Processing techniques for optical retinal imaging.

Abstract

To provide a computer-implemented method for generating an individual index of a tissue velocity at a position along an axial direction in an OCT image.SOLUTION: There is provided a computer-implemented method for processing each of a plurality of OCT image sets in a sequence of an OCT image of a part of the retina and generating an individual index of a tissue velocity at a position along an axial direction in the OCT image. The computer-implemented method generates an individual index of a tissue velocity at a position along an axial direction in the OCT image by executing: calculation of an individual comparison value on the basis of the quality or the similarity of a first OCT image and a second OCT image in a set for each set; use of phase information in the first OCT image and the second OCT image for calculating an individual index of a tissue velocity at a position when the comparison value is a threshold or more; and exclusion of phase information from the calculation when the comparison value is smaller than the threshold.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] Exemplary aspects of the present specification relate generally to the field of optical coherence tomography (OCT), and more particularly to techniques for processing optical retinal imaging (optoretinography) data generated by a Fourier-domain OCT imaging system to generate OCT data indicative of the physiological response of the retina of a subject's eye to optical stimuli. [Background technology]

[0002] Optical coherence tomography (OCT) is an imaging technique based on low-coherence interferometry that is widely used to obtain high-resolution two- and three-dimensional images of light-scattering media such as biological tissue.

[0003] Depending on how the depth range is achieved, OCT imaging systems can be classified as time-domain OCT (TD-OCT) or Fourier-domain OCT (FD-OCT) (also called frequency-domain OCT). In TD-OCT, the optical path length of the reference arm of the imaging system's interferometer is varied in time during the acquisition of a reflectance profile of a scattering medium (referred to herein as an "imaging target") being imaged by the OCT imaging system; the reflectance profile is commonly referred to as a "depth scan" or "axial scan" ("A-scan"). In FD-OCT, the spectral interferogram resulting from the interference of light in the reference arm with light in the sample arm of the interferometer at each A-scan position is Fourier transformed to simultaneously acquire all points along the depth of the A-scan without requiring a change in the optical path length of the reference arm. FD-OCT can enable much faster imaging than scanning the sample arm mirror in an interferometer, because all backreflections from the sample are measured simultaneously. Two common types of FD-OCT are spectral-domain OCT (SD-OCT) and swept-source OCT (SS-OCT). In SD-OCT, a broadband light source delivers many wavelengths to the imaging target, and a spectrometer is used as the detector to measure all wavelengths simultaneously. In SS-OCT (also called time-encoded frequency-domain OCT), the light source is swept across a range of wavelengths, and the temporal output of the detector is converted to spectral interference.

[0004] OCT imaging systems can also be classified as point scanning (also known as "point detection" or "scanning point"), line scanning, or full-field scanning, depending on how the imaging system is configured to acquire OCT data at locations on the imaging target. Point-scanning OCT imaging systems acquire OCT data by scanning a focused sample beam across the surface of the imaging target, typically along a single line (which may be straight, for example, or alternatively curved to define a circle or spiral), or along a set of (usually substantially parallel) lines on the surface of the imaging target, acquiring axial depth profiles (A-scans) for each of multiple points along the line(s), one point at a time, to construct OCT data comprising a one- or two-dimensional array of A-scans representing a two-dimensional (i.e., B-scan) or three-dimensional (i.e., C-scan or volumetric scan) reflectivity profile of the sample.

[0005] Line-scanning OCT imaging systems acquire OCT data by scanning a focused beam of light across the surface of an imaging target. The measured reflectance from the imaging target is used to generate OCT data that includes a two-dimensional reflectance profile (i.e., a B-scan) of the sample. By scanning the focused beam of light across multiple locations on the imaging target, OCT data that includes a three-dimensional reflectance profile (i.e., a C-scan or volumetric scan) of the sample can be obtained. Typically, the focused beam of light is straight and scanned in a direction perpendicular to it, but in some cases it may be curved and scanned with a scan direction adjusted accordingly. Full-field OCT imaging systems acquire OCT data that includes a three-dimensional reflectance profile (i.e., a C-scan or volumetric scan) of the sample by projecting a light beam onto the imaging target.

[0006] OCT imaging systems can also be classified as phase-resolved, in which both the intensity and phase of light reflected from the imaging target are measured as a function of axial depth. Modern FD-OCT imaging systems often have a degree of phase stability that allows them to function as phase-resolved OCT imaging systems.

[0007] Optical retinal imaging (ORG) generally refers to the detection of the physiological response of the retina of the eye to optical stimuli (i.e., functional activity of the retina induced by light). ORG techniques involve noninvasive optical imaging of this physiological response of the retina. For example, OCT imaging systems can be used to image retinal neurons that exhibit dimensional (size) changes in response to excitation by optical stimuli. These dimensional changes (typically changes in the length of the photoreceptor outer segment (OS) in the retina, i.e., the difference in depth between the inner-outer segment (IS / OS) junction and the cone outer segment tip (COST) of the cone photoreceptor) result in a sufficiently large phase change in the light waves returned from the eye by a phase-resolved OCT imaging system to be detectable. Phase is highly sensitive to motion in tissue, and many phase-resolved OCT imaging systems can resolve displacements of less than 10 nm, which can be much smaller than the axial resolution of the system or the wavelength of the light used for imaging. Summary of the Invention

[0008] According to a first exemplary aspect of the present specification, there is provided a computer-implemented method for processing each set of OCT images, including a first OCT image and a second OCT image, of a plurality of different sets of OCT images in a sequence of OCT images of a common portion of a retina of an eye acquired by a Fourier domain OCT imaging system to generate a separate measure of tissue velocity at an axial position in the OCT image. The method includes performing the following processes for each set of OCT images: calculating an individual comparison value, the individual comparison value being one of (i) an image quality metric value calculated based on at least one of a first OCT image in the set and a second OCT image in the set; and (ii) an image similarity metric value providing a measure of similarity between the images, calculated based on the first OCT image in the set and a second OCT image in the set; calculating an individual measure of tissue velocity at the axial position using a phase value at the axial position in the A-scan of the first OCT image in the set and a phase value at the axial position in the corresponding A-scan of the second OCT image in the set; comparing the comparison value with a threshold to determine whether the comparison value is greater than or equal to the threshold; retaining the individual measure of tissue velocity at the axial position if it is determined that the comparison value is greater than or equal to the threshold; and discarding the individual measure of tissue velocity at the axial position if it is determined that the comparison value is not greater than or equal to the threshold.

[0009] According to a second exemplary aspect of the present specification, there is provided a computer-implemented method for processing each set of multiple different OCT images in a sequence of OCT images of a common portion of a retina of an eye acquired by a Fourier domain OCT imaging system, the set including a first OCT image and a second OCT image, to generate a separate measure of tissue velocity (v) at an axial position in the OCT image. The method includes performing the following processes: for each set of OCT images, calculating an individual comparison value, the individual comparison value being one of (i) a value of an image quality metric calculated based on at least one of a first OCT image in the set and a second OCT image in the set; and (ii) a value of an image similarity metric providing a measure of similarity between the images, calculated based on the first OCT image in the set and a second OCT image in the set; comparing the comparison value to a threshold to determine whether the comparison value is greater than or equal to the threshold; if it is determined that the comparison value is greater than or equal to the threshold, using a phase value at an axial position in the A-scan of the first OCT image in the set and a phase value at an axial position in the corresponding A-scan of the second OCT image in the set in calculating an individual measure of tissue velocity at an axial position; if it is determined that the comparison value is not greater than or equal to the threshold, excluding a phase value at an axial position in the A-scan of the first OCT image in the set and a phase value at an axial position in the corresponding A-scan of the second OCT image in the set in calculating an individual measure of tissue velocity at an axial position.

[0010] In the above method, the comparison value can be the maximum cross-correlation calculated between the first and second OCT images. Other similarity measures that can be used instead include, for example, the sum of squared differences, mutual information, normalized mutual information, and the Kullback-Leibler distance.

[0011] The computer-implemented method of the first exemplary embodiment or the second exemplary embodiment may further include generating a concatenation of the generated indices of tissue velocity such that the concatenation indicates how tissue velocity at an axial position changes over time, and integrating the concatenation to generate data indicative of optical path length change at an axial position over time. If it is determined that the comparison value is not greater than or equal to the threshold, the individual indices of tissue velocity at the axial position are set to indicate zero velocity, such that the generated data includes one or more sets of equal consecutive values, and the method may further include smoothing the generated data by replacing one or more values ​​in one set of the one or more sets of equal consecutive values ​​with one or more estimated values ​​calculated based on adjacent values ​​adjacent to the set of equal consecutive values ​​in the generated data.

[0012] Additionally, in any of the foregoing, the computer-implemented method may further include, prior to calculating the individual measures of tissue velocity at axial positions, processing a phase value at an axial position in an A-scan of a first OCT image in the set and a phase value at an axial position in a corresponding A-scan of a second OCT image in the set to compensate for bulk motion of a common portion of the retina during acquisition of the sequence of OCT images by the Fourier-domain OCT imaging system.

[0013] According to a third exemplary aspect of the present specification, there is provided a computer program comprising computer-readable instructions that, when executed by a processor, cause the processor to perform the method of the first exemplary aspect or the second example, or any of the variations thereof described above. The computer program may be stored on a non-transitory computer-readable storage medium (e.g., a computer hard disk or a CD, etc.) or may be carried by a computer-readable signal.

[0014] According to a fourth exemplary aspect of the present specification, there is provided a data processing apparatus including a processor and computer readable instructions that, when executed by the processor, cause the processor to perform the method of the first exemplary aspect or the second example, or any of the variations thereof described above.

[0015] According to a fifth exemplary aspect of the present specification, there is provided a data processing device arranged to process each set of optical coherence tomography (OCT) images, including a first OCT image and a second OCT image, of a set of multiple different OCT images in a sequence of OCT images of a common portion of a retina of an eye (20) acquired by a Fourier domain OCT imaging system to generate a respective measure of tissue velocity at an axial position in the OCT image. The data processing device is configured to perform the following processes for each set of OCT images: calculating an individual comparison value, which is one of (i) an image quality metric value calculated based on at least one of a first OCT image in the set and a second OCT image in the set, and (ii) an image similarity metric value providing a measure of similarity between the images, calculated based on the first OCT image in the set and a second OCT image in the set; calculating an individual measure of tissue velocity at the axial position using a phase value at the axial position in the A-scan of the first OCT image in the set and a phase value at the axial position in the corresponding A-scan of the second OCT image in the set; comparing the comparison value with a threshold to determine whether the comparison value is greater than or equal to the threshold; retaining the individual measure of tissue velocity at the axial position if it is determined that the comparison value is greater than or equal to the threshold; and discarding the individual measure of tissue velocity at the axial position if it is determined that the comparison value is not greater than or equal to the threshold.

[0016] According to a sixth exemplary aspect of the present specification, there is provided a data processing device arranged to process each set of optical coherence tomography (OCT) images, comprising a first OCT image and a second OCT image, of a set of multiple different OCT images in a sequence of OCT images of a common portion of a retina of an eye acquired by a Fourier domain OCT imaging system, to generate an individual measure of tissue velocity at an axial position in the OCT image. The data processing device is configured to perform the following processes: for each set of OCT images, calculating an individual comparison value, the individual comparison value being one of (i) an image quality metric value calculated based on at least one of a first OCT image in the set and a second OCT image in the set; and (ii) an image similarity metric value providing a measure of similarity between the images, calculated based on the first OCT image in the set and a second OCT image in the set; comparing the comparison value to a threshold to determine whether the comparison value is greater than or equal to the threshold; if it is determined that the comparison value is greater than or equal to the threshold, using a phase value at an axial position in the A-scan of the first OCT image in the set and a phase value at an axial position in the corresponding A-scan of the second OCT image in the set in calculating an individual measure of tissue velocity at an axial position; and if it is determined that the comparison value is not greater than or equal to the threshold, excluding a phase value at an axial position in the A-scan of the first OCT image in the set and a phase value at an axial position in the corresponding A-scan of the second OCT image in the set in calculating an individual measure of tissue velocity at an axial position.

[0017] The comparison value may be a maximum value of cross-correlation calculated between the first OCT image and the second OCT image. Additionally or alternatively, the data processor may be further configured to generate a concatenation of the generated indices of tissue velocity such that the concatenation indicates how tissue velocity at an axial position varies over time, and to integrate the concatenation to generate data indicative of optical path length change at an axial position over time. If it is determined that the comparison value is not greater than or equal to the threshold, the data processor may be configured to set the individual indices of tissue velocity at the axial position to indicate zero velocity, and the generated data may include one or more sets of equal consecutive values, and the data processor may be further configured to smooth the generated data by replacing one or more values ​​in the one or more sets of equal consecutive values ​​with one or more estimated values ​​calculated based on neighboring values ​​adjacent to the set of equal consecutive values ​​in the generated data.

[0018] The data processing device may be further arranged to process the phase values ​​at the axial positions in the A-scan of a first OCT image in the set and the phase values ​​at the axial positions in the corresponding A-scan of a second OCT image in the set prior to calculation of the individual measures of tissue velocity at the axial positions to compensate for bulk motion of a common portion of the retina during acquisition of the sequence of OCT images by the Fourier domain OCT imaging system.

[0019] According to a seventh exemplary aspect of the present specification, there is provided a Fourier domain OCT imaging system comprising any of the data processing devices described above.

[0020] Exemplary embodiments will now be described in detail, by way of non-limiting example only, with reference to the accompanying drawings in which: like reference numbers appearing in different drawings may, unless otherwise indicated, indicate identical or functionally similar elements, and in which: [Brief explanation of the drawings]

[0021] [Figure 1]FIG. 1 is a schematic diagram of a system including a Fourier domain OCT imaging system 30 and a data processing apparatus 100, according to an exemplary embodiment of the present disclosure. [Figure 2] FIG. 2 is a schematic diagram of an exemplary implementation of data processing device 100 in programmable signal processing hardware. [Figure 3] FIG. 3 is a flow diagram illustrating a method for processing OCT images to generate a measure of tissue velocity according to a first exemplary embodiment herein. [Figure 4A] FIG. 4A shows the first pair of highly correlated B-scans. [Figure 4B] FIG. 4B shows the calculated cross-correlation of the B-scans of FIG. 4A. [Figure 5A] FIG. 5A shows a second pair of highly correlated B-scans. [Figure 5B] FIG. 5B shows the calculated cross-correlation of the B-scans of FIG. 5A. [Figure 6A] Figure 6A shows sinusoidal bulk vibrations added to the oscillatory motion of two model retinal layers. [Figure 6B] FIG. 6B shows the effect of removing bulk motion on the time course of the phase angle of light reflected from two model retinal layers. [Figure 7] FIG. 7 shows an example of a concatenation of velocity profiles that can be generated by the data processing device 100. [Figure 8] FIG. 8 shows a plot of the change in optical path length ΔOPL over time calculated by the data processing device of the exemplary embodiment for selected positions within the velocity profile of a concatenation of velocity profiles. [Figure 9] FIG. 9 shows a plot of an exemplary ORG signal generated by a data processing device of an exemplary embodiment herein. [Figure 10] FIG. 10 shows the result of smoothing the ORG signal of FIG. [Figure 11] FIG. 11 is a flow diagram illustrating an optional process that may be performed by the data processing apparatus 100 of the embodiments described herein. [Figure 12]FIG. 12 is a flow diagram illustrating a method for processing OCT images to generate a measure of tissue velocity according to a second exemplary embodiment herein. DETAILED DESCRIPTION OF THE INVENTION

[0022] A core part of the process of generating ORG data from a set of repeated B-scans (or other OCT images) acquired by an FD-OCT imaging system after application of optical stimulation is processing each subset of B-scans to generate individual indices of tissue velocity at one or more axial locations in the B-scan. The inventors have recognized that some of the tissue velocity indices calculated using conventional ORG processing techniques may be unreliable and may adversely affect the ORG data or other final processed results. The inventors have further devised a scheme to recognize the source B-scans (or other OCT images) that give rise to such unreliable indices and prevent the unreliable indices of tissue velocity from entering the processing pipeline in order to improve the ORG data or other final processed results. Embodiments will now be described in detail with reference to the accompanying drawings.

[0023] First Exemplary Embodiment 1 is a schematic diagram of a data processing device 100 according to a first exemplary embodiment. The data processing device 100 is arranged to process complex OCT data generated by a phase-resolved Fourier-domain OCT (FD-OCT) imaging system. More specifically, the data processing device 100 is arranged to process each set of OCT images, including a first OCT image 10-1 and a second OCT image 10-2, of a set of multiple different OCT images in a sequence of OCT images 10 of a common portion of the retina of an eye 20 imaged by a Fourier-domain OCT imaging system 30 to generate a separate measure of tissue velocity v at one or more locations along an axial direction z in the OCT image 10 (i.e., along either direction of the OCT image (e.g., B-scan or C-scan) in which elements of the component A-scans of the OCT image are arranged).

[0024] The FD-OCT imaging system 30 may be a swept-source OCT (SS-OCT) system, as in this exemplary embodiment. However, the FD-OCT imaging system 30 need not be provided in this form and may take alternative forms, such as spectral-domain OCT (SD-OCT). More generally, the exemplary embodiment may be provided as any form of phase-resolved FD-OCT imaging system capable of generating complex OCT data, i.e., the Fourier transform of each spectral interferogram representing the complex A-scan information obtained for each scan position at which OCT measurements are taken during a scan. Such complex OCT data encodes phase information from the acquired OCT measurements, which can be used by the data processing device 100 described herein to calculate tissue velocities in a common portion of the retina.

[0025] The FD-OCT imaging system 30 may include well-known components, including a scanning system, a photodetector, OCT data processing hardware, and a light beam generator (not shown). The scanning system may be configured to perform one-dimensional and / or two-dimensional point scans of a light beam across the retina and collect light scattered by the retina during the point scans. Thus, the scanning system is configured to acquire A-scans at each scan location distributed across the surface of the retina by sequentially irradiating the scan locations with a light beam, one scan location at a time, and collecting at least a portion of the light scattered by the retina at each scan location. The scanning system may acquire OCT images in the form of repeated B-scans by performing point scans using a linear scan pattern that follows a set of overlapping scan lines during the scan. While in this exemplary embodiment, the scanning system is configured to acquire repeated B-scans by performing point scans, in other embodiments, the scanning system may be configured to acquire repeated B-scans by performing line scans using hardware known to those skilled in the art. The FD-OCT imaging system 30 may alternatively be arranged to acquire OCT images in the form of a C-scan by performing a point scan or line scan using techniques well known to those skilled in the art, or by using a full-field setup.

[0026] The first OCT image described above may be in the form of a first OCT B-scan 10-1, as shown in FIG. 1 , as in this exemplary embodiment, and the second OCT image may be in the form of a second OCT B-scan 10-2. However, the forms of the first and second OCT images are not particularly limited; for example, the first and second OCT images may each be an OCT C-scan. The first B-scan 10-1 and second B-scan 10-2 of a common portion of the retina may be adjacent to each other in the set of B-scans of the sequence of OCT B-scans 10, as in this exemplary embodiment, or may be separated by n intervening B-scans (n≧1, but small enough to allow the B-scans to have a sufficiently high phase correlation with each other due to the lack of significant relative movement of the eye 20 with respect to the FD-OCT imaging system 30).

[0027] As will be explained in more detail below, the data processing device 100 is arranged to process at least a portion of the phase component of the first B-scan 10-1 and at least a portion of the phase component of the second B-scan 10-2, for example as shown in FIG. 1 , to calculate a velocity profile P including velocity v values, where the variation with position in the velocity profile indicates the distribution of velocities v between points along the axial direction z of a common portion of the retina. Thus, the velocity profile P may be, as in this exemplary embodiment, a two-dimensional data structure in which values ​​of velocity v are associated with corresponding values ​​of position. More generally, the velocity profile P may be a two-dimensional data structure in which values ​​indicative of velocity v (e.g., of calculated changes in phase or optical path length) are associated with corresponding values ​​of position. However, the data processing device 100 need not calculate a velocity profile P, but may be limited to calculating a separate measure of tissue velocity v at a single common location along the axial direction z for each of a plurality of sets of two or more B-scans.

[0028] The data processing apparatus 100 may be provided in any suitable form, for example, as programmable signal processing hardware 200 of the type shown schematically in FIG. 2. The programmable signal processing hardware 200 comprises a communications interface (I / F) 210 for receiving B-scans (or other forms of OCT images, such as C-scans) from the FD-OCT imaging system 30 and outputting calculated tissue velocity indices and / or graphical representations thereof on a display, such as a computer screen. The signal processing hardware 200 further comprises a processor (e.g., a central processing unit CPU and / or a graphics processing unit GPU) 220, a working memory 230 (e.g., random access memory), and an instruction store 240 for storing a computer program 245 containing computer-readable instructions that, when executed by the processor 220, cause the processor 220 to perform the various functions of the data processing apparatus 100 described herein. The working memory 230 stores information used by the processor 220 during execution of the computer program 245. The instruction store 240 may comprise a ROM (e.g., in the form of an electrically erasable programmable read-only memory (EEPROM) or flash memory) pre-loaded with computer-readable instructions. Alternatively, the instruction store 240 may comprise a RAM or similar type of memory, and the computer-readable instructions of the computer program 245 may be input to the instruction store from a non-transitory computer-readable storage medium 250 in the form of a CD-ROM, DVD-ROM, etc., or a computer program product such as a computer-readable signal 260 carrying the computer-readable instructions. In either case, the computer program 245, when executed by the processor 220, causes the processor 220 to perform the functions of the data processing apparatus 100 described herein.Thus, the data processing apparatus 100 of an exemplary embodiment may comprise a computer processor 220 and a memory 240 storing computer readable instructions that, when executed by the processor 220, cause the processor 220 to process each set of OCT images of a plurality of different sets of OCT images in a sequence of OCT images 10 of a common portion of the retina of the eye 20 to generate an individual index indicative of tissue velocity v at a position along the axial direction z in the OCT image 10 (i.e., in either direction of the OCT image (such as a B-scan or C-scan) in which the elements of the component A-scans of the OCT image are arranged).

[0029] However, it should be noted that the data processing apparatus 100 may alternatively be implemented with non-programmable hardware such as an ASIC, FPGA, or other integrated circuit dedicated to performing the functions of the data processing apparatus 100 described herein, or a combination of such non-programmable hardware with programmable signal processing hardware 200, as described above with reference to FIG. 2.

[0030] The data processing device 100 may be provided as a standalone product or as part of a system 1000 that includes an FD-OCT imaging system 30, which is typically arranged to acquire a sequence of OCT images 10 of a common portion of the retina of the eye 20 both before and after stimulation of the common portion with an optical stimulus 40 generated by a light source 45 (e.g., an LED). The data processing device 100 is arranged to process at least a portion of the phase component of a first OCT image 10-1 and at least a portion of the phase component of a second OCT image 10-2 of the sequence of OCT images 10 acquired by the FD-OCT imaging system 30 using the techniques described herein.

[0031] FIG. 3 is a flow diagram illustrating how the data processing device 100 processes each set of B-scans, e.g., a pair of B-scans including a first B-scan 10-1 and a second B-scan 10-2, from a plurality of sets of B-scans in the sequence of B-scans 10 of a common portion of the retina shown in FIG. 1 to generate a separate measure of tissue velocity v at position along the axial direction z in the B-scan 10.

[0032] In process S10 of FIG. 3 , the data processing device 100 calculates a comparison value based on the first OCT image (B-scan 10-1 in this exemplary embodiment) and the second OCT image (B-scan 10-2 in this exemplary embodiment). The comparison value can be the value of an image quality metric or attribute calculated based on at least one of the amplitude components of the first B-scan 10-1 and the second B-scan 10-2. The image quality metric (attribute) can take the form of dynamic range, contrast, signal-to-noise ratio, or a sharpness function value from the B-scans. A measure of success in intensity-based segmentation of the B-scans is another example of an image quality metric that can be used. Alternatively, the comparison value can be the value of an image similarity metric that provides a measure of similarity between the images, where the value of the image similarity metric is calculated based on the first B-scan 10-1 and the second B-scan 10-2. The image similarity metric, as in this exemplary embodiment, can take the form of a maximum value of cross-correlation between the first B-scan 10-1 and the second B-scan 10-2. Other similarity measures that may be used instead include sum of squared differences, mutual information, normalized mutual information, and Kullback-Leibler distance, among others.

[0033] f * The cross-correlation between two complex functions f(t) and g(t) of a real variable t, denoted by g, is given by f * g=fbar(t)*g(t)(in the formula, * denotes convolution, and fbar(t) is the complex conjugate of f(t).

[0034] The maximum cross-correlation between two B-scans is a good metric for how closely the B-scans are related to each other. For example, the two B-scans shown in Figure 4A are highly correlated with each other, and the calculated cross-correlation between these B-scans is approximately 3·10, as shown in Figure 4B. 4 It has a high peak value of

[0035] For comparison, Figure 5A shows two different B-scans. This dissimilarity can be caused by factors such as eye movement, leading to poor image registration. For the exemplary B-scans shown in Figure 5A, the maximum calculated cross-correlation between the B-scans is much lower than for the B-scans shown in Figure 5A, approximately 1·10 4 is.

[0036] 3 , in process S20, the data processing device 100 calculates a measure of tissue velocity v at a position along the axial direction z using a phase value at an axial position in the A-scan of the first OCT image (B-scan 10-1) and a phase value at the same position along the axial direction z in the corresponding A-scan of the second OCT image (B-scan 10-2). The corresponding A-scan of B-scan 10-2 has the same position along the x-axis along which the A-scan of B-scan 10-2 is arranged as the position of the A-scan considered in B-scan 10-1. Processing each set of B-scans according to S20 of FIG. 3 enables the data processing device 100 to calculate an individual measure of tissue velocity v at a single common position along the axial direction z for each of the set of B-scans, or, as in this exemplary embodiment, an individual velocity profile P for each of the set of B-scans, which velocity profile P indicates the distribution of tissue velocity v along the axial direction z in a common portion of the retina.

[0037] 3, the data processing device 100 processes the phase components of the first B-scan 10-1 and the second B-scan 10-2 to calculate a velocity profile P indicative of the distribution of velocities v along the axial direction z in the common portion of the retina. Thus, the data processing device 100 calculates the velocity profile P, for example, where the variation with position in the velocity profile includes values ​​of velocity v indicative of the distribution of velocities v in the portion of the retina (in the common portion) at each point on an axially aligned line, or values ​​of phase or optical path length change ΔOPL. The velocity profile P is thus a two-dimensional data structure in which values ​​of tissue velocity v (or values ​​of a variable indicative thereof, such as ΔOPL or phase change Δφ) are associated with corresponding values ​​of position, and this data structure can be visualized in the form of a plot of velocity v (or Δφ or ΔOPL) as a function of position z, as shown schematically in FIG.

[0038] Prior to calculating the measure of tissue velocity v in process S20 of FIG. 3 , the data processing device 100 may process the phase values ​​at positions along the axial direction z in the A-scan of the first B-scan 10-1 and the phase values ​​at positions along the axial direction z in the corresponding A-scan of the second B-scan 10-2 to compensate for bulk motion of the common portion of the retina during acquisition of the sequence of OCT images 10 by the Fourier domain OCT imaging system 30.

[0039] Figure 6A shows a sinusoidal bulk vibration added to the oscillatory motion of two model retinal layers. In this example, the bulk motion is modeled by adding a phase exp(iφ), where φ represents a fraction of the wavelength (half-wavelength motion is 2π for double-pass reflection). Figure 6A shows how the phase angle of light reflected from a first model retinal layer (Layer 1) and a second model retinal layer (Layer 2) changes over time while the layers are subjected to common bulk motion as well as individual vibrations (i.e., the plots labeled "Layer 1 + Bulk" and "Layer 2 + Bulk" in Figure 6A). Figure 6A also shows how the phase angle of light reflected from either model retinal layer changes over time while the layer is subjected to bulk motion only (i.e., the plot labeled "Bulk" in Figure 6A). Figure 6B shows the effect of removing bulk motion on the change in the phase angle of light reflected from the two model retinal layers over time.

[0040] The MATLAB® code used to generate the plots of FIGS. 6A and 6B is as follows:

[0041] t=(0:1000);

[0042] A = 0.95;

[0043] B=0.8;

[0044] RetinaBulkMovement=exp(j * cos(t * pi / sqrt(700)));

[0045] Layer1Movement=exp(j * cos(t * pi / sqrt(30)));

[0046] Layer2Movement=exp(j * cos(t * pi / sqrt(50)));

[0047] Layer1totalMovement=Layer1Movement.*RetinaBulkMovement;

[0048] Layer2totalMovement=Layer2Movement. * RetinaBulkMovement;

[0049] %%%%%%%%%%%%%%%%%

[0050] Layer1MovementDeduced=angle(Layer1totalMovement. * conj(RetinaBulkMovement));

[0051] Layer2MovementDeduced=angle(Layer2totalMovement. * conj(RetinaBulkMovement));

[0052] subplot(211);

[0053] plot(angle(Layer1totalMovement));hold

[0054] on;plot(angle(Layer2totalMovement));plot(angle(RetinaBulkMovement));legend(’Layer1+Bulk’,’Layer2+Bulk’,’Bulk’);

[0055] xlim([300 400]);

[0056] subplot(212);

[0057] plot(Layer1MovementDeduced);hold on;plot(Layer2MovementDeduced);shg;hold off;

[0058] xlim([300 400]);

[0059] legend('Layer1','Layer2');

[0060] We now describe an example of a process by which the data processing device 100 can perform S20 of Figure 3, which is based on the velocity-based ORG technique described in Kari V. Vienola et al., "Velocity-based optoretinography for clinical applications," Optica 9, 1100-1108 (2022), the contents of which are incorporated herein by reference in their entirety.

[0061] In this exemplary embodiment, the data processing device 100 can first flatten the first B-scan 10-1 and the second B-scan 10-2 so that the IS / OS reflections and COST reflections are at substantially the same height for each A-scan in each B-scan. Next, the data processing device 100 can align the second B-scan 10-2 with the first B-scan 10-1. Next, the phase data of the two B-scans for each spatial coordinate pair (i.e., phase data at the same (x,z) coordinate in the B-scan) can be unwrapped in the temporal dimension to minimize the magnitude of the phase difference between the data sets of the two B-scans 10-1 and 10-2. After unwrapping and processing the phase components of the first 10-1 and second 10-2 B-scans to compensate for bulk motion of the retina during imaging, the difference between corresponding phase values ​​of the B-scans 10-1 and 10-2 is calculated for each spatial location specified by the corresponding coordinate pair, and the difference between the acquisition times of the B-scans is then used to calculate the instantaneous velocity of the spatial location. These instantaneous velocities can, as in this exemplary embodiment, be averaged across the lateral dimension (i.e., along the x-axis) to provide a measure of the instantaneous depth dependence of velocity along the z-axis, i.e., a one-dimensional velocity profile P of the type shown schematically in FIG. 1 , which shows the distribution of velocity v along the axis (z-axis) in the common portion of the retina covered by the first and second 10-1 and 10-2 B-scans. (Optionally) B-scan amplitude can also be averaged across the lateral dimension (x-axis) to provide a measure of the instantaneous depth dependence of backscatter. It should be noted that averaging of the lateral dimension is not necessary; instead, a two-dimensional velocity profile may be generated showing the distribution of velocity v along both the axial (z-axis) and lateral (x-axis) directions in the common portion of the retina covered by the first B-scan 10-1 and the second B-scan 10-2.

[0062] The tissue velocity v within the portion of the velocity profile P corresponding to the photoreceptor OS in the retina is calculated by the first position z of the velocity profile P corresponding to the IS-OS junction at a given location on the retina, as shown schematically in Figure 1. max Maximum speed at v maxfrom the second position z of the velocity profile P corresponding to the COST at that position on the retina min The minimum velocity at v min The frequency varies (usually monotonically) from

[0063] The velocity profile P (or in other exemplary embodiments, an index of tissue velocity v at a single common location along the axial direction z) calculated by processing the OCT images in each set of OCT images from the sequence of OCT images 10 according to process S20 of FIG. 2 is useful for various purposes. For example, the velocity profiles (or the aforementioned index of tissue velocity) may be concatenated, and a cumulative sum of the values ​​in the resulting concatenation may be calculated to generate ORG data indicative of the response of the retina to an applied optical stimulus, as described below. Additionally or alternatively, the location(s) of one or both of the boundaries of the OS or other layers L of the retina (e.g., the location of the rod outer segment tips (ROST) or the retina pigment epithelium (RPE)) that change in response to the applied optical stimulus may be determined by processing the velocity profile (or the aforementioned index of tissue velocity) using a velocity-based approach to segmentation described in co-pending European Patent Application No. 253989. For example, the velocity profile P may be calculated based on the relationship z > z, which corresponds to the change in rod length in response to the applied stimulus. min , there may be further maxima followed by further minima along the z-axis of FIG. 1. The data processing device 100, as in this exemplary embodiment, first calculates all combinations of candidate first and second positions (z i ,z j ) the first position z of the candidate in the velocity profile P i and the second position z of the candidate in the velocity profile P j For each combination of and, the combination (z i ,z j ) the first position z of the candidate in i Individual velocities v at i and the second position of the candidate z j Individual velocities v at j Difference Dij , and calculate the individual values ​​of the first position z in at least one of the calculated velocity profiles P. max and the second position z min Then, the data processing device 100 determines both the calculated difference D ij The first position z of one of the candidate combinations that maximizes the value of i and the second position z of the candidate j the first position z max and the second position z min Identify as.

[0064] Once the boundaries of the OS are identified as described above, their respective velocities can be extracted from the velocity profile P, and the difference between the extracted velocities provides the velocity of contraction / extension of the OS at the time B-scans 10-1 and 10-2 were acquired by the FD-OCT imaging system 30, which can be used to generate ORG data indicative of the response of the common retina to the applied stimulus.

[0065] Regardless of the purpose for which the velocity profile P or the aforementioned indices of tissue velocity are subsequently used, it is useful to measure how reliable they are and to remove unreliable velocity profiles (or unreliable indices of tissue velocity) from the processing pipeline to prevent them from adversely affecting the ORG data or other final results. The reliability of the velocity profile P or the aforementioned indices of tissue velocity is affected by the image quality of the first B-scan 10-1 and the second B-scan 10-2, as well as how closely the B-scans correlate with each other. The maximum calculated cross-correlation between two B-scans provides a good metric for how closely the B-scans are correlated, i.e., the reliability of the velocity profile P or the aforementioned indices of tissue velocity determined by the data processing device 100.

[0066] Referring again to Figure 3, in process S30, the data processing device compares the comparison value calculated in process S10 of Figure 3 with a threshold to determine whether the comparison value is greater than or equal to the threshold. In this exemplary embodiment, some pairs of B-scans result in comparison values ​​greater than or equal to the threshold, and some other pairs of B-scans result in comparison values ​​less than the threshold. It should be noted that although process S20 of Figure 3 is illustrated and described herein as being performed after process S10, the order in which these processes are performed may be reversed, or they may be performed simultaneously. Alternatively, process S20 may be performed simultaneously with process S30 of Figure 3.

[0067] The value of the threshold in process S30 may depend on the subsequent use of the velocity profile P or the aforementioned indicator of tissue velocity, and may be selected by trial and error so that retaining only velocity profiles P (or the aforementioned indicator of tissue velocity) associated with comparison values ​​equal to or exceeding the selected threshold reduces or prevents degradation of ORG data or other processing results that would otherwise occur.

[0068] For example, if one or more of the velocity profiles are used in the velocity-based segmentation technique described in co-pending European Patent Application No. 253989, the threshold may be set by comparing the segmentation result from the data processing device 100 (i.e., the boundary of layer L determined by the data processing device 100) with the layer segmentation resulting from inspection of the source B-scans by a user, while taking into account the maximum cross-correlation value calculated for the source B-scans. In this way, it may be determined that pairs of B-scans between which the maximum cross-correlation calculated is below a certain threshold tend to yield unreliable values ​​for the index of layer boundaries when the B-scans are processed by the data processing device 100 as described herein, while pairs of B-scans between which the maximum cross-correlation calculated is above that threshold tend to yield reliable values ​​for the index of layer boundaries.

[0069] For example, it can be seen that the B-scans shown in Figure 4A provide indications of layer boundaries that compare favorably with the results of manual segmentation of the OS performed by inspection of these B-scans, and that the B-scans shown in Figure 5A provide indications of layer boundaries that are significantly different from the results of manual segmentation of the OS performed by inspection of these B-scans. In this case, the threshold was set to the maximum value of the cross-correlation shown in Figures 4B and 5B (i.e., approximately 3·10 4 and 1·10 4 ), e.g., approximately 1.4 10 as shown by the horizontal black line in Figure 5B. 4 It may be appropriate to set this to a value of . B-scans resulting in a maximum cross-correlation value below this threshold may be deemed not to allow reliable segmentation to be performed.

[0070] If it is determined in process S30 of FIG. 3 that the comparison value is greater than or equal to the threshold value ("Yes" in S40 of FIG. 3), the data processing device 100 retains (stores) the velocity profile P, or more generally, an indication of the tissue velocity v at the position along the axial direction z calculated in process S20 of FIG. 3.

[0071] On the other hand, if it is determined in process S30 of FIG. 3 that the comparison value is not greater than or equal to the threshold value, the data processing device 100 discards the velocity profile P, or more generally, the index of tissue velocity v at a position along the axial direction z calculated in process S20 of FIG. 3, and therefore does not use it in subsequent processing operations. The data processing device 100 may, for example, erase or overwrite the calculated index (or velocity profile, as the case may be). If, as in the present exemplary embodiment, the velocity profile P is calculated in process S20 of FIG. 3, the data processing device 100 may, following a negative determination in S40 of FIG. 3, retain in a storage device (i.e., memory) for subsequent use a null velocity profile indicative of zero velocity v at all positions along the axial direction z, or in more general cases, an index indicative of zero tissue velocity at a position along the axial direction, instead of the calculated velocity profile P (or, more generally, the index of calculated tissue velocity at a position along the axial direction z).

[0072] Following process S50 of Figure 3, the data processing apparatus 100, as in this exemplary embodiment, may concatenate the retained velocity profile P, or more generally, the retained measure of tissue velocity calculated in process S20 of Figure 3, with any null velocity profile (or, more generally, a measure of zero tissue velocity) generated after the negative determination in S40 of Figure 3. The resulting concatenation indicates how the tissue velocity v changes over time.

[0073] FIG. 7 shows an example of a concatenation 300 of velocity profiles generated by the data processing device 100. Each velocity profile P extends along the y-axis (labeled "OCT depth layer") of FIG. 7, and velocity profiles calculated for pairs of adjacent B-scans in the sequence of B-scans 10 are arranged along the x-axis (labeled "time (ms)") of FIG. 7. FIG. 7 shows the change in optical path length ΔOPL (providing an indication of velocity v) with position in each velocity profile P derived from B-scans acquired after application of an optical stimulus at time t=200 ms. The change in optical path length ΔOPL varies with position (i.e., along the y-axis of FIG. 7) in each velocity profile from a minimum ΔOPL of approximately −300 (arbitrary units) to a maximum ΔOPL of approximately +250 (arbitrary units).

[0074] In this exemplary embodiment, the data processing device 100 also integrates each portion of the concatenation of velocity profiles P having the same position along the axial direction z in the concatenation of velocity profiles to generate ORG data indicative of the change in optical path length at a position along the axial direction z over time. More generally, the data processing device 100 can integrate (i.e., calculate a cumulative sum or total) the measure of tissue velocity at the position retained in process S50 of FIG. 3 and any surrogate measure of zero velocity generated after a negative determination in S40 of FIG. 3 to generate ORG data indicative of the change in optical path length at a position along the axial direction z over time. The data processing device 100 can process each velocity profile P in the set of concatenated velocity profiles by selecting portions (segments or data elements) of the velocity profile P located at predetermined positions along the axial direction z of the velocity profile P and then integrating the selected (commonly located) portions by calculating a cumulative sum (or total) of velocity values ​​in these portions that indicates how the optical path length of the optical path of the OCT sample beam ending at a position in the retina corresponding to the predetermined position in the velocity profile P changes over time.

[0075] Figure 8 shows a plot of the change in optical path length ΔOPL over time calculated by the data processor 100 for selected positions within the velocity profile of a concatenation of velocity profiles. The plot in Figure 8 shows how ΔOPL at a common position varies from one B-scan to the next in a sequence of B-scans 10.

[0076] Figure 9 shows a plot of the cumulative change in optical path length over time, known as the ORG signal (or ORG data). The ORG signal is obtained by calculating the cumulative sum of the optical path length changes for selected locations within a concatenation of velocity profiles. As shown in Figure 9, the ORG signal becomes negative immediately after the stimulus is applied, then increases to become positive, and finally levels off at a value of approximately 250 nm in this example. Figure 9 also shows the ORG signal having several plateaus caused by the substitution of velocity profiles calculated using B-scans that do not yield sufficiently high maximum cross-correlation values ​​with velocity profiles exhibiting zero velocity (i.e., null velocity profiles), as discussed above.

[0077] Thus, the ORG data may include one or more sets of equal consecutive values, and the data processing device 100 may be arranged to smooth the generated ORG data by replacing one or more values ​​in at least one of the one or more sets of equal consecutive values ​​with one or more estimated values ​​calculated based on neighboring values ​​adjacent to the set of equal consecutive values ​​in the generated ORG data. The data processing device 100 may perform this smoothing, for example, by using a moving median or moving average, calculating the median or average value of a specified number of points on either side of the missing data point(s) and then assigning this to the missing point(s). Alternatively, the data processing device 100 may, for example, detect each set of equal consecutive values ​​in the ORG data and replace the value of each detected set with a corresponding estimated value obtained by interpolating (for example, linearly) between a first value in the ORG data adjacent to the first value of the detected set of equal consecutive values ​​and a second value in the ORG data adjacent to the last value of the detected set of equal consecutive values.

[0078] The results of smoothing the ORG signal using a moving average are shown in Figure 10. Before smoothing is applied, the majority of the data represents zero OPL change, as shown in Figure 9, and therefore generally under-represents the actual values ​​that would be observed. Using the methods described herein, the equivalent values ​​are replaced with estimated values ​​(shown as circles in Figure 10) that more accurately represent the change in OPL.

[0079] Optional further processing that may be performed by the data processing device 100 of this exemplary embodiment is summarized in FIG.

[0080] In process S70 of FIG. 11, the data processing device 100 generates a concatenation 300 of the indices of the generated tissue velocities v such that the concatenation 300 indicates how the tissue velocity v at a position along the axial direction z changes over time.

[0081] In process S80 of FIG. 11, the data processing device 100 integrates the concatenation 300 to generate data indicative of the change in optical path length over time with position along the axial direction z.

[0082] If in process S40 of FIG. 3 (or process S130 of FIG. 12 in the second exemplary embodiment described below) it is determined that the comparison value is not greater than or equal to the threshold value, the individual indicator of tissue velocity v at the position along the axial direction z may be set to indicate zero velocity, as in this exemplary embodiment, and the generated data may include one or more sets of equal consecutive values, as described above.

[0083] In process S90 of FIG. 11, the data processing device 100 smooths the generated data by replacing one or more values ​​in one set of one or more sets of equal consecutive values ​​with one or more estimated values ​​calculated based on adjacent values ​​adjacent to the set of equal consecutive values ​​in the generated data.

[0084] Although the data processing device 100 processes pairs of adjacent B-scans in the sequence of B-scans 10 sequentially (i.e., one adjacent pair followed by another in the sequence) to generate the velocity profile P, or more generally, an indication of tissue velocity at position along the axial direction z, the data processing device 100 may instead generate these by processing pairs of adjacent B-scans in parallel, thereby significantly speeding up processing. Also, although pairs of adjacent B-scans in the sequence of B-scans 10 (i.e., consecutive B-scans in the sequence) are processed, the data processing device 100 may instead process pairs of B-scans in a sequence in which the B-scans of each pair are separated from each other by one or more intervening B-scans, for example.

[0085] Furthermore, although the data processing apparatus 100 of this exemplary embodiment is described as processing pairs of B-scans, it may more generally be configured to process sets of more than two B-scans in a sequence of B-scans 10, where the B-scans may be consecutive B-scans in the sequence separated from each other by one or more intervening B-scans that do not form part of a set, or individual B-scans in the sequence. Each set (e.g., five) of B-scans may be selected by sliding a window selection function along the sequence of B-scans 10, where in each execution of the process of Figure 3 the window selection function may, for example, select all B-scans in a windowed portion of the sequence of B-scans 10, or every other B-scan in the windowed portion, and the selected sets may have one or more B-scans in common.

[0086] In a modification of process S10 of FIG. 3, the data processor 100 calculates a comparison value for one set of a plurality of different sets of consecutive B-scans in the sequence of B-scans 10. The comparison value can take any of the different forms described above. The data processor 100 then processes the phase components of at least some of the B-scans in the set according to a modification of process S20 of FIG. 3 to generate a velocity profile P indicative of a distribution, or more generally, an indication of tissue velocity v at position along the axial direction z. Here, the data processor 100 can first flatten the set of B-scans so that the IS / OS and COST reflections are at substantially the same height for each A-scan in each B-scan. The data processor 100 can then align the B-scans with each other. The phase data cube θ(x,z,t) is then calculated by |θ(x p, z q, t r )-θ(x p, z q, t r-1 )|, θ(x p, z q, t r ) where t r and t r-1 represents a continuous phase B-scan. This process involves calculating a pair of spatial coordinates (x p, z q) for each scan. The rate of phase change is calculated for each coordinate pair by performing a least-squares linear fit to t, giving Δθ / Δt(x,z) rad / s. From this, the instantaneous velocity at each spatial location can be calculated as Δz / Δt(x,z)=Δθ / Δt(x,z)·λ / 4πn', where λ is the wavelength of the OCT light being used and n is the nominal refractive index of the eye 20. These instantaneous velocities can be averaged in the lateral dimension (i.e., along the x-axis) to give a measure of the instantaneous depth dependence of velocity along the z-axis, i.e., a one-dimensional velocity profile P of the type shown schematically in Figure 1, which shows the distribution of velocity v along an axis (z-axis) in the common portion of the retina covered by the repeated B-scans. The B-scan amplitude can also be averaged in the lateral dimension (x-axis) (if desired) to obtain a measure of the instantaneous depth dependence of backscatter. Note that averaging in the lateral dimension is not required; instead, a two-dimensional velocity profile may be generated that shows the distribution of velocity v along both the axial (z-axis) and lateral (x-axis) directions in the common portion of the retina covered by the repeated B-scans. The process then proceeds according to S30 or S60 of FIG. 3.

[0087] Second Exemplary Embodiment In the first exemplary embodiment described above, for each set of multiple B-scans in the sequence of B-scans 10, a separate measure of tissue velocity at one or more locations along the axial direction z is calculated, and each calculated measure is either retained or discarded depending on how the associated comparison value compares to a threshold. However, in this exemplary embodiment, the calculation of the measure of tissue velocity v is performed selectively (i.e., not necessarily for all sets of B-scans) depending on how the comparison value calculated for the set of OCT images compares to a threshold, as will be described in more detail below with reference to FIG. 12 .

[0088] FIG. 12 is a flow diagram illustrating how the data processing device 100 of the exemplary embodiment processes each set of B-scans, e.g., a pair of B-scans including a first B-scan 10-1 and a second B-scan 10-2, from a plurality of sets of B-scans in the sequence of B-scans 10 of a common portion of the retina shown in FIG. 1 to generate an individual measure of tissue velocity v at a position along the axial direction z in the B-scan 10.

[0089] Processes S110 to S130 in FIG. 12 are the same as processes S10, S30, and S40 in FIG. 3, respectively, which were described in detail above.

[0090] If the data processing device 100 determines that the comparison value is greater than or equal to the threshold value ("Yes" in S130 of FIG. 12), then in process S140 of FIG. 12, the data processing device 100 uses the phase value of the OCT data element at that position along the axial direction z in the A-scan of the first B-scan 10-1 in the set and the phase value at the same position along the axial direction z in the corresponding A-scan of the second B-scan 10-2 in the set in calculating an individual index of tissue velocity v at that position along the axial direction z.

[0091] If the data processing device 100 determines that the comparison value is not greater than or equal to the threshold value ("No" at S130 of Fig. 12), then in process S150 of Fig. 12, the data processing device 100 excludes the phase value at that position along the axial direction z in the A-scan of the first B-scan 10-1 in the set and the phase value at the same position along the axial direction z in the corresponding A-scan of the second B-scan 10-2 in the set in calculating the individual index of tissue velocity v at that position along the axial direction z. Following a negative determination at S130 of Fig. 12, the data processing device 100 may generate a null velocity profile indicative of zero velocity v at all positions along the axial direction z, or in the more general case, an index indicative of zero tissue velocity at positions along the axial direction.

[0092] As in the first exemplary embodiment, the data processing apparatus 100 can then concatenate the calculated velocity profile P, or more generally, the measure of tissue velocity calculated in process S140 of Figure 12, with any null velocity profile (or in the more general case, a measure of zero tissue velocity) generated after the negative determination in S130 of Figure 12. The resulting concatenation indicates how the tissue velocity v changes over time.

[0093] In this exemplary embodiment, the data processing device 100 also generates ORG data indicative of optical path length change over time at a position along the axial z line by integrating respective portions of the concatenation of velocity profiles P within the concatenation of velocity profiles that have the same position along the axial z line (or, more generally, by integrating the tissue velocity indicator calculated in process S140 of FIG. 12 and any zero velocity indicator that may have been generated after a negative determination in process S130 of FIG. 12). The data processing device 100 can process each velocity profile P within the set of concatenated velocity profiles by selecting portions (segments or data elements) of the velocity profile P that are located at predetermined positions along the axial z line of the velocity profile P, and then integrating the selected (commonly located) portions by calculating a cumulative sum (or total) of the velocity values ​​in these portions that indicates how the optical path length of the optical path of the OCT sample beam that ends at a position in the retina corresponding to the predetermined position in the velocity profile P changes over time.

[0094] Thus, the ORG data may include one or more sets of equal consecutive values, and the data processing device 100 may be arranged to smooth the generated ORG data by replacing one or more values ​​in at least one of the one or more sets of equal consecutive values ​​with one or more estimated values ​​calculated based on neighboring values ​​adjacent to the set of equal consecutive values ​​in the generated ORG data. The data processing device 100 may perform this smoothing, for example, by using a moving median or moving average, calculating the median or average value of a specified number of points on either side of the missing data point(s) and then assigning this to the missing point(s). Alternatively, the data processing device 100 may, for example, detect each set of equal consecutive values ​​in the ORG data and replace the value of each detected set with a corresponding estimated value obtained by interpolating (for example, linearly) between a first value in the ORG data adjacent to the first value of the detected set of equal consecutive values ​​and a second value in the ORG data adjacent to the last value of the detected set of equal consecutive values.

[0095] It will be understood that at least some of the modifications of the first embodiment described above may be made to this embodiment.

[0096] In the foregoing description, exemplary aspects have been described with reference to several exemplary embodiments. Accordingly, the present specification should be considered illustrative rather than restrictive. Similarly, the diagrams shown in the drawings that highlight the functionality and advantages of exemplary embodiments are presented for illustrative purposes only. The architecture of the exemplary embodiments is sufficiently flexible and configurable so that it can be utilized in ways other than those shown in the accompanying figures.

[0097] Some aspects of the examples presented herein, such as the processing methods described with reference to Figures 3, 11, and 12, may be provided as computer programs or software, e.g., one or more programs having instructions or instruction sequences contained or stored on an article of manufacture, such as a machine-accessible or machine-readable medium, instruction store, or computer-readable storage device, which, in one example embodiment, may be non-transitory. The programs or instructions on the non-transitory machine-accessible medium, machine-readable medium, instruction store, or computer-readable storage device may be used to program a computer system or other electronic device. Machine or computer-readable medium, instruction store, and storage device may include, but is not limited to, floppy diskettes, optical disks, and optical-magnetic disks, or other types of media / machine-readable medium / instruction store / storage device suitable for storing or transmitting electronic instructions. The techniques described herein are not limited to any particular software configuration. They may find applicability in any computing or processing environment. As used herein, the terms "computer-readable," "machine-accessible medium," "machine-readable medium," "instruction store," and "computer-readable storage device" are intended to include any medium capable of storing, encoding, or transmitting instructions or sequences of instructions for execution by a machine, computer, or computer processor, which cause the machine / computer / computer processor to perform any one of the methods described herein. Furthermore, it is common in the art to speak of software, in one form or another (e.g., program, procedure, process, application, module, unit, logic, etc.), as performing an action or causing a result. Such expressions are merely a shorthand way of stating that execution of the software by a processing system causes the processor to perform an action and produce a result.

[0098] Some or all of the functionality of OCT data processing hardware 130 may also be implemented by the preparation of application specific integrated circuits, field programmable gate arrays, or by interconnecting an appropriate network of conventional component circuits.

[0099] The computer program product may be provided in the form of one or more storage media, instruction store(s), or storage device(s) having stored thereon instructions that can be used to control or cause a computer or computer processor to perform any of the procedures of the example embodiments described herein. The storage media / instruction store / storage device may include, by way of example and not limitation, optical disks, ROM, RAM, EPROM, EEPROM, DRAM, VRAM, flash memory, flash cards, magnetic cards, optical cards, nanosystems, molecular memory integrated circuits, RAID, remote data storage devices / archives / warehousing, and / or any other type of device suitable for storing instructions and / or data.

[0100] Some implementations include software stored on one or more computer-readable media, instruction store(s), or storage device(s) for controlling both the system hardware and enabling the system or microprocessor to utilize the results of the exemplary embodiments described herein to interact with a human user or other mechanism. Such software can include, but is not limited to, device drivers, operating systems, and user applications. Finally, such computer-readable media or storage device(s) further include software for performing exemplary aspects of the present invention, as described above.

[0101] The programming and / or software of the system includes software modules for performing the procedures described herein. In some exemplary embodiments herein, the modules include software, while in other exemplary embodiments herein, the modules include hardware or a combination of hardware and software.

[0102] While various exemplary embodiments of the present invention have been described above, it should be understood that they are presented by way of example, not limitation. Various changes in form and detail will be apparent to those skilled in the art. Therefore, the present invention should not be limited by any of the above-described exemplary embodiments, but should be defined only in accordance with the following claims and their equivalents.

[0103] While this specification contains details of many specific embodiments, these should not be construed as limitations on the scope of any invention or what may be claimed, but rather as descriptions of features unique to the particular embodiments described herein. Certain features described herein in the context of separate embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented separately in multiple embodiments or in any suitable subcombination. Furthermore, even if features are described above as acting in a particular combination and initially claimed as such, one or more features from a claimed combination may, in some cases, be deleted from the combination, and the claimed combination may also be directed to a subcombination or variations of the subcombination.

[0104] In certain circumstances, multitasking and parallel processing may be advantageous. Furthermore, the separation of various components in the above-described embodiments should not be understood as requiring such separation in all embodiments, and it should be understood that the described program components and systems may generally be integrated together in a single software product or packaged in multiple software products.

[0105] Having now described several exemplary embodiments and implementations, it should be apparent that the foregoing is presented by way of example, not limitation. In particular, while many of the examples presented herein involve particular combinations of device or software elements, those elements may be combined in other ways to achieve the same purpose. Operations, elements, and features discussed only in connection with one embodiment are not intended to exclude them from a similar role in other embodiments or implementations.

Claims

1. 1. A computer-implemented method for processing each set of optical coherence tomography (OCT) images, including a first OCT image (10-1) and a second OCT image (10-2), of a plurality of different sets of OCT images in a sequence of OCT images (10) of a common portion of a retina of an eye (20) acquired by a Fourier domain OCT imaging system (30), to generate a separate measure of tissue velocity (v) at a position along an axial direction (z) in the OCT images (10), the method comprising: a value of an image quality metric calculated based on at least one of the first OCT image (10-1) in the set and the second OCT image (10-2) in the set; and and an image similarity index value that provides a measure of similarity between the images, the image similarity index value being calculated based on the first OCT image (10-1) in the set and the second OCT image (10-2) in the set (S10). calculating (S20) the individual measure of tissue velocity (v) at the position along the axis (z) using a phase value at the position along the axis (z) in an A-scan of the first OCT image (10-1) in the set and a phase value at the position along the axis (z) in a corresponding A-scan of the second OCT image (10-2) in the set; comparing the comparison value with a threshold to determine whether the comparison value is greater than or equal to the threshold (S30); if the comparison value is determined to be greater than or equal to the threshold, retaining (S50) the individual measure of tissue velocity (v) at the position along the axial direction (z); If it is determined that the comparison value is not greater than or equal to the threshold, discarding (S60) the individual measure of tissue velocity (v) at the position along the axial direction (z).

2. The computer-implemented method of claim 1, wherein the comparison value is the maximum value of the cross-correlation calculated between the first OCT image (10-1) and the second OCT image (10-2).

3. generating (S70) a concatenation (300) of the generated indices of tissue velocity (v) such that the concatenation (300) indicates how tissue velocity (v) at said position along said axial direction (z) changes over time; Integrating (S80) the concatenation (300) to generate data indicative of optical path length change over time at the position along the axis (z); The computer-implemented method of claim 1 , further comprising:

4. If it is determined that the comparison value is not greater than or equal to the threshold value, setting the individual measure of tissue velocity (v) at the position along the axial direction (z) to a zero velocity, the generated data includes one or more sets of equal consecutive values; the computer-implemented method further comprising smoothing (S90) the generated data by replacing one or more values ​​in the set of one or more equal consecutive values ​​with one or more estimated values ​​calculated based on neighboring values ​​adjacent to the set of equal consecutive values ​​in the generated data. The computer-implemented method of claim 3 .

5. The computer-implemented method of claim 1, further comprising, prior to the calculation of the individual indices of tissue velocity (v) at the positions along the axial direction (z), processing the phase values ​​at the positions along the axial direction (z) in the A-scan of the first OCT image (10-1) in the set and the phase values ​​at the positions along the axial direction (z) in the corresponding A-scan of the second OCT image (10-2) in the set to compensate for bulk motion of the common portion of the retina during acquisition of the sequence of OCT images (10) by the Fourier-domain OCT imaging system (30).

6. A computer program (245) comprising computer-readable instructions that, when executed by a processor (220), cause the processor (220) to perform the method of claim 1.

7. A data processing device (100) arranged to process each set of optical coherence tomography OCT images comprising a first OCT image (10-1) and a second OCT image (10-2) of a set of multiple different OCT images in a sequence of OCT images (10) of a common portion of the retina of an eye (20) acquired by a Fourier domain OCT imaging system (30) to generate an individual measure of tissue velocity (v) at a position along an axial direction (z) in said OCT images (10), wherein for each set of said OCT images: a value of an image quality metric calculated based on at least one of the first OCT image (10-1) in the set and the second OCT image (10-2) in the set; and an image similarity index value that provides a measure of similarity between the images, calculated based on the first OCT image (10-1) in the set and the second OCT image (10-2) in the set; calculating the individual measure of tissue velocity (v) at the position along the axial direction (z) using a phase value at the position along the axial direction (z) in an A-scan of the first OCT image (10-1) in the set and a phase value at the position along the axial direction (z) in a corresponding A-scan of the second OCT image (10-2) in the set; comparing the comparison value to a threshold to determine whether the comparison value is greater than or equal to the threshold; if the comparison value is determined to be greater than or equal to the threshold, retaining the individual measure of tissue velocity (v) at the location along the axial direction (z); and discarding the individual measure of tissue velocity (v) at said position along said axial direction (z) if it is determined that the comparison value is not greater than or equal to said threshold value.

8. A data processing device (100) as described in claim 7, wherein the comparison value is the maximum value of the cross-correlation calculated between the first OCT image (10-1) and the second OCT image (10-2).

9. The data processing device (100) generating the concatenation (300) of the generated indices of tissue velocity (v) such that the concatenation (300) indicates how tissue velocity (v) at the position along the axial direction (z) changes over time; and further configured to integrate the coupling (300) to generate data indicative of optical path length change over time at the position along the axial direction (z). A data processing device (100) according to claim 7.

10. The method of claim 1, wherein if it is determined that the comparison value is not greater than or equal to the threshold value, the data processing device (100) sets the individual measure of tissue velocity (v) at the position along the axial direction (z) to a zero velocity; the generated data includes one or more sets of equal consecutive values; the data processing device (100) is further arranged to smooth the generated data by replacing one or more values ​​in the set of one or more equal consecutive values ​​with one or more estimated values ​​calculated based on adjacent values ​​adjacent to the set of equal consecutive values ​​in the generated data. A data processing device (100) according to claim 9.

11. A data processing device (100) as described in claim 7, further arranged to process, prior to the calculation of the individual index of tissue velocity (v) at the position along the axial direction (z), the phase value at the position along the axial direction (z) in the A-scan of the first OCT image (10-1) in the set and the phase value at the position along the axial direction (z) in the corresponding A-scan of the second OCT image (10-2) in the set to compensate for bulk motion of the common portion of the retina during acquisition of the sequence of OCT images (10) by the Fourier domain OCT imaging system (30).

12. A Fourier domain optical coherence tomography imaging system (30) comprising a data processing device (100) as described in claim 7.

Images