Signal processing device, signal processing method, and signal processing program

The method improves bulk phase error estimation in OCT by calculating phase gradients and integrating along multiple paths with weighted averages, addressing accuracy issues in existing methods and enhancing image quality.

JP7587847B2Active Publication Date: 2024-11-21UNIV OF TSUKUBA
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
JP2021556177
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2019-11-14
Filing Date
2020-11-13
Publication Date
2024-11-21
Estimated Expiration
2040-11-13

AI Technical Summary

Technical Problem

Existing methods for estimating bulk phase errors in optical coherence tomography (OCT) suffer from accuracy issues due to error accumulation during phase integration, leading to decreased estimation accuracy and potential failure in phase error estimation.

Method used

A method for calculating phase gradients in a plane intersecting the light irradiation direction of OCT signals, integrating phase errors along multiple paths to determine bulk phase errors, and using weighted averages based on path lengths to improve estimation accuracy.

Benefits of technology

The proposed method enhances the accuracy of bulk phase error estimation by reducing random error components and maintaining consistent estimation across different paths, resulting in improved OCT image quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007587847000008
    Figure 0007587847000008
  • Figure 0007587847000009
    Figure 0007587847000009
  • Figure 0007587847000010
    Figure 0007587847000010
Patent Text Reader

Abstract

According to the present invention: a phase gradient calculating unit calculates a phase gradient on a plane intersecting the light radiation direction of an optical coherence tomography signal representing the state of a specimen, for each sample point arranged on the plane; and for each of a plurality of routes from a starting point, which is a sample point of which the bulk phase error has been determined, to an end point, which is a sample point of which the bulk phase error has not been determined, a bulk phase error calculating unit integrates the phase gradient at each sample point along each route to calculate a route-specific phase error at the end point, and defines the bulk phase error at the end point by combining the route-specific phase errors between the plurality of routes.
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a signal processing device, a signal processing method, and a signal processing program. This application claims priority to Japanese Patent Application No. 2019-206506, filed on November 14, 2019, the contents of which are incorporated herein by reference. [Background technology]

[0002] Optical coherence tomography (OCT) is a technology that uses the coherence of light to obtain cross-sectional images of a sample (mainly a living body). OCT makes it possible to obtain images that show not only the surface of a sample but also its internal structure with high spatial resolution. The development of OCT technology has made it possible to take images at higher speeds and with higher resolution, and its functions are being improved. For example, Non-Patent Document 1 describes digital defocus in Fourier Domain OCT (FD).

[0003] Generally, bulk phase errors (BPE) occur in images obtained by OCT due to the movement of the sample during measurement and environmental fluctuations such as the atmosphere and temperature through which light propagates. Bulk phase errors refer to phase fluctuations that occur between multiple depth-direction signal acquisitions (A-scans). Therefore, bulk phase errors can be a factor that hinders the sophistication of functions. The method described in Non-Patent Document 1 is merely an approximate method that uses a single tomographic image (B-scan) whose measurement time is so short that bulk phase errors can be ignored. Non-Patent Document 2 describes a method of performing digital refocusing by performing phase correction on a three-dimensional OCT image (three-dimensional OCT volume) using a phase contrast method. The phase contrast method is a method of calculating differential phase fields in two directions ((x, y) directions) on the front surface of the three-dimensional OCT volume and estimating bulk phase errors by accumulating the differential phase fields. [Prior art documents] [Non-patent literature]

[0004] [Non-Patent Document 1] Yasuno et al., “Non-iterative numerical method for laterally superresolving Fourier domain optical coherence tomography,” OPTICS EXPRESS Vol.14, No. 3, p.1006-1020, February 6, 2006 [Non-Patent Document 2] Shemonski et al., “Three-dimensional motion correction using speckle and phase for in vivo computed optical interferometric tomography,” BIOMEDICAL OPTICS EXPRESS Vol.5, No.12, p.4131-4143, December 1, 2014. Summary of the Invention [Problem to be solved by the invention]

[0005] However, in the phase difference method, errors accumulate in the process of integrating the differential phase field, which tends to reduce the accuracy of the final estimated bulk phase error. Non-Patent Document 2 attempts to reduce the error by performing a smoothing process on the estimated bulk phase error. This method lacks theoretical grounds for reducing the error, and can instead be a factor in reducing the estimation accuracy or a factor in failing to estimate the bulk phase error.

[0006] The present invention has been made in consideration of the above points, and an object of the present invention is to provide a signal processing device, a signal processing method, and a signal processing program that can improve the estimation accuracy of the bulk phase error. [Means for solving the problem]

[0007] (1) The present invention has been made to solve the above-mentioned problems, and one aspect of the present invention is a signal processing device that includes: a phase gradient calculation unit that calculates a phase gradient in a plane that intersects with the light irradiation direction of an optical coherence tomography signal representing a state of a sample, for each sample point arranged on the plane; and a bulk phase error calculation unit that integrates the phase gradient for each sample point along each of a plurality of paths from a starting point that is a sample point for which a bulk phase error has been determined to an end point that is a sample point for which a bulk phase error has not been determined, to calculate a path-specific phase error at the end point, and combines the path-specific phase errors among the plurality of paths to determine the bulk phase error at the end point.

[0008] (2) Another aspect of the present invention is the signal processing device of (1), wherein the bulk phase error calculation section calculates an average value of the path-specific phase errors between paths as the bulk phase error.

[0009] (3) Another aspect of the present invention is a signal processing device as described in (2), wherein the bulk phase error calculation unit may calculate, as the average value, a weighted average value of the phase errors for each path using a weighting coefficient based on the path length of each of the multiple paths.

[0010] (4) Another aspect of the present invention is a signal processing device according to any one of (1) to (3), wherein each of the multiple paths may be included within a predetermined range from the end point.

[0011] (5) Another aspect of the present invention is a signal processing device of any of (1) to (4), wherein each of the multiple paths includes an nth fragment (N is an integer of 2 or more) that is a section from the nth (n is an integer greater than or equal to 1 and less than or equal to N-1) sample point to the n+1th sample point, and the n+1th sample point may be a sample point within a range equal to or less than the spatial resolution of an optical system that acquired the optical coherence tomography signal from the nth sample point.

[0012] (6) Another aspect of the present invention is a signal processing device according to any one of (1) to (5), which may pre-set a plurality of route patterns indicating the plurality of routes in correspondence with the positional relationship between the starting point and the end point, and select a plurality of routes indicated by the route pattern that satisfies the positional relationship, with a target sample point for which a bulk phase error is to be calculated as an end point.

[0013] (7) Another aspect of the present invention is the signal processing device of (6), wherein the bulk phase error calculation unit predetermines an order of calculating bulk phase errors such that one sample point on the plane is used as a reference point and sample points closer to the reference point are given priority, and when the mth sample point (m is an integer greater than or equal to 1) in the order is used as the target sample point, the order may start from the (m-1)th sample point and select the multiple routes indicated by a route pattern that satisfies the positional relationship.

[0014] (8) Another aspect of the present invention is a signal processing device according to any one of (1) to (7), which may include a signal correction unit that subtracts a bulk phase error at the end point from the phase of a signal value of the optical coherence tomography signal at a sample point arranged in a three-dimensional space that has a common position projected onto the plane as the end point.

[0015] (9) Another aspect of the present invention is a signal processing method in a signal processing device, comprising: a phase gradient calculation process for calculating a phase gradient in a plane intersecting a light irradiation direction of an optical coherence tomography signal representing a state of a sample for each sample point arranged on the plane; and a bulk phase error calculation process for integrating the phase gradient for each sample point along each of a plurality of paths from a starting point which is a sample point for which a bulk phase error has been determined to an end point which is a sample point for which a bulk phase error has not been determined to calculate a path-specific phase error at the end point, and combining the path-specific phase errors among the plurality of paths to determine a bulk phase error at the end point.

[0016] (10) Another aspect of the present invention is a signal processing program for executing a phase gradient calculation procedure for calculating a phase gradient in a plane intersecting a light irradiation direction of an optical coherence tomography signal representing a state of a sample for each sample point arranged on the plane, and a bulk phase error procedure for integrating the phase gradient for each sample point along each of a plurality of paths from a starting point which is a sample point for which a bulk phase error has been determined to an end point which is a sample point for which a bulk phase error has not been determined to calculate a path-specific phase error at the end point, and synthesizing the path-specific phase errors among the plurality of paths to determine a bulk phase error at the end point. Effect of the Invention

[0017] According to the present invention, the estimation accuracy of the bulk phase error can be improved. [Brief description of the drawings]

[0018] [Figure 1] 1 is a configuration diagram illustrating an example of an optical coherence tomography device according to an embodiment of the present invention. [Diagram 2] 1 is a block diagram showing an example of a functional configuration of a signal processing device according to an embodiment of the present invention; [Diagram 3] FIG. 2 is an explanatory diagram showing an overview of a calculation process for a bulk phase error according to the present embodiment. [Figure 4] FIG. 4 is an explanatory diagram showing an example of an integral path according to the embodiment; [Figure 5A] FIG. 4 is a diagram showing a first example of a route pattern according to the embodiment. [Figure 5B] FIG. 11 is a diagram showing a second example of a route pattern according to the embodiment. [Figure 5C] FIG. 11 is a diagram showing a third example of a route pattern according to the embodiment. [Figure 5D] FIG. 11 is a diagram showing a fourth example of a route pattern according to the embodiment. [Figure 5E] FIG. 11 is a diagram showing a fifth example of a route pattern according to the embodiment. [Figure 5F] FIG. 13 is a diagram showing a sixth example of a route pattern according to the embodiment. [Figure 5G] FIG. 13 is a diagram showing a seventh example of a route pattern according to the embodiment. [Figure 5H] FIG. 13 is a diagram showing an eighth example of a route pattern according to the embodiment. [Figure 5I] FIG. 13 is a diagram showing a ninth example of a route pattern according to the embodiment. [Figure 6A] FIG. 4 is a diagram showing a first example of a set of a start point and an end point of a route pattern according to the embodiment. [Figure 6B] FIG. 11 is a diagram showing a second example of a set of a start point and an end point of a route pattern according to the embodiment. [Figure 6C] FIG. 11 is a diagram showing a third example of a set of a start point and an end point of a route pattern according to the embodiment. [Figure 6D] FIG. 11 is a diagram showing a fourth example of a set of a start point and an end point of a route pattern according to the embodiment. [Figure 7A] FIG. 11 is a diagram showing a fifth example of a set of start points and end points of a route pattern according to the embodiment. [Figure 7B] FIG. 13 is a diagram showing a sixth example of a set of start points and end points of a route pattern according to the embodiment. [Figure 7C] FIG. 13 is a diagram showing a seventh example of a set of a start point and an end point of a route pattern according to the embodiment. [Figure 8A] FIG. 11 is a diagram showing an example of an initial state of iterative calculation of bulk phase errors according to the present embodiment; [Figure 8B] FIG. 11 is a diagram showing an example of a first iterative calculation of bulk phase errors according to the present embodiment. [Figure 8C] FIG. 11 is a diagram showing a second example of the iterative calculation of bulk phase errors according to the present embodiment. [Figure 9A] FIG. 11 is a diagram showing an example of a third iterative calculation of bulk phase errors according to the present embodiment. [Figure 9B] FIG. 11 is a diagram showing another example of an iterative calculation of bulk phase errors according to the present embodiment. [Figure 10A] FIG. 1 is a diagram showing an example of an OCT image. [Figure 10B] FIG. 1 shows an example of bulk phase error based on the phase difference method. [Figure 10C] FIG. 13 is a diagram showing an example of bulk phase error according to the present embodiment. [Figure 10D] FIG. 1 is a diagram showing an example of a spectrum of an OCT image. [Figure 10E] FIG. 1 shows an example spectrum of an OCT image corrected using bulk phase error based on the phase contrast method. [Figure 10F] FIG. 13 shows an example spectrum of an OCT image corrected using bulk phase error according to the present embodiment. [Figure 11A] FIG. 13 is a diagram illustrating an example of an average value in the x direction of a spatial spectrum. [Figure 11B] FIG. 13 is a diagram illustrating an example of the average value in the y direction of the spatial spectrum. [Figure 11C] FIG. 13 is a diagram illustrating an example of the standard deviation in the x direction of a spatial spectrum. [Figure 11D] FIG. 13 is a diagram illustrating an example of the standard deviation in the y direction of a spatial spectrum. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0019] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing an example of the configuration of an optical coherence tomography device 1 according to this embodiment. The optical coherence tomography device 1 constitutes an observation system for observing the state of a sample using OCT. The optical coherence tomography device 1 is a device that irradiates light onto a sample Sm, acquires interference light generated by the interference between the light reflected from the sample Sm and the reference light reflected by a reference mirror 40 (described later), and generates an image showing the surface and internal condition of the sample Sm from the acquired interference light. The object to be observed as the sample Sm may be, for example, a living human or animal body, or a non-living body. A living body may be the fundus of the eye, a blood vessel, a tooth, subcutaneous tissue, etc. A non-living body may be an artificial structure such as an electronic part or a machine part, a natural structure such as a stone or a mineral, or a substance without a specific shape.

[0020] The optical coherence tomography device 1 includes a light source 10, a beam splitter 20, collimators 30a, 30b, 50a, and 50b, a reference mirror 40, galvanometer mirrors 60a and 60b, a spectrometer 70, and a signal processing device 100. Among these components, the beam splitter 20, the collimators 30a, 30b, 50a, and 50b, the reference mirror 40, the galvanometer mirrors 60a and 60b, and the spectrometer 70 constitute an optical system called an interferometer. The interferometer illustrated in FIG. 1 is a Michelson interferometer equipped with an optical fiber F. More specifically, the light source 10 and the spectrometer 70, the collimators 30a and 50a are each connected to the beam splitter using the optical fiber F. The optical fiber F has a transmission band including the wavelength band of the light irradiated from the light source 10.

[0021] The optical coherence tomography 1 is, for example, a Fourier-Domain OCT (FD-OCT). The FD-OCT may be, for example, any of a Spectral-Domain OCT (SD-OCT) and a Swept-Source OCT (SD-OCT), but is not limited thereto. The optical coherence tomography 1 may be, for example, a Time-Domain OCT (TD-OCT). When the optical coherence tomography 1 is a TD-OCT, the position of the reference mirror 40 is not fixed, and a driving mechanism for varying the optical path length from the light source 10 to the reference mirror 40 is provided.

[0022] The light source 10 irradiates, for example, a probe light having a near-infrared wavelength (for example, 800 to 1000 nm). The light source 10 is, for example, a wavelength swept light source such as an ultrashort pulse laser or an SLD (Superluminescent Diode). The light irradiated from the light source 10 is guided inside the optical fiber F and enters the beam splitter 20.

[0023] The beam splitter 20 splits the incident light into light guided toward the collimator 30a (hereinafter, reference light) and light guided toward the collimator 50a (hereinafter, measurement light). The beam splitter 20 is, for example, a cube beam splitter.

[0024] The collimator 30a converts the reference light guided from the beam splitter 20 into parallel light, and emits the parallel light toward the collimator 30b. The collimator 30b collects the parallel light incident from the collimator 30a, and emits the collected reference light toward the reference mirror 40. The collimator 30b receives the reference light reflected by the reference mirror 40, changes it into parallel light, and emits the parallel light toward the collimator 30a. The collimator 30a collects the parallel light incident from the collimator 30b and guides the light toward the beam splitter 20.

[0025] On the other hand, the collimator 50a converts the measurement light guided from the beam splitter 20 into parallel light and emits the parallel light toward the galvanometer mirror 60a. The parallel light incident from the collimator 50a is reflected on the surfaces of the galvanometer mirrors 60a and 60b, respectively, and is emitted toward the collimator 50b. The collimator 50b collects the parallel light incident from the collimator 50a via the galvanometer mirrors 60a and 60b, and irradiates the collected measurement light onto the sample Sm. The measurement light irradiated onto the sample Sm is reflected on the reflecting surface of the sample Sm and enters the collimator 50b. The reflecting surface is not limited to, for example, the boundary surface between the sample Sm and the surrounding environment of the sample Sm (e.g., the atmosphere), but may be the boundary surface between materials or tissues with different refractive indices inside the sample Sm. Hereinafter, the light reflected on the reflecting surface of the sample Sm and incident onto the collimator 50b is referred to as reflected light.

[0026] The collimator 50b outputs the incident reflected light toward the galvanometer mirror 60b. The light is reflected by the surfaces of the galvanometer mirrors 60b and 60a, respectively, and is output toward the collimator 50a. The collimator 50a collects the parallel light incident from the collimator 50a via the galvanometer mirrors 60a and 60b, and guides the collected reflected light toward the beam splitter 20. The beam splitter 20 guides the reference light reflected by the reference mirror 40 and the light reflected by the sample Sm to the spectroscope 70 via the optical fiber F.

[0027] The spectrometer 70 includes a diffraction grating and a light receiving element therein. The diffraction grating separates the reference light and the reflected light guided from the beam splitter 20. The separated reference light and the reflected light interfere with each other to become interference light. The light receiving element detects the interference light and generates a signal based on the detected interference light (hereinafter, detection signal). The light receiving element is configured by arranging multiple pixels two-dimensionally on a predetermined imaging surface. The light receiving element outputs the generated detection signal to the signal processing device 100.

[0028] The signal processing device 100 acquires an OCT signal representing the state of the sample Sm based on the detection signal input from the spectrometer 70. The signal processing device 100 sequentially changes the observation point in a predetermined order, and sequentially accumulates the detection signals acquired at each observation point or collectively for a plurality of observation points forming a predetermined unit between the observation points to generate an OCT signal within a predetermined region. The signal processing device 100 converts the corrected signal value for each sample point into a luminance value for each pixel, and generates image data indicating the luminance value for each pixel corresponding to each individual sample point. In this embodiment, the signal processing device 100 performs the following process on the acquired OCT signal before generating image data. The signal processing device 100 calculates the phase gradient of the OCT signal for each sample point arranged on a predetermined plane. The signal processing device 100 sets a plurality of paths from the starting point to a certain end point, with a sample point for which the bulk phase error has already been determined as a starting point and a sample point for which the bulk phase error has not yet been determined as an end point. Then, for each set path, the signal processing device 100 integrates the phase gradient for each sample point on the path from the starting point to the end point along each path to calculate a bulk phase error (hereinafter, path-specific phase error). The signal processing device 100 combines the path-specific phase errors between the set paths to determine the bulk phase error at the end point. The signal processing device 100 executes a process of determining a bulk phase error for each sample point that is set as an end point on the plane. The bulk phase error means an error included in the phase of the OCT signal due to the movement of the sample and the fluctuation of the atmosphere and temperature through which the light propagates. The signal processing device 100 subtracts the bulk phase error from the phase of the signal value for each sample point indicating the OCT signal to correct the phase of the signal value.

[0029] Next, an example of the functional configuration of the signal processing device 100 according to this embodiment will be described. FIG. 2 is a block diagram showing an example of the functional configuration of the signal processing device 100 according to this embodiment. The signal processing device 100 includes a control unit 110 and a storage unit 180. The control unit 110 performs a function by, for example, reading a program stored in advance in the storage unit 180 by a processor such as a CPU (Central Processing Unit) and performing processing instructed by commands written in the read program. In this application, performing processing instructed by commands written in the program may be referred to as executing a program, running a program, etc. Part or all of the control unit 110 may be configured with hardware such as an LSI (Large Scale Integration) or an ASIC (Application Specific Integrated Circuit) or dedicated hardware.

[0030] The control unit 110 includes an optical system control unit 120 , a detection signal acquisition unit 130 , a phase gradient calculation unit 140 , a bulk phase error calculation unit 150 , a signal correction unit 160 , and an image processing unit 170 .

[0031] The optical system control unit 120 drives a driving mechanism that changes the positions of the galvanometer mirrors 60a and 60b, and scans a sample point, which is an observation point of the sample Sm. The sample point of the sample Sm is scanned in the depth direction of the sample Sm and in a direction intersecting the direction (e.g., vertical direction). In the following description, the depth direction of the sample is sometimes called the z direction in a three-dimensional orthogonal coordinate system, and the first and second directions mutually orthogonal to the z direction are sometimes called the x direction and y direction, respectively. Signal acquisition in the depth direction of the sample is sometimes called an A-scan, and signal acquisition in a direction intersecting the direction (e.g., x direction or y direction) is sometimes called a B-scan.

[0032] The detection signal acquisition unit 130 sequentially acquires detection signals from the spectrometer 70. Based on the acquired detection signals, the detection signal acquisition unit 130 calculates the distribution of the reflected light (reflected light) of the measurement light in the depth direction of the sample Sm as a signal value for each sample point arranged at a predetermined interval in the three-dimensional space where the sample Sm exists. The depth direction of the sample Sm corresponds to the incident direction (z direction) of the measurement light. The calculated refractive index distribution corresponds to the intensity distribution of the reflected light. The detection signal acquisition unit 130 repeats the process of calculating the intensity distribution of the reflected light for each observation point changed by signal acquisition along a plane (e.g., xy plane) intersecting the z direction. In this way, the detection signal acquisition unit 130 can acquire data (hereinafter, OCT signal) representing the distribution of the reflected light in a three-dimensional area that can be observed (hereinafter, observable area). The OCT signal indicates a signal value for each sample point distributed at a certain interval in the observable area. The OCT signal is three-dimensional data that represents the state of the sample Sm in the observable region, and is also called an OCT volume, a three-dimensional OCT signal (FIG. 3), etc. The detection signal acquisition unit 130 stores the acquired OCT signal in the storage unit 180. The plane intersecting the z direction does not necessarily have to be orthogonal to the z direction, and may not be parallel to the z direction. In addition, in a two-dimensional plane intersecting the z direction, each sample point does not necessarily have to be arranged on each lattice point of an orthogonal lattice. Each sample point may be arranged, for example, on each lattice point of an oblique lattice, or may be arranged non-periodically. Note that the interval between adjacent sample points is set to be equal to or less than the spatial resolution of the optical system. In other words, there is no significant difference in the phase component (sample phase) caused by the structure of the sample between adjacent sample points. As will be described later, by calculating the phase difference of the signal value between adjacent sample points as an element of the phase gradient, the sample phase component is offset, and the bulk phase error component remains. In the above description, the term "scan" refers to signal acquisition involving a change in the target sample point or collective signal acquisition for multiple sample points, but may not necessarily involve scanning by driving a member constituting the optical system (e.g., the collimator 50b, the support base for the sample Sm) or a detector. In that case, the detection signal acquisition unit 130 is responsible for the function related to signal acquisition, and is omitted in the optical system control unit 120. For example, when performing an A-scan at a point on the xy plane using FD-OCT as the OCT method, the detection signal acquisition unit 130 detects interference light based on reflected light having different frequency components depending on the depth of the sample Sm as a detection signal. The detection signal acquisition unit 130 can perform a Fourier transform on the detection signal to calculate a conversion coefficient for each frequency, and determine the conversion coefficient of the depth associated with the frequency as the signal value at that depth. Furthermore, when scanning by driving in the x direction or y direction is not required, the optical system control unit 120 can be omitted.

[0033] The phase gradient calculation unit 140 reads the OCT signal from the storage unit 180, performs phase unwrapping on the signal value of each sample point indicated by the read OCT signal using a known method, adjusts the phase of the signal value, and calculates an adjusted phase. Phase unwrapping is a process of removing the restriction on the phase restricted to a range of -π to π so as to ensure spatial continuity. The phase gradient calculation unit 140 calculates a phase gradient for each sample point arranged on a predetermined two-dimensional plane based on the adjusted phase for each sample point (FIG. 3: phase gradient analysis S01). The phase gradient is a two-dimensional vector including partial differentials with respect to a first direction (e.g., x direction) and a second direction (e.g., y direction) that are parallel to the two-dimensional plane and perpendicular to each other as elements. The two-dimensional plane may be a plane perpendicular to the z direction or a plane that is not perpendicular to the z direction, as long as it intersects with the z direction. The two-dimensional plane may be a plane that crosses the observable region, or may be a virtual plane, so long as it is a plane that can be projected from any point in the three-dimensional observable region. For example, the sample points arranged in the three-dimensional observable region are associated with points projected in a third direction (e.g., the z direction) perpendicular to the two-dimensional plane. The phase gradient is calculated for each sample point in the two-dimensional plane that intersects with the z direction because the bulk phase error does not depend on the z direction but mainly depends on the coordinates in the two-dimensional plane. In the following description, the region in which the sample points projected onto the two-dimensional plane are distributed may be referred to as a two-dimensional region. The phase gradient calculation section 140 stores data indicating the phase gradient calculated for each sample point in the two-dimensional region in the storage section 180 as phase gradient data (FIG. 3: two-dimensional phase gradient Pg01).

[0034] The bulk phase error calculation unit 150 reads the phase gradient data from the memory unit 180, and based on the phase gradient for each sample point indicated by the read phase gradient data, performs the processing described below to calculate the bulk phase error at each sample point arranged in a two-dimensional region. The bulk phase error calculation unit 150 sets a plurality of paths from the starting point to an end point, which is one of the sample points for which the bulk phase error has not yet been determined, starting from one or more sample points for which the bulk phase error has already been determined. For each of the set paths, the bulk phase error calculation unit 150 integrates the phase gradient for each sample point along each path from the starting point to the end point (FIG. 3: Phase estimation by path integration S02) and calculates the integral value at the end point as the path-specific phase error. Then, the bulk phase error calculation unit 150 determines a combined value calculated by combining the path-specific phase errors calculated for each path among the set paths as the bulk phase error at the end point (FIG. 3: bulk phase error Bp01). However, the bulk phase error calculation unit 150 sets one of the sample points on the two-dimensional plane as a reference point (for example, the origin of the two-dimensional area) and predetermines the bulk phase error at the reference point as a predetermined reference value (for example, 0). The bulk phase error calculation section 150 stores in the storage section 180 bulk phase error data indicating the bulk phase error for each of the determined sample points.

[0035] The signal correction unit 160 reads out the bulk phase error data and the OCT signal from the storage unit 180. The signal correction unit 160 performs a process of correcting the phase of the signal value for each sample point indicated by the OCT signal, based on the bulk phase error for each sample point indicated by the read bulk phase error data. The signal correction unit 160 identifies a sample point (hereinafter, three-dimensional sample point) in an observable region where the position of a sample point on a two-dimensional region indicated by the bulk phase error data (hereinafter, two-dimensional sample point) and the position of a point projected onto a two-dimensional plane are common as a corresponding three-dimensional sample point. The number of identified three-dimensional sample points can be plural for one two-dimensional sample point. The signal correction unit 160 corrects the phase of the signal value at the identified three-dimensional sample point by subtracting the bulk phase error of the corresponding two-dimensional sample point. The signal correction unit 160 may further perform a known correction process (for example, refocusing, speckle suppression, etc.) on the signal value whose phase has been corrected for each three-dimensional sample point. The signal correction unit 160 stores a corrected OCT signal indicating the signal value for each three-dimensional sample point obtained by the correction process in the storage unit 180.

[0036] The image processing unit 170 reads out the corrected OCT signal from the storage unit 180, and converts the signal value of each three-dimensional sample point into a luminance value using a predetermined conversion function. The converted luminance value is within a range that can be expressed by the bit depth of each pixel. Image data is generated that has the luminance value of each sample point on the two-dimensional plane of the observation target (hereinafter, the observation target plane) as the luminance value of each pixel. In this application, the sample point may be referred to as a pixel.

[0037] The image processing unit 170 may determine a cross section that crosses at least a part of the observable area as the two-dimensional plane to be observed based on an operation signal input from an operation input unit (not shown). The operation signal indicates requirements for the plane to be observed, such as the depth from the sample surface and the observation direction with respect to the sample surface. The operation input unit may be configured to include a member that receives a user operation and generates an operation signal according to the received operation, such as a button, a knob, a dial, a mouse, a joystick, etc. The operation input unit may be an input interface that receives an operation signal wirelessly or wired from another device (for example, a portable device such as a remote controller). The image processor 170 may output the generated image data to a display unit (not shown) and display an OCT image based on the image data. The display unit may be, for example, a liquid crystal display (LCD) or an organic electroluminescence display (OLED). The image processing unit 170 may store the generated image data in the storage unit 180, or may transmit the image data wirelessly or by wire to another device via a communication unit (not shown).

[0038] The storage unit 180 stores various data used in the processes executed by the control unit 110 and various data acquired by the control unit 110, in addition to the above programs. The storage unit 180 includes a non-volatile (non-transient) storage medium such as a ROM (Read Only Memory), a flash memory, a HDD (Hard Disk Drive), etc. The storage unit 180 includes a volatile storage medium such as a RAM (Random Access Memory), a register, etc.

[0039] (Phase gradient) Next, a calculation example of the phase gradient will be described. The phase gradient calculation unit 140 calculates a phase gradient (Bx, By) for each two-dimensional sample point (x, y) arranged in a two-dimensional region based on the adjusted phase φ(x, y, z) for each three-dimensional sample point (x, y, z). In the example shown in FIG. 3, the three-dimensional sample points (x, y, z) are distributed in the x direction, y direction, and z direction at a predetermined interval within the observable region of a rectangular parallelepiped. The x direction, y direction, and z direction are mutually orthogonal, and correspond to the directions of each side forming the outer edge of the observable region. The two-dimensional region is a two-dimensional xy plane parallel to the x direction and y direction. The coordinates of the two-dimensional sample points (x, y) obtained by projecting the three-dimensional sample points (x, y, z) onto the two-dimensional region can be obtained by ignoring the coordinate values ​​in the z direction among the three-dimensional sample points (x, y, z). In other words, a three-dimensional sample point (x, y, z) corresponding to a two-dimensional sample point (x, y) is a sample point whose coordinates are indicated by the x-coordinate value and the y-coordinate value that are common to the two-dimensional sample point (x, y), respectively.

[0040] The phase gradient calculation unit 140 calculates, for example, a complex conjugate A(x, y, z) of the spectrum A(x, y, z) of the signal value at a three-dimensional sample point (x+Δx, y, z) adjacent in the x direction as shown in Equation (1). * The argument of the sum of the multiplication value of (x+Δx,y,z) in the z direction in the observation area is the x-direction component B of the phase gradient at the corresponding two-dimensional sample point (x,y). x Calculated as (x,y).

[0041]

number

[0042] The phase gradient calculation unit 140 calculates the y-direction component B of the phase gradient at the corresponding two-dimensional sample point (x, y). y For (x, y), as shown in equation (2), the spectrum A(x, y, z) of the signal value at the three-dimensional sample point (x, y, z) and the complex conjugate A of the spectrum of the signal value at the three-dimensional sample point (x, y + Δy, z) adjacent in the y direction are *It can be calculated based on (x,y+Δy,z).

[0043]

number

[0044] In equations (1) and (2), Δx and Δy respectively indicate the intervals between three-dimensional sample points in the x and y directions, and also indicate the intervals between two-dimensional sample points. arg(...) indicates the argument of the complex number.... As shown in equations (1) and (2), by calculating the argument for the multiplication value between sample points distributed in the z direction perpendicular to the xy plane, random sample phase residuals occurring between individual two-dimensional sample points are not accumulated, but bulk phase errors are accumulated. Therefore, errors caused by sample phase residuals can be reduced. In addition, the effect of noise is reduced because the phase gradient is calculated with more importance given to sample points with high intensity than to sample points with low intensity, which tend to be affected by noise. The phase gradient B calculated for each two-dimensional sample point is used to calculate the bulk phase error as described above.

[0045] As described later, the integral path is not necessarily limited to a portion parallel to the x direction or the y direction, but may include a portion intersecting both the x direction and the y direction. In that case, the phase gradient calculation unit 140 calculates the next sample point (x+Δx, y+Δy) and the phase gradient B xy (x,y) can be calculated.

[0046]

number

[0047] (Path integral) Next, an example of the path integral of the phase gradient will be described. In the bulk phase error calculation unit 150, a plurality of integral paths are set in advance for each two-dimensional sample point (hereinafter also referred to as a phase determination target pixel) for which a phase error is to be determined. The starting point of each integral path is a two-dimensional sample point (hereinafter also referred to as a phase-determined pixel) for which a phase error has already been determined. However, the starting point is a two-dimensional sample point located within a predetermined range from the ending point. This limits the number of integral paths so that they do not become excessive.

[0048] The bulk phase error calculation unit 150 calculates, for example, as a path-specific phase error φ by performing path integration on the phase gradient B for each sample point (x, y) on the path Pt from the start point Op to the end point Dp of each path Pt as shown in Equation (4). e,Pt It is calculated as follows.

[0049]

number

[0050] In equation (4), B(x,y) represents the phase gradient, and the x-direction component B x (x,y) and y-direction component B y It is a two-dimensional vector containing (x, y) as an element. n(x, y) indicates a direction vector in the tangent direction of the path Pt. More specifically, as the direction vector n(x, y), a vector whose starting point is the sample point (x, y) and whose ending point is an adjacent sample point on the path Pt can be used. · indicates the dot product of the vectors. Then, the bulk phase error calculation unit 150 calculates the path-specific phase error φ e,Pt The bulk phase error φ at the end point Dp is calculated by combining the paths with the same end point Dp. e Calculate.

[0051] Next, an example of an integral path set on a two-dimensional plane will be described. FIG. 4 is an explanatory diagram showing an example of an integral path according to the present embodiment. In the example shown in FIG. 4, four paths Pt01-Pt04 are set for a phase-determined pixel Op00 as a starting point and a phase determination target pixel Dp11 as an end point. The coordinates of the phase-determined pixel Op00 and the phase determination target pixel Dp11 on a two-dimensional plane are (0,0) and (1,1), respectively. The end points are adjacent to the starting point in a diagonal direction intersecting both the x coordinate and the y coordinate. The x coordinate and the y coordinate are normalized so that the intervals Δx and Δy are 1, respectively. The paths Pt01-Pt04 are formed by sequentially connecting segments each having a sample point as a starting point and an adjacent sample point in the x direction or y direction as an end point. A vector indicating each segment is used as a directional vector shown in Equation (4) to calculate a phase error by path. In this example, the length of each fragment (hereinafter, fragment length) is 1, so the bulk phase error calculation section 150 can determine the number of fragments forming each path Pt01-Pt04 as the path length (hereinafter, path length). The path lengths of paths Pt01, Pt02, Pt03, and Pt04 are 2, 2, 4, and 4, respectively.

[0052] Therefore, the bulk phase error calculation unit 150 calculates the path-specific phase errors φ e,Pt01 , φ e,Pt02 , φ e,Pt03 , φ e,Pt04 B x (0,0)+B y (1,0), B y (0,0)+B x (0,1), B x (0,0)+B x (1,0)+B y (2,0)-B x (2,1), B y (0,0)+B y (0,1)+B x (0,2)-B y This can be calculated as (1,2).

[0053] Then, the bulk phase error calculation unit 150 calculates, for example, a simple average value between the paths as a bulk phase error obtained by combining the path-specific phase errors between the paths. In the example shown in Fig. 4, the process of calculating the bulk phase error includes a process of normalizing the sum of the path-specific phase errors by dividing it by the number of paths, 4.

[0054] However, the influence of the error superimposed on the phase gradient may differ depending on the path length, so the bulk phase error calculation section 150 may use a value that depends on the path length as a weighting coefficient for each path to calculate a weighted average of the phase errors for each path as the bulk phase error. For example, the bulk phase error calculation unit 150 may use a weighting coefficient for each path as p (a is a real number greater than 0 and less than 1, and p is the path length) and a real number normalized to 1 for the sum over the paths can be used. In that case, in the example shown in FIG. 4, the path-specific phase error φ e,Pt01 , φ e,Pt02 , φ e,Pt03 , φ e,Pt04 The weighting coefficients for 2 / 2(a 2 +a 4 ), a 2 / 2(a 2 +a 4 ), a 4 / 2(a 2 +a 4 ), a 4 / 2(a 2 +a 4 ) The bulk phase error calculation section 150 may use, as the weighting coefficient for each path, a value that is inversely proportional to the path length p and whose sum over the paths is normalized to 1. In that case, in the example shown in FIG. e,Pt01 , φ e,Pt02 , φ e,Pt03 , φ e,Pt04 The weighting coefficients for are 1 / 3, 1 / 3, 1 / 6, and 1 / 6, respectively. These methods can reduce the contribution of the path-specific phase error of a path with a long path length to the bulk phase error.

[0055] In addition, the bulk phase error calculation section 150 calculates the end point (x 0 ,y 0 ) bulk phase error φ b (t) (x 0 ,y 0 In formula (5), φ e,Pt (x 0 ,y 0 ) indicates the path-specific phase error for path Pt. Equation (5) calculates a sum vector, which is the sum of unit vectors between paths whose magnitude is 1 and whose argument is the path-specific phase error, and calculates the argument of the calculated sum vector as the bulk phase error φ b (t) (x 0 ,y 0 ) is defined as

[0056]

number

[0057] In addition, various integral paths can be set according to the positional relationship between the phase-determined pixel and the phase determination target pixel. In the following description, a set of multiple integral paths corresponding to the positional relationship between the phase-determined pixel and the phase determination target pixel is called a path pattern. The path pattern illustrated in FIG. 4 is called a path pattern Pn00. 5A-5I are diagrams showing other examples of route patterns according to this embodiment. However, the end point of each of the route patterns Pn01-Pn09 shown in each of FIGS. 5A-5I is the phase determination target pixel Dp11, and the position of the phase-determined pixel serving as the starting point is adjacent to the phase determination target pixel Dp11 in one or both of the negative direction of the x axis and the negative direction of the y axis. The number of phase-determined pixels serving as the starting point is not limited to one, but can be two or more. In the route patterns Pn00-Pn02, the number of phase-determined pixels serving as the starting point is one. In contrast, in the route patterns Pn03-Pn05, the number of phase-determined pixels serving as the starting point is two. In the route patterns Pn06-Pn09, the number of phase-determined pixels serving as the starting point is three.

[0058] 5A includes three paths Pt11-Pt13. Each of the paths Pt11-Pt13 starts from a phase-determined pixel Op01. The path lengths of the paths Pt11, Pt12, and Pt13 are 3, 1, and 3, respectively. 5B includes three paths Pt21-Pt23. Each of the paths Pt21-Pt23 starts from a phase-determined pixel Op10. The path lengths of the paths Pt21, Pt22, and Pt23 are 3, 1, and 3, respectively. 5C includes two paths Pt31 and Pt32. The paths Pt31 and Pt32 start from phase-determined pixels Op01 and Op02, respectively. The path lengths of the paths Pt31 and Pt32 are 1 and 3, respectively.

[0059] 5D has two paths Pt41 and Pt42. The paths Pt41 and Pt42 start from phase-determined pixels Op10 and Op20, respectively. The path lengths of the paths Pt41 and Pt42 are 1 and 3, respectively. 5E has four paths Pt51-Pt54. The paths Pt51, Pt52, Pt53, and Pt54 start from phase-determined pixels Op01, Op10, Op01, and Op10, respectively. The path lengths of the paths Pt51, Pt52, Pt53, and Pt54 are 1, 1, 3, and 3, respectively. 5F has three paths Pt61-Pt63. The paths Pt61, Pt62, and Pt63 start from phase-determined pixels Op00, Op01, and Op02, respectively. The path lengths of the paths Pt61, Pt62, and Pt63 are 2, 1, and 2, respectively.

[0060] 5G has three paths Pt71-Pt73. The paths Pt71, Pt72, and Pt73 start from phase-determined pixels Op00, Op10, and Op20, respectively. The path lengths of the paths Pt71, Pt72, and Pt73 are 2, 1, and 2, respectively. 5H has four paths Pt81-Pt84. The paths Pt81, Pt82, Pt83, and Pt84 start from phase-determined pixels Op01, Op10, Op01, and Op20, respectively. The path lengths of the paths Pt81, Pt82, Pt83, and Pt84 are 1, 1, 3, and 2, respectively. 5I includes four paths Pt91-Pt94. The paths Pt91, Pt92, Pt93, and Pt94 start from phase-determined pixels Op01, Op02, Op10, and Op10, respectively. The path lengths of the paths Pt91, Pt92, Pt93, and Pt94 are 1, 3, 1, and 3, respectively.

[0061] 4 and 5A to 5I show route patterns Pn00-Pn09 having only routes where both the x and y coordinates of the end point are equal to or greater than the x and y coordinates of the start point, respectively, and the number of start points is between 1 and 3, but the present invention is not limited to this. A route pattern may be set that has a route where the x coordinate of the end point is less than the x coordinate of the start point or the y coordinate of the end point is less than the y coordinate of the start point, or a route pattern with a greater number of start points may be set. 6A-6D and 7A-7C show examples of sets of starting points and end points of a route pattern according to this embodiment. 6A-6D and 7A-7C show all sets that can be set in an area in an xy plane of 3 samples by 3 samples with the end point at the center. However, a set in which the distribution of starting points of a certain set is symmetrical with respect to the x-axis or y-axis passing through the end point, and a set in which the distribution of starting points is rotationally symmetrical with respect to the end point are regarded as a common set and are not shown. Here, the starting points and end points are shown by diagonal lines and hatching, respectively. Also, the illustration of the route is omitted.

[0062] 6A-6D show sets with the number of origins being 1 to 4, respectively. When the number of origins is 1, 2, 3, or 4, 2, 4, 8, or 12 sets can be set, respectively. 7A to 7C show sets with the number of origins being 5 to 7. When the number of origins is 5, 6, or 7, 8, 4, or 2 sets can be set, respectively.

[0063] 4 to 7C show an example in which the start point is a phase-determined image adjacent to the end point in one or both of the x and y directions (diagonal directions), but this is not limited as long as it is within a predetermined range from the end point. For example, the start point may be a phase-determined pixel within two samples from the end point. Furthermore, as illustrated in Figures 4 and 5A-5I, the multiple integration paths are all included within a predetermined path setting range from the end point (e.g., Figures 8B, 8C, and 9: path setting range Pa), and are set so that no part or all of them fall outside the predetermined range, thereby preventing the integration paths from becoming excessively large.

[0064] 4 and 5A-5I show an example in which an integral path is formed by sequentially connecting fragments (links) consisting of sample points adjacent to a certain sample point in the x-direction or y-direction from a starting point to an end point, but this is not limited thereto. The starting point of each fragment may be a sample point located within the spatial resolution range of the optical system from the end point of the fragment. The fragment may be, for example, a section connecting two sample points adjacent in a diagonal direction intersecting both the x-direction and the y-direction, or a section connecting two sample points that are not adjacent to each other and that cross one or more sample points (skip). In addition, sample points on the integral path, that is, sample points on the integral path excluding the starting point and the end point, may be sample points for which a bulk phase error has already been determined, or may be sample points for which a bulk phase error has not yet been determined. However, for each possible end point position of a fragment, the phase gradient calculation unit 140 calculates, as a phase gradient, the phase difference of the OCT signal between the sample point as the starting point of the fragment and the end point of the fragment for each sample point in the two-dimensional region.

[0065] (Repetitive operations) The bulk phase error calculation unit 150 may be set in advance with a plurality of route patterns (for example, route patterns Pn00-Pn09). The bulk phase error calculation unit 150 selects, for each sample point in the two-dimensional region, a route pattern from the previously set route patterns that satisfies the positional relationship and correspondence relationship with the sample point for which the bulk phase error has already been determined as the starting point, with the sample point as the end point. The bulk phase error calculation unit 150 calculates the bulk phase error at the end point by the above-mentioned method, with each of the multiple routes indicated by the selected route pattern as an integral route. Then, the bulk phase error calculation unit 150 changes the target sample point for which the bulk phase error is to be calculated to another sample point for which the bulk phase error has not yet been determined. The bulk phase error calculation unit 150 may repeat the calculation of the bulk phase error and the change of the target sample point until the bulk phase error is determined for all sample points in the two-dimensional region. The target sample point corresponds to the above-mentioned phase determination target pixel.

[0066] Therefore, the bulk phase error calculation unit 150 may predetermine a calculation order (iteration) for calculating the bulk phase error such that one sample point in the two-dimensional region is set as a reference point, and a sample point closer to the reference point is given priority. Furthermore, when the bulk phase error calculation unit 150 sets the target sample point whose calculation order is the m-th (m is an integer equal to or greater than 1) sample point as the end point, the bulk phase error calculation unit 150 may select a route pattern corresponding to the positional relationship indicated by the end point and the start point of the m-1th sample point of the calculation order. Here, the reference point is set to the 0th sample point of the calculation order, and the bulk phase error is set as a predetermined reference value. Furthermore, the number of sample points having the same calculation order is not necessarily one, but may be two or more. The number of sample points having the same calculation order may be two or more, but the calculation order of the bulk phase error between two or more sample points of the mth may be a predeterminable order or may be randomly determined.

[0067] 8A-8C and 9A, the origin (0,0) of the two-dimensional region is set as a reference point Op00, and the calculation order is set in the bulk phase error calculation unit 150 in the order of the distance from the origin in the diagonal directions sandwiched between the x direction and the y direction. Sample points with the same calculation order are distributed on a line perpendicular to the diagonal direction. As shown in FIG. 8B, the sample points at (1,0) and (0,1) in the first calculation order are set as target sample points Dp10 and Dp01. At this time, the bulk phase error calculation unit 150 can select route patterns Pn01 and Pn02 for the target sample points Dp10 and Dp01, respectively. However, among the selected route patterns Pn01 and Pn02, the bulk phase error calculation unit 150 discards routes Pt11 and Pt21 as invalid routes because they pass outside the range of the two-dimensional area in which the sample points are arranged, and uses the remaining valid routes for calculating the bulk phase error. The weighting coefficient used when combining the route-specific phase errors between valid routes may be determined in advance based on the number of valid routes or the route length of the valid routes.

[0068] 8C, the sample points in the second calculation order are set as target sample points Dp11, Dp20, and Dp02, which are located at (1,1), (2,0), and (1,0), respectively. At this time, the bulk phase error calculation unit 150 can select path patterns Pn05, Pn01, and Pn02 for the target sample points Dp10 and Dp01, respectively. However, among the selected path patterns Pn01 and Pn02, the bulk phase error calculation unit 150 discards paths Pt11 and Pt21 as invalid paths because they pass outside the range of the two-dimensional region in which the sample points are arranged, and uses the remaining valid paths to calculate the bulk phase error. As shown in Fig. 9A, the sample points located at (2,1) and (1,2) are set as target sample points Dp21 and Dp12, respectively, as sample points in the third calculation order. However, in the example shown in Fig. 9A, the target sample point Dp22, which is the farthest from the reference point Op00 in the two-dimensional region, is also included in the sample points in the third calculation order. In this case, the bulk phase error calculation unit 150 can select the path patterns Pn05, Pn05, and Pn00 for the target sample points Dp21, Dp12, and Dp22, respectively. However, of the route patterns Pn05, Pn05, and Pn00 selected for the target sample points Dp21, Dp12, and Dp22, respectively, the bulk phase error calculation unit 150 discards routes Pt54, Pt53, Pt03, and Pt04 as invalid routes because they pass outside the range of the two-dimensional area in which the sample points are arranged, and uses the remaining valid routes to calculate the bulk phase error.

[0069] The calculation order between sample points in a two-dimensional region is not limited to the examples shown in Fig. 8A-8C and Fig. 9A. For example, a sample point that is a reference point may be set as the center of the two-dimensional region, and the calculation order may be determined such that the closer the sample is to the reference point, the higher the calculation order is. When the calculation order is set to the m-th (m is an integer equal to or greater than 1) sample point as the target sample point with the calculation order being the end point, the bulk phase error calculation section 150 may select a route pattern corresponding to the positional relationship indicated by the end point and the start point of the (m-1)th sample point in the calculation order.

[0070] 9B shows an example of the calculation sequence. When calculating a bulk phase error at a sample point A where a path Sq is set in a clockwise spiral shape centered on a reference point, the bulk phase error calculation unit 150 sets a 3×3 point area centered on the sample point A as a path setting range P A Set the route range P A5B. When calculating the bulk phase error at the sample point B, the bulk phase error calculation unit 150 defines a 3×3 area with the sample point B as a route setting range P B Set the route range P B Since there are four phase-determined pixels in the sample point B, namely, directly above, above left, to the left, and below left, the bulk phase error calculation section 150 selects the path pattern illustrated at the left end of the first row in FIG. 6D.

[0071] In this way, the calculation order for each sample is determined so that the closer the distance from the reference point, the higher the priority, and the path integration is performed starting from the previous calculation order, so the number of calculations until the final bulk phase error is calculated in the two-dimensional region is reduced, and therefore the accumulation of errors is suppressed in the entire two-dimensional region. In addition, there may be multiple route patterns that end at the target sample point and start at a sample point for which the bulk phase error has already been determined, depending on the positional relationship. In this case, the bulk phase error calculation unit 150 may predetermine one of the multiple route patterns, or may randomly select one of the multiple route patterns.

[0072] The order of calculation between sample points in a two-dimensional region is not limited to this. For example, the bulk phase error calculation unit 150 repeats the process of calculating the bulk phase error for each phase-determined pixel using the path pattern shown in FIG. 5A as a calculation process of the bulk phase error in each row (hereinafter, line processing) until the phase-determined pixel reaches the sample point at the end of the two-dimensional region, and changing the phase-determined pixel to an adjacent sample point in the x-axis direction. After that, the bulk phase error calculation unit 150 calculates the bulk phase error at the sample point at the start of the next row adjacent in the y-axis direction using the path pattern shown in FIG. 5B. Then, the bulk phase error calculation unit 150 may repeat the line processing until the row to be determined for the bulk phase reaches the end of the two-dimensional region.

[0073] (Example of phase correction) Next, an example of correcting the phase of the OCT signal based on the calculated bulk phase error will be described. 10A to 10F are diagrams showing examples of correction of the phase of an OCT signal according to this embodiment. 10A shows an example of an OCT image based on an OCT signal. The OCT image shown in FIG. 10A shows the state of an object plane that is a part of the observable area. The OCT image is based on a raw OCT signal that has not yet been compensated for bulk phase errors. FIG. 10D shows a spectrum obtained by converting the OCT signal of FIG. 10A into the spatial frequency domain. The vertical and horizontal axes indicate spatial frequencies in the x and y directions, respectively, with the center indicating the origin. The intensity of the spectrum is expressed by shading. Brighter areas indicate higher intensity, and darker areas indicate lower intensity. The intensity is asymmetric with respect to the origin, with the center of the frequency band with the highest intensity displaced from the origin in the x direction. In addition, the spread of the high intensity area is relatively large.

[0074] Figure 10B shows an example of the bulk phase error calculated using the conventional phase contrast method. The phase is represented by shading. Lighter areas indicate a phase closer to π, and darker areas indicate a phase closer to -π. The bulk phase error varies relatively regularly in the x direction, but varies more irregularly in the y direction than in the x direction. Figure 10E shows a spectrum corrected using the bulk phase error shown in Figure 10B. The intensity of the spectrum is symmetrical with respect to the y axis passing through the origin in the x direction, and the spread in the x direction is reduced. However, the spread of the spectral intensity in the y direction is rather expanded. This indicates that the correction in the y direction is inappropriate.

[0075] Fig. 10C shows an example of the bulk phase error calculated by the signal processing device 100 according to the present embodiment. The phase is represented by shading. This bulk phase error is calculated using the method described with reference to Figs. 8A-8C and 9. The bulk phase error varies regularly in both the x and y directions and shows a relatively smooth stripe pattern. FIG. 10F shows the spectrum of the OCT signal corrected using the bulk phase error shown in FIG. 10C. The signal correction section 160 can remove the bulk phase error contained in the OCT signal by correcting the OCT signal using equation (6). Equation (6) expresses the bulk phase error φ from the phase of the spectrum A(x, y, z) of the signal value at the three-dimensional sample point (x, y, z). b (t) (x, y) is subtracted to calculate the spectrum A'(x, y, z) of the signal value after correction. The intensity of the spectrum is symmetrical with respect to the origin in both the x and y directions, and the spread is reduced. This shows that this embodiment can reduce the directional dependency of the phase error compared to the conventional phase difference method, and can achieve more appropriate spectrum correction.

[0076]

number

[0077] Next, another example of correction of the OCT signal for each sample will be described. Figures 11A, 11B, 11C, and 11D show the spatial frequency f x , the average value of the spatial frequency f in the y direction y Average value of spatial frequency f x standard deviation of spatial frequency f y and standard deviations are shown. These average values ​​and standard deviations were calculated for the original data before correction, the case where correction was made with the bulk phase error calculated by simple integration, and the case where correction was made with the bulk phase error calculated by the method proposed in this embodiment. Sm01 and Sm03 respectively show cases where different parts of chicken breast muscle were used as samples. Sm02 and Sm04 respectively show cases where different parts of porcine heart tissue were used as samples. Avg shows the average value between Sm01 and Sm04.

[0078] Figures 11A and 11B show that for the raw data, the average values ​​vary significantly between samples, while for the simple integral and proposed methods, the average values ​​are nearly 0, which is significantly reduced from the raw data. y Regarding the average value of , the simple integral reduces the variation more than the proposed method. On the other hand, Fig. 11C shows that the spatial frequency f x The standard deviation of the spatial frequency f is smaller than that of the original data. y The standard deviation of the original data increases compared to the original data. This indicates that the spectrum is broadened in the y direction more than in the x direction. In contrast, the proposed method has a spatial frequency f x , f y No significant change in the standard deviation of is observed, indicating that the proposed method can correct the bulk phase error without introducing any spectral bias in a specific direction.

[0079] The signal correction unit 160 may refocus the corrected OCT signal (digital refocusing). In the refocusing, the signal correction unit 160 executes the following steps of processing. (Step S11) The signal correction unit 160 performs a two-dimensional Discrete Fourier Transform (DFT) on the corrected two-dimensional OCT signal (e.g., en face OCT slice) to calculate a complex spatial frequency spectrum in the spatial frequency domain. (Step S12) The signal correction unit 160 calculates the spatial frequency (f x ,f y ) for each conversion coefficient, the phase filter H -1 (f x ,f y ) to calculate the filtered spectrum including the multiplied value obtained by multiplication. -1 (f x ,f yAn example of this is shown in equation (7).

[0080]

number

[0081] In equation (7), f r f denotes the spatial frequency in the radial direction. r is the x-space f x and the spatial frequency in the y direction f y It corresponds to the square root of the sum of the squares of λ c z denotes the central wavelength of the incident light. 0 indicates the amount of defocus. 0 generally depends on the optical system. Phase filter H -1 (f x ,f y ) can reduce or eliminate defocusing by compensating for phase shifts that are more pronounced at higher spatial frequencies, which typically appear in OCT images as "blur." (Step S13) The signal correction unit 160 performs a two-dimensional inverse Fourier transform (IDFT: Inverse Discrete Fourier Transform) on the filtered spectrum to generate a filtered OCT signal as a refocused OCT signal. Then, the signal correction unit 160 outputs the refocused OCT signal to the image processing unit 170. The image processing unit 170 can acquire a clearer OCT image based on the OCT signal input from the signal correction unit 160.

[0082] As described above, the signal processing device 100 according to this embodiment includes a phase gradient calculation section 140 and a bulk phase error calculation section 150. The phase gradient calculation section 140 calculates the phase gradient of the OCT signal representing the state of the sample for each sample point arranged on a predetermined plane (for example, a two-dimensional region). The bulk phase error calculation section 150 integrates the phase gradient for each sample point along each of a plurality of paths from a starting point, which is a sample point for which the bulk phase error has been determined, to an end point, which is a sample point for which the bulk phase error has not yet been determined, to calculate a path-specific phase error at the end point, and combines the path-specific phase errors between the plurality of paths to determine the bulk phase error at the end point. The bulk phase error calculation section 150 calculates, for example, an average value of the path-specific phase errors between the paths as the bulk phase error. With this configuration, random components of the error that may vary from one path to another are combined and relatively reduced, so that the components that are not path-specific are estimated as a bulk phase error more accurately than if the phase gradients were simply accumulated.

[0083] Furthermore, in this embodiment, the bulk phase error calculating section 150 may calculate, as the average value between paths, a weighted average value of the phase errors for each path using a weighting coefficient based on the path length of each of the multiple paths. According to this configuration, it is possible to calculate the bulk phase error more accurately by taking into account the influence of the error depending on the path length. For example, by making the weighting coefficient smaller for a path with a longer path length, it is possible to reduce the contribution of the path-specific phase error of the path when the influence of the error is stronger as the path length increases. Conversely, by making the weighting coefficient larger for a path with a longer path length, it is possible to increase the contribution of the path-specific phase error of the path when the influence of the error is weaker as the path length increases.

[0084] In addition, in this embodiment, each of the multiple paths used to calculate the bulk phase error may be included within a predetermined range from the end point. According to this configuration, each of the multiple paths is limited within a predetermined range, so that the number of paths involved in the calculation of the bulk phase error is prevented from becoming excessive.

[0085] Furthermore, in this embodiment, each of the multiple paths used to calculate the bulk phase error includes an nth fragment (N is an integer of 2 or more) which is the section from the nth (n is an integer greater than or equal to 1 and less than or equal to N-1) sample point to the n+1th sample point, and the n+1th sample point may be a sample point within a range equal to or less than the spatial resolution of the optical system that acquired the OCT signal from the nth sample point. As a result, each of the multiple paths is formed by concatenating one or more fragments consisting of two sample points at which no significant difference occurs in the signal value of the OCT signal, thereby suppressing the contribution of the signal components of the OCT signal to the calculated bulk phase error.

[0086] In addition, in this embodiment, the bulk phase error calculation unit 150 may preset multiple route patterns indicating multiple routes corresponding to the positional relationship between the starting point and the end point, and select multiple routes indicated by route patterns that satisfy the positional relationship with the target sample point for which the bulk phase error is to be calculated as the end point and the sample point for which the bulk phase error has been determined as the starting point. This allows multiple paths to be selected for use in calculating the bulk phase error for each target sample point within a two-dimensional region more efficiently than searching each time in relation to sample points for which the bulk phase error has already been calculated.

[0087] Furthermore, in this embodiment, the bulk phase error calculation section 150 may predetermine an order for calculating bulk phase errors such that one sample point in a two-dimensional plane (e.g., a two-dimensional region) is used as a reference point, and sample points closer to the reference point are given priority. When the predetermined order is for the mth sample point (m is an integer greater than or equal to 1) to be the target sample point, the predetermined order may select multiple routes indicated by a route pattern that satisfies a positional relationship starting from the (m-1)th sample point. This reduces the effect of errors accumulating in the sequentially calculated bulk phase errors more than when the calculation order is arbitrary, thereby enabling the bulk phase errors to be calculated with higher accuracy for the entire two-dimensional plane.

[0088] Moreover, the signal processing device 100 according to the present embodiment may include a signal correction unit 160. The signal correction unit 160 may subtract the bulk phase error at the end point from the phase of the signal value of the OCT signal at a sample point arranged in a three-dimensional space that has the same position as the end point projected onto the plane. This compensates for the bulk phase error contained in the three-dimensional OCT signal, reducing the spectral bias and broadening caused by the bulk phase error and enabling the production of higher quality optical coherence tomographic images.

[0089] Although the embodiment of the present invention has been described in detail above with reference to the drawings, the specific configuration is not limited to the above, and various design changes, etc. are possible within the scope that does not deviate from the gist of the present invention.

[0090] For example, in the above description, the signal processing device 100 is a part of the optical coherence tomography device 1, but the present invention is not limited to this. The signal processing device 100 may be a single device independent of the optical coherence tomography device 1 and not equipped with an optical system. In this case, the optical system control unit 120 may be omitted from the control unit 110 of the signal processing device 100. The detection signal acquisition unit 130 is not limited to an optical system, and may acquire detection signals or OCT signals from other devices such as a data storage device or a PC in a wired or wireless manner, for example, via a network.

[0091] Furthermore, the signal processing device 100 may include the above-mentioned operation input unit and display unit, or one or both of them may be omitted. The image processing unit 170 may be omitted in the control unit 110 of the signal processing device 100. In that case, the control unit 110 of the signal processing device 100 may output the corrected OCT signal generated by the signal correction unit 160 to another device such as a data storage device, a PC, an image processing device, etc. The device to which the signal is output may have the same function as the image processing unit 170, that is, a function of generating image data based on the corrected OCT signal input from the signal processing device 100 and displaying an OCT image based on the generated image data.

[0092] In the above example, the two-dimensional area in which the sample points related to the calculation of the phase gradient or bulk phase error are arranged is set on the xy plane, but this is not limited thereto. It is sufficient that the two-dimensional area is set on a two-dimensional plane in which a correspondence relationship between the three-dimensional sample points arranged in the three-dimensional space forming the observable area and the two-dimensional sample points obtained by projecting the three-dimensional sample points can be formed. Such a two-dimensional plane may be, for example, a yz plane or a zx plane. When such a two-dimensional plane is the yz plane, the three-dimensional sample points (x, y, z) are associated with the two-dimensional sample points (y, z). In addition, an observation target plane in which sample points corresponding to pixels constituting an OCT image are arranged may be used as the two-dimensional area. The signal correction unit 160 may perform correction of the phase of the OCT signal based on the bulk phase error for each observation target plane.

[0093] In addition, a part of the signal processing device 100 in the above-mentioned embodiment, for example, all or a part of the control unit 110, may be realized by a computer. In that case, a program for realizing this control function may be recorded in a computer-readable recording medium, and the program recorded in the recording medium may be read into a computer system and executed to realize the control function. In addition, the "computer system" here refers to a computer system built into the signal processing device 100, and includes hardware such as an OS and peripheral devices. In addition, the "computer-readable recording medium" refers to a portable medium such as a flexible disk, an optical magnetic disk, a ROM, a CD-ROM, and a storage device such as a hard disk built into a computer system. Furthermore, the "computer-readable recording medium" may also include a medium that dynamically holds a program for a short time, such as a communication line when a program is transmitted via a network such as the Internet or a communication line such as a telephone line, and a medium that holds a program for a certain period of time, such as a volatile memory inside a computer system that is a server or client in that case. In addition, the above-mentioned program may be a program for realizing a part of the above-mentioned function, and may further be a program that can realize the above-mentioned function in combination with a program already recorded in the computer system.

[0094] In addition, a part or the whole of the signal processing device 100 in the above-mentioned embodiment and modified examples may be realized as an integrated circuit such as an LSI (Large Scale Integration). Each functional block of the signal processing device 100 may be individually processed, or may be integrated into a processor in part or in whole. The integrated circuit method is not limited to LSI, and may be realized by a dedicated circuit or a general-purpose processor. In addition, when an integrated circuit technology that replaces LSI appears due to the progress of semiconductor technology, an integrated circuit based on that technology may be used. [Explanation of symbols]

[0095] 1...optical coherence tomography, 10...light source, 20...beam splitter, 30a, 30b, 50a, 50b...collimator, 40...reference mirror, 60a, 60b...galvanometer mirror, 70...spectroscope, 100...signal processing device, 120...optical system control unit, 130...detection signal acquisition unit, 140...phase gradient calculation unit, 150...bulk phase error calculation unit, 160...signal correction unit, 170...image processing unit

Claims

1. a phase gradient calculation unit that calculates a phase gradient in a plane intersecting a light irradiation direction of an optical coherence tomography signal representing a state of a sample for each sample point arranged on the plane; For each of a plurality of paths from a starting point, which is a sample point for which the bulk phase error has already been determined, to an end point, which is a sample point for which the bulk phase error has not yet been determined, the phase gradient for each sample point is integrated along each path to calculate a path-specific phase error at the end point; a bulk phase error calculation unit that combines the phase errors by paths among the plurality of paths to determine a bulk phase error at the end point. Signal processing device.

2. The bulk phase error calculation unit Calculating the average of the phase errors for each path as the bulk phase error. The signal processing device according to claim 1 .

3. The bulk phase error calculation unit As the average value, a weighted average value of the phase errors for each of the paths is calculated using a weighting coefficient based on the path length of each of the plurality of paths. The signal processing device according to claim 2 .

4. Each of the plurality of routes is included within a predetermined range from the end point. The signal processing device according to any one of claims 1 to 3.

5. Each of the multiple paths includes an n-th segment (N is an integer of 2 or more) that is an interval from an n-th sample point (n is an integer of 1 or more and N-1 or less) to an n+1-th sample point, The n+1 sample point is a sample point within a range equal to or less than the spatial resolution of an optical system that has acquired the optical coherence tomographic signal from the n sample point. The signal processing device according to any one of claims 1 to 4.

6. The bulk phase error calculation unit includes: A plurality of route patterns are set in advance, each of which indicates the plurality of routes in association with a positional relationship between the starting point and the ending point; A plurality of routes are selected, each of which is indicated by a route pattern that satisfies the positional relationship, with the target sample point for which the bulk phase error is to be calculated as the end point. The signal processing device according to any one of claims 1 to 5.

7. The bulk phase error calculation unit includes: One sample point on the plane is set as a reference point, determining a sequence for calculating the bulk phase errors in advance such that a sample point closer to the reference point is given priority; When the sample point in the mth order (m is an integer equal to or greater than 1) is the target sample point, the plurality of routes indicated by a route pattern that satisfies the positional relationship are selected, starting from the sample point in the m-1th order. The signal processing device according to claim 6.

8. A signal correction unit that subtracts a bulk phase error at the end point from a phase of a signal value of the optical coherence tomography signal at a sample point arranged in a three-dimensional space that has a common position projected onto the plane as the end point. The signal processing device according to any one of claims 1 to 7.

9. 1. A method in a signal processing apparatus, comprising: a phase gradient calculation step of calculating a phase gradient in a plane intersecting a light irradiation direction of the optical coherence tomography signal representing a state of the sample for each sample point arranged on the plane; For each of a plurality of paths from a starting point, which is a sample point for which the bulk phase error has already been determined, to an end point, which is a sample point for which the bulk phase error has not yet been determined, the phase gradient for each sample point is integrated along each path to calculate a path-specific phase error at the end point; a bulk phase error calculation step of determining a bulk phase error at the end point by combining the phase errors for each path among the plurality of paths. Signal processing methods.

10. The signal processing device computer a phase gradient calculation step of calculating a phase gradient in a plane intersecting a light irradiation direction of an optical coherence tomography signal representing a state of a sample for each sample point arranged on the plane; For each of a plurality of paths from a starting point, which is a sample point for which the bulk phase error has already been determined, to an end point, which is a sample point for which the bulk phase error has not yet been determined, the phase gradient for each sample point is integrated along each path to calculate a path-specific phase error at the end point; a bulk phase error calculation step of determining a bulk phase error at the end point by combining the phase errors per path among the plurality of paths; A signal processing program for executing the above.

Citation Information

Patent Citations

  • Interferometer and method for measuring air refractive index and optical path length effects

    JP2002504671A

  • Swept source optical coherence tomograph and phase stabilization program for the same

    JP2013140077A

  • Systems and methods for sub-aperture-based aberration measurement and correction in interferometric imaging

    JP2016504947A

  • Systems and methods for self-referenced quantitative phase microscopy

    US20140375792A1