for estimating the refraction of an individual's eye

Abstract

The present invention relates to a method for estimating the refraction of an individual's eye, the method comprising the steps of: a) providing a set of N acquired decentered photorefraction images of the eye; b) setting values ​​of a set of parameters including at least a spherical power (110); c) providing a set of N simulated decentered photorefraction images (120); d) identifying an estimator of the difference between the set of N acquired decentered photorefraction images and the set of N simulated decentered photorefraction images (130); d) performing an optimization (140) using an optimization algorithm to optimize said estimator by adjusting values ​​of the set of parameters and repeating steps c) and d); and estimating an estimate of at least one refractive parameter of the eye.

Description

[Technical field]

[0001] The present invention relates to a method and system for assessing the refraction of an individual's eye.

[0002] More precisely, the invention relates to a method and a system for estimating the refraction of an individual's eye. The invention also relates to a computer program for estimating the refraction of an individual's eye. The method, system and / or computer program can be used for determining a prescription for an ophthalmic lens adapted to an individual or for manufacturing an ophthalmic lens according to the estimated refraction. The invention also provides refraction measurements that can be used as a starting point for further subjective refraction tests performed with another device. [Background technology]

[0003] Many documents describe devices and methods for measuring the refraction of an individual's eye. Subjective refraction is based on the interaction of the individual looking at various vision charts and the use of a set of lenses with different refractive corrections. Subjective refraction is based on the measurement of the optical properties of the eye taken into account. In particular, some methods and devices for measuring subjective refraction are based on decentered photorefraction or photoretinoscopy.

[0004] Eccentric photorefraction is used to perform subjective refraction tests by illuminating the user's eye with an eccentric light source and observing the image of the pupil with a camera. In most cases, the reflected light forms a bright shape on the pupil in the detected image, along with a complementary non-bright shape called the dark crescent. Analysis of the size, shape, and orientation of the bright and dark crescents allows the eye's refraction to be estimated according to the position of the eccentric light source. For example, publications (Non-Patent Document 1) or (Non-Patent Document 2) disclose analytical expressions of the bright areas. A relatively simple method to estimate the sphere, cylinder, and axis values ​​from measurements of the light gradient along the three meridians is described in (Non-Patent Document 3).

[0005] However, these methods do not take into account the higher order aberrations of the eye. Moreover, these publications are mainly theoretical, but do not disclose methods and systems that allow refractive measurements to be obtained quickly. Also, depending on the refractive error of the user, detection of the crescent can be difficult.

[0006] In developed countries, professional optometrists commonly use autorefractometers, however these systems are expensive and cumbersome. [Prior art documents] [Non-patent literature]

[0007] [Non-Patent Document 1] W. Wesemann, AMNorcia, D. Allen “Theory of eccentric photo refraction(photoretinoscopy): astigmatic eyes”, J.Opt.Soc.Am.A, Vol.8, No.12, 1991, pages 2038-2047 [Non-Patent Document 2] R. Kusel, U. Oechsner, W. Wesemann, S. Russlies, EMIrmer, and B. Rassow, “Light-intensity distribution in eccentric photorefraction crescents,” J.Opt.Soc.Am.A15, 1500-1511 (1998) [Non-Patent Document 3] Gekeler F, Schaeffel F, Howland HC, Wattam-Bell J, “Measurement of astigmatism by automated infrared photoretinoscopy”, Optometry and Vision Science: Official Publication of the American Academy of Optometry,1997,Jul;74(7):472-482.DOI:10.1097 / 00006324-199707000-00013 Summary of the Invention [Problem to be solved by the invention]

[0008] What is needed is a system and method that provides photorefraction estimation that is fast, compact, easy to use, and low cost. [Means for solving the problem]

[0009] It is therefore an object of the present invention to provide a method for estimating the refraction of an individual's eye, the method comprising the following steps: a) providing a set of N acquired decentered photorefraction images of an eye, where N is an integer greater than or equal to 1; b) initializing values ​​of a set of parameters including at least a sphere; c) providing a set of N simulated off-center photorefraction images using the simulation model according to values ​​of the set of parameters; d) determining an estimator of the difference between the set of N acquired eccentric photorefraction images and the set of N simulated eccentric photorefraction images; e) performing an optimization using an optimization algorithm to optimize said estimator by adjusting the values ​​of a set of parameters and repeating steps c) and d); and f) Deducing an estimate of at least one refractive parameter of the eye from the adjusted values ​​of the set of parameters.

[0010] According to a particularly advantageous embodiment, step a) comprises the following steps g) to i), each step c) comprising the following step j): g) positioning a device in front of the eye including an image capture device having an aperture and a plurality of M light sources, where M is an integer equal to or greater than 1, the plurality of M light sources being eccentrically positioned around the aperture of the image capture device at predetermined positions along at least two directions transverse to an optical axis of the image capture device, each light source of the M light sources adapted and configured to illuminate the eye with a light pulse; h) illuminating the eye using each light source of the plurality of M light sources; i) recording a set of N acquired eccentric photorefraction images of the eye using an image capture device, where N is less than or equal to M, and each image in the set of N acquired eccentric photorefraction images represents the eye illuminated by a light pulse of at least one light source of the plurality of M light sources; j) generating a set of N simulated off-center photorefraction images using the simulation model according to values ​​of the set of parameters.

[0011] According to a particularly advantageous aspect of the method, step b) comprises selecting a set of three acquired decentered photorefraction images from the set of N acquired decentered photorefraction images, where the set of three acquired decentered photorefraction images are recorded using three light sources positioned along at least three directions transverse to the optical axis of the image capture device, and processing each of the three acquired decentered photorefraction images to identify a size and a tilt angle of the dark crescent and infer therefrom three refractive power values ​​for the three directions.

[0012] According to one embodiment, M is equal to N, and step h) comprises illuminating the eye sequentially using each light source of the plurality of M light sources, and step i) comprises acquiring each image of the set of N off-center photorefraction images when the eye is illuminated with a light pulse of one of the plurality of M light sources.

[0013] According to another embodiment, N is less than M, and step h) comprises illuminating the eye using two light sources of the plurality of M light sources simultaneously, and step i) comprises acquiring one of the set of N off-center photorefraction images when the eye is illuminated with light pulses from the two light sources.

[0014] Advantageously, said simulation model is based on a geometric-optical model or a ray-tracing model of the light intensity distribution.

[0015] According to certain aspects, the simulation model further depends on hardware parameters, such as the power of each light source, ophthalmic parameters, such as corneal reflex, and / or operational parameters, such as the position of the image capture device relative to the eye.

[0016] According to another particular embodiment, the estimator is based on pixel-by-pixel differences between the set of N acquired eccentric photorefraction images and the set of N simulated eccentric photorefraction images, or the estimator is based on a comparison of the set of N pre-processed acquired eccentric photorefraction images and the set of N pre-processed simulated eccentric photorefraction images.

[0017] Advantageously, the optimization or minimization algorithm is based on gradient-free methods such as Simplex or Nelder-Mead or gradient-based methods such as Levenberg-Marquardt.

[0018] According to one embodiment, the method further comprises determining a distance between the image capture device and the eye and / or an orientation of the image capture device relative to the eye.

[0019] According to a particular advantageous aspect, the set of parameters further comprises cylinder, axis, interpupillary distance, high order aberrations, half interpupillary distance, direction of gaze, amount of red reflex, and at least one other parameter of the eye among the Stiles-Crawford parameters.

[0020] According to certain aspects, the method is performed at a first distance and a second distance between the image capture device and the eye.

[0021] Preferably, the plurality of M light sources comprises 3, 6, 9, or 12 light sources.

[0022] Advantageously, the plurality of M light sources comprises light emitting diodes.

[0023] It is a further object of the present invention to provide a system for estimating the refraction of an individual's eye, adapted to communicate with a mobile device or a remote computer.

[0024] The above objects are achieved according to the present invention by providing a system comprising an image capture device and a plurality of M light sources, M being an integer equal to or greater than one, the plurality of M light sources being eccentrically positioned around an aperture of the image capture device at predetermined positions along at least two directions transverse to an optical axis of the image capture device, the system being adapted and configured to illuminate an eye with a light pulse using each light source of the plurality of M light sources, the image capture device being configured to record a set of N acquired eccentric photorefraction images of the eye, N being an integer equal to or less than M, each image of the set of N acquired eccentric photorefraction images representing the eye illuminated by a light pulse of at least one light source of the plurality of M light sources, the system comprising a computation module configured to include a memory and a processor, the processor executing program instructions stored in the memory, k) initializing values ​​of a set of parameters including at least a sphere; l) generating a set of N simulated off-center photorefraction images using the simulation model based on values ​​of the set of parameters; m) calculating an estimator of the difference between the set of N acquired eccentric photorefraction images and the set of N simulated eccentric photorefraction images; n) using a minimization algorithm to minimize the estimator by adjusting the values ​​of a set of parameters and repeating steps l) and m); o) estimating an estimate of at least one refractive parameter of the eye from the adjusted values ​​of the set of parameters; The present invention is configured to:

[0025] Advantageously, the image capture device and the plurality of M light sources are mounted in an attachment that is removably attached to the mobile device.

[0026] According to one embodiment, the computing module is included in the accessory.

[0027] According to another embodiment, the computing module is included in a mobile device or a remote computer.

[0028] It is a further object of the present invention to provide a computer program product comprising one or more stored sequences of instructions accessible to a processor and which, when executed by the processor, causes the processor to perform at least the following steps: p) providing a set of N acquired eccentric photorefraction images, each image of the set of N acquired eccentric photorefraction images representing the eye illuminated by a light pulse of at least one light source of a plurality of M light sources eccentrically positioned around the image capture device at a predetermined position in at least two directions transverse to the optical axis of the image capture device, where M is an integer equal to or greater than 2 and N is an integer equal to or less than M; q) initializing values ​​of a set of parameters including at least a sphere; r) generating a set of N simulated off-center photorefraction images using the simulation model based on values ​​of the set of parameters; s) identifying an estimator of the difference between the set of N acquired eccentric photorefraction images and the set of N simulated eccentric photorefraction images; t) performing an optimization using an optimization algorithm to optimize said estimator by adjusting the values ​​of a set of parameters and repeating steps r) and s); and u) deducing an estimate of at least one refractive parameter of the eye from the adjusted values ​​of the set of parameters.

[0029] Detailed explanation of the example The following description, with reference to the accompanying drawings, will make clear what constitutes the present invention and how it can be achieved. The present invention is not limited to the embodiments shown in the drawings. Therefore, when reference signs follow features recited in the claims, it should be understood that these signs are included only to enhance the comprehension of the claims and do not limit the scope of the claims. For a more detailed understanding of the description provided herein and its advantages, reference is now made to the following brief description in conjunction with the accompanying drawings and detailed description, in which like reference signs represent like parts. [Brief description of the drawings]

[0030] [Figure 1A-1B] 1 shows a perspective view of a system for estimating the refraction of an individual's eye according to a first and a second embodiment, respectively; [Diagram 2] FIG. 2 is a front view of a photorefraction module with a camera and multiple light sources for estimating an individual's eye refraction, according to an embodiment of the present invention. [Diagram 3] 1 illustrates an exemplary configuration for acquiring a set of metamorphic photorefraction images of at least one eye of a subject. [Figure 4] 1 shows a set of nine decentered photorefraction images acquired of an individual's eye using systems and methods according to the present invention. [Diagram 5] 2 illustrates a schematic flow chart of a method for estimating the refraction of an individual's eye according to the invention; [Figure 6] Schematically represents an arrangement of three light sources positioned at equal distances from the camera on three distinct meridians. [Figure 7]13 shows an example of an acquired off-center photorefraction image with a dark crescent acquired using one light source along one meridian. [Figure 8] 13A-B show schematic diagrams of a method for determining initial values ​​of sphere, cylinder, and axis based on decentered photorefraction images using three light sources along three distinct meridians. [Figure 9] 1A and 1B show schematic diagrams of an example of an off-center photorefraction image acquired using two light sources simultaneously along two separate meridians. [Figure 10] 1 shows an initial set of N simulated eccentric photorefraction images at an initial step of the method based on initial values ​​of parameters and initial estimator values ​​for comparison with a set of 9 acquired eccentric photorefraction images. [Figure 11] Another set of N simulated eccentric photorefraction images compared to the same set of 9 acquired eccentric photorefraction images as in FIG. 10 are shown, along with intermediate results after one iteration using the corresponding values ​​of the estimators and parameters resulting from this simulation. [Figure 12] The final set of nine simulated off-center photorefraction images compared to the same set of nine photorefraction images and the final results after 16 iterations with the final values ​​of the estimators and parameters resulting from the minimization between the two sets of images are shown. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0031] In the following description and drawings, the figures may not necessarily be to scale, and certain features may be shown in generalized or schematic form for clarity and conciseness, or for reference purposes. In addition, while the making and using of various embodiments are discussed in detail below, it should be understood that, as described herein, many inventive concepts are provided that may be embodied in a variety of contexts. The embodiments discussed herein are merely representative and do not limit the scope of the invention. It will also be apparent to those skilled in the art that all technical features defined in relation to a process can be substituted into an apparatus, either individually or in combination, and conversely, all technical functions related to an apparatus can be substituted into a process, either individually or in combination.

[0032] Devices and methods 1A and 1B represent a system 50 for estimating the refraction of an eye 2 of a subject 1. The system 50 comprises a photorefraction module 10 or add-on linked to a mobile device 20, such as a smartphone, a tablet, a personal computer or a laptop computer. Advantageously, the photorefraction module 10 is mechanically fixed to the mobile device 20.

[0033] The photorefraction module 10 is connected to the mobile device 20 using either a direct plug and socket connection as shown in Figure 1A or a wire connection 11 as shown in Figure 1B. The photorefraction module 10 is configured to communicate with the mobile device 20 using either a wired or wireless connection.

[0034] The photorefractive module 10 shown in Figures 1A, 1B, and 2 includes a camera 30 and a number M of light sources, where M is an integer equal to or greater than 2. In other words, the minimum configuration is two separate meridians X A and X B The optical system includes at least two light sources 16-A and 16-B arranged along a line.

[0035] 2, the photorefractive module 10 includes nine light sources 16-A1, 16-A2, 16-A3, 16-B1, 16-B2, 16-B3, 16-C1, 16-C2, and 16-C3. The three light sources 16-A1, 16-A2, and 16-A3 are aligned in the same direction across the plane of the camera 30, i.e., the meridian X A Similarly, the three light sources 16-B1, 16-B2, and 16-B3, and the three light sources 16-C1, 16-C2, and 16-C3 are arranged along different directions, i.e., the meridian X. B , meridian X C are arranged along the meridian X, B and X C Both of them intersect the optical axis of the camera 30. The three meridians X A , X B , and X C correspond to three distinct directions in the same plane perpendicular to the optical axis of the camera 30. For example, the three meridians X A , X B , and X C are spaced apart from each other by an angle of 120 degrees.

[0036] As an option, the photorefraction module 10 further comprises another light source 18 (see FIG. 1A). Alternatively, the photorefraction module 10 further comprises another set of three light sources 18A, 18B, 18C arranged at another eccentricity and placed on three other meridians, as shown in FIG. 1B. According to another alternative, the photorefraction module 10 further comprises another set of three light sources 18A, 18B, 18C arranged on another same meridian and placed at a different eccentricity from the camera.

[0037] In one example, a set i of light sources 16-Ai, 16-Bi, 16-Ci emits light at a first wavelength, and at least another set of light sources 18, respectively 18A, 18B, 18C, emits light at a second wavelength different from the first wavelength. Typically, the first wavelength is in the near infrared or infrared range, for example about 850 nm, so that the user's pupils remain unchanged when the light sources are illuminated.

[0038] The respective positions of each light source 16, 18 relative to the camera 30 are predetermined and fixed. The light sources 16-Ai, 16-Bi, 16-Ci of each set i are at the same distance or eccentricity e from the optical axis of the camera 30. i The range of eccentricity is typically 0 to 20 mm. Advantageously, the range of eccentricity is typically 0.5 mm to 20 mm.

[0039] Advantageously, the light sources consist of light emitting diodes or LEDs. For example, the photorefraction module 10 includes 9 LEDs arranged at three different eccentricities along three different meridians. The camera 30 is adapted and configured to capture eccentric photorefraction images of the individual's eye 2 with each light source 16-Ai, 16-Bi, 16-Ci, i=1,2,3 shining in succession. This configuration allows to obtain a set N=9 eccentric photorefraction images per measurement.

[0040] In another example, the photorefraction module 10 comprises twelve light sources arranged on three meridians with four different eccentricities.

[0041] FIG. 3 shows an arrangement for estimating the refraction of an individual's eye 2 using a system 50 according to the invention. The optometrist 4 or optometrist holds the system 50, is at the height of the individual's face and positions the position and orientation of the optical axis of the camera 30 so that the plane of the light sources 16, 18 is vertical. The photorefraction module 10 is at a distance DistZ from the individual's eye 2. Typically, the distance DistZ is about 1 meter and can be between 30 cm and 2 m. In one example, two distance measurements are considered: 1 m and 0.5 m. Thus, in case of a measurement at 1 m, an accurate result is not returned and the measurement is performed at a distance of 50 cm. Advantageously, the system 50 comprises a means for measuring the distance DistZ, such as an autofocus device. Alternatively, the subject 1 holds the system 50 and looks into the camera, either directly or by reflection on a mirror surface, to take a picture of himself for the autorefraction measurement. The system 50 has a very large positional tolerance and is non-invasive.

[0042] The system 50 also comprises a calculation module including a memory and a processor configured to execute program instructions stored in the memory to implement the method of estimating the refraction of an individual's eye according to the invention. The calculation module is located inside the photorefraction module 10 or inside the mobile device 20. Alternatively, the calculation module is located inside a remote computer in communication with the photorefraction module 10. The calculation module generates a set of simulated images based on data of the light source used, the position of the photorefraction module 10 relative to the individual's eye 2 and ophthalmic refraction data of the eye 2.

[0043] In a preliminary step, the system 50 is used to acquire a set of N off-center photorefraction images, as shown in FIG. 4. For example, the system uses M=9 light sources, as detailed in connection with FIG. 2. In one embodiment, the system 50 is configured to capture one image with each off-center flash fired sequentially. For example, the light source 16-A1 emits a flash while all other light sources are off, and the system 50 captures a first off-center photorefraction image T1 of the pupil of the eye 2. Then, the light source 16-A2 emits a flash while all other light sources are off, and the system 50 captures a second off-center photorefraction image T2 of the pupil. By repeating this flashing and acquiring process, the system 50 acquires N=9 off-center photorefraction images T1, T2, . . . , T9 of the pupil of the eye 2, each image corresponding to one light source at a determined position. The system then stores in memory a set of 9 acquired decentered photorefraction images T1-T9 of the pupil of the eye 2. With 12 LEDs, it takes less than 0.5 seconds to capture a set of 12 acquired decentered photorefraction images T1-T12. This short acquisition time allows maintaining a constant pupil diameter of the eye 2, lowering the risk of eye accommodation changing between different acquisitions. Increasing the number of light sources allows for a wider range of refraction measurements. Advantageously, the system 50 can be configured to simultaneously acquire images of both eyes of the subject, thus capturing a separate set of N acquired decentered photorefraction images for each eye. Each stored image is associated with a light source at a determined position relative to the subject's eye 2 and the camera 30.

[0044] The system 50 is easy to use. Capturing a set of images is as simple as taking a photograph, as shown in FIG.

[0045] Then, as shown in Fig. 5, the method is based on comparing the set of N simulated images with the set of N acquired decentered photorefraction images of the eye and adjusting a combination of parameters so that the set of simulated images matches the set of acquired decentered photorefraction images. More precisely, the method comprises an initialization step 100 for initializing parameters such as sphere, cylinder, axis values ​​of the eye 2; a step 110 for setting values ​​of parameters (parameters related to the pursued eye, including at least sphere, cylinder and axis, and optionally measurement parameters such as high order aberrations or HOAs, hardware related parameters and measurement distances); a step 120 for generating a set of simulated images based on the values ​​of the parameters set in step 110; a step 130 for comparing the set of simulated images with a set of target images, i.e. a set of acquired decentered photorefraction images of the pursued eye; and a step 140 for optimizing or minimizing an estimator of the difference between the set of simulated images and the set of target images. Steps 110, 120, 130, and 140 are repeated until the minimum is found.

[0046] The method illustrated in FIG. 5, like any optimization or minimization algorithm, requires a starting point or initialization step 100 .

[0047] When using an optimization algorithm, one seeks to optimize the merit function of the algorithm, which is typically minimized.

[0048] When using a minimization algorithm, one seeks to minimize an estimator of the difference between a set of simulated images and a set of target images.

[0049] The initialization step 100 may rely on data from the subject's previous prescription, making it possible to initialize the sphere, salinity and axis values ​​of the eye being pursued.

[0050] Alternatively, the initialization step 100 is based on conventional known methods such as meridians, which will be explained in relation to Fig. 6. The different processing steps are listed below. The ocular aberrations can be addressed in the first order by the spherical (S), cylindrical (C) and axial (A) components of this cylinder, which can be represented according to this diagram of Fig. 6. The dots 16-A, 16-B, 16-C are aligned with each meridian X A , X B , and X C The highest eccentricities on the three meridians correspond to the infrared LEDs, respectively. At least three meridians are needed to identify S, C, and A. Therefore, at each meridian, the pupil diameter and the rising or falling front, i.e., the sharp transition between high and low gray levels, are calculated for the corresponding image T j , thus translating the dark crescent typical in photorefraction measurements.

[0051] Figure 7 shows, for example, meridian X A The straight line represents the meridian X. A Circle 42 represents a simulation of the pupil of the eye, with a pupil diameter D P By image processing, a dark crescent 41 and a light crescent 51 can be distinguished. Curve 43 represents a theoretical crescent based on a pure sphere value S. Curve 44 represents a theoretical crescent based on a combination of a sphere value S and a cylinder value C. The intersections of curves 43 and 44 with the meridian give the size D of the dark crescent 41. CR Alternatively, the size of the bright crescent 51 can be used. The refractive power along the considered meridian is derived from the following formula [1]:

number

[0052] Therefore, the three meridians X A , X B , and X C The refractive power is measured at each of the meridian X and X = 0. This allows at least three refractive powers to be determined from three different angles θ, as shown in FIG. A is at an angle of 30 degrees, and meridian X B is at an angle of 150 degrees, and meridian X C is at an angle of 270 degrees. The sphere (S), cylinder (C), and axis (A) data are derived from the following formula [2]: P(θ)=S+C·sin 2 (θ-A)

[0053] Equation [1] is theoretically restricted to eyes where the refractive error is purely spherical. Nevertheless, the method provides an initial value for the simulator parameters, thereby providing a usable starting point for the subsequent minimization algorithm.

[0054] The step of generating a set of simulated images 120 aims to generate images similar to those provided by the camera. This step starts with initial values ​​of the parameters derived from the previous step 110.

[0055] To perform step 120, the processor uses a simulation model. For example, the simulation model is based on a geometric-optical model of the light intensity distribution. For example, the publications R. Kusel, U. Oechsner, W. Wesemann, S. Russlies, E Miermer, and B. Rassow, “Light-intensity distribution in eccentric photorefraction crescents”, J. Opt. Soc. Am. A15, 1500-1511 (1998) and A. Roorda, MCW Campbell, and W. R Bobier, “Geometrical theory to predict eccentric photorefraction intensity profiles in the human eye”, J. Opt. Soc. Am. A12, 1647-1656 (1995) describe such a geometric-optical model. Alternatively, the simulation model is based on a ray tracing model, e.g., based on Zemax optical design software, as detailed in YL. Chen, B. Tan, and JWL Lewis, "Simulation of eccentric photorefraction images", Optics Express 11, 1628-1642 (2003). Nevertheless, the simulation model can be based on other models among many others, and is not based solely on a geometric-optical model or a ray tracing model.

[0056] The simulation model takes into account different types of parameters. More precisely, the parameters belong to the following categories: 1 - ophthalmic parameters related to the eye being pursued, 2 - hardware related parameters, and 3 - parameters linked to the measurement protocol.

[0057] More precisely, the eye-related ophthalmic parameters sought include at least one of the following parameters: - Pupil radius (R), - Spherical power (S), which is the low order aberration (LOA), - Cylinder (C), which is the LOA, - Cylinder axis (A), which is the LOA, - Higher Order Aberrations (HOA), represented by Zernike coefficients A6w to A14w, - pupil half deviation, - gaze direction, - "Red Reflection" (RR), which represents the bright offset caused by multiple diffuse reflections on the retina, - The Stiles-Crawford effect (SC), and / or - Purkinje reflex of the eyes.

[0058] The hardware-related parameters may include at least one of the following parameters: - the position and apparent size of the light source, e.g. LEDs, - LED wavelength, - the optical power of a light source, e.g. an LED, represented for all light sources by the G and R matrices; - the relative luminance of different light sources, - camera sensor gain and optionally camera noise, - camera exposure parameters, the position and size of the aperture of the camera's optical system, in particular the position and size of the camera's entrance pupil (dxCam, dyCam); - the focal length of the camera's optics, - Pixel size, dynamics, - The point spread function of the camera's optics.

[0059] The parameters linked to the measurement protocol may include at least one of the following parameters: - measuring distance, - The center position of the eye in the image.

[0060] The positions of the set of M light sources are crucial parameters for the simulation step 120 and the minimization or optimization step 140. However, the positions of the set of M light sources are fixed and are not optimized in step 140.

[0061] According to the simulation model used, the combination of all these parameters allows to generate one image per light source (here per LED). In the example shown in Fig. 10, the simulation model generates nine images 0S1 to 0S9 in this initial step of the simulation.

[0062] FIG. 10 shows a set of simulated images 0S1, 0S2, ..., 0S9 obtained with the initial values ​​of the parameters shown on the left. FIG. 10 also shows a set of target images T1, T2, ..., T9. The target images T1, T2, ..., T9 correspond to the acquired decentered photorefraction images captured and stored in memory as detailed in connection with FIG. 4. More precisely, the target images correspond to the central part inside the pupil of the acquired decentered photorefraction images, while the area around the pupil is set to a black background. FIG. 10 corresponds to iteration n° 0, i.e. the parameters are nominally 0 at the HOA and the same optical power (G and R matrices) for all LEDs. The sphere, cylinder and axis values ​​are filled in, for example, using the meridian method. The pupil radius R value is also filled in by measuring the best circle with respect to the gradients present in the image, representing the pupil edge, by image processing methods.

[0063] The calculation module calculates the difference between each of the simulated images 0S1, 0S2, ..., 0S9 at iteration n°0 and each of the corresponding target images T1, T2, ..., T9. Figure 10 also shows the generated difference images 0D1, 0D2, ..., 0D9 together with the numerical difference corresponding to each difference image. The difference images are calculated here based on the pixel-by-pixel intensity difference of the area inside the pupil image. An estimator, denoted E, of the difference between the two sets of images is calculated based on the following formula:

number

[0064] In the example of Fig. 10, an estimator of the difference between a set of nine simulated images 0S1, 0S2,...,0S9 and a set of nine target images T1, T2,...,T9 is calculated as the sum of nine numerical differences calculated for each difference image. In the example of Fig. 10, the value of the global estimator E0 at iteration n°0 is estimated to be 210.810 (in arbitrary units, i.e., au).

[0065] Alternatively, the acquired and simulated images are pre-processed, for example by a Fourier filter or other filter, and then compared with each other.

[0066] Then, at iteration n°1, the minimization algorithm starts a backpropagation loop. In practice, the input parameters of the model need to be adjusted to minimize the value of the estimator. For that purpose, various minimization methods are available: gradient-free methods such as the Simplex method or Nelder-Mead; gradient-enabled methods such as Levenberg-Marquardt, to name a few. In the following example, we use the Levenberg-Marquardt method. Minimization methods are standalone algorithms and can be used as black boxes.

[0067] For each input parameter, the calculation module calculates the gradient, i.e. the local derivative at that point. The gradient of each parameter then allows the calculation of the displacement vector of each parameter. Thus, in step 140, the minimization algorithm adjusts the values ​​of a set of input parameters used by the simulation model, such as ophthalmic parameters, hardware parameters, and / or parameters linked to the assumed protocol. FIG. 11 shows the result of an iteration step n°1 on the same set of target images T1 to T9. The adjusted values ​​of the input parameters are displayed on the left side of FIG. 11. FIG. 11 shows a set of simulated images 1S1, 1S2, . . ., 1S9 obtained with the adjusted parameter values ​​in iteration n°1. The calculation module calculates the difference between each of the simulated images 1S1, 1S2, . . ., 1S9 in iteration n°1 and each of the corresponding target images T1, T2, . . ., T9. FIG. 11 also shows the generated difference images 1D1, 1D2, . . ., 1D9, together with the numerical values ​​corresponding to each difference image. In the example of Fig. 11, the value of the global estimator E1 at iteration n° 1 is estimated to be 70.655 (au). Thus, a rapid drop in the global estimator of the difference between the set of simulated images and the set of target images from the first iteration is observed.

[0068] By repeating this procedure several times until the estimator E no longer converges, the computational unit reaches a local minimum.

[0069] The minimization algorithm is repeated until a stopping condition is reached. Each minimization algorithm can use a different stopping criterion. For example, a stopping condition can be when the estimator becomes smaller than a predefined value. Another stopping condition can be based on the residual between the S, C, A values, e.g., the minimization stops if the difference between two consecutive iterations is less than 0.01 diopters. Another stopping condition can be a maximum number of iterations of 20.

[0070] Continuing with the same example as in Fig. 10 and Fig. 11, the minimization algorithm is repeated until iteration N° 16, at which it is stopped. Fig. 12 shows the result of iteration step n° 16 on the same set of target images T1 to T9. The adjusted values ​​of the input parameters are displayed on the left side of Fig. 12. Fig. 12 shows a set of simulated images 16S1, 16S2, ..., 16S9 obtained with the adjusted parameter values ​​in iteration n° 16. The calculation module calculates the difference between each of the simulated images 16S1, 16S2, ..., 16S9 in iteration n° 16 and each of the corresponding target images T1, T2, ..., T9. Fig. 12 also shows the generated difference images 16D1, 16D2, ..., 16D9 with the numerical values ​​corresponding to each difference image. In the example of Fig. 12, the value of the global estimator E16 in iteration n° 16 is estimated to be 4.176 (au). Thus, a rapid drop in the overall estimator of the difference between the set of simulated images and the set of target images from the first iteration is observed. The method allows accurate estimation of ophthalmic parameters close to the target values ​​for the Zernike coefficients of low and high order aberrations such as sphere, cylinder, and axis.

[0071] The computation time to obtain these results is on the order of a few seconds. The method described herein allows obtaining accurate results for sphere, cylinder, and axis, and further allows estimating high order aberrations without the need for bulky equipment.

[0072] The example detailed in detail is based on acquiring each image while one light source is on and all other light sources are off. Alternatively, two light sources may be used simultaneously to acquire each decentered photorefraction image of the pupil of the considered eye. For example, FIG. 9 shows an image captured with two LEDs lit at the same time. The two LEDs are positioned along two different meridians to avoid the two crescents overlapping. Meridian X A A first bright crescent 51 generated by a first LED arranged along meridian X B4. Observe the second bright crescent 52 produced by a second LED positioned along the line 41. The first bright crescent 51 and the second bright crescent 51 are separated in the image by abrupt transition curves 41, 42 between high and low grey levels. It is therefore possible to determine the sizes of the two dark crescents from a single off-center photorefraction image capture.

[0073] More generally, a set of N images are taken with two LEDs lit simultaneously to generate two crescents in the pupil. Preferably, the two light sources are located on two different meridians. For example, a first image Im1 is captured using two LEDs n11, n12, a second image Im2 is captured using two other LEDs n21, n22 and a third image Im3 is captured using two other LEDs n31, n32.

[0074] A simulation step is fitted by generating a set of N corresponding images with the same configurations Im1 (LEDs n11, n12), Im2 (LEDs n21, n22), Im3 (LEDs n31, n32) for a given set of parameters (Sph, Cyl, axis...).

[0075] During the minimization step, the same principle as in optimization is applied: comparison of one or more Imj's.

[0076] According to another embodiment, two or more LEDs from the same meridian illuminate simultaneously, in which case the captured images are processed to determine the light gradient in the light distribution along this meridian, instead of determining the exact shape of the crescent.

[0077] The above method is implemented in a computer program, which may also be implemented as a smartphone application that is loaded onto a smartphone equipped with an add-on photorefraction module 10, as mentioned above.

[0078] The combination of the photorefraction module 10 capturing a set of decentered photorefraction images and corresponding applications for simulation and minimization is easily adapted for any customer around the world who owns a smart phone who wants a low-cost tool that quickly provides a good starting point before proceeding to subjective refraction testing.

[0079] The disclosed inventions enable equipping people in rural and / or emerging markets. The disclosed inventions enable non-professional optometrists, such as teachers, health workers, NGO volunteers, etc., to implement mass screening in schools, government programs worldwide to detect refractive errors.

[0080] The systems and methods disclosed herein may also be integrated into other medical devices as a technology block for eye analysis, for example for subjective refraction measurements.

[0081] Although representative processes and articles have been described in detail herein, those skilled in the art will recognize that various substitutions and modifications can be made without departing from the scope of what is described and defined by the appended claims. [Explanation of symbols]

[0082] 2 eyes 4. Optometrist 10 Photorefraction module 11 Wire Connection 16-A light source 16-B Light source 16-C light source 18-A light source 18-B Light source 18-C light source 20. Mobile Devices 30 Camera 50 Systems

Claims

1. 1. A method for estimating the refraction of an eye of an individual, comprising: a) providing a set of N acquired decentered photorefraction images of the eye, where N is an integer greater than or equal to 1; b) initializing values ​​of a set of parameters including at least a sphere; c) providing a set of N simulated off-center photorefraction images using a simulation model according to said values ​​of said set of parameters; d) determining an estimator of the difference between the set of N acquired eccentric photorefraction images and the set of N simulated eccentric photorefraction images; e) performing an optimization using an optimization algorithm to optimize the estimator by adjusting the values ​​of the set of parameters and repeating steps c) and d); f) inferring an estimate of at least one refractive parameter of the eye from the adjusted values ​​of the set of parameters; Including, Said step a) comprises: g) positioning in front of the eye a system including an image capture device having an aperture and a plurality of M light sources, where M is an integer equal to or greater than 1, the plurality of M light sources being eccentrically positioned around the aperture of the image capture device at predetermined positions along at least two directions transverse to an optical axis of the image capture device, each light source of the M light sources adapted and configured to illuminate the eye with a light pulse; h) illuminating the eye using each light source of the plurality of M light sources; and i) recording the set of N acquired eccentric photorefraction images of the eye using the image capture device, N being less than or equal to M, each image in the set of N acquired eccentric photorefraction images representing the eye illuminated by the light pulse of at least one light source of the plurality of M light sources; Including, The step c) comprises: j) generating the set of N simulated off-center photorefraction images using the simulation model according to the values ​​of the set of parameters; A method comprising:

2. 2. The method of claim 1 , wherein step b) comprises: selecting a set of three acquired decentered photorefraction images from the set of N acquired decentered photorefraction images, where N is 3 or more, and the set of three acquired decentered photorefraction images are recorded using three light sources positioned along three directions transverse to the optical axis of the image capture device; and processing each of the three acquired decentered photorefraction images to identify a size and tilt angle of a dark crescent and infer therefrom three refractive power values ​​for the three directions.

3. 3. The method of claim 1 or 2, wherein M is equal to N, and wherein step h) comprises illuminating the eye using each light source of the plurality of M light sources in sequence, and wherein step i) comprises acquiring each image of the set of N off-center photorefraction images when the eye is illuminated with the light pulse of one light source of the plurality of M light sources.

4. 2. The method of claim 1 , wherein N is less than M, and wherein step h) comprises illuminating the eye using two light sources of the plurality of M light sources simultaneously, and wherein step i) comprises acquiring one image of the set of N off-center photorefraction images when the eye is illuminated by the light pulses of two light sources.

5. The method according to claim 1 or 2, wherein the simulation model is based on a geometric-optical model or a ray tracing model of the light intensity distribution.

6. The method of claim 1 or 2, wherein the simulation model is further based on hardware parameters, such as the power of each light source, ophthalmic parameters, such as corneal reflex, and / or operational parameters, such as the position of the image capture device relative to the eye.

7. 3. The method of claim 1 or 2, wherein the estimator is based on pixel-by-pixel differences between the set of N acquired eccentric photorefraction images and the set of N simulated eccentric photorefraction images, or the estimator is based on a comparison of a pre-processed set of N acquired eccentric photorefraction images and a pre-processed set of N simulated eccentric photorefraction images.

8. The method according to claim 1 or 2, wherein the optimization or minimization algorithm is based on gradient-free methods such as Simplex or Nelder-Mead or gradient-based methods such as Levenberg-Marquardt.

9. The method of claim 1 or 2, further comprising determining a distance between the image capture device and the eye and / or an orientation of the image capture device relative to the eye.

10. 3. The method of claim 1 or 2, wherein the set of parameters further comprises cylinder, axis, interpupillary distance, high order aberrations, half interpupillary distance, direction of gaze, red reflex, and at least one other parameter of the eye among Stiles-Crawford parameters.

11. 1. A system for estimating an eye refraction of an individual, the system being adapted to communicate with a mobile device or a remote computer, the system comprising: an image capture device and a plurality of M light sources, M being an integer greater than or equal to 2, the plurality of M light sources being eccentrically positioned around the periphery of the image capture device at predetermined positions along at least two directions transverse to an optical axis of the image capture device; - the system is adapted and configured to illuminate the eye with light pulses using each light source of the plurality of M light sources; - said image capture device is configured to record a set of N acquired eccentric photorefraction images of the eye, N being an integer less than or equal to M, each image of said set of N acquired eccentric photorefraction images representing the eye illuminated by the light pulses of at least one light source of said plurality of M light sources; the system comprises a computing module including a memory and a processor, the processor executing program instructions stored in the memory to: k) initializing values ​​of a set of parameters including at least a sphere; l) generating a set of N simulated off-center photorefraction images using a simulation model based on values ​​of the set of parameters; m) calculating an estimate of the difference between the set of N acquired eccentric photorefraction images and the set of N simulated eccentric photorefraction images; n) minimizing the estimator using a minimization algorithm by adjusting the values ​​of the set of parameters and repeating steps l) and m); o) inferring an estimate of at least one refractive parameter of the eye from the adjusted values ​​of the set of parameters; A system configured to:

12. The system of claim 11 , wherein the image capture device and the plurality of M light sources are mounted on an attachment that is removably attached to the mobile device.

13. The system of claim 11 or 12, wherein the computing module is included in the mobile device or the remote computer.

14. The method includes the steps of: p) providing a set of N acquired eccentric photorefraction images, each image of the set of N acquired eccentric photorefraction images representing the eye illuminated by a light pulse of at least one light source of a plurality of M light sources eccentrically positioned around the image capture device at a predetermined position along at least two directions transverse to an optical axis of the image capture device, where M is an integer greater than or equal to 2 and N is an integer less than or equal to M; q) initializing values ​​of a set of parameters including at least a sphere; r) generating a set of N simulated off-center photorefraction images using a simulation model based on said values ​​of said set of parameters; s) determining an estimator of the difference between the set of N acquired eccentric photorefraction images and the set of N simulated eccentric photorefraction images; t) optimizing the estimator using an optimization algorithm by adjusting the values ​​of the set of parameters and repeating steps r) and s); and u) inferring an estimate of at least one refractive parameter of the eye from the adjusted values ​​of the set of parameters; A computer program product comprising one or more stored sequences of instructions that cause a computer to execute a program.

Images