An ocular motion capture and tear film detection system and apparatus

Abstract

The application relates to an eye movement capture and tear film detection system and device, comprising an illumination module for illuminating the eyes of a subject, a collection module for recording images of the eyes of the subject, and a processing module for analyzing and processing the image information of the eyes of the subject, the illumination module comprising a first light source capable of periodically producing specular reflection in the tear film area of the eyes and a second light source capable of producing a light spot in the eyeball part of the eyes; the collection module collects a first image storing tear film information and a second image storing eyeball movement information; the processing module calculates the lipid layer thickness of the tear film of the subject based on the first image, and simulates the eyeball movement trajectory of the subject based on the second image; the processing module can also screen out the first image combined with the second image in which the eyeball is in a preset position based on the time sequence of the first image and the second image, so that the calculated thickness information is more accurate.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, and in particular to an eye motion capture and tear film detection system and device. Background Technology

[0002] The eye is the most important sensory organ in humans, and approximately 80% of the knowledge the brain acquires is through the eyes. Reading, recognizing characters, viewing pictures, observing people, and appreciating scenery all require the use of the eyes. The tear film is a protective membrane on the surface of the eye that separates the cornea from the air. The tear film consists of three layers: an inner layer of mucin, a middle layer of tear fluid, and an outer lipid layer. The lipid layer is generally unevenly and irregularly distributed on the tear film surface, and is dynamically changing. By blinking, the tear fluid and lipid layer are spread onto the ocular surface, which secretes mucin, forming the tear film. Under normal circumstances, the tear film ruptures in about ten seconds; after significant rupture, blinking is necessary to rebuild the tear film. The tear film is the outermost barrier of the human eye and an important component of the ocular surface immune defense system. Instability of the tear film will disrupt the integrity of the tear film barrier, allowing bacterial toxins and antigenic substances to easily come into direct contact with the corneal epithelium, leading to immunopathological reactions and damage to the cornea. The function of the tear film lipid layer is to prevent the aqueous layer of tears from directly contacting the air, thus reducing evaporation. The thickness of the lipid layer in the tear film plays an important role in the stability of the tear film, so measuring the thickness of the lipid layer in the tear film of the eye is of great significance.

[0003] Prior art includes instruments for imaging and measuring one or more thicknesses of the tear film layer, as disclosed in patent document CN103251375B; and ocular surface interferometry (OSI) apparatus, systems, and methods for measuring the tear film thickness (TFLT) of the tear film in the eye, wherein the TFLT includes lipid layer thickness (LLT) and / or aqueous layer thickness (ALT). The measured TFLT can be used to diagnose dry eye syndrome (DES). In some disclosed embodiments, a multi-wavelength light source can be controlled to illuminate the tear film. Light emitted from the multi-wavelength light source undergoes optical wave interference interactions within the tear film. The imaging device can be focused on the lipid layer of the tear film to capture the optical wave interference interactions of specularly reflected light from the tear film combined with background signals (one or more) in a first image. The imaging device can also be focused on the lipid layer of the tear film to capture a second image containing background signals (one or more) appearing in the first image.

[0004] A prior art tear film lipid layer thickness detection device, as disclosed in patent document CN111110185A, includes an illumination module, an image acquisition module, an image processing module, and an image analysis module. The image processing module, connected to the patent document, performs color analysis on the reflective area to determine the tear film lipid layer thickness of the subject. The image processing module includes a grayscale submodule, a normalization submodule, an image stretching submodule, a filtering submodule, a recognition submodule, and an extraction submodule. The advantage of this invention is that the provided tear film lipid layer thickness detection device measures the thickness of the tear film lipid layer by detecting the color distribution of the pattern generated by the interference of the tear film lipid layer with a light source.

[0005] A prior art patent document, CN114098628A, discloses an ocular surface tear film lipid layer illumination device attached to a corneal topography map. The device includes a corneal topography map and is characterized by: an ocular surface tear film lipid layer illumination device mounted on the corneal topography map, located below the subject's eye, illuminating the lower half of the corneal surface for observation of the tear film lipid layer; the device includes a rotating part and an illumination body, the rotating part being hinged to the corneal topography optical body, and the illumination body mounted on the rotating part, the illumination body being a surface-emitting light source; the light emitted by the ocular surface tear film lipid layer illumination device is tilted relative to the axial direction of the eye (the direction of the axial direction when the subject's eye is looking straight ahead), with an tilt angle ≥20° to reduce noise interference from transparent areas such as the iris behind the cornea, thereby reducing noise and improving the clarity of observation and imaging of the ocular surface lipid layer.

[0006] However, in the existing technical solutions, most devices used for detecting and analyzing the tear film are relatively large. During the examination, the examinee brings their head close to the detection device and keeps their eyes in a specific position on the device. The detection device emits a light source from inside to illuminate the examinee's eyeball, and then the light is reflected by the eyeball to the imaging device of the detection device, thereby collecting information about the tear film. The collected image information is then analyzed and calculated to obtain the thickness of the lipid layer. In existing imaging schemes, the imaging camera and the illumination device for emitting the light source are on the same side relative to the lens in the imaging scheme. In other words, with the lens of the imaging device as the dividing line, the eye is on one side of the lens, and the illumination device and the imaging camera are on the other side. Under this imaging scheme, there are strict requirements on the straight-line distance between the subject's eye, the illumination device, and the imaging device. Therefore, it is impossible to design a lightweight detection device. It can only be designed as a large device that can be placed on a table or desktop. Furthermore, since the subject needs to maintain a specific position at the front of the detection device, and the subject's physical strength or endurance is limited, they cannot maintain the same posture in the same position for a long time. Therefore, when using existing detection devices, the subject can only complete image acquisition for tens of seconds at a time. However, since long-term data detection is required to ensure accurate analysis of the thickness of the lipid layer in the tear film, it is necessary to repeatedly keep the eye at the front of the detection device.

[0007] Furthermore, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the inventors studied a large number of documents and patents when making this invention, but due to space limitations, not all details and contents were listed in detail. However, this does not mean that the present invention does not possess the features of these prior art. On the contrary, the present invention already possesses all the features of the prior art, and the applicant reserves the right to add relevant prior art to the background art. Summary of the Invention

[0008] For tear film layer detection, accurately understanding the tear film state cannot be achieved by simply examining a single image of the patient's eye. The lipid layer of the tear film is dynamically changing and related to eye movement. Therefore, it is necessary to simultaneously acquire images of the patient's eye and their eye movement trajectory to obtain a more comprehensive understanding of the tear film state. However, even with solutions that use acquisition devices to collect reflected light after illumination to the eye to obtain an image, and then simultaneously process and calculate the tear film layer image and the eye movement trajectory image, unfortunately, obtaining a good tear film layer image requires illuminating the eye with white or near-white light, as this provides better and clearer crystalline reflection. However, because the vitreous humor of the human eye has a non-planar structure, the scattered light during eye movement affects the determination of the image edges, thus impacting the judgment of the eye movement trajectory.

[0009] Additionally, due to limitations in processor and communication capabilities, existing medical products used for detection, unless they are large-scale medical devices listed in the background section of this solution, are limited by cost, communication capabilities, and energy consumption. This makes it difficult to utilize high-performance processors and large bandwidths. Consequently, when processing tear film layer images, human eye movement images, and high-precision image transmission simultaneously, the processor's processing power is limited, leading to problems such as delays, frame drops, and data errors in the processing of related image data. This significantly impacts the accuracy of the data, posing considerable application difficulties for wearable, relatively lightweight, and simplified tear film detection devices.

[0010] Therefore, in view of the shortcomings of the existing technical solutions, this application proposes an eye motion capture and tear film detection system, including: an illumination module for illuminating the subject's eyes; an acquisition module for recording images of the subject's eyes; and a processing module for analyzing and processing the subject's eye image information.

[0011] Preferably, the illumination module is capable of periodically generating a first light source that provides specular reflection in the tear film region of the eye and a second light source that generates light spots in the eyeball region of the eye.

[0012] The first light source provides uniform or substantially uniform intensity in all emission directions. The first light source is configured such that light emitted from it is specularly reflected from the tear film and undergoes constructive and destructive optical wave interference interactions (also referred to as "interference interactions") within the tear film of the eye. An acquisition module focuses on one or more sites of interest on the lipid layer of the tear film, and this acquisition module has spectral detection including detecting the spectrum of the light source. The acquisition module captures the interference interactions of light specularly reflected from the illuminated tear film, which are converged due to the focusing action of the acquisition module in a first image. The acquisition module then captures the optical wave interference signal representing the interference interactions of the light specularly reflected from the tear film, and generates an output signal (one or more) representing the interference signal in the first image. The first image may contain the interference signals of one or more pixels of a given image of the lipid layer captured by the acquisition module.

[0013] The second light source is fixed relative to the subject's head, making the position of the light spot fixed relative to the patient's eyes. For example, by determining the relative position between the markers inside the eye and the light spot, the eye movement can be determined. This invention does not use absolute position, but only relative position, which can greatly reduce communication and computing costs, and is especially suitable for small wearable devices (which work offline in most cases).

[0014] This invention advantageously employs a dual-light source arrangement. Specifically, the first and second light sources can be configured to produce different effects on the eye due to differences in wavelength and frequency. Specifically, the first light source for capturing tear film images and the second light source for capturing eye movement trajectories are configured to emit light at different wavelengths. This prevents mutual interference between the two sources. Specifically, when using a single light source to capture both tear film images and eye movement trajectories, the clarity of the tear film image and the accuracy of the eye movement trajectory may be reduced. Furthermore, when simultaneously analyzing the tear film image and eye movement trajectory on a single image, the accuracy of the results obtained is lower than the clarity of a single tear film image under a single light source and the accuracy of a single eye movement trajectory image under a single light source.

[0015] Preferably, the acquisition module is capable of continuously acquiring a first image storing tear film information based on reflected light from the first light source and a second image storing eye movement information based on the light spot formed on the eye by the second light source over a long period of time. The acquisition module is positioned on the same side of the protective lens as the light source illuminates the eye (in the prior art, the protective lens is designed as protective transparent glass for a protective device; in this design, it is designed as the lens of eyeglasses, i.e., the plane where the ring frame is located). The acquisition module and the eye are on the same side of the protective lens, while the illumination device is on the other side of the protective lens. Preferably, the first light source flashes at a first frequency, and the second light source flashes at a second frequency. The first frequency is configured to be at least greater than 60Hz, and the second frequency is configured to be greater than or equal to the first frequency.

[0016] Preferably, when observing the power supply waveforms of the two light sources, the peak of the first light source is configured to correspond exactly to the trough of the second light source, the peak of the second light source is configured to correspond exactly to the trough of the first light source, and there may be more than one peak of the other light source within the trough range of one of the light sources.

[0017] Furthermore, the electronic shutter of the first acquisition device used to capture the first light source is configured to be adapted to the first frequency, and the electronic shutter of the second acquisition device used not as the second light source is configured to be adapted to the second frequency, so that the images acquired by the acquisition devices are always presented when their corresponding light rays are at their peak.

[0018] Furthermore, in existing technologies, the illumination module and the acquisition module are typically positioned on the opposite side of the subject's eye. This design results in a relatively large or inconvenient-to-wear detection device, hindering long-term testing and conflicting with the need for extended data collection for tear film diagnosis. In this invention, the acquisition device and the subject's eye are positioned on the same side, while the illumination device is on the opposite side. This design allows for a lightweight detection device, such as eyeglasses, that can be worn directly on the subject's eye, thus meeting the requirements for long-term testing without affecting the subject's normal activities.

[0019] Preferably, the processing module calculates the thickness of the tear film lipid layer of the subject based on the first image according to the partitioning unit and algorithm unit of the processing module, and the processing module simulates the eye movement trajectory of the subject based on the second image according to the extraction unit and line drawing unit of the processing module.

[0020] Most existing technologies for calculating the lipid layer estimate the entire lipid layer observed on the ocular surface. However, the distribution of the lipid layer within the tear film is uneven, irregular, and dynamically changing. A simple overall estimate is inaccurate. Therefore, segmenting the tear film image and calculating the lipid layer separately can more accurately reflect its actual state. Secondly, in terms of eye tracking, simply tracking eye movements is not very meaningful. A data-driven description of the eye movement trajectory is needed to give the eye movements medical significance.

[0021] Preferably, the processing module can also filter out the first image based on the time series of the first image and the second image, combining the first image where the eyeball is in a preset position in the second image, so that the calculated thickness information is more accurate.

[0022] During the detection process, the human eyeball may move involuntarily, and the eyeball movement will affect the imaging information of the tear film layer. Therefore, based on the time series, the first image in the adjacent time series of the second image where the eyeball is in the central region can be selected. In this way, the offset rate of the tear film image of the selected first image will be reduced.

[0023] Furthermore, using this method, the eye position can be divided into five regions based on the eye position in the second image. Specifically, a Cartesian coordinate system is established at the eye. The processing module uses the pupil position under normal conditions as the origin of the Cartesian coordinate system, draws a central region with the origin as the center (with a radius of a certain distance), and uses the perpendicular and horizontal lines emanating from it as the horizontal and vertical coordinates. The regions in the other four quadrants, after excluding the central region, are respectively divided into the first region, the second region, the third region, and the fourth region.

[0024] Preferably, the processing module can also consider "the line connecting the inner and outer canthi as the horizontal axis," and "the vertical axis is constructed by the perpendicular line from the midpoint of the upper eyelid to this horizontal axis." This allows for the determination of the quadrant where the pupil is located, or by first binarizing the eye image to obtain a black and white image (a preferred method for creating a second image). In this case, the pupil and iris (black) are on an image with other parts of the eye (white) as the background. The quadrant of the current image can then be determined by analyzing the proportion of black areas in each quadrant. Since binarization has minimal computational overhead, and a resolution of 1024*768 is sufficient to determine the quadrant of the pupil's current time series, and it avoids the complex calculations caused by a moving pupil and moving iris, this minimization of computation significantly improves the battery life of wearable offline devices. Furthermore, as a black and white second image, it significantly reduces interference when performing subsequent calculations on the first image, thus improving the efficiency and accuracy of tear film analysis combined with the first image.

[0025] The four-quadrant division is highly beneficial for tear film analysis because there is a significant difference in the number of meibomian glands in the upper and lower eyelids. For example, there are approximately 25-30 meibomian glands in the upper eyelid and about 20 in the lower eyelid, and their distribution varies from person to person. As the largest sebaceous glands in the body, the meibomian glands have a total of about 50 openings, and their blockage varies. Therefore, dividing the glands into four quadrants allows for targeted identification of which quadrant truly contains meibomian glands that need to be unblocked. The blockage of the glands cannot be seen with the naked eye, especially in the meibomian glands of the upper eyelid, which are long and thin, reaching more than two-thirds of the height of the tarsal plate. The blockage inside these long, thin ducts is caused by the solidification of lipids, and it is difficult to pinpoint the location by touch. In other words, visual observation combined with palpation of the meibomian glands cannot assess which quadrant has dry eye syndrome or its severity. Therefore, the approach of this invention is of great guiding significance for precise treatment, especially since current treatment interventions for meibomian glands involve painful, even quite painful, mechanical massage lasting for tens of minutes. Moreover, if a meibomian gland is functioning normally but is subjected to external force, it may actually be detrimental to eye function, and the time-consuming nature of treatment due to the shortage of treatment equipment is also a problem.

[0026] According to this invention, the quadrant division method, which uses the line connecting the inner and outer canthi as the horizontal axis and the perpendicular line from the midpoint of the upper eyelid to this horizontal axis as the vertical axis, is advantageous for patients who cannot readily follow doctor's instructions (such as those with severe dry eye syndrome leading to frequent blinking and eye movement). While the division method using the pupil as the origin is coarser, it can still provide a rough diagnostic result for patients who follow doctor's instructions, allowing for efficient testing and further reducing testing time. This invention effectively provides both high-precision accurate testing and coarse testing modes, which is significant in scenarios with long patient waiting times.

[0027] At a specific time point within a time series determined by the processing module, the processing module, based on the region where the pupil is located in the current second image, "divides" at least one first image from a time point directly adjacent to the current second image within the time series to the region where the current second image is located. That is, the processing module assigns at least one first image to each second image, wherein the first image is directly adjacent to the second image in time within the time series determined by the processing module. The first image related to the tear film, obtained by illumination from the first light source, is used to analyze the function of the meibomian glands, and the second image related to the pupil position, obtained by illumination from the second light source, is used to determine the current position of the eyeball.

[0028] At this point, the first and second images, which are directly adjacent in time, are acquired by the same acquisition module from different light sources. Eye movement is inevitable during the acquisition time interval between the first and second images, leading to differences in calculation results. For example, when determining two directly adjacent first images centered on the time point of the second image, the processing module will inevitably determine differing calculation results. The pairing of the second and first images is only stored for subsequent analysis when the difference between the two images is below a preset threshold. When the "difference in calculation results caused by eye movement" exceeds the preset threshold, since the amount of binarized second image data is very small, the second round of computation can directly load two other second images that are temporally adjacent to the two first images but different from the second images exceeding the preset threshold without clearing the two first images in memory. This allows for the determination of a valid series of first images in the shortest possible time, enabling the acquisition of the required number of first images for tear film analysis with significantly less time and extremely limited CPU and battery power. As a result, tests that originally lasted several hours can be reduced to one hour or less. This improves the efficiency of patient diagnosis, which is crucial given the ongoing shortage of ophthalmologists due to long training cycles. Consequently, the overall cost of treatment is reduced.

[0029] Furthermore, the line connecting the inner and outer canthi is used as the horizontal axis, and the vertical axis is constructed by the perpendicular line from the midpoint of the upper eyelid to this horizontal axis. By dividing the eye into quadrants using coordinates, the tear film thickness in different quadrants is measured, allowing for specific analysis of the meibomian glands in different quadrants and the specific location of meibomian gland disease. For example, if the meibomian glands in the first quadrant are in a state of severe dry eye, while the meibomian glands in the adjacent second and fourth quadrants have a milder state of dry eye, and the meibomian glands in the opposite third quadrant have an even milder state, then the meibomian glands in the first quadrant can be identified as the main disease area. In other words, by analyzing and comparing the state of the meibomian glands in different quadrants, the main area of ​​meibomian gland disease can be identified, allowing for individual treatment of the main disease area and thus reducing patient suffering.

[0030] Preferably, the light source of the illumination module is directed toward the subject's eye via a polarizing component, a lens component, and a one-way coating in sequence. The polarizing component filters out interfering light received by the eye. The one-way coating is configured such that light rays from the side away from the eye pass through the one-way coating and are directed toward the eye. The one-way coating reflects the reflected light from the eye again, which is then collected by the acquisition module located on the same side as the eye.

[0031] The polarization component is mainly used to filter interfering light. The unidirectional coating enables the acquisition module and the subject's eye to be arranged on the same side. The unidirectional coating allows external light to enter the eye, thereby meeting the subject's normal visual field requirements. The light reflected from the eye cannot pass through the unidirectional coating, thus ensuring that the acquisition module on the same side as the subject's eye can acquire eye image information.

[0032] Preferably, the lighting module generates the first light source and the second light source in a periodic alternation, and the acquisition module is configured with a first acquisition unit capable of collecting the first image and a second acquisition unit capable of collecting the second image, the first acquisition unit and the second acquisition unit being matched with the alternation period of the first light source and the second light source.

[0033] The alternating period setting ensures that the first image and the second image do not interfere with each other, and the alternation period is very small compared to the total detection time, so the time series of the first image and the second image within the same period can be regarded as the same during calculation.

[0034] Preferably, the partitioning unit divides the first image into partitions centered on the pupil and applies a label to each partition. The algorithm unit calculates the thickness of the lipid layer in the eye based on the partitions divided by the partitioning module. The algorithm unit stores the lipid layer thickness result of each partition into the storage module of the system according to the partition label.

[0035] Preferably, the extraction unit can coordinate the second image and extract the coordinates of the pupil position in each second image that can be determined based on the light spot. The drawing unit connects the pupil coordinates extracted by the extraction unit according to the time sequence to simulate the movement trajectory of the eyeball within a preset time.

[0036] Preferably, the system further includes a display module capable of displaying the images and detection results collected by the acquisition module. The display module can at least display the maximum thickness, minimum thickness, average thickness, maximum thickness region, minimum thickness region, and thickness distribution map of different partitions of the lipid layer of the entire ocular surface. The display module can also display an ocular motion trajectory map arranged in a time series.

[0037] Preferably, the system's storage module can save the data collected by the acquisition module and the data processed by the processing module according to time parameters. The storage module can periodically delete the saved data. The storage module can also perform deletion or ignoring operations on specific data according to instructions.

[0038] Preferably, the processing module can also generate a corneal topography map reflecting eye information based on the data in the storage module. The corneal topography map is used to digitize and analyze the corneal morphology through a computer image processing system, and the obtained information is represented by color images with different features.

[0039] This application also proposes an eye motion capture and tear film detection device, which can be worn at least in the eye area of ​​a subject, wherein the device includes a support structure for holding the device and a lens structure for placing the illumination module and the acquisition module.

[0040] Preferably, the support structure is designed according to the positional structure of the human ear, eye and nose, with two annular frames that can hold the lens structure at the front of the subject's eye and a strip frame connecting the two ends of the two annular frames, and the two annular frames are connected by an arc frame that can fit against the upper end of the bridge of the nose.

[0041] Preferably, the lens structure includes at least a polarizing layer for placing the polarizer, a lens layer for placing ordinary lens components, and a coating layer for placing a unidirectional coating.

[0042] Preferably, the device further includes an external device for housing the processing module and the display module. The external device is wirelessly connected to the lighting module and the acquisition module inside the bracket structure and the lens structure. The display screen of the external device can be used to set the alternation period of the first light source and the second light source of the lighting module.

[0043] By configuring the aforementioned system onto a specific device, the device equipped with the system can be worn on the subject's eyes. While meeting the need for acquiring test data over a long period of time, it will not affect the subject's normal vision or movement. Medical staff do not need to repeatedly operate the instrument; they only need to set parameters such as the alternation period of the first and second light sources and the total test duration in the initial stage. During the test, the real-time situation can be observed through the display screen, and any unexpected situations can be detected in a timely manner. Attached Figure Description

[0044] Figure 1 This is a simplified relational structure diagram of the system of the present invention;

[0045] Figure 2 This is a simplified overall structural diagram of the device of the present invention;

[0046] Figure 3 This is a side view anatomical diagram of the eye area of ​​the present invention;

[0047] Figure 4 This is a schematic diagram of the actual anatomical structure of the eye area in this invention.

[0048] List of reference numerals

[0049] 100: Illumination module; 101: First light source; 102: Second light source; 200: Acquisition module; 201: First image; 202: Second image; 300: Processing module; 301: Partitioning unit; 302: Algorithm unit; 303: Extraction unit; 304: Line drawing unit; 400: External device; 401: Storage module; 402: Display module; 500: Support structure; 600: Annular frame; 700: Eye area; 701: Lacrimal gland; 702: Accessory lacrimal gland; 703: Lipid layer; 704: Tear film layer; 705: Mucin layer; 706: Inner canthus; 707: Outer canthus. Detailed Implementation

[0050] Dry eye syndrome is the most common and prevalent ocular surface disease. Its incidence rate in China is 20-36%, and in recent years, due to increased frequency and duration of eye use, its incidence has shown a significant upward trend. Researching this disease is of practical importance to human health. The causes of dry eye syndrome can be categorized into three areas: problems with the mucin layer 705, tear film layer 704, and lipid layer 703. Problems with lipid layer 703, specifically insufficient or abnormal lipid layer 703, prevent proper coverage of the tear surface, leading to premature tear film evaporation and rupture, thus failing to retain moisture and causing dry eye syndrome. For the diagnosis and treatment of dry eye syndrome, it is necessary to use a detection device to determine the thickness and rupture time of the lipid layer 703. However, existing detection methods rely on the subject manually holding the device in front of the imaging equipment. Since analyzing the state of the lipid layer 703 requires a long period of data collection, and the human body cannot remain in front of the device for extended periods, the subject needs to alternate between holding the device for a period and resting in a comfortable position during the long data collection process. This is both time-consuming and physically and mentally tiring for the subject. Therefore, the inventors have proposed the following several implementation methods. (See attached diagram.) Figure 1-4 The various embodiments of the present invention will be described in detail below.

[0051] Example 1

[0052] Figure 1 An eye motion capture and tear film detection system is shown, comprising: an illumination module 100 for illuminating the subject's eyes; an acquisition module 200 for recording images of the subject's eyes; and a processing module 300 for analyzing and processing the subject's eye image information.

[0053] Preferably, the illumination module 100 is capable of periodically generating a first light source 101 that provides specular reflection in the tear film region of the eye and a second light source 102 that generates a light spot in the eyeball region. That is, the light emitted by the first light source 101 is mainly used to illuminate the tear film layer 704 of the eye, and the light emitted by the second light source 102 can form a light spot at the pupil position of the eye, thereby acquiring different information in the eye.

[0054] Specifically, the first light source 101 may be a Lambertian emitter, which provides uniform or substantially uniform intensity in all emission directions. The second light source 102 may employ infrared light, and in this configuration, the device used to collect the second image 202 is matched with infrared light.

[0055] Preferably, the acquisition module 200, configured on the same side as the subject's eye, can continuously acquire, over a long period, a first image 201 storing tear film information based on reflected light from the first light source 101 and a second image 202 storing eye movement information based on a light spot formed on the eye by the second light source 102. The acquisition module 200 is equipped with a polarizing lens to acquire the first image 201 under the first light source 101, and an infrared lens to acquire the second image 202 under the second light source 102. The working time of the polarizing lens and the infrared lens is coordinated according to the alternating cycle of the first light source 101 and the second light source 102, ensuring that the polarizing lens works synchronously with the first light source 101 and the infrared lens works synchronously with the second light source 102.

[0056] The first light source 101 and the second light source 102 can be configured to produce different effects on the eye by relying on different wavelengths and frequencies of light. The first light source 101 used to capture tear film images and the second light source 102 used to capture eye movement trajectories are respectively configured to prevent mutual interference between the two. That is, when the same light source is used to capture tear film images and eye movement trajectories, the clarity of the tear film image and the accuracy of the eye movement trajectory may be reduced. In other words, the clarity of the tear film image and the accuracy of the eye movement trajectory on the same image are not as good as the clarity of the tear film image and the accuracy of the eye movement trajectory under a single light source.

[0057] Furthermore, in existing technologies, the illumination module 100 and the acquisition module 200 are typically positioned on the opposite side of the subject's eye. This design results in a relatively large or inconvenient-to-wear detection device, which is unsuitable for prolonged testing and conflicts with the need for long-term data collection for tear film diagnosis. In this invention, the acquisition device and the subject's eye are positioned on the same side, while the illumination device is on the opposite side. This design allows the detection device to be configured as a lightweight device, such as glasses, worn directly on the subject's eye, thus meeting the requirements for prolonged testing without affecting the subject's normal activities.

[0058] Preferably, the processing module 300 calculates the thickness of the tear film lipid layer 703 of the subject based on the first image 201 according to the partitioning unit 301 and the algorithm unit 302 of the processing module 300, and the processing module 300 simulates the eye movement trajectory of the subject based on the second image 202 according to the extraction unit 303 and the drawing unit 304 of the processing module 300.

[0059] Specifically, partitioning unit 301 divides the entire tear film lipid layer 703 on the ocular surface into partitions, centered on the center point of the image of the ocular surface tear film lipid layer 703 in the first image 201. Partitioning unit 301 divides the tear film lipid layer 703 into horizontal and vertical small regions according to n equidistant vertical lines and m equidistant horizontal lines. Then, algorithm unit 302 performs measurement processing on all horizontal and vertical small regions respectively. Alternatively, partitioning unit 301 divides the ocular surface into ring-shaped regions according to n straight lines with equal deflection angles passing through the center and m equidistant circular arcs. Then, algorithm unit 302 performs measurement processing on all ring-shaped small regions respectively, where n≥1 and m≥1.

[0060] Specifically, the extraction unit 303 performs coordinate processing on the entire ocular surface, extracts and records the coordinates of the light spot representing the position of the eyeball in the second image 202, and saves the extracted coordinate information locally in the processing module 300 or sends it to the system's storage module 401 according to the time sequence. The drawing unit 304 can draw each coordinate point on the two-dimensional image based on the cursor coordinates obtained by the extraction unit 303, and smoothly connect each coordinate point according to the time sequence, thereby simulating the movement trajectory of the eyeball.

[0061] Preferably, the processing module 300 can also consider the line connecting the inner canthus 706 and the outer canthus 707 as the horizontal axis, and the vertical axis is constructed by the perpendicular line from the midpoint of the upper eyelid to this horizontal axis. This allows the quadrant of the pupil to be determined, or by first binarizing the eye image to obtain a black and white image (a preferred method for obtaining a second image). In this case, the pupil and iris (black) are on an image with other parts of the eyeball (white) as the background. The quadrant of the current image can be determined by analyzing the proportion of black areas in each quadrant. Since binarization has minimal computational overhead, and a resolution of 1024*768 is sufficient to determine the quadrant of the current pupil time series, and avoids the complex calculations caused by the moving pupil and moving iris, this minimization of computation significantly improves the battery life of wearable offline devices. Furthermore, as a black and white second image, it significantly reduces interference when performing subsequent calculations on the first image, thus improving the efficiency and accuracy of tear film analysis combined with the first image.

[0062] Most existing technologies calculate the lipid layer 703 by estimating the entire lipid layer 703 observed on the ocular surface. However, the distribution of the lipid layer 703 within the tear film layer 704 is uneven, irregular, and dynamically changing. A simple overall estimation is inaccurate. Therefore, segmenting the tear film image and calculating the lipid layer 703 separately can more accurately reflect its actual state. Secondly, regarding eye tracking, simply tracking eye movements is not very meaningful; a data-driven description of the eye movement trajectory is needed to give the eye movements medical significance.

[0063] Specifically, according to Figure 3 and Figure 4 The diagram shows a partial structural distribution of the anatomical structure of the human eye region 700. The eye region 700 includes the lacrimal gland 701 and accessory lacrimal gland 702. The tear film layer 704 of the eye region 700 is formed by the secretions of the lacrimal gland 701 and accessory lacrimal gland 702. Outside the tear film layer 704 is the lipid layer 703, and inside the tear film layer 704 is the mucin layer 705. The inner corner of the eye region 700 near the bridge of the nose is called the inner canthus 706, and the inner corner away from the bridge of the nose is called the outer canthus 707. The coordinate axes are established based on the inner canthus 706 and outer canthus 707 of the eye region 700.

[0064] Preferably, the processing module 300 can further filter the first image 201 in the second image 202, based on the first image 201 and the second image 202 in the time series, to select the first image 201 in which the eyeball is located in a preset position in the second image 202, so that the calculated thickness information is more accurate. In other words, the first image 201 and the second image 202 in adjacent time series are bound together, and then the first image 201 is classified according to the specific coordinates of the pupil in the second image 202 based on the light spot measurement. That is, when the pupil falls in different areas in different time series, the subject's eye state is different, and the display state of its lipid layer 703 is also different. Classifying the first images 201 of the eye state in the same area into one category is more conducive to the accurate calculation of the lipid layer 703. For example, the first images 201 of the subject looking to the left in different time series can be classified into one category, and the time series of looking to the right can be classified into another category, etc.

[0065] During the detection process, the human eyeball may involuntarily rotate, which will affect the imaging information of the tear film layer 704. Therefore, based on the time series, the first image 201 in the adjacent time series of the second image 202 where the eyeball is located in the central region can be selected. In this way, the offset rate of the tear film image of the selected first image 201 will be reduced. In addition, based on the eyeball position of the second image 202, the eyeball position can be divided into five regions. Specifically, a rectangular coordinate system is established at the eye, with the pupil position in the normal state as the origin of the rectangular coordinate system. A central region is drawn with the origin as the center and a certain distance as the radius. The regions in the other four quadrants are divided into the first region, the second region, the third region, and the fourth region, respectively, after excluding the central region. Based on the time series, the first image 201 in the adjacent time series of the eyeball position falling in different regions is assigned to the corresponding region.

[0066] Preferably, the partitioning unit 301 divides the entire ocular surface of the first image 201 into partitions centered on the pupil and applies a label to each partition. The algorithm unit 302 calculates the thickness of the lipid layer 703 of the eye based on the partitions divided by the partitioning module. The algorithm unit 302 stores the thickness result of the lipid layer 703 of each partition into the system's storage module 401 according to the partition label.

[0067] Specifically, partitioning unit 301 divides all images within the same region using the same partitioning method. For example, partitioning unit 301 divides all first images 201 in the central region into horizontal and vertical small regions, and partitioning unit 301 divides all first images 201 in the first region into ring-shaped small regions. Alternatively, partitioning unit 301 uses completely different partitioning methods for images in different regions. For example, the first image 201 in the central region indicates that the subject's eyes are looking straight ahead without any eye movement, so the entire lipid layer 703 can be partitioned and calculated. Another example is the first image 201 in the first region, which is located in the first quadrant, indicating that the subject is looking to their upper left, causing eye movement. In this case, the tear film structure in the upper left part of the eye is obscured, so partitioning unit 301 partitions the lower right region of the tear film structure as a valid image and submits it to algorithm unit 302 for calculation. The selection method for valid images in other regions is the same. Within the storage module 401, the algorithm unit 302 records the calculation results for each region. The algorithm unit 302 can also integrate the calculation results for each region into a complete result for the tear film.

[0068] Specifically, the algorithm of algorithm unit 302 can use the color / pattern algorithm commonly used in the prior art to calculate the specific thickness of lipid layer 703.

[0069] Preferably, the extraction unit 303 can coordinate the second image 202 and extract the coordinates of the pupil position on the second image 202 based on the light spot measurement in each second image 202. The drawing unit 304 connects the pupil coordinates extracted by the extraction unit 303 according to the time sequence to simulate the movement trajectory of the eyeball within a preset time. The preset time can be twelve hours, twenty-four hours, etc.

[0070] Preferably, the light source of the illumination module 100 passes through the polarization component, the lens component, and the one-way coating in sequence and is directed toward the subject's eye. The polarization component filters out the interference light received by the eye. The one-way coating is configured so that light from the side away from the eye passes through the one-way coating and is directed toward the eye. The one-way coating reflects the reflected light from the eye again and is collected by the acquisition module 200, which is located on the same side as the eye.

[0071] Specifically, the first light source 101 and the second light source 102 of the illumination module 100 alternately pass through the polarization assembly, the lens assembly, and the unidirectional coating to reach the subject's eye according to instructions. The polarization assembly is specifically configured as a polarizer, the lens assembly is made of resin lens, the polarizing lens is attached to the side of the resin lens away from the subject's eye, and the unidirectional coating is applied to the side of the resin lens closer to the subject's eye. The coating is made of nano-chromium aluminum, and the nano-chromium aluminum layer enables an ordinary bidirectional light-transmitting lens to become a unidirectional light-transmitting lens.

[0072] The polarization component is mainly used to filter interference light. The one-way coating enables the acquisition module 200 and the subject's eye to be arranged on the same side. The one-way coating allows external light to enter the eye, thereby meeting the subject's normal visual field requirements. The light reflected from the eye cannot pass through the one-way coating, thus ensuring that the acquisition module 200 on the same side as the subject's eye can acquire eye image information.

[0073] Preferably, the lighting module 100 generates a first light source 101 and a second light source 102 in a periodic alternation. The acquisition module 200 is configured with a first acquisition unit capable of collecting a first image 201 and a second acquisition unit capable of collecting a second image 202. The first acquisition unit and the second acquisition unit are matched with the alternation period of the first light source 101 and the second light source 102.

[0074] The alternating period setting ensures that the first image 201 and the second image 202 do not interfere with each other, and the alternation period is very small compared to the total detection time, so the time series of the first image 201 and the second image 202 within the same period can be regarded as the same during calculation.

[0075] Specifically, the alternation cycle of the first light source 101 and the second light source 102 of the lighting module 100 can be configured through the connected display module 402. The alternation cycle can be adjusted according to the actual needs of medical staff. When adjusting the cycle, the working cycle of the first light source 101 and the polarizing lens of the acquisition module 200 that acquires the first image 201 is adjusted together, and the working cycle of the second light source 102 and the infrared lens of the acquisition module 200 that acquires the second image 202 is also adjusted together. The working cycles of the first light source 101 and the second light source 102 can be the same or different.

[0076] Preferably, the system further includes a display module 402 capable of displaying the images collected by the acquisition module 200 and the detection results. The display module 402 can at least display the maximum thickness, minimum thickness, average thickness, maximum thickness region, minimum thickness region, and thickness distribution map of different partitions of the lipid layer 703 of the entire ocular surface. The display module 402 can also display an ocular motion trajectory map arranged in a time series.

[0077] Specifically, all the above-mentioned data are obtained through the corresponding calculations performed by the processing module 300. The data can be selectively displayed on the display module 402 via touch control. The displayed data can be either data that has undergone visualization processing by the processing module 300 or basic data.

[0078] Preferably, the system's storage module 401 can save the data collected by the acquisition module 200 and the data processed by the processing module 300 according to time parameters. The storage module 401 can periodically delete the saved data. The storage module 401 can also perform deletion and ignore operations on specific data according to instructions.

[0079] Specifically, to ensure that the system does not malfunction due to excessive stored data, the processing module 300 will delete the data in the storage module 401 according to a certain period. If there is important data, medical staff can apply special markings to the data through the display module 402, such as whitelisting, so that the processing module 300 will skip such data when periodically deleting data.

[0080] Preferably, the processing module 300 can also generate a corneal topographic map reflecting eye information based on the various data in the storage module 401. The corneal topographic map uses a computer image processing system to digitize and analyze the corneal morphology, and the obtained information is represented by color images with different features.

[0081] Specifically, corneal topography can accurately measure and analyze the curvature of any point on the entire anterior corneal surface and detect corneal refractive power, making it a systematic and comprehensive quantitative analysis method for studying the morphology of the anterior corneal surface. A corneal topography system consists of three parts: ① a Placido disk projection system; ② a real-time image monitoring system; and ③ a computer image processing system. The Placido disk projection system projects 16–34 concentric rings uniformly onto the corneal surface from the center to the periphery. The diameter of the central ring can be as small as 0.4 mm, and the rings can cover the entire cornea. In the scheme of this application, a partitioning module can be used instead of the Placido disk projection system, an acquisition module 200 can replace the real-time image monitoring system, and a processing module 300 can replace the image processing system to complete the construction of the corneal topography. The corneal topography system can be used to examine dry eye conditions, such as examining the tear film lipid layer 703.

[0082] Example 2

[0083] like Figure 2 As shown, this application also proposes an eye motion capture and tear film detection device based on Embodiment 1. The device can be worn at least in the eye area of ​​the subject. The device includes a support structure 500 for holding the device and a lens structure for placing the illumination module 100 and the acquisition module 200.

[0084] According to the system scheme in Embodiment 1, the system can be integrated into a smart device. For example, in this embodiment, the present invention will design a smart glasses based on the system of Embodiment 1. Integrating the system into a lightweight smart device, such as smart glasses, allows the subject to wear the smart glasses for extended periods. The smart glasses can continuously collect eye information to complete the collection of test data.

[0085] Preferably, the support structure 500 is designed according to the positional structure of the human ear, eye and nose, with two annular frames 600 that can hold the lens structure at the front of the subject's eye and a strip frame connecting the two ends of the two annular frames 600. The two annular frames 600 are connected by an arc frame that can fit against the upper end of the bridge of the nose.

[0086] Specifically, the polarizing lens used to collect tear film information in the acquisition module 200 is set inside the support structure 500, which is supported on both sides of the subject's head and receives support from the subject's ears and the sides of the head, i.e. a common eyeglass frame. The difference from a common eyeglass frame is that one or more polarizing lenses and infrared cameras are configured on the inner side of each frame.

[0087] The illumination module 100 emits a first light source 101 using a Lambertian emitter, which provides uniform or substantially uniform intensity in all emission directions. A second light source 102 is emitted using an infrared emitter, and in this configuration, the device used to collect the second image 202 is matched with the infrared light.

[0088] The lighting module 100 can be configured at the top of two annular frames 600, with the emitters of the first light source 101 and the second light source 102 arranged alternately or side-by-side at the top of the annular frames 600. Alternatively, the lighting module 100 can be arranged around the annular frames 600, with a short annular ring extending from the end of the annular frame 600 away from the eye. The emitters of the first light source 101 and the second light source 102 are alternately distributed on the annular ring, and the first light source 101 and the second light source 102 emitted by both emitters have a certain angle to ensure that the light source accurately illuminates the eye.

[0089] Preferably, the lens structure includes at least a polarizing layer for placing a polarizer, a lens layer for placing a common lens assembly, and a coating layer for placing a unidirectional coating.

[0090] Specifically, the annular frame 600 sequentially holds the polarizing layer, the lens layer, and the coating layer inside the annular frame 600. The polarizing component in the polarizing layer is specifically configured in the form of a polarizing film. The lens component in the lens layer is a resin lens, and the polarizing lens is attached to the side of the resin lens away from the examinee's eye. The unidirectional coating in the coating layer is applied to the side of the resin lens closer to the examinee's eye. The coating is made of nano-chromium aluminum, and the nano-chromium aluminum layer can transform an ordinary bidirectional light-transmitting lens into a unidirectional light-transmitting lens.

[0091] Preferably, the device further includes an external device 400 for placing the processing module 300 and the display module 402. The external device 400 is wirelessly connected to the lighting module 100 and the acquisition module 200 inside the bracket structure 500 and the lens structure. The display screen of the external device 400 can be used to set the alternation period of the first light source 101 and the second light source 102 of the lighting module 100.

[0092] The external device 400, the acquisition module 200, and the lighting module 100 are wirelessly connected via Bluetooth, Wi-Fi, or LAN. The external device 400 controls the acquisition module 200 and the lighting module 100. The acquisition module 200 transmits the acquired image information to the processing module 300 via wireless connection. The processing module 300 processes the data in the aforementioned manner and then displays it visually on the screen.

[0093] By configuring the aforementioned system onto a specific smart device, the smart device equipped with the system can be worn on the eyes of the examinee. While meeting the need to acquire test data for a long time, it will not affect the examinee's normal vision or normal movement. Medical staff do not need to operate the instrument repeatedly. They only need to set parameters such as the alternation period of the first light source 101 and the second light source 102 and the total test duration in the initial stage. During the test, the real-time situation can be observed through the display screen, and any unexpected situations can be detected in time.

[0094] It should be noted that the specific embodiments described above are exemplary, and those skilled in the art can devise various solutions inspired by the disclosure of this invention. These solutions all fall within the scope of this invention and its protection. Those skilled in the art should understand that this specification and its accompanying drawings are illustrative and not intended to limit the scope of the claims. The scope of protection of this invention is defined by the claims and their equivalents.

Claims

1. An eye motion capture and tear film detection system, comprising: An illumination module (100) is used to illuminate the eyes of the examinee; The acquisition module (200) is used to record images of the subject's eyes; The processing module (300) is used to analyze and process the eye image information of the examinee; Its features are, The lighting module (100) is capable of periodically generating a first light source (101) that produces specular reflection in the tear film region of the eye and a second light source (102) that produces light spots in the eyeball region of the eye; The acquisition module (200) can acquire a first image (201) based on the reflected light of the first light source (101) to collect and store tear film information, and a second image (202) based on the light spot formed in the eye by the second light source (102) to collect and store eye movement information; The processing module (300) calculates the thickness of the tear film lipid layer (703) of the subject based on the first image (201), and the processing module (300) simulates the eye movement trajectory of the subject based on the second image (202); The processing module (300) is also able to filter out the first image (201) from the first image (201) and the second image (202) based on the time series, which is adjacent to the second image (202) where the eyeball is in the central region.

2. The eye motion capture and tear film detection system according to claim 1, characterized in that, The light source of the illumination module (100) is directed toward the subject's eyes through a polarization component, a lens component, and a one-way coating in sequence. The one-way coating reflects the reflected light from the eyes again, which is then collected by the acquisition module (200).

3. The eye movement capture and tear film detection system according to claim 1 or 2, characterized in that, The lighting module (100) generates the first light source (101) and the second light source (102) in a periodic alternation, and the acquisition module (200) acquires the first image (201) and the second image (202) in an alternating period.

4. The eye motion capture and tear film detection system according to claim 3, characterized in that, The partitioning unit (301) of the processing module (300) divides the first image (201) into partitions centered on the pupil and applies a label to each partition. The algorithm unit (302) calculates the thickness of the lipid layer (703) of the eye based on the partition.

5. The eye motion capture and tear film detection system according to claim 4, characterized in that, The processing module (300) performs coordinate processing on the second image (202) based on the pupil position of the eye measured by the light spot.

6. The eye motion capture and tear film detection system according to claim 5, characterized in that, The system also includes a display module (402) capable of displaying the images collected by the acquisition module (200) and the detection results.

7. The eye motion capture and tear film detection system according to claim 6, characterized in that, The system's storage module (401) can save the data collected by the acquisition module (200) and the data processed by the processing module (300) according to time parameters.

8. The eye motion capture and tear film detection system according to claim 7, characterized in that, The processing module (300) can also generate a corneal topography map reflecting eye information based on the data in the storage module (401).

9. A device equipped with an eye motion capture and tear film detection system according to any one of claims 1 to 8, characterized in that, The device can be worn at least in the eye area of ​​the subject. in, The device includes a support structure (500) for holding the device and a lens structure for placing the lighting module (100) and the acquisition module (200). The support structure (500) is designed according to the positional structure of the human ear, eye and nose, and includes two annular frames (600) that can hold the lens structure at the front of the subject's eye, and a strip frame connecting the two ends of the two annular frames (600). The two annular frames (600) are connected by an arc-shaped frame that can fit against the upper end of the bridge of the nose. The lens structure includes at least a polarizing layer for placing a polarizer, a lens layer for placing ordinary lens components, and a coating layer for placing a unidirectional coating.

10. The device according to claim 9, characterized in that, The device also includes an external device (400) for housing the processing module (300) and the display module (402), the external device (400) being connected to the lighting module (100) and the acquisition module (200) inside the bracket structure (500) and the lens structure.

Images