System and method for digital measurement of stereopsis - Patents.com

Abstract

A system and method for stereopsis measurement is provided. A three-dimensional visual stimulus is generated, the visual stimulus being composed of a plurality of spatially filtered dot elements configured to isolate stereoscopic capabilities. The visual stimulus is then presented to a user via a display device, and input data is received in response to the user viewing the presented visual stimulus. The user's stereoscopic capabilities are then determined from the input data.

Description

[Technical field]

[0001] Related Applications This application claims priority to U.S. Provisional Patent Application No. 62 / 793,632, filed January 17, 2019, which is incorporated herein by reference.

[0002] The present disclosure relates to the field of vision, and more particularly to stereopsis measurement. [Background technology]

[0003] In normal binocular vision, the brain uses inputs from both eyes together. Due to the distance between the eyes, the image projected on the retina of one eye differs slightly from that projected on the other eye. These differences depend on the relative distance between the person looking at the eye and the feature (e.g., object) that forms the retinal image. The brain can use the horizontal disparity between features seen by both eyes to judge the distance to that feature. Such abilities are often assessed as part of a battery of tests used to monitor the health of the visual system. Poor performance against some accepted criteria may indicate i) deterioration of the retinal images captured by one or both eyes, ii) a defect in the processing used to combine the images from both eyes, or iii) a defect in extracting disparity information.

[0004] In some patients with impaired binocular vision, the visual cortex selects the image seen by the dominant eye and suppresses the image of the weaker eye. In particular in cases of amblyopia, the amblyopic eye is called the inhibited eye and the fellow eye is called the dominant eye. Suppression of the amblyopic eye may continue even if the optics of that eye are corrected (e.g., by a corrective lens) to obtain acceptable visual acuity. Thus, when the fellow eye is covered (e.g., by an eye patch), the image of the amblyopic eye can be used to perceive the world in the same way as the fellow eye. However, when the eye patch is removed and both eyes are used for vision, the suppression generally returns. This results in a loss of stereopsis, in addition to a decrease in the quality of vision.

[0005] The images used in stereopsis testing generally provide the information that the subject must use in responding, and the response can be measured to determine the subject's ability to perceive solids. An exemplary test image may be one in which multiple different objects are presented at different disparities, and the subject is then asked which object appears closest. For subjects with defective stereopsis, the inability to perceive different depths can be frustrating, and there is a temptation, consciously or unconsciously, to judge the answer in another way. If not carefully designed, stereopsis testing may be permeated with non-stereo cues that allow patients with stereoblindness to achieve an acceptable level of performance. To introduce disparity into the stimuli, features presented in depth need to be moved in different directions in the images presented to both eyes. This movement may result in a change in the local arrangement or density of features that indicate the area of ​​the display that has been manipulated. In a sharp-edged stimulus presented with red / green glasses, the very edges of the features to which the disparity is applied may appear colorful. Also, attempts to present sharp-edged stimuli at exact disparity (where their location falls between two pixels) can result in a "blurred" appearance where the manipulated features are visible. As a final example, the image of the weaker eye can be perceived by closing the fellow eye, so that the subject can effectively perceive the change in disparity of the object in the image pair by blinking with the fellow eye. This allows the subject to know what answer to give even if they do not involve stereopsis in their answer. A robust stereopsis test should minimize the performance of patients who are not performing the task based on stereoscopic disparity depth perception.

[0006] Furthermore, conventional stereoscopic tests may not be useful for patients with poor vision in one or both eyes. This is especially problematic for amblyopic patients, whose weaker eye's optics may be corrected but whose vision is still not good. Images presented in conventional stereoscopic tests generally have hard edges. These hard edges are in details (high spatial frequencies) that may not be perceived by patients with low vision. Hard edges are perceived by patients with low vision as being "blurry". The resulting loss of information from the stereoscopic test stimuli puts these subjects at a disadvantage compared to those with well-corrected vision. This therefore confounds the measurements. For this reason, it may not be clear whether a relatively poor performance in a stereoscopic test is due to a problem with stereoscopic vision or is simply a result of the subject's poor vision in one or both eyes.

[0007] Currently available techniques for testing stereopsis do not have a measure of error associated with the measured sensitivity. Also, existing techniques are limited in the range of disparities that can be tested. This reduces the usefulness of these techniques for measuring stereopsis, especially when stereopsis is impaired. Patients with poorer stereopsis may fall outside the testable range and may be erroneously considered "stereoblind." Conversely, patients may have such good stereopsis that they cannot be accurately measured with the current tests. This is the case when the smallest disparities are presented and the test can still be performed. In addition, due to the limited test range and the inability to measure errors, existing techniques cannot be used as the primary outcome measure in clinical trials where improvement of stereopsis is required as a result of treatment. Furthermore, current techniques present subjects with stimuli from books with a limited set of disparities. This does not allow accurate measurement of stereopsis in subjects whose abilities fall between the given levels.

[0008] Thus, currently, no stereoscopic test is available that accurately measures stereoacuity over a wide range of disparities and the associated errors. This is particularly problematic for measuring subjects who have poor stereoacuity and require large disparities to perceive depth, and in situations where changes in stereoacuity as a result of disease or treatment must be evaluated. Thus, there is room for improvement. Summary of the Invention [Means for solving the problem]

[0009] According to a first broad aspect, there is provided a system for stereopsis measurement, the system including a processing unit and a non-transitory memory communicatively coupled to the processing unit, the non-transitory memory including computer readable program instructions executable by the processing unit for generating a three-dimensional visual stimulus, the visual stimulus being comprised of a plurality of spatially filtered dot elements configured to isolate stereopsis function, presenting the visual stimulus to a user via a display device, receiving input data in response to the user viewing the presented visual stimulus, and determining the user's stereopsis ability from the input data.

[0010] According to a second broad aspect, there is provided a computer-implemented method for stereopsis measurement, the computer-implemented method including: generating, in a computing device, a three-dimensional visual stimulus, the visual stimulus consisting of a plurality of spatially filtered dot elements configured to isolate stereopsis function; presenting the visual stimulus to a user via a display device; receiving input data in response to the user viewing the presented visual stimulus; and determining the user's stereopsis ability from the input data.

[0011] According to a third broad aspect, there is provided a non-transitory computer-readable medium having stored thereon program code executable by at least one processing unit for generating a three-dimensional visual stimulus, the visual stimulus being comprised of a plurality of spatially filtered dot elements configured to isolate stereoscopic function; presenting the visual stimulus to a user via a display device; receiving input data in response to the user viewing the presented visual stimulus; and determining the user's stereoscopic vision capability from the input data.

[0012] Further features and advantages of the present invention will become apparent from the following detailed description taken in conjunction with the accompanying drawings. [Brief description of the drawings]

[0013] [Figure 1] 1 is a flowchart of a method for digital measurement of stereopsis, according to an exemplary embodiment. [Diagram 2] FIG. 1 is a schematic diagram illustrating a three-dimensional visual stimulus in accordance with an exemplary embodiment. [Diagram 3] 3A-3B show an exemplary isotropic log Gabor dot and the corresponding intensity profile of a horizontal cross section through the image, according to an exemplary embodiment. [Figure 4] 4A-4F show the effect of blurring on the log Gabor dots of FIGS. 3A and 3B compared to hard-edge dots of similar size, according to an example embodiment. [Diagram 5] 5A-5D show proposed designs of areas that can be targeted for stimulation, according to an exemplary embodiment. [Figure 6] 6A-6C are exemplary psychometric function plots, exemplary extrapolation plots, and exemplary synthesis function plots, respectively, according to exemplary embodiments. [Figure 7] 2 is a schematic diagram of an exemplary computing system for implementing the method of FIG. 1, according to an exemplary embodiment. [Figure 8-9]8A-8C are isotropic log Gabor dots according to the present disclosure, and 9A-9C are Gaussian difference dots according to the present disclosure. [Figure 10] FIG. 1 is a schematic diagram of a Gaussian difference configuration according to the present disclosure. [Figure 11] 11A-11B are bandpass rings according to the present disclosure. [Figure 12-13] 12A to 12C are bandpass shapes according to the present disclosure, and FIGs 13A to 13B are diagrams showing the configuration of a spatial bandpass filter. [Figure 14] 1 is a proposed design of a visual stimulus according to the present disclosure. [Figure 15-1] 15A-15B show a visual stimulus according to the present disclosure. [Figure 15-2] 15C-15D show visual stimuli according to the present disclosure. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] It should be noted that throughout the accompanying drawings, like features are designated with like reference numerals.

[0015] Described herein are systems and methods for digital measurement of stereopsis. Using the systems and methods described herein, measurements of stereoacuity can be used as an endpoint measure, allowing clinical trials targeting binocular vision function. The systems and methods described herein can also provide a measure of confidence (measurement of error) of the stereoacuity threshold. In fact, the systems and methods described may allow testing of a large continuous range of disparities. Visual stimuli may be generated at any disparity within the disparity range, allowing accurate measurement of the stereopsis threshold.

[0016] Thus, the systems and methods described herein can be used by medical professionals (e.g., by ophthalmologists and optometrists in clinics) to assess a patient's stereopsis. The systems and methods described herein can also be used as a primary outcome measure in clinical trials aimed at treating abnormal binocular function. In addition, the systems and methods described herein can also be used in school health screenings (designed to detect the presence of visual problems that impair stereopsis in one or both eyes) and / or career-related assessments to test the ability to perform tasks that depend on visual abilities (e.g., driving, flying).

[0017] A method 100 for digital measurement of stereopsis according to one embodiment is now described with reference to Fig. 1. In step 102, a three-dimensional (3D) visual stimulus composed of spatially filtered dot elements is generated. Then, in step 104, the visual stimulus is presented to a subject (also referred to herein as a user) using a suitable 3D display device and / or system, with a different image being presented to each eye of the subject. Then, in step 106, data related to the subject is collected and analyzed to realize a stereopsis measurement of the subject.

[0018] As shown in FIG. 2, the stimulus 200 includes a number of dots 202, which are spatially bandpassed circular wavelets that are less susceptible to blur. The arrangement of dots is presented within a "fusion fixation" frame 204. The purpose of this frame is to encourage proper convergence of the eyes in the plane of the screen. A high contrast patterned border should be easily fused between the eyes. The frame 204 may be presented identically to both eyes. The frame may thereby encourage the eyes to converge the image properly. Although the frame 204 is shown as a high contrast patterned border, it may take other forms, such as a solid border, a dashed border position, a zigzag border, or a curved border. Additionally, the frame 204 may be separated from the dots 202 by a space. This space may prevent the patient from comparing features (such as the dots 202) against the frame 204 to more easily find the disparity of the target. The patient may otherwise consciously or unconsciously compare the disparity in the frame and the dots, or compare the monocularly visible parts of the features in the visual stimulus to each other. The visual stimuli 200 are generated to isolate the patient's stereoscopic vision. The stimuli are designed to minimize any other cues that the subject may use to perform a given stereoscopic task. As used herein, the term stereoscopic task refers to a stereoscopic test in which the subject is presented with and prompted to respond to the stimuli 200, as described in more detail below.

[0019] In particular, the stimulus 200 is generated in step 102 by tiling dots 202 in a square grid with a defined average spacing (not shown) between the dots 202. Each dot 202 is randomly assigned to be either a black or white dot before being placed on the grid. The x and y coordinates that control the placement of each dot 202 are perturbed by a random sample from a uniform distribution to place the dots 202 within a displacement range of the dot from a given original position, thereby breaking up the impression of a regular grid structure. Ensuring that the displacement range is less than half the dot spacing allows for adjacent dots 202 to not overlap, at least before parallax is introduced.

[0020] While FIG. 2 shows a mixture of dots, including dots with both "black" and "white" appearances, one of skill in the art will appreciate that stimuli 200 according to the present disclosure may, in some embodiments, include dots of only one shade. The dots 202 may be colored such that viewing with colored glasses (e.g., red / green glasses) affects how the dots appear. A stereoscopic presentation may be achieved by presenting the left and right eye images in different colors (anaglyph). Thus, color filters may be placed in front of each eye to ensure that only the appropriate image for each eye is visible. In some embodiments, the color of the dots 202 may not be readily apparent when viewed with the naked eye.

[0021] As described above, the visual stimulus may be composed of spatially filtered dot elements. In general, a spatially filtered dot element may be any element composed of one or more dots generated using a spatial filter. A spatially filtered dot element may include modulation of a pixel having a center and one or more edges. A bandpass amplitude spectrum may give different regions of the dot a blurred appearance. A spatially filtered element may be configured such that its center may be located at a position between pixels of the display device on which it is displayed. It will be recognized by those skilled in the art that a spatially filtered element may take a variety of forms. Some exemplary forms are described below, with Figures 3A-3B and 8A-8C illustrating log Gabor dots, Figures 9A-9C and 10 illustrating difference of Gaussians (DoG dots), Figures 11A-11C illustrating annular dots, and Figures 12A-12C and 13 illustrating spatially filtered dot elements of any shape. However, any type of spatially filtered element may be used without departing from the scope of the present disclosure.

[0022] In one embodiment, the dots 202 are isotropic log Gabor dots defined in the Fourier domain to have equal energy in each orientation and a bandpass spatial frequency spectrum, as shown in Figures 3A-3B. This spatial frequency spectrum is a Gaussian spectrum on the logarithmic frequency axis. There are several advantages to using log Gabor dots compared to using hard-edged square or circular dots. First, stimuli with step-edge luminance boundaries are vulnerable to any blurring introduced by the visual system of subjects with substandard visual acuity. This is illustrated in Figures 4A-4F, where the top row shows a stimulus constructed from the proposed bandpass log Gabor dots (Figure 4A) and the bottom row shows an even stimulus constructed from hard-edged circular dots (Figure 4D). The second row (Figures 4B and 4E) shows the effect of blurring the stimuli presented in the first row. The third row shows the effect of blurring on the amplitude spectrum of the stimuli (Figures 4C and 4F). The spectrum shows that the log Gabor stimuli are less susceptible to blurring. Preliminary tests using the bandpass log Gabor dots disclosed herein further indicate that stereo thresholds are unaffected by removing all spatial frequency information above 4c / deg.

[0023] Figures 3A and 3B show a pair of log Gabor dots. Figures 8A-8C show a single log Gabor dot. The dots can be spatial band-pass dots, which refers to the spatial frequency present in the dot, or the variation in luminance over space.

[0024] Figure 3A shows two log Gabor dots in the spatial domain. The left dot has a negative peak to low luminance and may appear black in the center, while the right dot has a positive peak to high luminance and may appear white in the center. These dots may specifically be isotropic log Gabor dots defined in the Fourier domain to have equal energy in each orientation and a bandpass spatial frequency spectrum.

[0025] FIG. 3B shows the same log Gabor dot in the frequency domain. A Fourier transform can be used to convert the dot between the spatial and frequency domains. The bandpass spatial frequency spectrum can be clearly seen in FIG. 3B. A highpass filter can remove low frequencies, a lowpass filter can remove high frequencies, and a bandpass filter can include both highpass and lowpass filters. As can be seen from FIG. 3B, a bandpass filter can be applied in this case. Intensities higher than 1 and lower than -1 can be filtered out. It will be recognized by those skilled in the art that bandpass filters with different high and low values ​​can also be used without departing from the scope of the present disclosure.

[0026] In the spatial domain, a low-pass filtered image may appear blurry, while a high-pass filtered image may contain only smaller details such as edges. A band-pass filtered image may combine these two effects. These effects can be seen in Figure 8C, where the log Gabor magnitude spectrum has a peak with amplitude drops both above and below the peak.

[0027] In other words, the dot is spatially filtered because some control is being exercised over the contribution of different spatial frequencies to its construction. In the particular case of spatial bandpass, there is a particular range of spatial frequencies that will form the dot. Furthermore, the dot may be a log Gabor dot, which has a particular meaning for the shape of the amplitude spectrum, i.e. the amplitude per spatial frequency. A logarithmic transformation may be performed on the frequency axis, and the amplitude spectrum may be defined as a Gaussian distribution spectrum on that logarithmic frequency axis.

[0028] Figures 8A-C show a single log Gabor dot in the spatial domain, in the Fourier domain and as an amplitude spectrum taken as a radial slice from the center to the edge of Figure 8B, respectively. The spatial domain shown in Figure 8A is what might be seen in a visual stimulus constructed using log Gabor dots.

[0029] As shown in Figure 8B, a log Gabor dot can be constructed in the Fourier domain. The value of each point in the Fourier transform shown here can indicate the contribution of spatial frequency (f) and orientation (θ) to the image. The Fourier transform (F) can be expressed as:

number

[0030] In equation (1), f0 may be the peak spatial frequency and b may control the bandwidth. Transformation from the Fourier domain shown in Figure 8B to the spatial domain shown in Figure 8A gives log Gabor dot elements, which can then be used to construct visual stimuli.

[0031] Note from equation (1) that when f=0, the amplitude must be infinitesimally small. For practical purposes, the amplitude may be considered to be zero. This is important because it means that for log Gabor dots, there should be no overall change in the brightness of the screen as a result of adding the dot to the display. Areas of increased brightness are balanced by areas of decreased brightness. This may be particularly advantageous for visual displays. This means that no matter how spatial bandpass dots are positioned on the screen, they do not affect the overall brightness. For example, when dots are shifted in position to create parallax in an image, the density of the dots may become locally higher or lower. If the dots change the local brightness of the stimulus, the density change will make that part of the stimulus significantly brighter or darker than other parts. This may provide a non-stereoscopic cue that allows the patient to identify areas in the display where parallax is being manipulated.

[0032] In addition to the log Gabor dots described above, other spatially filtered dots can be used to construct the visual stimuli. In fact, the log Gabor dots may be only one of a huge number of spatially filtered dot options. There may be a special interest in developing the visual stimuli used herein while controlling their amplitude spectrum. This may allow the developer to design visual stimuli with desired characteristics. For example, high spatial frequency information that is not useful for low vision patients may be removed. It may be possible to develop such stimuli in the Fourier domain or in the spatial domain.

[0033] Another example of a spatially filtered dot may be a Difference of Gaussian (DoG) dot. Figures 9A-9C show a DoG dot in the spatial domain, in the Fourier domain and as an amplitude spectrum taken as a radial slice from the center to the edge of Figure 9B, respectively. As can be seen in Figure 9B, a DoG element may include a central portion and a peripheral portion.

[0034] DoG is a well-known function with a history of use for spatial filtering of images. Similar to the log Gabor described above, DoG elements can be used to directly replace dots in a visual stimulus. Alternatively, stimuli can be generated with dots that can approximate a delta function, and those dots can be filtered with the DoG function.

[0035] A DoG element can be created by subtracting one Gaussian distribution from another. The two Gaussian distributions can have the same mean but different standard deviations. Figure 10 shows a Gaussian difference (solid line) created by subtracting the surrounding Gaussian distribution (dashed line) from the central Gaussian distribution (gray line).

[0036] The radial distance of each pixel from the center of the generated dot can be given by the following equation (2). x,y =0.

number

[0037] The central Gaussian distribution can be defined by equation (3) below, and the surrounding Gaussian distribution can be defined by equation (4) below.

number

[0038] σ 周囲 is σ 中心 and the two sigmas may control the shape of the DoG. The DoG may be created by taking the difference between a central Gaussian and a surrounding Gaussian using equation (5). d(x,y)=g 中心 (r x,y ,σ 中心 )-a×g 周囲 (r x,y ,σ 周囲 ) (5)

[0039] In equation (5), a may control the relative contribution of the surrounding Gaussian distribution. The value of a may be set to result in a DoG with no global change in luminance by choosing a value that has a minimum amplitude at f=0. This is shown in Figures 9A-9C. It may therefore be possible to achieve a DoG dot with a similar amplitude spectrum (see Figure 9C) as the log Gabor dot.

[0040] While the above examples only show isolated spatially filtered dots in a grid configuration, it should be understood that these dots can be configured in other shapes as well. These can then be used to create spatial bandpass shapes as shown in Figures 13A and 13B. Spatial bandpass shapes can be considered as spatially filtered dot elements. This is mathematically equivalent to either i) arranging spatially filtered dots to form a particular shape (Figure 13B), or ii) starting with an image of sharp lines of the desired shape and convolving a single spatially filtered dot with that image (Figure 13A). A further method for generating spatial bandpass shapes is to mathematically define the outline of the shape and then render a cross section of that outline to have a particular shape. Two possible forms of these cross sections utilize the log Gabor cross section shown in Figure 8B or the DoG cross section shown in Figure 9B. Other functions can also be used, as described below.

[0041] Stimulus elements having a spatial bandpass shape according to the present disclosure can be created by fitting the fourth derivative of a Gaussian function. The spatial frequency of such a function can be set by σ, as given by Equation (6) below.

number

[0042] In equation (6), f0 can control the peak spatial frequency of the amplitude spectrum. Then, for each position x, the squared distance D of the cross section to the center x0 sq may be defined by equation (7).

number

[0043] Squared distance D sq may be used to define the luminance at each position of the cross section C. C may be defined by equation (8).

number

[0044] 11A-11D show an example of a spatially filtered dot element with an annular shape made with a cross section defined by Equation 8. FIG. 11A shows the annular dot in the spatial domain as it would appear in a stimulus made with such an annular dot. The annular dot has the appearance of two narrow dark rings flanking a wider light ring. This can also be seen in FIG. 11C, a horizontal cross section through the center of FIG. 11A. The positive luminance deviation of the central light ring is balanced by the negative luminance deviation of the adjacent dark rings. As a result, no overall luminance offset occurs when this stimulus is added to a display. The transition between the light and dark regions appears smooth because there are no high spatial frequencies that would give the impression of sharp edges.

[0045] Figure 11B shows the Fourier transform of Figure 11A. As with the spatially filtered dot, the stimulus energy is arranged in a circular band centered at the origin. The amplitude spectrum is presented in Figure 11D. This has a similar bandpass morphology as shown earlier for the log Gabor and Gaussian difference stimuli. Thus, the annular dot may offer similar advantages to the other dots mentioned above.

[0046] In some embodiments, annular dots may be generated by taking the coordinates of each pixel relative to the center of the annular dot. In some embodiments, the coordinates of each pixel may be given in Cartesian coordinates, and the center of the annular dot may be (0,0). Using these coordinates, the radial distance from the center of the dot may be calculated for each pixel using equation (9), and the intensity at each radial position may be related to the intensity at the corresponding Cartesian position by equation (10).

number

[0047] The above described method for creating annular dots can be extended to generate dot elements of any shape. For example, Figures 12A-12C show annular dots, circular square elements, and circular cross elements. The outline of the annular dot may or may not be mathematically defined, for example as a Fourier-like series of radial frequency pattern components.

[0048] The circular square shown in Figure 12B can be created similarly to the annular function outlined above using equations (9) and (10). The equation for the radius can be slightly modified from that given in equation (9) and the following equation (11) can be used instead.

number

[0049] The circular cross shown in Figure 12C is an example that can be realized with radial frequency pattern components. Since its radius at a given pixel depends on the angular position around the center of the image of that pixel, its angle needs to be calculated from the orthogonal x and y coordinates as shown in equation (12) below. θ x,y =atan2(y,x) (12)

[0050] Therefore, like the ring in equation (10), its shape is defined by its radius, where the radius is the θ x,y The circular cross is generated by the following formula (13): R = 1.4 + 0.6 × sin(3 × θ x,y )+0.2×sin(9×θ x,y )-0.12×sin(15×θ x,y ) (13)

[0051] Following these examples, one skilled in the art will recognize that a variety of shapes can be generated from the sinusoidal components.

[0052] It is possible to generate spatially filtered elements even for shapes that cannot be expressed using a sum of ring components as shown in equation (13). FIG. 13A shows how such dots can be generated. A desired outline can be generated, which in this example is a triangle. The outline can serve as a source image. The outline can then be filtered with a spatial band-pass filter kernel. The log Gabor function or DoG function described above can be used as the filter kernel. Convolving the source image with the filter kernel can result in a spatial band-pass ring shape with a predetermined outline. It will be recognized by those skilled in the art that a wide variety of dots can be generated in this manner. In some embodiments, the dots can be designed for a specific function.

[0053] Any of the dot element types discussed above may be used to create a visual stimulus 200, for example, as shown in Figure 2. Visual stimuli according to the present disclosure are described in more detail below.

[0054] In some embodiments, the visual stimulus 200 may be created by first rendering an image filled with placeholder dots. The placeholder dots may approximate a delta function. The image may then be convolved with one of the functions mentioned above, for example the log Gabor function. In the convolution, every placeholder dot may be replaced at its location with a spatially filtered dot element 202. This method may be particularly efficient. Alternatively, multiple spatially filtered dot elements 202 may be rendered first and then placed at random locations. These steps may produce a visual stimulus 200 with randomly placed spatially filtered dot elements 202.

[0055] Next, a sense of disparity may need to be created between what is presented to the left eye and what is presented to the right eye when viewing visual stimulus 200. Visual stimulus 200 may be modified to generate such a sense of disparity using the following steps.

[0056] To create a sense of parallax, the final x-coordinate of the dots 202 in the grid 204 may be further adjusted depending on the desired parallax (or depth) of the dots 202 relative to the surface of the screen on which the stimuli are presented, also referred to herein as the fixation plane. Shifting the dots inward for the left and right eyes (i.e., moving the dots to the right for the left eye and the dots to the left for the right eye) can create the sense of crossed parallax (i.e., the dots are closer than the fixation plane). Shifting the dots 202 outward can create the sense of uncrossed parallax (i.e., the dots are further away than the fixation plane). Figure 5A shows the design of four possible target areas in the stimulus, showing that the parallax of the dots 202 can be manipulated within the boundaries of a wedge shape. This creates the sense of the wedge floating. A task given to a patient viewing a visual stimulus may include identifying where in the stimulus the wedge appears to be at a different depth than the rest of the stimulus. The task can be made easier by increasing the parallax to create a larger depth difference, or made more difficult by decreasing the depth difference.

[0057] Those skilled in the art will recognize that Figure 5A illustrates only an example embodiment of a control pattern for dot disparity. As described above with respect to the wedge 502 shown in Figure 5A, the dots 202 in the grid 204 can be shifted to generate any shape known in the art.

[0058] 5B-5D show other embodiments of the target area. FIG. 5B shows four symbols: a pentagon, a star, a triangle, and a square. In some embodiments, the target area can be given the shape of one of these symbols or any other symbol known in the art. The patient can be asked to identify the symbol, to indicate the location of the symbol, or to determine whether the symbol appears to float in front of or behind the rest of the visual stimulus. FIG. 5C shows the letter E in four different orientations. In some embodiments, the target area can be given the shape of a letter or other symbol whose orientation is clearly recognizable. The patient can be asked to identify the location and / or orientation of the symbol, or to determine whether the symbol appears to float in front of or behind the rest of the visual stimulus. FIG. 5D shows four different letters. In some embodiments, the target area can be given the shape of a letter or number, and the patient can be asked to identify and / or locate the letter or number, or to determine whether the symbol appears to float in front of or behind the rest of the visual stimulus.

[0059] 15A-15D show exemplary visual displays according to the present disclosure. The displays shown in these figures may be similar to the display shown in FIG. 2. Specifically, the visual display may include a number of elements of dots that are first arranged on a grid, then offset by random amounts, and finally offset in subsets to create disparity between what is seen by both eyes. FIG. 15C shows the visual display as it would be displayed to a patient on a display device. The patient views the visual display wearing glasses with a red lens on the left eye and a green or blue lens on the right eye. In some embodiments, different glasses may be used, for example glasses with different colored lenses. FIGS. 15A and 15B show what the left and right eyes see individually, respectively. FIG. 15D shows what a patient with stereoscopic vision would see when wearing glasses and viewing FIG. 15C. That is, the patient sees a wedge-shaped portion of the visual display that is offset in front of the frame of the display device.

[0060] Note from Figures 15A-15D that cheating on such a test would be difficult. The wedge-shaped region that extends from the page when viewed by a stereoscopic patient wearing tinted glasses cannot be discerned by any other means when viewed in Figures 15A-15C. Rather, all areas of the visual stimulus appear to contain randomly placed dot elements, and no difference can be easily detected between the left eye stimulus shown in Figure 15A and the right eye stimulus shown in Figure 15B.

[0061] Some conventional stereopsis tests presented on digital displays have limited themselves to presenting stimuli only at disparities that can be achieved by shifting the stimulus features by an integer number of pixels (e.g., Hess, To, Zhou, Wang & Cooperstock, 2015. Stereo vision: the haves and have-nots. i-Perception 6, 1-5; Rodriguez-Vallejo, Llorens-Quintana, Montagud, Furlan & Monsoriu, 2016. Fast and reliable stereopsis measurement at multiple distances with iPad. arXiv 1609.06669). This significantly limits the range and number of disparities that can be tested. This also means that the disparities that can be tested depend on the characteristics of the device used, which is a disadvantage for creating a test that can be run on multiple devices. In order to generate more accurate disparities, it may be desirable to shift the dots 202 by an amount that is not an integer number of pixels. For example, very small disparities may require the dots 202 to be shifted by less than one pixel. To achieve this, sub-pixel interpolation can be used. However, this has the side effect of blurring dots presented in depth in the hard-edge stimuli used in conventional testing. This non-stereoscopic cue can allow patients to cheat. The design proposed herein avoids this problem in two ways. First, the inherent "fuzzy" appearance of the spatially filtered dots 202 of the present disclosure makes it much harder to perceive any slight blur introduced by sub-pixel interpolation. Second, depth can be incorporated into the proposed stimulus by moving the target wedge 502 forward (crossed parallax) and the background backward (uncrossed parallax) by the same amount, both. This means that the same degree of interpolation is applied to all dots 202 in the stimulus.

[0062] When a stimulus is presented to a subject (at step 104 in FIG. 1), the image passing through the subject's eye is filtered by the optical transfer function of the subject's eye before reaching the subject's retina. This has the effect of blurring any input. Subpixel interpolation takes advantage of this by presenting a stimulus that appears (after optical blur) to be located between two pixels on the screen. If the optical blur is relatively wide compared to the pixel spacing on the display, this allows the presentation of disparities below the resolution limit of the screen. It has been found that this interpolation, when applied to the proposed log Gabor dots 202, better approximates an image presented at the desired disparity (the disparity that subpixel interpolation is emulating) than when applied to hard-edge pixel dots.

[0063] Those skilled in the art will recognize that the above type of spatially filtered dot elements with blurred edges may be particularly advantageous for achieving sub-pixel interpolation. Sub-pixel interpolation may allow the dots 202 to be shifted more finely than would be possible using hard-edge visual stimuli. In particular, the centers of the spatially filtered dot elements may be able to be moved in sub-pixel increments. Such advantages may be achieved using any type of spatially filtered dot element. In contrast, hard-edge elements may only be moved in pixel increments, thereby being limited by the resolution of the display being displayed.

[0064] Now, referring again to FIG. 1, step 104 of presenting visual stimuli to the subject will be further described. The visual stimuli may be rendered in any suitable binocular split vision configuration that allows showing different images to each eye of the subject. In particular, the visual stimuli may be presented using a display system provided to the subject. For example, the visual stimuli may be rendered on an electronic device (e.g., on a screen of a desktop computer, a portable computer, a laptop computer, a tablet computer, on a head-mounted display or on a lenticular display, etc.). In some embodiments, the visual stimuli can be viewed using dedicated glasses. The electronic device exemplarily controls the presentation of the stimuli, and more specifically the disparity presented to the subject, using an adaptive procedure to identify the subject's stereoscopic threshold. In particular, the device presents stimuli composed of spatially filtered dot elements to reduce the effect of visual acuity on the obtained stereoscopic threshold. The device then analyzes the data obtained from the subject after the presentation of the stimuli using a maximum likelihood based fitting procedure that also obtains a measure of the reliability (measurement of error) of the threshold calculated from the subject's data (step 106 of FIG. 1).

[0065] To create a sense of depth from stereoscopic disparity, in step 104, different images are shown to the subject's eyes using a suitable 3D display device and / or system. In one embodiment, this can be achieved by presenting stimuli on a screen associated with a computer in step 104, as described above. For example, a liquid crystal display (LCD) computer monitor with a refresh rate of 120 Hz can be used. The screen can be synchronized with dedicated glasses that the subject uses to view the presented visual stimuli through any suitable system, such as the NVIDIA 3D Vision® 2 system. In one embodiment, wireless shutter glasses are used. However, it should be understood that other suitable glasses (e.g., ferroelectric liquid crystal glasses) can be applied, and that these glasses can be connected to the screen wirelessly or not. The subject's eyes can be stimulated separately by stimuli shown on the same screen by alternating the left and right eye images on the screen in accordance with the "opening and closing" of the glasses for the left and right eyes. Such stereopsis tests can be written using any suitable programming language (Matlab / Octave, Python, etc.) and the presentation of stimuli on the screen can be controlled using any suitable toolbox (such as Psychtoolbox for Matlab / Octave) or by custom-made software.

[0066] In another embodiment, the stereoscopic test can be presented on a handheld or portable computing device (e.g., tablet computer), and the presentation of the stimuli can be realized using red / green anaglyphs. The stimuli can be generated in left and right eye versions by removing information from the red or green color channel. The two resulting images can then be superimposed using appropriate computer software. When the subject watches with red / green filter 3D glasses, both of the subject's eyes can see separate images (with some crosstalk). It should be understood that other embodiments can be applied. For example, the stereoscopic test can be presented on a lenticular, polarized or dichroic tablet or using a dichroic projection system.

[0067] In yet another embodiment, the stereoscopic test may be presented on a head mounted display. For example, an Oculus Rift virtual reality headset may be used to present stimuli to both eyes of the subject in step 104. In this embodiment, no specialized glasses are required to view the presented visual stimuli.

[0068] It should be understood that the stereopsis test described herein may be developed using static (as described above) or dynamic stimuli. In the case of dynamic stimuli, the dot (reference number 202 in FIG. 2) disappears after being presented (e.g., on the screen of the device) for a random period (e.g., 100-300 milliseconds). Each time the dot 202 disappears, it is replaced by a new dot 202 that appears in a different position (with the appropriate disparity depending on its position in the stimulus). This dynamic stimulus may contain more information than a static stimulus, thereby assisting the subject to perform a given stereopsis task. In addition, a dynamic stimulus may lack non-stereo cues that would facilitate locating the target or reduce the subject's ability to scan the stimulus. This further ensures that the measured threshold reflects only the subject's sensitivity to stereoscopic disparity.

[0069] In one embodiment, the stereoscopic task performed by the subject is introduced with additional non-stereo cues to allow subjects with poor stereoscopic vision to learn how the task works before the test begins. Additional trials with non-stereo cues may be randomly introduced during the test, which may, for example, motivate children with poor stereoscopic vision to continue performing the task. Also, for subjects with malposition, there may be built-in adjustments to align the stimuli.

[0070] The stereopsis test described herein consists of a series of trials in which the subject is presented with a stimulus containing a target in one of four locations. In each trial, the subject may be prompted to select in which location the target, defined by the modulation of the disparity of the dots, is located (reference numeral 202 in FIG. 2). In trials in which the subject is unsure of the location of the target, the subject may be instructed to guess. In the desktop computer version of the test (above), the subject may respond by pressing a button on a keypad. In the tablet computer version of the test (above), the subject may tap the screen at the location of the target. It should be understood that the test may also be extended to receive verbal responses from the subject (e.g., "up", "down", "left" and "right").

[0071] After responding, the subject is presented with the next stimulus. The disparity of the stimuli can be controlled by a pair of staircase routines that respond to the subject's successes by making the stereoscopic task more difficult (smaller disparity) and to failures by making the stereoscopic task easier (larger disparity). After each trial, the disparity of the stimuli is recorded, as well as whether the target's location was successfully identified or not. This is later used to calculate the disparity at which the subject can reliably identify the target. It is also possible to obtain an online rough estimate of the stereoscopic threshold calculated from the staircase reversal (the disparity at which the subject goes from incorrect to correct or correct to incorrect).

[0072] In the first few trials of each test session, the stimuli may feature additional cues to the location of the target. The contrast magnitude of the dots 202 in the non-target areas is reduced so that the target is still visible even without stereoscopic sensitivity. This may be done to facilitate training the subject on how to perform the tasks required of them in the experiment. Also, as mentioned above, trials may have additional cues randomly introduced during testing to keep the stereoblind subject motivated. Data from these trials may be discarded before performing the analysis of stereoscopic function.

[0073] Each staircase records the number of trials the subject completes and the number of reversals (when the task switches from easier to harder or from harder to easier). The staircase ends when either the maximum number of trials or the maximum number of reversals is reached. Once both staircases have finished, the main part of the data collection is complete.

[0074] It may also be possible to perform a further test of the subject's ability to discriminate between crossed and uncrossed disparities at some multiple of the threshold (e.g., 3 times the threshold). In this case, wedges with disparity are present at all four target locations (as in FIG. 5), and the background dots have zero disparity. The disparity of three of the wedges points in one direction (crossed or uncrossed), while the disparity of the remaining foreign wedges points in the other direction. The subject then identifies the foreign wedge. This is tested until the subject reaches a certain number of answers (e.g., 12) or a lower number of correct answers (e.g., 6). The subject's behavior can then be converted into a measure of performance on this crossed / uncrossed disparity discrimination task.

[0075] The data collected during the experiment (at step 106 of FIG. 1) may then be presented in any suitable manner. FIGS. 6A-6C show one way of presenting and analyzing the collected data. In one embodiment, the collected data forms a table with three columns: the disparity of the stimuli (in arc seconds) in the first column, the number of trials tested at each disparity in the second column, and the number of correct responses at each disparity in the third column. The data may then be fitted with a suitable psychometric function (such as a logistic function) 602, as shown in FIG. 6A, using any suitable fitting routine as described herein above. The fitting of the psychometric function 602 outputs an estimate of the subject's disparity threshold. Bootstrapping methods may also be used to calculate the standard error and confidence interval of the threshold estimate.

[0076] Furthermore, the thresholds and errors can be used to assess when the normal analysis fails and a measure from a more robust secondary analysis should be used instead. This secondary analysis can be calculated over the accuracy scores achieved by the patient at each disparity to give a measure of performance that forms a composite measure with the range of thresholds that can be obtained from fitting a psychometric function. As performance improves, the patient can move from that part of the scale to a part where the patient's performance can be characterized by a psychometric function.

[0077] In other words, it should be understood that for some subjects, it may not be possible to successfully constrain the fitting of the psychometric function as in 602. In this case, the error associated with the estimate of the threshold value will be large. However, even in such cases, it may be useful to be able to provide some measure of the subject's performance. As mentioned above, in such cases, a robust analysis may be required. In some embodiments, a suitably robust measure may be derived by calculating the area between the subject's accuracy data (including an additional right-most point extending the polygon horizontally to any high x-value) and a line representing chance-level performance (also referred to herein as the 25% "guess" line). This area may be calculated using trapezoidal integration. FIG. 6B illustrates a robust area-based measurement method. In FIG. 6B, the area may be calculated after log2 transformation of the x-axis. To prevent outliers from inappropriately affecting the area calculation, a non-parametric bootstrap method may be performed from the raw data for each trial, and the median area calculated from those resampled bootstrap data sets may be used. The calculated areas were found to be highly correlated with the fitted psychometric function thresholds. The thresholds and errors from the fitting of the psychometric functions can be used to assess when the usual analysis (such as fitting a logistic function) fails and measures from a more robust analysis should be used instead. Those skilled in the art will recognize that the trapezoidal integration method shown in FIG. 6B is merely an exemplary extrapolation method. Other methods can be used to analyze data for which a function cannot be fitted without departing from the scope of the present disclosure.

[0078] Furthermore, this secondary scale can be utilized to create a composite performance scale (or composite performance axis). The scale can have a first part that corresponds to the thresholds estimated from the psychometric function of 602 (up to a given maximum value) and a second part where the scale is extended to include values ​​from a more robust analysis. Figure 6C shows this composite scale, where the first part gives the level of performance that can be measured by fitting the psychometric function. The second part of the scale extrapolates the thresholds and determines the subject's location on that part of the scale using an enclosed area method. As performance improves, the patient can move from that part of the scale to a part where the patient's performance can be characterized by the psychometric function.

[0079] Thus, in one embodiment, the overall results of the analysis of the subject data (as performed in step 106 of FIG. 1 and described above in this specification) may consist of: (1) a graph of a psychometric function such as 602 that allows the test administrator to keep track of the subject's performance; (2) an estimate of the threshold for detecting stereoscopic disparity and its associated standard error and confidence interval; (3) a measure derived from the area between the subject's data points and the 25% guess line, used as a back-up indicator of stereoscopic ability for subjects whose data could not be fitted; and (4) a score on the depth orientation discrimination task ranging from 6 / 6 (i.e., perfect) to 0 / 12 (i.e., a score below chance level such as 3 / 12).

[0080] Those skilled in the art will recognize that the method shown in Figures 6A-6C and described above may be used for any type of testing of stereopsis. That is, data may be collected at different disparity levels to indicate the patient's performance at each disparity level. The data may then be fitted with a suitable logistic psychometric function 602, as shown in Figure 6, using any suitable fitting routine described herein above. The collected data may be from a stereopsis test described herein or from any other type of stereopsis test known in the art. Such analysis may enable more accurate data to be collected from patients using new and traditional testing methods.

[0081] The system and method described herein can be further extended to perform measurements of stereoscopic vision at various positions in the visual field. This can serve as a diagnostic tool for diseases that cause localized loss of sensitivity to disparity. For example, a disease that affects the retina of one eye can prevent a subject from using disparity information at the affected position. This can allow retinal deformations to be identified at an earlier stage than would otherwise be possible.

[0082] As mentioned above, the present disclosure measures a patient's ability to detect depth from image disparity. To perceive depth from disparity, consistent input from corresponding regions of the retina of both eyes is required. When retinal disease reduces the quality of the output of one eye (or for some reason does not match the output of the other eye), stereopsis at the reduced position is impaired. Examples of diseases that can deteriorate the retina in this way include macular degeneration, glaucoma, and diabetic retinopathy. Proper treatment of these diseases can benefit from both early detection and the ability to monitor the progression of the disease. In current clinical practice, the health of various regions of the visual field can be assessed in a variety of ways. These include the Amsler grid, where patients are presented with a grid of squares and asked to report any abnormalities in their appearance. Preferential hyperacuity perimetry works on a similar principle, where patients are presented with a row of dots and asked to point out deviations from a straight line. Both of these tests rely on the patient's subjective report of their vision. There are also visual field tests in which a patient is presented with a target (a flashing light or other luminance contrast stimulus) at different locations in the visual field and is scored on their ability to detect the target at each location. These tests do not identify when the perceived object is seen but distorted, or when subsequent brain regions "perceptually fill in" for distorted or non-responsive areas. Furthermore, each eye must be tested separately.

[0083] The method according to the present disclosure uses the sensitivity to binocular disparity to identify areas of the visual field where stereoacuity is significantly worse than expected with normal binocular vision. This may indicate that a retinal abnormality in one or both eyes is affecting the visual acuity at that location. The positional shift that results in the perception of depth is very small, so the test should detect relatively small retinal changes that prevent the shift. The method is based on an optimized algorithm to quickly test the target area of ​​the visual field. The test also has the advantage of testing both eyes simultaneously, further reducing the time required to identify whether the patient needs further screening. The map of sensitivity obtained across the visual field does not rely on the patient's report of the location where the abnormally perceived object was seen. Instead, the test is a more objective test of the ability to see stereoscopic disparity at each location.

[0084] Testing of stereopsis across the visual field can be performed at a single disparity after verifying that the subject can reliably detect that disparity on the first test. In one embodiment, the targets can be disks presented at 10 locations around a fixation point. In each trial, the subject can be shown 1-4 disks. The subject can respond with the number of disks seen. An entropy minimization algorithm can then determine which disks the subject consistently fails to count on a trial-by-trial basis. This allows the program to focus on those locations (i.e., the locations where there is a suspected defect) for further testing.

[0085] In some embodiments, the systems and methods for measuring stereoscopic vision at different positions may be used to measure macular degeneration, diabetic retinopathy, or other diseases that cause asymmetric degeneration of the eye.

[0086] In some embodiments, the test of the entire visual field may be performed at multiple disparities. For example, the entire test method using the disk described above may be performed at multiple disparities. Such a test may determine the disparity that the patient can perceive at each position of the visual field. This may therefore allow a map of the stereoscopic vision ability of the eye to be made. This map may allow identification of weak areas where degeneration may occur. In some embodiments, the test of the entire visual field may be performed at a single disparity. Such tests may include other variations, such as, for example, the size of the target or the length of time the target is displayed. These tests may also be able to establish thresholds of performance in different areas. Thus, the present system and method can not only identify areas of weakness / strength, but also measure weakness / strength for different areas of the entire visual field of the eye.

[0087] In some embodiments, the test may identify areas where the eyes have different intensities. This may indicate that one eye is damaged / injured / degenerated at that location. The test may indicate which eye is degenerated, or other imaging or testing procedures may be used to determine which eye is degenerated.

[0088] 7 is an exemplary embodiment of a computing device 700 for performing the method 100 described above with reference to FIG. 1. The computing device 700 includes a processing unit 702 and a memory 704 having computer-executable instructions 706 stored therein. The processing unit 702 may include any suitable device configured to cause the computing device 700 or other programmable apparatus to perform a sequence of steps such that the instructions 706, when executed by the computing device 700 or other programmable apparatus, may cause the computing device 700 to perform the functions / acts / steps specified in the methods described herein. The processing unit 702 may include, for example, any type of general-purpose microprocessor or microcontroller, a digital signal processing (DSP) processor, a CPU, an integrated circuit, a field programmable gate array (FPGA), a reconfigurable processor, other suitably programmed or programmable logic circuitry, or any combination thereof.

[0089] The memory 704 may include any suitable known or other machine-readable storage medium. The memory 704 may include a non-transitory computer-readable storage medium, such as, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. The memory 704 may include any suitable combination of any type of computer memory located internal or external to the device, such as, for example, random access memory (RAM), read only memory (ROM), electro-optical memory, magneto-optical memory, erasable programmable read only memory (EPROM), electrically erasable programmable read only memory (EEPROM), or ferroelectric RAM (FRAM). The memory 704 may include any storage means (e.g., device) suitable for retrievably storing machine-readable instructions 706 executable by the processing unit 702.

[0090] Although shown in the block diagram as a group of separate components communicating with each other via different data signal connections, those skilled in the art will appreciate that the present embodiment is provided by a combination of hardware and software components, some components are implemented by a given function or operation of a hardware or software system, and many of the illustrated data paths are implemented by data communications within a computer application or operating system. Thus, the illustrated structure has been provided for efficiency in teaching the present embodiment.

[0091] Figure 14 shows an alternative embodiment of a visual stimulus that may use any of the shapes described above. In particular, the visual stimulus may use annular dots or dots of any shape as shown in Figures 11-12.

[0092] In a variant of the stereopsis test, annular dots as described in Figures 11A-11B may be used. In a variant of the stimulus test, four rings may be presented to the subject using binocular presentation as shown in Figure 14. A "left eye image" may be presented to the left eye and a "right eye image" may be presented to the right eye. This may be done by using two screens, by using specialized glasses as described above, or by using any other means known in the art. The position of the rings may be shifted to present them with stereoscopic disparity. Both images may include a patterned ring surrounding the stimulus ring, which may be set to zero disparity. The patterned ring may define a reference plane. The patient looking at the rings may be tasked with detecting which ring is in a different depth plane than the other rings. In the exemplary embodiment shown in Figure 14, the leftmost ring may be the target. Any shape of dot may also be used in this stereopsis test, as shown in Figures 12A-12C.

[0093] This stereopsis test can be used to investigate problems that are difficult to investigate using the dots described above. For example, the ring dots can be generated on a smaller screen than required for the random dot stereopsis test, and the correspondence of the rings seen by the two eyes can be less ambiguous than in the case of the random dot stimuli. Measurement procedures such as the staircase routine and analyses such as fitting and bootstrapping to obtain confidence intervals can be similar to those used in the random dot version.

[0094] The above-mentioned method and system may provide several advantages. In particular, the method and system may prevent fraud in stereoscopic vision testing. As mentioned above, the random dot test may include a number of dots that are mostly the same between the eyes, but a small percentage of the dots are shifted. Furthermore, the dots may have blurred edges. Similarly, the above-mentioned rings also have unclear edges and average zero luminance. Thus, in both the random dot test and the ring test, it may be difficult to recognize the target answer by any means other than the patient's stereoscopic vision. Thus, patients who do not have stereoscopic vision will not be able to answer correctly, and patients who have stereoscopic vision will answer correctly. This may prevent false negatives or false positives, making the test more accurate. This may also increase the value of the test in remote medical or screening applications where a medical professional or other operator is not supervising the test. This may increase the range of patients who can use the test.

[0095] Furthermore, the methods and systems disclosed herein may be more suitable for patients with low vision than conventional systems and methods. As shown in Figures 4A-4F, patients with low vision experience less deprivation when viewing images of spatially filtered dots than when viewing images of hard-edged dots. Because the edges of the spatially filtered dots are already blurred, patients are not penalized by the loss of high-frequency information. This allows them to be used for patients with low vision without correction or compensation for low vision. This may broaden the potential patients for whom the systems and methods may be used and / or may be used without additional steps for vision correction. The bandpass nature of the dots further opens up the possibility of testing at various "peak" spatial frequencies to assess stereopsis at different spatial scales.

[0096] Furthermore, the methods and systems disclosed herein can shift elements of the visual stimulus by sub-pixels, which may allow for more accurate measurements of stereoscopic vision. In particular, it may allow for identifying the point at which a patient transitions from being able to see stereoscopically to being unable to see stereoscopically. The analysis methods disclosed herein may further allow for such precise measurements to be made in a variety of situations and using a variety of tests.

[0097] It should be noted that the present invention may be practiced as a method, embodied as a system, and / or embodied on a computer readable medium. The embodiments of the present invention described above are intended to be exemplary only. Accordingly, the scope of the present invention is intended to be limited only by the scope of the appended claims.

Claims

1. 1. A computer-implemented method for stereopsis measurement, comprising: in a computing device including a processing unit, a display device, and a memory communicatively coupled to the processing unit, the method comprising: providing a three-dimensional visual stimulus, the three-dimensional visual stimulus being comprised of one or more spatially filtered elements configured to provide a stereoscopic function in a binocular separation configuration, the one or more spatially filtered elements being filtered using a bandpass filter that removes high and low spatial frequency regions, wherein the providing of the three-dimensional visual stimulus comprises: placing one or more placeholder dots on a grid; generating a three-dimensional visual stimulus with blurred dot edges by moving each of the one or more placeholder dots by a random amount and convolving the placeholder dots with a function to generate the one or more spatially filtered elements; displaying the three-dimensional visual stimulus to a user via the display device; While displaying the three-dimensional visual stimulus to the user, displaying a task corresponding to the three-dimensional visual stimulus via the display device; receiving the one or more sets of inputs corresponding to the three-dimensional visual stimulus for performing the displayed task; in response to receiving the one or more sets of inputs for performing the displayed task corresponding to the three-dimensional visual stimulus, storing data associated with the one or more sets of inputs for performing the displayed task; storing data associated with the one or more sets of inputs for performing the indicated task; determining a stereoscopic vision capability of the user based on the stored data, the determining the stereoscopic vision capability of the user including: determining a threshold at which the user can see stereoscopically based on a predetermined function; and determining an error associated with the determined threshold; outputting a result of the determination of the stereoscopic vision ability of the user; The computer-implemented method comprising:

2. The step of providing a three-dimensional visual stimulus includes: generating the three-dimensional visual stimulus; and obtaining the three-dimensional visual stimulus from the memory. The method of claim 1 , comprising one or more of:

3. The method of claim 1 or 2, wherein the spatially filtered elements are spatially filtered dot elements.

4. 4. The method of claim 3, wherein the spatially filtered dot elements include one or more elements selected from the group consisting of isotropic log Gabor dot elements, Gaussian difference dot elements, annular dot elements, a convolution of a spatial filter with an image object, and at least two tone elements.

5. 5. The method of claim 3 or 4, wherein the step of providing the three-dimensional visual stimulus comprises the step of generating the three-dimensional visual stimulus by arranging the spatially filtered dot elements on a grid and then offsetting each of the spatially filtered dot elements by a random amount.

6. The method of claim 5 , wherein one or more of the spatially filtered dot elements have centers offset from a pixel of the display device to present the element with sub-pixel parallax.

7. 7. The method of claim 1, wherein portions of the one or more spatially filtered elements appear at a first location for a first eye and at a second location for a second eye, and the portions are arranged in a single region of the three-dimensional visual stimulus.

8. A method according to any preceding claim, wherein one or more of the one or more spatially filtered elements appear offset from two or more of the one or more spatially filtered elements.

9. providing the three-dimensional visual stimuli, wherein all of the three-dimensional visual stimuli have a first disparity; displaying the three-dimensional visual stimulus to the user; displaying a task associated with the generated three-dimensional visual stimulus; receiving the one or more sets of inputs for performing the displayed task associated with the generated three-dimensional visual stimulus; storing data associated with the one or more sets of inputs for performing the represented task; determining the stereoscopic vision ability of the user based on the stored data; The method of claim 2 , further comprising repeating:

10. the predetermined function is a logistic psychometric function; The method of claim 1 , further comprising storing the stereoscopic performance of the user at each disparity and fitting the logistic psychometric function to a plot of the disparity versus the stereoscopic performance at each disparity.

11. The method of claim 1 , wherein the predetermined function is a logistic psychometric function and the associated error is determined through a bootstrap method.

12. A computer readable medium having stored thereon program code executable by at least one processor configured to perform the method according to any one of claims 1 to 11.

13. 1. A system for stereopsis measurement, comprising: A processing unit; A memory communicatively coupled to the processing unit, the memory including computer readable program instructions executable by the processing unit to perform the method of any one of claims 1 to 11; A system including:

Images