System, method, and apparatus for absorption, phase, and dark field imaging of the retina by oblique illumination

Abstract

A system, method, and apparatus are provided for absorption, phase, and dark-field imaging of the retina with oblique illumination. [Solution] A method for imaging ocular tissue, comprising the steps of: providing oblique illumination to the eye with a plurality of light-emitting regions of a light-transmitting device, the plurality of light-emitting regions being independently controllable and positioned to direct light toward at least one of the retina and iris of the eye; generating an output beam from light backscattered from at least one of the retina and iris by the oblique illumination; capturing the output beam using an imaging system to provide a series of images of the fundus; and extracting phase and absorption contrast images from the series of images of the fundus, wherein the series of images of the fundus in the capturing step are obtained by sequentially turning on one or more of the plurality of light-emitting regions at a time in the step of providing oblique illumination.

Description

Technical Field

[0001] Cross - reference to Related Applications This application claims priority to International Patent Application PCT / IB2016 / 052787, filed May 13, 2016, and International Patent Application PCT / IB2016 / 056806, filed Nov. 11, 2016, which are incorporated herein by reference.

[0002] The present invention relates to high - resolution quantitative and qualitative absorption, phase, and dark - field imaging of the retina by use of oblique illumination.

Background Art

[0003] Standard photography relies on the absorption differences of different feature parts that provide contrast. This also applies to conventional fundus examinations that observe blood vessels, photoreceptors, and other retinal structures by different reflectance values that bring about intensity modulation on the sensor surface. This does not apply to most of the cells in the inner retina (ganglion, nuclear, and plexiform layers), and their absorption and scattering values are so low that they show little contrast even at high resolution. Furthermore, the intensity modulation of these feature parts is so small that it can be ignored with respect to the background modulation signal (due to the underlying feature parts) combined with noise. Even when using optical coherence tomography (OCT), due to the weak contrast of these cells, the retina appears as a smooth layer with few feature parts.

[0004] Non-Patent Document 1 showed that a single-phase image of a sample can be obtained using images of phase samples (samples having weakly absorptive characteristic portions) acquired by illumination at different angles. In this method of illuminating by transmission, a phase image is reconstructed using an algorithm based on a Wiener filter. Such a method cannot be directly applied to living biological media because transmission illumination is generally impossible. A solution to this problem has been proposed by Non-Patent Document 2, where light from the deep layer of the sample is used as a secondary light source that illuminates through the top layer. To provide oblique illumination, Non-Patent Document 2 showed that light is irradiated on the sample at point p 1 and a region located at a distance d is illuminated obliquely from the rear. By subtracting the images obtained at two opposing illumination points, a phase contrast image of the sample could be reconstructed.

[0005] The dark-field images of the retina when illumination passes through the pupil have been studied. In this study, an enhanced contrast of the vascular system was shown (Non-Patent Document 3, Non-Patent Document 4).

[0006] When combining several dark-field images of a sample, an image containing the partial derivative function of phase information can be obtained by subtracting two images captured from asymmetric illumination. This has been shown in transmission of a microscope (Non-Patent Document 5), or reflection using oblique back-illumination of an endoscope (Patent Document 1, Non-Patent Document 2). Furthermore, when the illumination angle of the sample, i.e., its spectrum, is known, quantitative phase information can be recovered using a weak object transfer function model (Non-Patent Document 6, Non-Patent Document 1), or a Fourier iterative algorithm (Patent Document 2, Non-Patent Document 7). A similar method can also be used for the reconstruction of 3D images of the sample (Non-Patent Document 8).

[0007] Finally, by processing the image, it is also possible to correct the aberrations of the eye and optimize the Fourier characteristics of the image (Patent Document 3, Non-Patent Document 9).

[0008] There have been few reports of attempts to obtain a higher-contrast dark-field image obtained by the above-described trans-pupillary illumination.

[0009] One study is to obtain a higher-contrast dark-field image of the retina using trans-scleral illumination, i.e., the light applied to the fundus through the sclera (Non-Patent Document 10). In FIG. 21 of Patent Document 4, Patent Document 5, and FIG. 22 of Patent Document 6, trans-scleral illumination having several different wavelengths (red, green, blue) is used to create one image having several wavelengths simultaneously. The image obtained thereby is used to diagnose choroidal tumors. In contrast to full-field illumination (referred to as trans-pupillary illumination) through the eye lens, trans-scleral illumination can collect only the light generated from below the first layer of the retina having a thickness of 100 μm or less. This is because the high reflectance (specular reflectance or near-specular reflectance) generated from the surface is blocked by the pupil of the eye. Since tumors absorb more light than healthy tissues because of the intense activity of the cells and vascular system of the tumor tissue, the trans-scleral image of the tumor has a stronger contrast than the image obtained by trans-pupillary illumination, and thus, the spatial extent of the tumor can be diagnosed better (Non-Patent Document 10).

[0010] The above-described trans-scleral method is described as using one illumination point, or in some cases, two illumination points where two point light sources provide illumination simultaneously. Here, the point means "point light source-like" such as a small area. This point can be an area larger than the area indicated by the diffraction limit.

[0011] However, none of the above techniques provide quantitative or non-quantitative phase image information of the interior of the eye without using a scanning system. This phase contrast includes, but is not limited to, the fundus and the retina. Therefore, by obtaining phase information from the biological material above the photoreceptor, it is necessary to obtain improved contrast, improved image resolution, and further extract the existing functional information from large-scale studies of biological quantitative phase imaging.

[0012] Quantitative phase imaging refers to an imaging method for the well-known relationship (linear or logarithmic, although not limited thereto) that exists between the grayscale pixel values of a camera and the corresponding phase imparted to the light passing through a physical sample by that sample. A phase unwrapping method can also be used to remove the effects of phase periodicity and obtain a more detailed image.

[0013] In contrast, the absence of an absolute relationship between the grayscale pixel values of a camera and the corresponding phase imparted to the light passing through a physical sample by that sample is referred to as non - quantitative or qualitative phase imaging. An example of qualitative phase imaging is phase gradient contrast, as shown in Non - Patent Document 2.

[0014] The refinement of retinal images has been achieved in different ways. One of the main features of the human retina is the presence of cone cells. At high magnification, cone cells appear as bright spots on a dark background, enabling a guiding star reconstruction algorithm (Non - Patent Documents 11 and 12). Another method utilizes the natural movement of the eye (saccades). Most research attempts to suppress this phenomenon, which leads to a lower resolution in the case of averaging, but Non - Patent Document 13 uses averaging after finding the relative shift of each image. This is done using image correlation, and all images can be aligned while performing averaging. Due to the eye movement, each image shows a different point spread function (PSF) of aberration, and the resulting average image shows an average PSF, excluding the highest - order aberration.

[0015] Thereafter, a PSF much more similar to the diffraction - limited PSF can be obtained. By performing deconvolution of images with different PSFs and estimating the best PSF, the effect of eye aberration can be further reduced. This process has been carried out by Non - Patent Document 14. In this research, the deconvolution is based on different It is obtained by the PSF. Using entropy as a measure of image quality, the best correction is estimated and a high-resolution image is generated. Furthermore, the phase and absorption reconstruction using the method of Non-Patent Document 1 already consists of a deconvolution process. Using this method, phase and absorption information can be obtained whether or not the aberration pupil is known (Non-Patent Document 9).

[0016] High-resolution imaging of the living retina is performed by aberration correction. This operation is performed using a computer as shown in Patent Document 3 or using a hardware device. In-vivo imaging of the retina is usually performed by a scanning system (Patent Document 7), optionally by a scanning system (Patent Document 8) connected to an adaptive optics, or, for example, by a camera flood illumination system that may also be connected to an adaptive optics as in Patent Documents 9 and 10.

[0017] The resolution of optical imaging of the eye is mainly limited by three factors: the numerical aperture of the pupil (maximum 0.24 (Non-Patent Document 15)), lens aberrations, and intraocular scattering. Compensation for the latter two effects has been proposed, leading to the so-called adaptive optics confocal scanning laser ophthalmoscope (AOCSLO) (Non-Patent Document 16).

[0018] This system (Non-Patent Document 17), which may be combined with optical coherence tomography (OCT), provides a lateral resolution of about 1.5 μm and an axial resolution of 2 μm. This value is limited by the numerical aperture provided by the pupil of the eye. Non-Patent Document 18 uses a wavefront shaping method to focus light through a highly scattering medium. Further research has shown how the scattering medium can be used as an optical element to provide a high numerical aperture (NA = 0.85). (Non-Patent Documents 19 and 20). Using the memory effect, this spot can be scanned and an image of the scanned object can be reconstructed (Non-Patent Document 21). Another technique consists of directly scanning a sample using a speckle pattern and reconstructing the original image by a phase retrieval algorithm (Non-Patent Documents 22 and 23).

[0019] Thus, as described above, in the background art of the ophthalmic imaging apparatus, the retina is constantly illuminated by passing through the lens of the eye. Thereafter, the reflected light is collected again by the lens. Also, as described above, this exhibits several drawbacks and complex problems, and a new and substantially improved ophthalmic method, system, and device are desired.

Prior Art Documents

Patent Documents

[0020]

Patent Document 1

Patent Document 2

Patent Document 3

Patent Document 4

Patent Document 5

Patent Document 6

Patent Document 7

Patent Document 8

Patent Document 9

Patent Document 10

Non-Patent Documents

[0021]

Non-Patent Document 1

Non-Patent Document 7

Non-Patent Document 8

Non-Patent Document 9

Non-Patent Document 10

Non-Patent Document 11

Non-Patent Document 17

Non-Patent Document 18

Non-Patent Document 19

Non-Patent Document 20

Non-Patent Document 21

[0022] According to one aspect of the present invention, a method for imaging eye tissue is provided. Preferably, the method includes a step of providing oblique illumination to the eye by a plurality of light - emitting regions of a light delivery device, wherein the plurality of light - emitting regions are independently controllable and are arranged to direct light at least to one of the retina and the iris of the eye; a step of generating an output beam from light back - scattered from at least one of the retina and the iris by the oblique illumination; and a step of capturing the output beam using an imaging system to provide a series of images of the fundus.

[0023] Furthermore, the method includes a step of extracting phase and absorption contrast images from a series of images of the fundus oculi, and it is more preferable that the series of images of the fundus oculi in the step of capturing are obtained by sequentially turning on one or more of a plurality of light-emitting regions at a time in the step of providing oblique illumination.

[0024] According to another aspect of the present invention, the eye tissue is a part of a living human or animal eye, and the oblique illumination is at least one of trans-pupillary illumination, trans-scleral illumination, and trans-epidermal illumination. In addition, the light delivery device is configured for at least one of the following illumination modalities, namely, the light delivery device does not contact the face of the eye patient, the light delivery device contacts the skin around the eye, the light delivery device contacts the sclera of the eye, and the light delivery device contacts the cornea of the eye.

[0025] According to still another aspect of the present invention, a system for imaging eye tissue is provided. The system preferably includes a light delivery device having a plurality of light-emitting regions directed toward the eye tissue to provide oblique illumination and an output beam generated by the oblique illumination from the plurality of light-emitting regions that is backscattered from the fundus oculi, and an imaging system configured to capture the output beam and provide a series of images of the fundus oculi.

[0026] Furthermore, the system includes a controller configured to individually control the plurality of light-emitting regions of the light delivery device and sequentially turn on one of the plurality of light-emitting regions at a time to capture a series of images by the imaging system, and it is more preferable that the imaging system is further configured to extract a quantitative phase contrast image, a quantitative absorption image, a qualitative phase and absorption image, a qualitative phase contrast image, a qualitative absorption image, a qualitative phase and absorption image, and a dark field image from the fundus oculi.

[0027] According to yet another aspect of the present invention, the imaging system further includes a scanning system and a detector. The scanning system has a collection pupil that is centered or moved with respect to the center of the pupil of the eye. The detector preferably includes at least one of a one-pixel detector, a line camera, a two-dimensional multi-pixel device, and a split detector.

[0028] By considering the following description with reference to the accompanying drawings showing some preferred embodiments of the present invention, the above and other objects, configurations, and advantages of the present invention, and the methods for realizing them will become clear, and the present invention itself will be best understood.

[0029] The accompanying drawings incorporated herein and forming a part of this specification show the preferred embodiments of the present invention at present, and together with the above general description and the following detailed description, serve to explain the configuration of the present invention.

Brief Description of the Drawings

[0030]

Figure 1

Figure 2

Figure 3

Figure 4A

Figure 4B

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Figure 23

Figure 24

Figure 25

Figure 26

Figure 27

Figure 28A

Figure 28B

Figure 29

Figure 30

Figure 31

Figure 32

Figure 33-1

Figure 33-2

Figure 34

Figure 35

Figure 36

Figure 37

Figure 38

Figure 39

Figure 40-1

Figure 40-2

Figure 41

Figure 42

Figure 43-1

Figure 43-2

Figure 44

Figure 45

Figure 46

Figure 47

Figure 48

Figure 49

Figure 50

Figure 51

Figure 52

Figure 53

Figure 54

Figure 55

Mode for Carrying Out the Invention

[0031] In this specification, wherever possible, the same reference numerals are used to denote the same elements common to the drawings. Also, the images are simplified for illustrative purposes and may not be shown to scale.

[0032] According to one aspect of the present invention, there is provided a retinal imaging device capable of establishing phase and absorption contrast images by oblique illumination of a retinal layer. According to an aspect of the present invention, the device can be used for in vitro and in vivo imaging. In the first part of this section, an overview of in vitro implementation will be described, and in the second part, in vivo implementation will be described in detail.

[0033] Examples of phase and absorption contrast images include, but are not limited to, quantitative phase contrast images, quantitative absorption images, qualitative phase and absorption images, qualitative phase contrast images, qualitative absorption images, qualitative phase and absorption images, and dark field images. Also, examples of phase and absorption contrast images include, but are not limited to, one-dimensional images, two-dimensional images, three-dimensional images, or multi-dimensional images.

[0034] Examples of in vitro samples of the eye include, but are not limited to, whole eyes, untreated eye pieces, fixed eye pieces, stained eye pieces, and in-tube samples.

[0035] Regarding in vitro imaging, instead of using direct illumination of the sample according to an aspect of the present invention, phase contrast is obtained by oblique illumination caused by scattering in the deep layer of the eye. In the first embodiment, the sample is illuminated by a light source and an oblique angle. In the second embodiment, the light source is arranged on the same side as the imaging system to create a reflection configuration. In the third embodiment, the scattering layer is arranged behind the phase sample to provide backscattered illumination. In the fourth embodiment, the scattering layer that provides back illumination is the choroid of the eye. In the fifth embodiment, the light source is scattered by a diffuser plate before reaching the sample. In the sixth embodiment, a back reflection layer is added below the sample. For the reconstruction process, see the following.

[0036] Examples of light sources include, but are not limited to, light-emitting diodes, superluminescent diodes, quantum dot light sources, lamps, blackbody radiation sources, low temporal coherence light sources, low spatial coherence light sources, and laser sources.

[0037] Regarding in-vivo imaging, instead of using direct illumination of the fundus, phase contrast is obtained by oblique illumination caused by scattering in the deep layer of the eye. The configurations of the method, device, and system may be simplified in the following categories: type of illumination, light delivery device, image acquisition system, and reconstruction process.

[0038] The type of illumination can be shown by the following non-limiting and non-exclusive embodiments:

[0039] In a first embodiment, light passes through the sclera, choroid, and retina. The transmitted and scattered light illuminates the fundus. In this variant, little or no light enters the pupil-lens. The light delivery device is in contact with the sclera. In other variants, a combination of trans-scleral illumination and trans-pupillary illumination may be used.

[0040] In a second embodiment, light passes through the skin layer near the eye, the sclera, the choroid, and the retina. The transmitted and scattered light illuminates the fundus. In this variant, little or no light enters the pupil-lens.

[0041] In a third embodiment, light passes through the pupil and is directed towards the side of the eye. Here, the light rays are scattered and reflected towards the fundus.

[0042] In a fourth embodiment, light passes through the pupil and is directed towards the fundus at an angle, generating backscattered light at a similar angle.

[0043] In a fifth embodiment, light passes through the pupil and irradiates a site near the imaging region. The light is scattered in the deep layer, providing angled illumination behind the imaging region.

[0044] In a sixth embodiment, light passes through the temporal bone. The transmitted and scattered light illuminates the fundus. No light enters the pupil-lens.

[0045] In the seventh embodiment, light passes through the pupil or the skin and sclera and directly illuminates the imaging site of the retina, but does not illuminate the background, thus providing dark-field contrast.

[0046] In the eighth embodiment, the wavefront is manipulated before entering the eye. The feedback light is collected through the eye's crystalline lens. And several methods are possible, for example, obtaining a spot size smaller than the eye-pupil resolution (0.24NA) by focusing light on the fundus to compensate for scattering, obtaining focusing by an iterative process, using the feedback light passing through the eye-pupil as a criterion for optimization, and scanning the fundus using the well-known memory effect of the scattering medium by adding a phase gradient to the wavefront. The scan pattern can include an optimized focus spot or a speckle pattern.

[0047] When the scan data is recorded, this data is processed by a phase retrieval algorithm or other digital means to reconstruct a super-resolution image, i.e., an image having the resolution of the illumination speckle pattern.

[0048] The light delivery device according to an aspect of the present invention can be designed for contact and non-contact use. The contact light delivery device is shown in the following embodiments:

[0049] The light delivery device is made from a flexible electronic circuit incorporating a light-emitting device and an electronic wire that generates a drive signal for the light delivery device.

[0050] The light delivery device contacts the skin or the sclera, and the light delivery device in contact with the skin has a removable protective patch (one for each patient).

[0051] Members that contact the face of the patient for analyzing the eye include, but are not limited to, a head holder and a jaw holder. The light delivery device is covered by a removable disposable member. Examples of the removable member include, but are not limited to, a layer of paper, stacked layers of paper, and a polymer layer.

[0052] The shape of the flexible electronic circuit of the light-emitting device is ergonomically designed.

[0053] The flexible electronic circuit-patch is held by a head-mounted frame (e.g., a glass frame) placed on the subject.

[0054] To suppress backscattered light, an absorption or reflection layer is placed behind the illumination device.

[0055] A light-emitting device for non-contact illumination is shown in the following embodiments:

[0056] The non-contact light-emitting device is made from a point light source, and an image of the point light source is formed on the illumination surface (cornea, sclera, or skin).

[0057] The non-contact light-emitting device consists of a circular light source, and its light irradiates the illumination surface (cornea, sclera, or skin).

[0058] The non-contact light-emitting device is annular, and its light irradiates the illumination surface (cornea, sclera, or skin).

[0059] Illumination is provided by, but not limited to, one light source or a combination of light sources in the wavelength range of 400 nm to 1200 nm, such as pulsed or continuous laser light sources, light-emitting diodes, superluminescent diodes, quantum dot light sources, lamps, blackbody radiation sources, and laser light sources. The light is delivered by placing the light source in direct contact with the tissue (sclera or skin), or guided from the light source to the tissue, or forming an image of the light source on the illumination surface (cornea, sclera, or skin). As the waveguide member, a multimode fiber, a capillary Examples include, but are not limited to, a waveguide, a multimode fiber with a lens, a single-mode fiber, and a photonic crystal fiber. The light beam may be focused, diffused, or collimated, depending on the selected illumination method. The light may be linearly polarized, circularly polarized, unpolarized (meaning it does not exhibit a known preferential polarization), or a mixture of different polarizations, but is not limited thereto.

[0060] In the Fourier domain, tilted illumination with a plane wave corresponds to a movement towards higher spatial frequencies, meaning higher spatial resolution. Additionally, by irradiating the fundus with a larger angle even a rear illumination with a larger tilt is generated, resulting in a higher contrast.

[0061] The image acquisition process varies depending on the required imaging modality: dark field or phase / absorption. In the case of dark field, imaging can be performed without image processing using only one illumination point. A wider field of view can be obtained by stitching together images obtained from different imaging sites.

[0062] For the image acquisition system, the following operating steps or methods can be performed: An image of the retina is formed on the camera by a series of lenses and mirrors. Translating the lens or mirror to change the focal plane within the retina. Using a tunable lens to change the focal plane within the retina. Rotating and translating a cylindrical lens to compensate for astigmatism in the eye. Translating two independent cylindrical lenses to compensate for astigmatism in the eye. Using the patient's prescription glasses to compensate for the aberration of the eye. Using a deformable mirror to compensate for the aberration of the eye. Conjugating a wavefront sensor to the pupil plane to measure the aberration of the eye. Conjugating the camera to the pupil plane to measure the illumination function. Conjugating the aperture to the retina to select a small area of the retina and measure the illumination function. Conjugating the camera to the cornea to observe if the patient's eye is in the correct position. Additionally, a polarization optical system can be used to stop the back reflection of light at different surfaces.

[0063] For reconstruction, the following operating steps can be performed: The illumination contour of the retina (backscattered light) is not a uniform function in the collection range NA. The light used for illumination passes through the retina, and its phase and intensity are affected by its optical properties. For different illumination functions, modulated light is recorded by the camera. The images are processed together to reconstruct the phase and absorption images. The image quality is improved by enhancing the resolution and contrast through image processing. Anatomical features are extracted and analyzed to detect possible abnormalities.

[0064] For phase imaging, at least two images captured at two different illumination points are required. By using a reconstruction algorithm, qualitative or quantitative phase images can be obtained. Such reconstruction algorithms known in the art include, but are not limited to, L. Tian and L. Waller, "Quantitative differential phase contrast imaging in an LED array microscope", Opt. Exp. 23, 9, pp. 11394 - 11403 (2015), Z. Phillips, M. Chen, L. Waller, "Quantitative Phase microscopy with Simultaneous Aberration Correction", Optics in the Life Sciences, 2017), or those described in International Patent Application Publication WO2015 / 179452, S. B. Mehta and C. J. R. Sheppard, "Qantitatvie phase - gradient imaging at high resolution with asymmetric illumination - based differential phase contrast", Opt. Lett. 34, 13, pp. 1924 - 1926 (2009). The proposed illumination method described in the exemplary embodiment may be combined to record the images to be processed.

[0065] Phase imaging is based on the interference of a beam with itself due to the phase difference on the object plane. This interference causes intensity modulation in the camera layer in the case of non-uniform illumination. By recording two images with opposite illumination profiles (S 1 (u)=S 2 (-u)) and subtracting their intensities, the background signal is removed and only the phase information remains. An image with differential phase contrast is obtained (Z. Liu, S. Liu, and L. Waller, "Real-time brightfield, darkfield, and phase contrast imaging in a light emitting diode array microscope", J. of Biomed., Opt. 19, 10, 106002 (2014), T. N Ford, K. K Chu, and J. Mertz, "Phase-gradient microscopy in thick tissue with oblique back-illumination", Nat. methods, 9, 12 (2012)). The principle is shown in FIGS. 3, 4A, and 4B.

[0066] Let the two complementary illumination angle images be I0 and I1. Then the differential phase contrast image is Idiff=(I0 - I1) / (I0 + I1) (1) calculated by.

[0067] The intensity value of the image is related to the phase gradient of the image plane. Since this technique requires the illumination beam to pass through the sample, it seems impossible for thick biological media.

[0068] A method is known for providing back-illumination of a sample of a certain kind using the properties of a scattering medium such as biological tissue (T.N Ford, K.K Chu, and J.Mertz, "Phase-gradient microscopy in thick tissue with oblique back-illumination", Nat.methods, 9, 12 (2012)). In fact, when a light beam is irradiated perpendicularly to a scattering medium, the backscattered light shows different angular distributions depending on the distance from the beam incident position (see Fig. 3). Due to this non-uniform angular distribution, an inclined average illumination is obtained, which can be used to provide oblique illumination. As shown in Figs. 34 and 35, a similar effect is observed when the irradiation beam is not perpendicular to the surface, and the backscattered beam shows directivity.

[0069] This effect can be used, but not limited to, in an eye such as the human eye: when light is irradiated onto the fundus at an angle (for example, when passing through the sclera), the light passes through the transparent retinal layer and scatters in a deeper layer (for example, the pigment epithelium and choroid). Here, the light scatters backward, maintains an inclined direction, and passes through the upper layer of the retina, being affected by retinal absorption and phase.

[0070] By sequentially acquiring images at at least two different complementary illumination angles (for example, +90 and -90 degrees), a phase-gradient imaging algorithm (T.N Ford, K.K Chu, and J.Mertz, "Phase-gradient microscopy in thick tissue with oblique back-illumination", Nat.methods, 9, 12 (2012), S.B.Mehta and C.J.R.Sheppard, "Qantitatvie phase-gradient imaging at high resolution with asymmetric illumination-based differential phase contrast", Opt.Lett.34, 13, 1924 - 1926 (2009), L.Tian and L.Waller, "Quantitative "Differential phase contrast imaging in an LED array microscope", Opt. Exp. 23, 9, pp. 11394 - 11403 (2015), see, for example, Figure 4), or the Fourier ptychographic algorithm (International Patent Application Publication WO2015 / 179452, G. Zheng, R. Horstmeyr, and C. Yang, "Wide - field, high - resolution Fourier ptychographic microscopy") can be used to reconstruct qualitative or quantitative phase gradient images. Then, this reconstruction process can be generalized to any kind of illumination that introduces the phase and absorption transfer functions already incorporated by Tian and Waller. In practice, the intensity of the camera can be expressed as follows: [Number]

[0071] δ is the Dirac delta function, B 1 , H 1 and G 1 are transfer functions that depend only on S. If S is known, the only unknowns are μ and φ, and by acquiring at least two images, two unknown functions can be obtained. Figure 49 shows the principle of two different illumination points. A human in - vivo fundus image is captured by epi - scleral illumination. The shadow in Figure 40A shows the effect of a "double" image of the vascular tree due to a substantially transparent layer between the blood vessels and the layer that projects the shadow, as also shown in images 55 and 56.

[0072] Next, some details of the mathematical background of the aspects of the present invention are shown. When image I 1 and a second image I 2 are acquired with two different illumination patterns, these images are related, without limitation, in different ways: [Number] are renormalized using.

[0073] Here, N i {~} is the i-th folding method, and Low{~} is a 2D low-pass filter. At the same time, in many cases, it is convenient to subtract the resulting image by its average to remove the zero component in the Fourier space.

[0074] As a normalization method, N 1 the relationship defined by N 2 the relationship defined by N 3 the relationship defined by N

[0075] Regarding the reconstruction of phase and absorption, from the research of Tian and Waller (L. Tian and L. Waller, "Quantitative differential phase contrast imaging in an LED array microscope", Opt. Exp. 23, 9, pp. 11394 - 11403 (2015)), the image can be represented in the Fourier domain as follows:

Equation

Equation

Equation

Number

Number

[0076] Regarding the estimation of the illumination function, three different variants are shown. In the flat approximation method, as described above, reconstruction is possible only by knowing different H and G. This is obtained from the studies of Tian and Waller by knowing the pupil function (aberration) and the illumination profile.

[0077] A method for obtaining phase and absorption using the DPC images obtained by equations (9) and (10) and the transfer function (11) is herein called the modified Waller method (MWM). B 1 =B 2 、H 1 =H 2 、and G 1 =-G 2 In the special case where P ij =I i -I j (14) The transfer function is, T ij =2G i (15)

[0078] Each of these methods requires a deconvolution step of deconvolving the DPC image with a transfer function. This can be done in different ways, including but not limited to: direct inversion, Wiener filtering, conjugate gradient minimization, maximum likelihood method, blind deconvolution method, etc.

[0079] The methods for extracting these phases and absorption are herein referred to as "phase and absorption recovery algorithms". Examples of phase and absorption recovery algorithms include, but are not limited to, the Waller method with or without convolution, the modified Waller method with or without convolution, and phase recovery algorithms.

[0080] The first method consists of an approximation S(u)=h(au x +bu y ), where h(u i ) represents the Heaviside function that depends only on the coordinate u i , and a and b are two coefficients selected to determine which half-space is equal to 1 . This method completely ignores the angular distribution of the backscattered light and enables reconstruction without conducting other investigations on the illuminated surface. Subsequently, the reconstruction can be performed by, including but not limited to, inverse filtering, least squares filtering, constrained least squares filtering, Tikhonov regularization, blind deconvolution, iterative filtering, and can be applied to both the spatial or Fourier domains. The image in Figure 49 is obtained by this approximation and subsequent reconstruction of the phase image by Tikhonov regularization.

[0081] Regarding the lamp approximation method, this method is similar to what was obtained previously, but instead of using the Heaviside approximation, the illumination function is selected as S(u)=au x +bu y +q, where a, b, and q are arbitrary values. Subsequently, the image can be reconstructed in a similar manner as proposed by the flat approximation method.

[0082] Next, regarding the angular scattering information method, this method is based on accurate knowledge of the function S(u). This function can be obtained by, but is not limited to, Monte Carlo scattering simulations or experimental measurements. The experimental results can be obtained by, but are not limited to, a camera conjugated to the pupil plane. The reconstruction is obtained by using the same technique as shown in Method 1. The image shown in Fig. 51 is obtained using the function S(u) obtained by Monte Carlo simulation, and then the phase image is reconstructed by Tikhonov regularization. The parameters used in the simulation are from Rovati et al. (L. Rovati, S. Cattini, N. Zambelli, F. Viola, and G. Staurenghi, "In-vivo diffusing-wave-spectroscopy measurements of the ocular fundus", Optics Express Vol. 15, Issue 7, pp. 4030 - 4038 (2007)), as well as Curcio et al. (C. A. Curcio, J. D. Messinger, K. R. Sloan, A. Mitra, G. McGwin, and R. F. Spaide, "Human Chorioretinal Layer Thicknesses Measured in Macula-wide, High-Resolution Histologic Sections", Invest Ophthalmol Vis Sci. 2011 Jun;52(7):3943 - 3954).

[0083] After the image reconstruction, the resolution can be improved by, but is not limited to, the guiding star method of Meitav or Shemonski, the moving and adding method of Meitav, the deconvolution-entropy method of Hillman, the blind deconvolution algorithm, or iterative filtering. In the case of the guiding star algorithm, the deconvolution function is first estimated in the photoreceptor layer by physical focusing. Subsequently, the phase retinal layer is focused, an image is acquired, and the same filter is applied to improve the image quality. The filter can be applied to the original image or the DPC(I DPC ) image.

[0084] The image quality can also be improved in the reconstruction process. In fact, when the aberration of the pupil plane is known, the original image is restored by reconstruction. When the aberration is unknown, as described in Phillips (Z. Phillips, M. Chen, L. Waller, "Quantitative Phase microscopy with Simultaneous Aberration Correction", Optics in the Life Sciences, 2017), the aberration can also be estimated using a blind deconvolution method.

[0085] Once the phase image is extracted, an improved pattern recognition algorithm can be executed for feature extraction. Feature extraction can be applied to, but is not limited to, different retinal layers, such as the inner limiting membrane (ILM), retinal nerve fiber layer (RNFL), ganglion cell layer (GCL), inner plexiform layer ( IPL), inner nuclear layer (INL), outer plexiform layer (OPL), outer nuclear layer (ONL), outer limiting membrane (ELM), acting on the characteristic parts of the retina, such as cells, nuclei, and microvascular systems present therein.

[0086] Feature extraction can be performed, but is not limited to, by edge detection, corner detection, blob detection, ridge detection, scale-invariant feature transform, Hough transform, for example, based on deep learning software. In such cases, the deep learning software can be trained with pathological and non-pathological images (in vivo or in vitro). Furthermore, feature extraction can be applied to medical information extraction to assist clinicians in analyzing the data.

[0087] Regarding the measurement of the illumination function, the illumination function can be estimated by placing a camera on a surface conjugate to the pupil plane in the case of a curved surface (such as the eye), or on a surface conjugate to the Fourier plane when the sample surface is flat (such as in the case of an in vitro sample mounted flat). By means of an aperture on the surface conjugate to the sample surface, the scattering profile can be selected better. In practice, the measured illumination function is averaged over the sample region, but the aperture can limit this region and the illumination function can be measured locally. Furthermore, a small aperture can approximately make a curved surface locally flat. With this idea, the difference between the case of placing the camera on the pupil and the case of placing it on the Fourier plane disappears.

[0088] With this configuration, the image obtained by the pupil camera is the illumination function averaged over the region limited by the aperture diaphragm. When it is necessary to measure the illumination function point-by-point over different regions of the sample, the aperture can be replaced by a lens array. In this configuration, each lens creates an image with a camera of the local illumination function.

[0089] Regarding aberration measurement, the limiting factors of image quality are given by optical aberrations. These limiting factors can be due to the aberrations of the eye and the optical system. Due to aberrations, usually high-frequency damping occurs, resulting in a decrease in the resolution of the image. To compensate for this effect, for example, a wavefront sensor can be made part of the system by physical correction or post-processing. This operation can be performed using many different devices, but mainly used are the Shack-Hartmann wavefront sensor and the Tscherning wavefront sensor.

[0090] In both cases, light is sent to the retina and the camera is conjugated to this surface. In the case of Shack Hartmann, light is sent to create points, while in the case of Tscherning, an image of a pattern similar to a grid of points is created on the retina. In Shack Hartmann, the lens array is then placed on the surface conjugated to the pupil to generate several images of the same points of the camera. Since the lens array is conjugated to the pupil plane, the image of each spot is translated from the center of the lens by a distance proportional to the local gradient of the wavefront. In this way, the wavefront can be reconstructed by measuring the spot field. The Tschering wavefront sensor performs the same measurement, but directly generates the spot field on the retina.

[0091] Such a system requires illumination of the retina, so if both are used simultaneously, this light may interfere with the retinal image used for reconstruction. Therefore, it is convenient to use different wavelengths: one wavelength for illumination and the other for wavefront detection. Then, filters and dichroic mirrors are used to prevent the light of one system from entering the other.

[0092] Regarding physical aberration correction, the aberration can be compensated in post-processing, but if the damping effect is too strong, the dynamic range of the camera may not be sufficient to record the information. In that case, it is more convenient to physically compensate for the effect of the aberration when the aberration of the eye is too strong. Physical correction can be performed in different ways.

[0093] Changing the focal length of the method, system, and device is useful in both cases where the eye is out of focus (nearsighted / farsighted) and where the change in the focal plane is required for the performance of different stacked surfaces, for example, but not limited to. This operation can be performed in different ways, for example, but not limited to, the translational movement of the focusing element (lens or curved mirror), the translational movement of the mirror to increase the path (Badal system), the change in the focal length of the tunable lens, and the change in the variable shape mirror.

[0094] Furthermore, in measurements using the patient's prescription glasses or contact lenses, the patient's myopia / hyperopia can be corrected.

[0095] Another aberration commonly seen in the human eye is due to astigmatism. This consists of a difference in the focal lengths of the lens along two different axes. Therefore, for example, the translation movement of two independent cylindrical lenses, a deformable mirror, the patient's prescription glasses or contact lenses can be used to compensate for astigmatism.

[0096] Referring to FIG. 48, by placing a deformable mirror on the plane conjugate to the pupil plane, both low-order and higher-order aberrations can be compensated. This configuration becomes more robust when connected to a wavefront sensor with a sensor placed behind the deformable mirror. In this way, the wavefront sensor can be used in a closed loop with the deformable mirror to compensate for aberrations.

[0097] Furthermore, dark-field illumination can be used in the present method, system, and device. Regarding dark-field illumination, an illumination method adopted to illuminate a range of illumination angles different from the light collection range is used. In this way, trans-scleral illumination associated with trans-pupillary light collection is regarded as dark-field illumination.

[0098] When light is incident on the fundus, it is scattered by many different layers. The first retinal layer results in relatively strong backscattering. By using different angles for illumination and light collection, most of the light scattered by this first retinal layer is not collected by the pupil. Deeper layers scatter the light, as seen in OCT images or non-coaxial light collection (T.Y.P. Chui, D.A. Van Nasdale, and S.A. Burns, "The use of "Forward scatter to improve retinal vascular imaging with an adaptive optics scanning laser ophthalmoscope", Biomed. Opt. Exp. 3, 10, 2537-2549 (2012)). In this paper, dark field illumination through the pupil is used during illumination and light collection to observe the microvascular systems of the retina and choroid.

[0099] To provide oblique retroillumination of the fundus, the following illuminations can be used: transscleral (Figs. 9A, 9B), transcutaneous (Figs. 10A, 10B), pupil oblique illumination (Figs. 11A, 11B), pupil direct illumination (Figs. 12A, 12B), or illumination through the temporal side (Fig. 13).

[0100] Regarding dark field light collection, a light collection method that collects only the light diffracted by the sample and does not collect the light from the background will be described. Dark field imaging of the retinal layer can be obtained by avoiding illuminating the background of the imaging site with large-angle illumination or by filtering a part of the background layer in the conjugate plane. This can be done by flood illumination, a scanning method, and a mixture of these two, for example, but not limited to, a scanning system for flood illumination and light collection.

[0101] Examples of scan acquisition systems include, but are not limited to, confocal scan acquisition, optical coherence tomography acquisition, moving pupil scan acquisition, split detector acquisition, and lock-in scan acquisition. Examples of detectors include, but are not limited to, 1-pixel detectors, line cameras, and 2D detectors. Examples of 1-pixel detectors include, but are not limited to, photodiodes, avalanche photodiodes, photomultiplier tubes, microchannel plates, lock-in 1-pixel detectors, and split detectors composed of 1-pixel detectors. Examples of 2D detectors include, but are not limited to, lock-in multi-pixel detectors, CMOS cameras, sCMOS cameras, CCD cameras, and 2D split detectors.

[0102] ​ Next, different illumination methods and systems will be described using different configurations that serve as non-limiting and non-exclusive embodiments.

[0103] Configuration 1 (trans-scleral) Referring to FIGS. 9A and 9B, these figures show a trans-scleral illumination method: light 16 is directly irradiated onto the scleral tissue 9. The light may be delivered either by direct contact with a light source or a light waveguide member, or non-contact by a light beam (a collimated beam, a focused beam, a diffused beam, or one having structured illumination). Some examples of the light positions are shown by the disk 45. The scattering characteristics of the sclera 9 and the underlying layers 10, 11 generate a diffused beam 19 that illuminates the fundus at a large angle. When physically contacting the sclera, local anesthesia may be used to make the measurement more comfortable for the patient.

[0104] Examples of structured illumination include, but are not limited to, a sinusoidal phase pattern, a sinusoidal intensity pattern, a light pattern intensity-modulated by a micromirror array, a phase-modulated, and / or a light pattern modulated by a spatial light modulator.

[0105] Examples of waveguides and waveguide members include, but are not limited to, single-mode fibers, multi-mode fibers, capillary waveguides, multi-mode fibers with lenses, and photonic crystal fibers.

[0106] Configuration 2 (trans-epithelial) Referring to FIGS. 10A and 10B, these figures show a trans-epithelial illumination method: light 16 is irradiated onto the upper eyelid 14 and / or the lower eyelid 15, and scattered 19 from there through different layers to the inside of the eye 1. The light may be delivered either by direct contact 27 (with a light source or a waveguide member) or by a light beam 16 (collimated or non-collimated). Some examples of the light positions are shown by the disk 45. Many spaced-apart point light sources provide illumination at different angles. Also, since no anesthesia is required, contact with the skin can be more comfortable for the patient.

[0107] Configuration 3 (transpupillary from the side) Referring to FIGS. 11A and 11B, these figures show a pupil illumination method: Light 17 is irradiated onto the inner layer of the eye after passing through the pupil 4 and the lens 5. After being retroreflected from the focal point 28, the light is scattered inside the eye and illuminates the fundus. The light may be delivered by direct contact with the cornea 3 (of the light source or the waveguide member), or may be delivered non - contact using a light beam (collimated or non - collimated).

[0108] Configuration 4 (transpupillary direct bright - field) Another illumination method is based on direct illumination of the fundus. When light reaches the fundus, the backscattered light is modulated by the retina and then collected for imaging purposes. In this configuration, the light may illuminate the background of the imaging site (bright - field) or only the side of the imaging site (dark - field).

[0109] Configuration 5 (transpupillary direct dark - field) Referring to FIGS. 12A and 12B, these figures show that light is sent directly through the pupil to the imaging retinal site However, the light is sent at an angle such that it does not reach the retinal pigment epithelium (RPE) behind the imaging retinal site. In this way, the background appears dark. The light collected by this retinal site is given not by modulation of the background light, but by diffraction of the retinal feature portions.

[0110] Configuration 6 (Mertz - like) Light is irradiated through the pupil without directly irradiating the imaging retinal site or its background, and is focused on the RPE. The light is scattered inside the RPE and the choroid and reaches the layer behind the imaging retinal site. From here, the light is backscattered and irradiates the retina to provide illumination.

[0111] Configuration 7 (trans - temporal) Referring to FIG. 13, this figure shows that the fundus can be illuminated by trans - temporal illumination. Light is irradiated onto the patient's temple and scattered into the eye from here.

[0112] Configuration 8 (beam shape) Referring to FIGS. 14A, 14B, and 14C, these figures show that, although not limited to, the beam shape can be modified by optical methods, wavelength selection, and wavefront shaping using Configurations 1-7.

[0113] Next, different exemplary illumination methods and systems will be described as non-limiting and non-exclusive embodiments.

[0114] Configuration 1 (Contact PCB) Referring to FIGS. 23, 24, 25, 26, 27, and 28A, a transcutaneous illumination method is shown. Light 16 is irradiated onto the upper eyelid 14 and the lower eyelid 15 simultaneously or sequentially, or in any combination, and is scattered 19 from there through different layers towards the inside of the eye 1. The light is delivered by the light source 27 contacting the skin 15 directly. A transparent or scattering medium can be between the light source and the skin to expand the illumination site and reduce the power density on the skin. An exemplary example of the light source position is shown by the disc 45. Many spaced-apart point light sources provide illumination at different angles to the inside of the eye. Also, in contrast to the case of contacting the eye (sclera, cornea) with the light source, since no anesthetic lubricant is required, contact with the skin can be more comfortable for the patient.

[0115] In the devices of FIGS. 25 and 28A, a transscleral illumination system is connected to the master driver substrate. The substrate provides drive signals for all the connected LEDs and trigger signals for the imaging device to synchronize the illumination with the acquisition system. By turning on different LEDs, different illumination spectra can be generated with respect to both the emission wavelength and the angular spectrum. By changing the drive current, the total intensity, the shape of the power spectrum, and the spatial distribution of the light can be changed.

[0116] Configuration 2 (Contact PCB) Referring to FIG. 24, the same illumination principle as Configuration 1, i.e., trans-epidermal illumination before passing through the sclera, is shown. The light-emitting device and its flexible member have a continuous light source along the arc shape of the eyelid, above and below the eye. The continuous light source is composed of pixels, and each pixel can be independently turned on or off.

[0117] Configuration 3 (Contact LED) Referring to FIG. 28B, the same illumination principle as Configuration 1, i.e., trans-epidermal illumination before passing through the sclera, is shown. The light-emitting device is an LED and has an encapsulated transparent material (such as, but not limited to, epoxy and polydimethylsiloxane) with a diameter of several millimeters. The LED is arranged in contact with the skin of the patient's eyelid. The number of LEDs is not limited to 4.

[0118] Configuration 4 (Non-contact) The illumination is provided non-contact, and the beam illuminating the eye or surrounding tissue can be focused, collimated, or diffused.

[0119] Configuration 5 (Wheel) In the devices of FIGS. 19, 21, and 22, the light directed towards the scattering tissue is provided by, but not limited to, the light beam 16 and the rotating wheel 39 with small holes 41 drilled therein. The light beam illuminates the entire surface of the wheel such that the light passes only through the holes 40 and illuminates only one point on the sclera 9. Alternatively, referring to FIGS. 4, 8, and 9, the illumination point may be on the skin 14, 15 surrounding the eye, or even outside the eye.

[0120] Configuration 6 (Wheel and Fiber) In the apparatus of FIG. 20, light on the scattering tissue is provided by a fiber 18 which can be a single-mode fiber or a multi-mode fiber, although not limited. The rotating wheel 39 holds the fiber 18 and a lens 22 that focuses light on the sclera 9 or the skin 14, 15. The fiber holder is designed such that the fiber can rotate freely without causing stress on the fiber. Further, the disk holding the fiber rotates for only a limited time to prevent the fiber from being wound around the rotating arm. Another solution to prevent winding consists of rotating the disk from the side (thereby removing the rotating arm).

[0121] Another embodiment is a series of light sources arranged in a fixed structure such as, although not limited to, a circle (ring). Alternatively, the light beam 74 can be split as shown in FIGS. 16 and 18. In these aforementioned examples, referring to FIGS. 10A - 14C, the apparatus is configured to send light non - contact.

[0122] Configuration 7 (patch) In the apparatus of FIG. 23, light contacts the patient's skin by a patch 46 and is sent to the skin. The patch is connected to several fibers that bring several illumination points 45 to the patient's eyelid. One optical fiber 18 is required for each illumination point. The patch can be composed of a consumable protection member that contacts the skin. The patch is connected to the split light source by an optical connector 47.

[0123] Configuration 8 (contact sclera) In the apparatus of FIG. 17, light is sent in contact with the patient's sclera 9. The light exits from a plurality of optical fibers 18 or optical waveguides. In addition, the objective lens of the imaging system 21 is in substantially contact with the cornea 3 and a refractive index - matching biocompatible gel 65 is between them. Note that the principle shown in FIG. 17 differs from FIG. 2 in the number of illumination points (three or more) and the illumination method. Here, the beams are turned on sequentially, while in FIG. 2, two points are irradiated simultaneously.

[0124] FIG. 54 is an exemplary schematic diagram of a scanning system for inspecting an eye 541 according to an aspect of the present invention. By using a scanning system to collect signals, the depth selectivity of the imaging layer of a sample can be improved. This system can be used to collect both phase / absorption or dark field information. The system uses scanning elements such as two-axis scanning mirrors 543 and 545 to scan a condensing beam along the imaging site of the eye 541. Other mirrors 540, 544, 548, and 549 are used to reflect light towards the detector. Then, an aperture or pinhole 546 can be placed in a plane conjugate to the imaging plane. In this way, depth selection is improved. Compared with a standard SLO system, since the illumination 542 is provided through the sclera and does not pass through the pupil, only the detection arm is required. The use of the scanning system can be combined with a hardware compensation optical system. The signals can be collected by a one-pixel detector 547 such as a photodiode, an avalanche photodiode, a photomultiplier tube (PMT), a micro PMT, etc., which can be used for standard acquisition or in lock-in mode.

[0125] Referring to FIG. 55, lock-in acquisition can be efficiently incorporated into the system. Lock-in acquisition can be performed using, for example, a lock-in camera with flood illumination or a single detector of the scanning system. And the output of the camera / detector is a DPC image. In addition, several DPC signals can be integrated to have an output that is the average value of the DPC signals. Thereby, the signal-to-noise ratio can be increased. Finally, the digital resources are used more efficiently by removing the background at the initial stage of the readout chain.

[0126] With reference to FIGS. 29, 30, and 31, wavefront shaping will be further described. Different illumination methods are used to provide a speckle pattern on the retinal surface with a speckle grain size smaller than the diameter of photoreceptors, which is several micrometers (D. Mustafi, A., and H. Engel, Krzysztof Palczewski, "Structure of cone photoreceptors", Progress in Retinal and Eye Research, Vol. 28, No. 4, pp. 289-302, 2009). After the high-resolution speckle pattern is irradiated on the fundus, high-resolution images of the fundus can be reconstructed in some embodiments.

[0127] The first embodiment involves looking through the pupil of the eye (and thus with much lower resolution) at the speckle pattern (i.e., collecting a digital image). Due to the memory effect of the scattering medium, the speckle pattern can be moved and an image can be collected each time it is moved. Subsequently, the images collected in this way are used with a phase retrieval algorithm to reconstruct an image of the fundus at the same resolution as the original high-resolution projected speckle pattern. Another embodiment is to place a marker on the surface of the eye, such as by embedding it in a contact lens or by displaying a non-contact marker or any high-resolution eye tracker, in order to obtain feedback and provide a constant speckle pattern by always illuminating the same site.

[0128] Another embodiment of image reconstruction is based on scanning a single focal spot. By wavefront shaping, the speckle pattern can be converted into a single spot (of the same size as the original speckle). Similarly, in this method, the memory effect can be used to scan the spot. By collecting the reflected intensity of each point, the intensity profile of the entire fundus image can be reconstructed. The main problem with this embodiment lies in the focusing part. This is because the resolution caused by the pupil is limited, so the transmission matrix cannot be measured. An alternative, although not limited, is to use an iterative process such as a genetic algorithm (GA). (D. Conkey, A. Brown, A. Caravaca-Aguirre, and R. Piestun, "Genetic algorithm optimization for focusing through turbid media in noisy environments", Opt. Express, vol. 20, pages 4840 - 4849, 2012). The GA can bring about focusing by maximizing a parameter that measures how close the pattern is to the ideal case (perfect focusing). This type of algorithm is the most efficient for the targeted application because it has a fast convergence time, i.e., only about 1000 iterations are required to focus light with an acceptable contrast (I. M. Vellekoop, "Feedback-based wavefront shaping", Opt. Express 23, 12189 - 12206 (2015)).

[0129] In the case of the fundus, the unique properties of this tissue can be used to provide parameters for focusing. Cone photoreceptors (with a diameter of 1 - 1.25 μm (D. Mustafi, A. and H. Engel, Krzysztof Palczewski, "Structure of "cone photoreceptors" (B. Vohnsen, "Photoreceptor waveguides and effective retinal image quality", J. Opt. Soc. Am. A 24, 597 - 607 (2007); B. Vohnsen, I. Iglesias, and P. Artal, "Guided light and diffraction model of human - eye photoreceptors", J. Opt. Soc. Am. A 22, 2318 - 2328 (2005); Progress in Retinal and Eye Research, Vol. 28, No. 4, pp. 289 - 302, 2009)) appear much brighter than the background due to their waveguide properties. Their sparse distribution can be used to break the resolution limit. It can be easily seen that maximizing the total reflectivity coincides with focusing light onto the brightest photoreceptors. Another parameter that can be maximized is the ratio of the total intensity of a site containing only one photoreceptor to the background intensity. When two or more points contribute to the generation of the PSF, its maximum value / energy ratio is smaller than in the ideal case.

[0130] Regarding interference imaging, Fig. 53 shows the optical principle scheme of optical coherence tomography measurement by scattered light. A broadband light source 53 (e.g., SLD) is split into a reference 50 and an object 51 arm. The reference 50 illuminates a mirror 54 that can translate after interference to scan the sample in the depth direction. The object arm illuminates the eye 1 through the skin 8 and / or sclera 9, choroid 10, and retina 11. After illuminating the fundus, the back - scattered light is collected by the pupil 4 and interferes with the reference beam. After recombining the two beams by a beam splitter 42, the interference beam passes through the imaging optical system block 49 and is recorded by a detector 48.

[0131] The elements of the system can be combined with other imaging modalities, such as, but not limited to, OCT, fluorescence imaging, magnetic resonance imaging (MRI), on a single platform to obtain and integrate medical information that aids in diagnosis and establish a multimode retinal imaging platform. In particular, by using a scanning system for image acquisition, the system becomes more compatible with scanning laser ophthalmoscopy and OCT technology.

[0132] Phase imaging of the retina according to aspects of the present invention may be performed with infrared light. The human eye does not perceive infrared light. As a result, in the case of a living retina, for example, through the pupil or sclera, by stimulating the retina with visible wavelengths and performing functional retinal imaging, some retinal functions can be imaged. For example, the response of photoreceptors to different wavelengths can be examined. The functional analysis method, as used herein, can include, but is not limited to, deep learning algorithms.

[0133] Examples of ophthalmic imaging systems include, but are not limited to, optical coherence tomography systems, fundus imaging systems, slit illumination imaging systems, fluorescence angiography imaging systems, indocyanine green angiography imaging systems, fundus autofluorescence imaging systems, corneal topography imaging systems, endothelial cell layer photography systems, and specular microscopy systems for performing multimode imaging of eye tissues.

[0134] The same imaging method can be applied to imaging of the anterior eye segment. Examples of anterior eye segment tissues include, but are not limited to, the lens, endothelium, and cornea. Light scattered from the fundus or pupil passes through these layers and the intensity and phase are modulated. By using an imaging system whose focal plane is not the retina but the anterior layer of the eye (e.g., corneal endothelium), an image including phase and absorption information can be recorded. Similarly, by using different illuminations to record two or more of these images, the absorption and phase profiles can be reconstructed. or more of them, the absorption and phase profiles can be reconstructed.

[0135] Next, with respect to FIG. 38, a general method of describing an imaging protocol for the eye will be explained. According to the embodiments of FIGS. 9A-29, the optical system is positioned on the patient. Next, the imaging system is aligned with the patient's eye. After this step, each point (one or more together at a time) is illuminated with one or more wavelengths selected in the spectrum of about 400 nm to 1200 nm, and an image of the fundus is acquired through the crystalline lens of the eye. Each image is acquired sequentially. The patient's pupil may or may not be dilated. In the case of a dark-field image (at least one illumination point), the captured image directly becomes a dark-field image. If the selected method is phase imaging (at least two illumination points), the acquired image needs to be processed first to obtain a qualitative or quantitative phase image. Once all the images are acquired, the images can be post-processed

[0136] A series of experiments were conducted, demonstrating operability, proof of principle, and results that were substantially improved over the background art

[0137] For measurements related to in vivo phase imaging, to show proof of principle, an indirect ophthalmoscope was made as shown in FIG. 46. With the aspherical lens 30, a field of view of about 60° is possible at the fundus. The objective lens 33 of the camera is focused on the image plane of the lens 30, and the camera 32 records the inverted image 31 of the fundus. FIGS. 49, 50, and 51 show two dark-field images 35, 36 for two humans and the corresponding phase gradient image 37 obtained by subtracting the two dark-field images using the relationship of Equation (1). The illumination wavelength was 643 nm. The left side of FIG. 49 shows the image captured transcutaneously from the illumination point below the center, the center shows the image captured transcutaneously from the illumination point below the left, and the right side shows the two differences showing phase contrast. FIG. 50 shows, for example, the phase gradient image 37 showing the gradient of the optic nerve head and the spectral line profile 38. Next, in FIG. 51, the left side shows the image captured transcutaneously from the illumination point below the right, the center shows the image captured transcutaneously from the illumination point below the left, and the right side shows the two differences showing phase contrast

[0138] Referring to FIG. 47, in the second step, another ophthalmoscope is made to obtain a small field of view. This ophthalmoscope includes a stage for adjusting the focus, a fixed target for the patient, and two telescopes formed by an eyepiece lens including a first lens, a second lens, and a third lens from left to right. In addition, a diaphragm is placed on the pupil plane to filter the beam. Finally, a high-sensitivity camera records the retinal image. FIG. 52 shows an example of a 2×2 mm 2 field of view of the retina under transcutaneous LED (peak wavelength of 870 nm) illumination.

[0139] Referring to FIG. 48, in the third step, the system is designed to correct aberrations using a closed-loop wavefront sensor and a deformable mirror with an aberration correction method.

[0140] Examples of aberration correction methods include, but are not limited to, a deformable mirror, a spatial light modulator, a Burch system, a tunable lens, a series of cylindrical lenses, the Waller method, the modified Waller method, and a blind deconvolution algorithm when used herein.

[0141] Regarding the measurements related to in vitro phase imaging, referring to FIG. 42, a microscope having the same parameters as in the case of in vivo imaging is made. This microscope includes oblique illumination first scattered by a scattering plate, a reflective layer that provides backscattered light, an objective lens of the microscope, an imaging lens, and a camera. FIG. 43 shows the in vitro measurement results together with a comparison with a digital holographic microscope that provides an image evaluation of resolution, an image obtained by the present invention, and a quantitative phase image. FIG. 43 also shows a comparison with an image obtained by a confocal microscope that provides an intensity image. FIGS. 44 and 45 show the depth direction scan results of a porcine retina together with different layers of the retina, and calculate the cell density of the nuclear layer. density.

[0142] Next, measurements were taken to demonstrate operability and proof of principle using wavefront shaping. To demonstrate proof of principle, a static sample of microbeads and a liquid-crystal-based spatial light modulator were used to optimize the feedback light. This system is shown in FIG. 32. A linearly polarized collimated laser beam illuminates the SLM before passing through a 400-μm-thick scattering layer. Next, the scattered beam illuminates the sample through an objective lens with a 0.25 NA. The reflected light is collected by the objective lens, passes through a diaphragm to artificially reduce the detection NA and mimic the limited resolution of the eye's pupil. A sample of microbeads is used to reproduce the situation of high-reflectivity features that the detection system cannot resolve.

[0143] FIG. 33 shows the results of the focusing process performed for one bead with a diameter of 10 μm and a detection NA of 0.02. The procedure is as follows: After recording the image at maximum resolution (the curve before optimization in FIG. 33D), the diaphragm is closed to optimize the wavefront (FIG. 33C). The low-resolution PSFs before and after optimization are shown in FIG. 33C. Finally, the diaphragm is opened to record the optimized high-resolution PSF (the curve after optimization in FIG. 33D). Two-dimensional images before (FIG. 33A) and after (FIG. 33B) optimization are shown. For a sample of several beads, the focusing is not as good as for one bead, so a method for identifying the beads based on the shape of these PSFs is developed. When two beads are closer than the resolution distance, the collected image is similar to the PSF. In any case, the ratio of the maximum value to the total energy changes according to the distance between the centers. This parameter can be used to distinguish between the case of one bead and the case of multiple beads.

[0144] Various applications can be performed by the devices, systems, and methods of the present invention. Examples of applications include quantitative phase imaging of retinal layers on photoreceptors, retinal layers between the inner and outer limiting membranes, such as the inner limiting membrane (ILM), retinal nerve fiber layer (RNFL), ganglion cell layer (GCL), inner plexiform layer (IPL), inner nuclear layer (INL), outer plexiform layer (OPL), outer nuclear layer (ONL), and outer limiting membrane (ELM).

[0145] Next, the proposed method can provide dark-field images of the choroid and retinal pigment epithelium (RPE), and can image choroidal tumors and the microvasculature of the choroid with enhanced contrast.

[0146] Finally, by recording two dark-field images, phase gradient information of the retinal layer can be obtained.

[0147] In short, according to an aspect of the present invention, the visual process of the eye is determined by the first layer of the retina. The light entering the eye needs to pass through a layer about 100 μm thick of ganglion and nerve cells that form the retina before reaching the photoreceptor cells. Since these cells are phase objects, they are difficult to observe by standard imaging methods. In fact, phase imaging methods usually require the illumination system to be placed on the opposite side of the sample with respect to the imaging system, and phase imaging cannot be performed in vivo. However, it has been shown that it is possible to perform phase imaging from one side using the characteristics of the scattering medium.

[0148] According to an aspect of the present invention, a system for performing qualitative and quantitative imaging of the fundus by oblique illumination is proposed. By using different illumination points directly through the pupil, on the sclera itself, or on the skin covering the sclera, oblique back-illumination is provided through the scattering characteristics of the eye, and phase-contrast images are obtained. Using these phase-contrast images, images containing only phase or absorption information can be reconstructed. Furthermore, using the same illumination method, dark-field images can be collected from the pupil. Furthermore, by using incoherent illumination, the resolution of the restored image can be doubled compared to coherent imaging. - rent imaging and the resolution of the restored image can be doubled.

[0149] In this application, it has been shown that phase contrast can be obtained, and how absolute absorption and phase profiles can be obtained for two-dimensional (2D) and three-dimensional (3D). According to some embodiments, different illumination modes are shown that provide phase contrast and different devices for providing this illumination and acquiring images. Algorithms for reconstructing 2D and 3D phase and absorption profiles have been described. In addition, the secondary information obtained by this technique and different improvements has been described.

[0150] The present invention has been disclosed with reference to certain preferred embodiments, but many modifications, alterations, and changes can be made to the described embodiments and their equivalents without departing from the scope of the invention. Accordingly, the invention is not limited to the described embodiments, but is to be accorded the broadest reasonable interpretation in accordance with the language of the appended claims.

[0151] References ·U.S. Pat. No. 7,387,385 ·U.S. Pat. Pub. No. 2007 / 0159600 ·U.S. Pat. Pub. No. 2007 / 0030448 ·U.S. Pat. No. 3,954,392 ·U.S. Pat. No. 4,200,362 ·Medibell Medical Vision Technologies Ltd.- Panoret 1000-Wide-Angle Digital Retinal Camera, printed at least as early as Oct. 2002. ·A. Schalenbourg, L. Zografos “Pitfalls in colour photography of choroidal tumours.” Eye. 2013;27(2):224-229 ·Devrim Toslak, Damber Thapa, Yanjun Chen, Muhammet Kazim Erol, R. V. Paul Chan, and Xincheng Yao, “Trans-palpebral illumination: an approach for wide-angle fundus photography without the need for pupil dilation," Opt. Lett. Vol. 41, pp. 2688-2691 (2016) ·D. Scoles, Y. N. Sulai and A. Dubra “In vivo dark-field imaging of the retinal pigment epithelium cell mosaic,” Biomed. Opt. Exp. Vol. 4, 9, pp. 1710-1723 (2013) ·T. Y. P. Chui, D. A. VanNasdale, and S. A. Burns, “The use of forward scatter to improve retinal vascular imaging with an adaptive optics scanning laser ophthalmoscope,” Biomed. Opt. Exp. Vol. 3, 10, pp. 2537-2549 (2012) ·T. Y. P. Chui, T. J. Gast, and S. A. Burns, “Imaging of Vascular Wall Fine Structure in the Human Retina Using Adaptive Optics Scanning Laser Ophthalmoscopy,” Invest Ophthalmol VisSci. vol. 54, pp. 7115-7124 (2013) ·T. N Ford, K. K Chu and J. Mertz, “Phase-gradient microscopy in thick tissue with oblique back-illumination,” Nat. Methods, Vol. 9, 12 (2012) ·S. B. Mehta and C. J. R. Sheppard, “Quantitative phase-gradient imaging at high resolution with asymmetric illumination-based differential phase contrast,” Opt. Lett. 34, 13, pp. 1924-1926 (2009) ·L. Tian and L. Waller, “Quantitative differential phase contrast imaging in an LED array microscope,” Opt. Exp. 23, 9, pp. 11394-11403 (2015) ·Z. Liu, S. Liuand, L. Waller “Real-time brightfield, darkfield, and phase contrast imaging in a light emitting diode array microscope,” Journal of Biomed. Opt. 19,10, 106002 (2014) ·Int. Pat. Pub. No. WO 2013 / 148360 ·Int. Pat. Pub. No. WO 2015 / 179452 ·G. Zheng, R. Horstmeyer, and C. Yang, “Wide-field, high-resolution Fourier pt ychographic microscopy,” Nature photonics, Vol. 7, Iss. 9, 2013, pp. 739-745. ·U.S. Pat. No. 8,731,272 ·U.S. Pat. Pub. No. 2004 / 0189941 ·European Pat. No. EP 1427328 ·U.S. Pat. No. 7,364,296 ·M. Vellekoop and A. P. Mosk, "Focusing coherent light through opaque strongly scattering media," Opt. Lett.32, 2309-2311 (2007) ·H. Yilmaz, E. G. van Putten, J. Bertolotti, A. Lagendijk, W. L. Vos, and A. P. Mosk, "Speckle correlation resolution enhancement of wide-field fluorescence imaging," Optica 2, 424-429 (2015) ·U.S. Pat. No. 8,717,574

Claims

1. 1. A method for imaging ocular tissue, comprising: providing oblique illumination to the eye with a plurality of light-emitting regions of a light delivery device, the plurality of light-emitting regions being independently controllable and positioned to direct light to at least one of the sclera, the retina, and the iris of the eye; generating an output beam from light scattered from at least one of the retina and the iris by oblique illumination; capturing the output beam with an imaging system to provide a series of images of the fundus; wherein the oblique illumination has a range of illumination angles that is different from the range of collection angles of the output beam; the series of images of the fundus in the capturing step are obtained by sequentially turning on one or more of the plurality of light emitting regions at a time in the providing oblique illumination step; The method, wherein the imaging system includes a wavefront sensor and a deformable mirror at a plane conjugate to the pupil plane, the wavefront sensor being positioned after the deformable mirror, and the wavefront sensor being used in a closed loop with the deformable mirror to compensate for aberrations.

2. The ocular tissue is a part of a living human or animal eye, the oblique illumination is at least one of transpupillary illumination, transscleral illumination, and transepidermal illumination; The light delivery device supports the following illumination modalities: There is no contact between the light delivery device and the patient's face; The light delivery device contacts the skin around the eye; The light delivery device contacts the sclera of the eye; and The light delivery device contacts the cornea of ​​the eye. The method of claim 1 , configured for at least one of:

3. The method of claim 1 , wherein the ocular tissue is an ex vivo sample of a human or animal eye.

4. Oblique illumination is a type of illumination that can be divided into divergent beam, collimated beam, focused beam, and structured illumination. The method of claim 1 formed by at least one

5. the ocular tissue comprises at least one of a human in vivo retina, a human ex vivo retina, and an animal in vivo retina; The method of claim 1 , wherein the capturing step includes at least one modality of dark field illumination, dark field collection, focused coherent illumination with wavefront shaping, and tilted optical coherence tomography with a low coherence source.

6. The method of claim 1 , wherein in the retrieving step, the reconstructed phase and absorption images are obtained by a phase and absorption retrieval algorithm.

7. The method of claim 1 , wherein in the capturing step, the series of images is captured by at least one of 2D single frame acquisition and 2D lock-in acquisition.

8. the ocular tissue is an anterior segment tissue, The method of claim 1 , wherein in the step of providing oblique illumination, the illumination is obtained by back reflection from at least one of a fundus and an iris of the eye.

9. measuring at least one of an ocular aberration and an illumination function using at least one of a wavefront sensor and a pupil camera; The method of claim 1 , further comprising the step of: correcting the aberrations by an aberration correction method.

10. The method of claim 1 , wherein functional information is extracted from data of the series of images.

Images