Quantitative optical coherence tomography angiography

Abstract

By acquiring sequences of data associated with a fast scan direction at a multiple of slow scan locations, a three-dimensional (3D) spatio-temporal representation of a sample can be obtained. Based on the 3D spatio-temporal representation, absolute particle velocities can be determined and / or absolute flow rates in flow channels associated with the particles. In an example, the particles are red and / or white blood cells and the flow channels are blood vessels.

Description

[0001] QUANTITATIVE OPTICAL COHERENCE TOMOGRAPHY ANGIOGRAPHY

[0002] CROSS REFERENCE TO RELATED APPLICATION

[0003] This application claims the benefit of U.S. Provisional Patent Application 63 / 605,079, filed December 1, 2024 which is incorporated herein by reference.

[0004] FIELD

[0005] The disclosure pertains to sample evaluation by optical coherence tomography.

[0006] BACKGROUND

[0007] Blood flow is inextricably linked to metabolic function and plays a role in many diseases. For example, in the retina, some diseases have a direct cause or effect on the blood supply such as diabetic retinopathy leading to angiogenesis and retinopathy of prematurity. Other diseases, like glaucoma, have a tight linkage between cellular mechanisms and capillary blood flow. For these reasons, different approaches to making blood flow measurements have been proposed or implemented. One approach to retinal blood flow imaging is optical coherence tomography angiography (OCTA). Unfortunately, OCTA (and other approaches) are unsuitable for making absolute blood flow measurements, particularly in capillaries. Such limitations of OCTA in retinal imaging are also disadvantages in imaging other biological or non-biological specimens, and improved approaches are needed, particularly those that can determine absolute particle velocities and flow rates in blood vessels and other flow channels.

[0008] SUMMARY

[0009] Methods, apparatus, and systems are disclosed that permit determination of particle velocity in a flow channel (such as red blood cells in a blood vessel) and absolute flow rate. In an example, methods for determining three-dimensional (3D) particle velocity comprise receiving a 3D spatiotemporal representation of a sample volume along with a corresponding OCTA volume, and, based on the 3D spatio-temporal representation and OCTA for vessel visualization, determining a velocity of particles in at least one flow channel in the sample volume. In typical examples, the 3D spatiotemporal representation of the sample volume is based on sequences of optical coherence tomography (OCT) scans in a first direction at consecutive second scan positions. The OCT scans can be obtained with an OCT system having a fast scan direction and a slow scan direction, wherein the first direction is the fast scan direction, and the second scan direction corresponds to the slow scan direction. In further examples, the velocity of the particles in at least one flow channel is determined by identifying associated streaks in the 3D spatio-temporal representation and measuring the flow channel morphology in the corresponding OCTA volume. The orientations of the associated streaks are obtained based on a three-dimensional Radon transform and flow channel morphology is characterized as angle and diameter. Typically, at least one flow channel in the sample volume is identified based on the 3D spatio-temporal representation to determine an orientation of the at least one morphological feature (e.g., retinal blood vessels).

[0010] In other examples, systems comprise an optical coherence tomography (OCT) system and a controller coupled to the OCT system and configured to acquire a 3D spatio-temporal representation of a sample by: (1) directing the OCT system to acquire sequences of scans along a first scan axis and process each scan to produce two-dimensional image data associated with a sample location along the first scan axis and a sample depth, and a sequence of scans is acquired at a respective location along a second scan axis, and (2) processing the 3D spatio-temporal representation to determine a velocity of at least one particle in the sample. In some examples, the processor is configured to identify a streak in the 3D spatio-temporal representation associated with the particle, wherein the velocity of the particle is determined based on the streak. In these and other examples, an orientation of the streak is identified based on a three-dimensional Radon transform. In still other variations, the processor is configured to identify at least one morphological feature of the sample associated with the particle based on the 3D spatio-temporal representation and determine an orientation and dimension of at least one morphological feature.

[0011] The foregoing and other features, and advantages of the disclosed technology will become more apparent from the following detailed description, which proceeds with reference to the accompanying figures.

[0012] BRIEF DESCRIPTION OF THE DRAWINGS

[0013] FIG. 1 illustrates a representative optical coherence tomography (OCT) system.

[0014] FIG. 2A illustrates acquisition of sequences of two-dimensional images having image data associated with a fast scan direction and a depth direction at different slow scan positions.

[0015] FIG. 2B illustrates a streak associated with a moving particle whose projections on XY and YZ planes produce spatio-temporal images.

[0016] FIG. 3 illustrates a representative method of determining absolute flow rate.

[0017] FIG. 4A illustrates a processed OCTA image.

[0018] FIG. 4B illustrates the image of FIG. 4A after Frangi filtering. FIG. 4C illustrates a 3D skeleton based on the image of FIG. 4B and identifying vessel centerlines.

[0019] FIG. 5A illustrates an OCT spatio-temporal image exhibiting streaks associated with particle movement.

[0020] FIG. 5B illustrates the OCT image of FIG. 5A after Sobel filtering.

[0021] FIG. 5C illustrates streak orientation identification based on a 3D Radon transform method.

[0022] FIG. 5D illustrates streak angle distribution across depth.

[0023] FIG. 6A illustrates a 3D Frangi-filtered image such as illustrated in FIG. 4B.

[0024] FIG. 6B illustrates a 3D velocity distribution associated with the image of FIG. 6A and which can be obtained with the disclosed methods and apparatus.

[0025] FIG. 6D illustrates a parabolic cross-sectional velocity profile within the vessel, which can be obtained at any lateral position along the vessel.

[0026] FIGS. 7A-7B illustrate blood cell velocities associated with various vessel sizes of a human retina.

[0027] FIG. 8 illustrates a method of determining absolute flow rates based on 3D spatio-temporal data obtained with OCT.

[0028] FIG. 9 illustrates a representative system for use in control of acquisition and processing of image data to obtain absolute velocities and / or flow rates.

[0029] FIG. 10 illustrates a representative geometry for description of OCT scan data acquisition.

[0030] FIGS. 11A-1 IB illustrate conservation of flow rates in a branched arteriole showed a close agreement with ~4% difference in flow rate between the parent and the sum of the daughter branches.

[0031] FIGS. 12A-12B illustrate pulsatile flow in microvasculature visualized by analyzing consecutive images of the same vessel of interest with modified slow scan parameters. Temporal resolution can be increased by trading off the spatial resolution in the slow scan direction to properly sample the human cardiac cycle (~60 beats / min).

[0032] FIGS. 13A-13D illustrate the ability to discriminate and quantify flow rate of overlapping vessels in depth which can be obtained with the disclosed methods and apparatus.

[0033] DETAILED DESCRIPTION

[0034] The disclosed methods and apparatus pertain to 3D tracking of particle movement in a specimen. In particular, the disclosed approaches can provide absolute particle velocities and absolute flow rates in flow channels associated with the particles, including directions of flow. In some examples, spatio-temporal streaks associated with particle motion are processed to determine streak angles to permit determination of particle velocity in three dimensions. As used herein, velocity refers to both speed and direction, generally in three dimensions. For convenient description, the methods and apparatus are described with reference to measurement of blood flow and movement of blood cells (both erythrocytes and leukocytes, more commonly referred to as red and white blood cells, respectively). Other particles and flows and flow channels can be similarly measured. If desired, an adaptive optics-enabled system (an AO-enabled system) can be used to increase resolution. Individual cells are observed. As noted above, characterization of retinal blood flow is only an illustrative example, and similar approaches can be used throughout the body or in other clinical or non-clinical applications.

[0035] General Terminology

[0036] Optical coherence tomography (OCT) uses multiple scans of a sample with an optical beam to produce a 3-dimensional image. Two-dimensional image data is obtained by scanning an optical beam with galvanometers, rotating prisms, resonant scanners, or other types of physical scanners that scan the optical beam over a sample in a first scan direction. The scanning permits acquisition of image data in the scan direction and optical coherence characteristics of the optical beam are used to produce image data in a depth or a depth dimension, typically, multiple scans in a first direction are made at multiple locations along a second scan direction to produce 3D image data. By acquiring a sequence of data associated with scans in the first direction, temporal variations in a sample can be evaluated in three dimensions, and the associated data referred to as 3D spatiotemporal data.

[0037] In OCT, depth data can be obtained with time domain OCT in which interference of an optical beam portion returned from a sample is mixed to interfere with a reference beam portion received along a reference beam path whose length is varied by scanning a reference reflector. Reference beam scanning tends to be impractically slow so that frequency domain OCT methods (spectral domain OCT and swept source OCT) are used in which the depth information of the tissue is encoded in the spectral characteristics of the OCT beam. These methods do not require physical scanning of a reference path and rapid frequency sampling of a suitable optical beam source can be achieved. Multiple depth scans along the fast scan associated with OCT are referred to as B -scans. Repeat B-scan data can be used to produce OCT angiographic data, and the application of OCT to produce angiographic data is referred to as OCTA. B-scan data collected in combination with lateral scanning at different slow scan positions produces C-scan or volumetric OCT data.

[0038] In OCTA, consecutive B-scans are acquired at the same location and a decorrelation signal between them based on differences in intensity and / or phase is used to generate angiographic data. There are several ways to generate angiographic data from OCT scan. Some algorithms use amplitude changes between scans while others use both amplitude and phase information to generate angiographic data.

[0039] As used herein, OCT systems are described generally as scanning a suitable optical beam (referred to herein as an OCT beam) along a first direction and then along a second direction to scan a sample region of interest. The first direction is typically associated with faster scanning than the second direction and is referred to as a fast scan direction while the other direction is referred to a slow scan direction. FIG. 10 illustrates a typical scanning sequence with reference to an XYZ- coordinate system 1000 in which an X- direction is referred to as a fast scan direction and a Y- direction is referred to as a slow scan direction. Depth data is associated with the Z-direction and is obtained based on the coherence characteristics of the OCT beam as discussed above. An OCT beam is shown at 1011-1015 as scanning a sample volume 1001 in XZ planes at respective Y- coordinates Y 1 - Y5 to produce image data as a function of X and Z for each Y -coordinate, thus producing a series of 2D images that can be combined to produce a 3D image. Such scans are typically referred to as “B-scans.” In FIG. 10, scanning is illustrated as raster scanning but other scan sequences such as vector scanning can be used. For convenient description, scanning in the first direction is referred to herein as a “fast scan” and the acquired data referred to as “fast scan data,” which corresponds to 2D image data (in an XZ-plane in FIG. 10). The fast scan data is typically arranged as image data but can be stored and arranged in other ways.

[0040] Fast scan data as a function of time can also be obtained. For each value of Y-coordinate, a sequence of XZ scans (B-scans) can be obtained as a function of a time interval between scans. The combination of these XZ sequences at multiple Y-coordinates produces time dependent, 3D image data, referred to herein as “a 3D spatio-temporal representation.” Because the sequences of fast scans at different locations are obtained at different times, optimal results are obtained with scan rates that are fast enough to sufficiently sample the decorrelation of any moving features of interest and typically such scans are performed as rapidly as the associated OCT scan hardware permits to achieve sufficient spatio-temporal sampling. If OCT scanning is sufficiently rapid, time-dependent scan data can be acquired with individual fast scans at different coordinates in a slow scan direction, and acquisition of multiple volumes for averaging is not needed. In addition, in some cases, delays between individual fast scans in a scan sequence (or between individual fast scans at the different slow scan coordinates) can be provided as may be suitable based on characteristics of a sample of interest.

[0041] In the following discussion, data acquired in response to a single scan (a fast axis scan at a fixed slow axis position) is referred to an image, a 2D image, a B-scan, or merely as scan data. The scan data can be stored in a processor readable memory or storage device as image data such as TIF, JPG or other format, or can be otherwise stored such as a data series. As used herein, image refers a visual presentation such as on a display device for viewing by technician or other user, or stored data used to produce such a visual presentation. In the examples, data is acquired as sequences of fast scans at multiple slow scan coordinates to take advantage of the higher temporal resolution available with a fast scan but other scan approaches can be used.

[0042] The disclosed approaches permit determination of flow velocity including flow direction and absolute flow rate over a cross-sectional area of a flow channel. Such flow velocities as a function of position in a flow channel cross-sectional area are referred to herein as “flow profiles” and can be determined for arbitrary profiles such as laminar flows such as laminar flows having a parabolic or other dependence on transverse coordinate in the cross-sectional area.

[0043] As used in this application and in the claims, the singular forms “a,” “an,” and “the” include the plural forms unless the context clearly dictates otherwise. Additionally, the term “includes” means “comprises.” Further, the term “coupled” does not exclude the presence of intermediate elements between the coupled items.

[0044] The systems, apparatus, and methods described herein should not be construed as limiting in any way. Instead, the present disclosure is directed toward all novel and non-obvious features and aspects of the various disclosed embodiments, alone and in various combinations and subcombinations with one another. The disclosed systems, methods, and apparatus are not limited to any specific aspect or feature or combinations thereof, nor do the disclosed systems, methods, and apparatus require that any one or more specific advantages be present or problems be solved. Any theories of operation are to facilitate explanation, but the disclosed systems, methods, and apparatus are not limited to such theories of operation.

[0045] Although the operations of some of the disclosed methods are described in a particular, sequential order for convenient presentation, it should be understood that this manner of description encompasses rearrangement, unless a particular ordering is required by specific language set forth below. For example, operations described sequentially may in some cases be rearranged or performed concurrently. Moreover, for the sake of simplicity, the attached figures may not show the various ways in which the disclosed systems, methods, and apparatus can be used in conjunction with other systems, methods, and apparatus. Additionally, the description sometimes uses terms like “produce” and “provide” to describe the disclosed methods. These terms are high- level abstractions of the actual operations that are performed. The actual operations that correspond to these terms will vary depending on the particular implementation and are readily discernible by one of ordinary skill in the art. In some examples, values, procedures, or apparatuses are referred to as “lowest”, “best”, “minimum,” or the like. It will be appreciated that such descriptions are intended to indicate that a selection among many used functional alternatives can be made, and such selections need not be better, smaller, or otherwise preferable to other selections nor to these terms imply any particular spatial orientation.

[0046] Example 1

[0047] Referring to FIG. 1, a simplified representative OCT system 100 includes a Fourier-domain mode-locked laser (“FDML”) source 102 (or other swept frequency source) that directs an optical beam to a first fiber coupler (FC) 104 that splits the light into two paths including a reference beam arm 106 and a sample beam arm 108. The reference beam arm 106 includes a retroreflector 112 and a stage 114 operable to adjust a reference optical path length. The retroreflector 112 returns the reference beam to a second fiber coupler 116. The sample beam arm 108 includes an adaptive optical element 120, a galvanometer scanner 122, and a resonant scanner (RS) 124 that deliver, shape, and scan a sample 126 with a sample beam. In this example, the sample 126 is a retina of a human subject. The resonant scanner 124 typically permits more rapid scanning than the galvanometer scanner 122 and these can be referred to as “fast” and “slow” scanners for convenience. In this example, a fixation and stimulus system 130 is situated so that the human subject can more readily stabilize the eye for measurement or perform functional measures of the eye.

[0048] A portion of the sample beam is returned from the sample 126 to the first fiber coupler 104 which directs a portion to the second fiber coupler 116. The second fiber coupler 116 provides portions of the reference beam and the returned sample beam to a balanced detector 132 that provides an electrical signal to a control and acquisition system 134 based on interference of the beam portions. A balanced detector tends to reduce common-mode variations in the detected electrical signal caused by variations in optical beam amplitude, but single detectors can be used. The first fiber coupler 104 and the second fiber coupler 116 can be selected to have split ratios as convenient; it is generally preferred to direct a relatively smaller amplitude optical beam to the sample 126 to maximize the coupling signal to be detected from the return beam back from the sample arm 108. In this example, the adaptive optical element is coupled to be controlled to shape the beam based on beam portions provided to a Shack-Hartmann wavefront sensor (SHWS) 140.

[0049] The control and acquisition system 134 is operable to control acquisition and processing of scan sequences to identify morphological features, alignment of such features, apply 3D Radon transforms, and determine absolute velocities and flow rates in various flow channels. The arrangement of FIG. 1 is abbreviated for clarity. Additional features of OCT systems are described in Liu et al., “Ultrahigh- speed multimodal adaptive optics system for microscopic structural and functional imaging of the human retina,” Biomedical Optics Express 13:5860-5878 (2022), which is incorporated herein by reference.

[0050] Example 2

[0051] FIG. 2A illustrates acquisition of a 3D spatio-temporal representation of a sample. In FIG. 2A, a sample volume 200 is irradiated by an optical beam 202 of an OCT system along a Z-axis (a depth direction) and the optical beam 202 is scanned in a fast scan direction (i.e., in an XZ-plane) with a fixed location in a slow axis (Y) direction. For example, the optical beam is scanned in an XZ plane 204 to obtain a 2D image that is a function of X and Z coordinates for a first value of Y- coordinate. To obtain temporal data, a sequence 206 of scans in the XZ planes, referred to in FIG. 2 A as B -scans, is acquired. FIG. 2A also shows an XZ plane 214 associated with a second slow scan coordinate Y2 and an associated sequence of XZ scans 216. Typically, sequences of fast scans (scans in XZ planes at a fixed Y coordinate) at multiple slow scan (Y) coordinates are obtained, producing spatial and temporal image data (a 3D spatio-temporal representation) associated with the volume 200.

[0052] Scanning can be unidirectional, bidirectional, or other scan patterns, and sequences can be obtained at fixed or variable time intervals. Particular coordinates and axes are provided for purpose of illustration in FIG. 2A. It will be understood that fast scans need not be limited to inplane scans, and in any case, deviations from planarity are typical in practical implementations.

[0053] Example 3

[0054] Referring to FIG. 3, a representative method 300 of determining absolute flow rates includes receiving or acquiring suitable OCT data such as discussed above at 302. The OCT data typically includes scan sequences associated with depth and fast axis scanning at multiple slow axis scan locations. At 304, the OCT data is subject to speckle variance or other processing to produce an OCTA volume and at 306, the OCTA data is Frangi filtered to aid in identification and quantification of flow channels such as blood vessels. At 308, vessel diameters are determined based on the processed OCTA data and at 310, vessel angular orientations are determined. If desired, both vessel diameter and angular orientations can be output. To establish absolute flow rates, the OCT data is processed at 320 to reduce noise and the reduced noise data is edge filtered at 322, as illustrated using a Sobel filter, but other edge filters can be used. The edge filtering enhances image streaks associated with particle (i.e., red blood cell) movement. At 324, a 3D Radon analysis is executed in which projections of streaks along multiple axes in three dimensions are used to establish particle directions of motion and 3D particle velocity. Axes associated with maximum (or relatively large) projected values are selected as corresponding to particle direction of motion. At 326, based on the determined particle directions of motion and vessel angular orientations obtained at 310, a 3D velocity or velocity distribution is obtained. While 3D spatiotemporal data alone permits identification and visualization of particle movement, particle direction of motion as determined at 324 is needed to provide absolute particle velocities at 326. With vessel diameters, orientations, and 3D velocity distributions, absolute flow rates are determined at 330.

[0055] Example 4

[0056] A modified 3D Radon transform is used in determining particle velocity. FIG. 2B illustrates a representative streak. The criteria for observing streaks include sufficient spatio-temporal resolution to resolve the target particles and a small-angle orientation of the flow channel relative to the fast scan direction, as detailed by Z. Zhong et a., “In vivo measurement of erythrocyte velocity and retinal blood flow using adaptive optics scanning laser ophthalmoscopy,” Optics Express, 16: 12746-12756 (2008).

[0057] To quantify individual particle velocity within each image, a gliding region of interest (ROI) defined by the longest streak length, is considered on two projection planes (XY and YZ). A circular mask on each ROI is implemented to avoid Radon artifacts. The Radon transform is implemented in an iterative manner for each ROI to achieve higher angle precision (step angle -0.001°) as described in P. Y. Chhatbar, and P. Kara, “Improved blood velocity measurements with a hybrid image filtering and iterative Radon transform algorithm,” Frontiers in Neuroscience, Volume 7, Article 106 (2013), which is incorporated herein by reference. The first iteration uses a large step angle across 0° to 180° with progressively smaller step angles to allow precise angle determination without longer computation times. The iterative Radon transform can achieve identical angle precision six times faster than the traditional Radon transform. The maximum variance angle provides the dominant streak orientation (0 in XY and cp in YZ). For computing absolute particle velocity, knowledge of the vessel orientation to the fast axis scan direction is required. This is derived from the segmented OCTA image by calculating the local gradient of the vessel centerline. Particle velocity can be calculated with the equation:

[0058] ^particle = [(^ * cot £* secj)xr+ [( / c * cot <p* sec / )rz] where, vparticieis the local individual particle velocity (mm / s), k is equal to the B-scan rate (Hz) multiplied by sampling density (pm / pixel) and x is the vessel orientation to the fast axis. In the equation the XZ plane projection is neglected as only the XY and YZ planes form space-time information whereas the XZ plane only forms space-space information.

[0059] Example 5

[0060] FIGS. 4A-4C, 5A-5D, and 6A-6D illustrate processing of the 3D spatio-temporal representations obtained with OCT as discussed above. FIGS. 4A-4C illustrate acquired and processed OCTA images obtained with an OCT system. FIG. 4A illustrates acquired data that is shown in FIG. 4B after application of a Frangi filter. In FIG. 4C, a centerline of the structures (vessels) of FIG. 4B is show, forming a 3D skeleton. FIG. 5A is a representative OCT intensity image as a function of image coordinate in slow and fast scan directions exhibiting streaks that can be associated with particle motion. FIG. 5B illustrates the image of FIG. 5A after edge filtering with a Sobel filter, in this example. FIG. 5C illustrates a portion 502 (in this case, a 20 pixel by 20 pixel portion in this example) containing a streak 503 shown as a stack of image portions 511-514 at different depths showing a streak center 516. Using a 3D Radon transformation, a streak angle with respect to the slow and fast scan directions as a function of depth can be obtained as shown in FIG. 5D. A streak angle at a center of a vessel is shown. Based on vessel dimensions and orientation and particle velocity based on streak angle as a function of depth and streak angle with respect to the fast and slow scan directions, a particle velocity and absolute flow rate can be determined in 3D. FIG. 6A illustrates a representative image of a sample volume 602 that includes an image portion 604 corresponding to a blood vessel such as illustrated in FIG. 4B. FIG. 6B shows a 3D velocity distribution and FIG. 6C shows a 3D flow map associated with the sample volume of FIG. 6A. FIG. 6D shows measured cross-sectional parabolic flow velocity distribution across the blood vessel.

[0061] Example 6

[0062] FIG. 7 A is a color fundus image of a human retina, one type of sample for which the disclosed methods and apparatus can be used. FIG. 7B is a graph of flow velocity as a function of vessel diameter. As shown in FIG. 7B, there is a substantial range of both cell velocities and vessel diameters, any and all of which are within ranges suitable for the disclosed approaches.

[0063] Example 7

[0064] Referring to FIG. 8, a representative method 800 includes selecting a number of images to be included in an OCT sequence by scanning in a fast scan direction at a fixed slow scan coordinate at 802. In a fast scan, depth data is acquired via optical beam coherence so that each of the images in a sequence is a function of position in a fast scan direction and a depth direction. At 804, a sequence of such images is acquired at a first slow scan coordinate. Each of these images is then a function of a fast scan coordinate and a depth coordinate. Typically, the number of images to be acquired is fixed and used at all slow scan coordinates, but different numbers can be used at different locations. In some cases, a number of particle movements are adequately captured. At 806, image acquisition at additional slow scan locations can be selected and the slow scan location incremented at 808, and a sequence of scans performed at the new slow scan location at 804.

[0065] Upon completion of scanning, morphological features of a sample based on the acquired scans can be obtained at 812. Such features include sizes, shapes, and orientations of structures such as blood vessels or other sample structures. In addition, at 814, streaks can be identified in the OCT scans and evaluated along with vessel orientation at 812 to determine particle velocities. Based on the particle velocities and morphological features, absolute flow rates are determined at 816.

[0066] Example 8

[0067] FIG. 9 illustrates a generalized example of a suitable computing environment 900 in which aspects of the described embodiments can be implemented. The computing environment 900 is not intended to suggest any limitation as to the scope of use or functionality of the disclosed technology, as the techniques and tools described herein can be implemented in diverse general- purpose or special-purpose environments that have computing hardware.

[0068] With reference to FIG. 9, the computing environment 900 includes at least one processing device 910 and memory 920. The processing device 910 (e.g., a central processing unit (CPU), a graphics processing unit (GPU), microprocessor, complex-programmable logic device (CPLD) or dedicated or general-purpose logic device) executes computer-executable instructions. In a multiprocessing system, multiple processing devices execute computer-executable instructions to increase processing power. The memory / storage 920 may be any variety of non-transitory storage device including volatile memory (e.g., registers, cache, RAM, DRAM, SRAM), non-volatile memory (e.g., ROM, EEPROM, flash memory), or some combination of the two. The memory / storage 920 stores processor-executable instructions for performing any of the disclosed methods and steps in providing velocity and absolute flow data in three dimensions. As shown in FIG. 9, portions of the memory / storage 920 are provided with processor-executable instructions, image or scan data, scan configuration information, or other data or instructions such as fast and slow scanning control 981, slow scan location increments at 982, number of fast scans in a scan sequence 983, OCT depth determination at 984, 3D Radon transformations at 985, morphological feature identification and quantification at 986, velocity and absolute flow determination at 987, and storage of scan data generally at 988. The computing environment 900 is coupled to an OCT system to direct scanning and to acquire OCT scan data but OCT scan data can be provided to other processing systems for sample evaluations, and OCT system control and OCT scan data processing need not be performed by the same processing systems.

[0069] The computing environment 900 can have additional features. For example, the computing environment 900 includes one or more input devices 950, one or more output devices 960, and one or more communication connections 970. An interconnection mechanism (not shown), such as a bus, controller, or network, interconnects the components of the computing environment 900. Typically, operating system software (not shown) provides an operating environment for other software executing in the computing environment 900, and coordinates activities of the components of the computing environment 900.

[0070] The memory / storage 920 can be removable or non-removable, and can include one or more magnetic disks (e.g., hard drives), solid state drives (e.g., flash drives), or any other tangible nonvolatile storage medium which can be used to store information and which can be accessed within the computing environment 900.

[0071] The input device(s) 950 can be a touch input device such as a keyboard, touchscreen, mouse, pen, trackball, a voice input device, a scanning device, or another device that provides input to the computing environment 900. The output device(s) 960 can be a display device (e.g. , a computer monitor, laptop display, smartphone display, tablet display, netbook display, or touchscreen), printer, speaker, or another device that provides output from the computing environment 900.

[0072] Example 9

[0073] FIGS. 11A-11B and FIGS. 12A-12B illustrate typical measurement results obtained with the disclosed methods. FIGS. 11A-11B illustrate conservation of flow rates in a branched arteriole comprising a parent vessel 1 that branches into daughter vessel 2 and daughter vessel 3. The flow rates summarized in the table below for flow in the parent vessel 1, the daughters 2-3, and the combined flow for both daughters (shown in the last row) show close agreement with ~4% difference in flow rate between the parent and the sum of the daughter branches.

[0074] Flow Rates for the Branched Arteriole of FIGS. 11 A- 1 IB

[0075] FIGS. 12A-12B illustrate pulsatile flow in microvasculature visualized by measuring velocity over a cardiac cycle. Proper pulsatile flow can be captured by increasing system temporal resolution (e.g., increasing volume rate), which is achieved by strategically trading off the spatial resolution of the imaging system. FIG. 12A illustrates images of the same vessel with pixel densities (which determine the digital resolutions) of 512 by 512 pixels, 256 by 512 pixels, 128 by 512 pixels, and 64 by 512 pixels as indicated. The higher the pixel density, the longer the time required for image acquisition. In this example, two images with 512 by 512 pixels can be acquired in 1.25 s, while 16 images with 64 by 512 pixels can be acquired in the same time. However, processing any of these images permits successful measurement of flow at a single point of the flow velocity plot shown in FIG. 12B in a single cardiac cycle, with the lower resolutions permitting superior measurement of temporal flow variations at various times in the cardiac cycle.

[0076] FIGS. 13A-13C illustrate the capability to discriminate and quantify overlapping blood vessels in depth and the type of blood vessel (artery or vein) and direction of flow based on the streak orientation. In these examples, blood flow rates and blood cell velocities are measured, but the other flow rates and velocities of other particles can be assessed in the same manner. FIG. 13A illustrates an overlapping arteriole and venule and includes the corresponding spatio-temporal representation of the region with overlapped vessels 1302. FIG. 13B illustrates the different streak orientations associated with each vessel and FIG. 13C is a volumetric view illustrating the venule (1 ) and the arteriole (2) of FIG. 13 A. FIG. 1 D illustrates flow rate as a function of depth associated with FIGS. 13A-13C in which flow in the venule and arteriole are distinguished and quantified.

[0077] Additional Examples

[0078] In some examples, an imager for optical coherence tomography (OCT) with high B-scan acquisition rate is used to resolve particle motion. In some examples, the imager includes adaptive optics (AO) for dynamically correcting wavefront aberration introduced by the optics of the imaging target and the imaging system itself, providing spatial resolution to visualize one or more moving particles. In these and other embodiments, the imager includes a computer to operate and acquire 3D spatio-temporal OCT scans in real time. In a further embodiment, the imager includes a multi-threaded CPU with multiple cores and a GPU adapted for parallel computation. In yet another embodiment, the pre-processing of acquired scan data is performed to correct non-linearity of the fast scanner, the slow scanner, and sample motion in one or more scans. In yet another embodiment, the calculation of an OCT angiography (OCTA) volume includes an angiography analysis algorithm to differentiate static tissue from moving particles for vessel visualization in the 3D spatio-temporal representation. In yet another embodiment, generating an OCTA image includes extracting morphological features such as vessel centerline, vessel diameter, and vessel orientation in 3D. In yet additional examples, the velocity determination step includes denoising scan data by thresholding in a shearlet domain to improve signal-to-noise ratio (SNR). In typical examples, velocity determination includes application of a modified 3D Radon transform to determine an orientation of one or more streaks associated with particle motion. In these and other examples, velocity determination includes computation of particle velocity based on a corresponding vessel orientation to the fast scan direction. Another example further includes absolute flow rate determination based on calculation of flow rate from the velocity measurement and the vessel morphology.

[0079] In still further examples, at least a portion of the 3D spatio-temporal representation and a representation of particle velocity and flow rate associated with multiple vessels are displayed. 3D visualization of velocity and flow rate distribution overlayed with an image of the pre-processed volume. The 3D visualization of velocity can include a depth-resolved velocity profile within the vessel and / or a plot of measured particle velocity along a length of the vessel to demonstrate pulsatile flow. In yet another embodiment, the displaying can include exhibiting a dynamic range of measured particle speed and flow rate in vessels over a wide size range including capillaries.

[0080] Representative Disclosure Clauses

[0081] Clause 1 is a method, including: receiving a 3D spatio-temporal representation of a sample volume; and based on the 3D spatio-temporal representation, determining absolute flow rates in a plurality of flow channels in the sample volume.

[0082] Clause 2 includes the subject matter of Clause 1, and further specifies that the flow channels of the plurality of flow channels are distributed throughout the sample volume in at least one spatial direction. Clause 3 includes the subject matter of any of Clauses 1-2, and further includes determining a flow velocity associated with at least one of the flow channels.

[0083] Clause 4 includes the subject matter of any of Clauses 1-3, and further includes determining flow velocity associated with each of the plurality of flow channels.

[0084] Clause 5 includes the subject matter of any of Clauses 1-4, and further specifies that the flow channels of the plurality of flow channels have effective diameters between 10 pm and 150 pm.

[0085] Clause 6 includes the subject matter of any of Clauses 1-5, and further specifies that the flow channels are blood vessels.

[0086] Clause 7 includes the subject matter of any of Clauses 1-6, and further specifies that the determined absolute flow rates are associated with pulsatile blood flow in a single cardiac cycle.

[0087] Clause 8 includes the subject matter of any of Clauses 1-7, and further includes determining absolute flow rate as a function of cross-section.

[0088] Clause 9 includes the subject matter of any of Clauses 1-8, and further specifies that a direction of flow discriminating arterial and venous flow in at least one of the plurality of flow channels is determined based on a streak angle associated with flow in the associated flow channel and a scan location.

[0089] Clause 10 includes the subject matter of any of Clauses 1-9, and further specifies that the 3D spatio-temporal representation of the sample volume is based on sequences of optical coherence tomography (OCT) scans in a first direction at a multiple of second scan positions.

[0090] Clause 11 includes the subject matter of any of Clauses 1-10, and further specifies that the OCT scans are obtained with an OCT system having a fast scan direction and a slow scan direction, wherein the first direction is the fast scan direction and the second scan positions correspond to positions along the slow scan direction.

[0091] Clause 12 includes the subject matter of any of Clauses 1-11, and further specifies that velocities of particles in the plurality of flow channels are determined by identifying associated streaks in the 3D spatio-temporal representation.

[0092] Clause 13 includes the subject matter of any of Clauses 1-12, and further specifies that orientations of a plurality of associated streaks are obtained based on a three-dimensional Radon transform.

[0093] Clause 14 includes the subject matter of any of Clauses 1-13, and further includes: identifying morphological features in the sample volume associated with flow channels based on the 3D spatio-temporal representation; and determining orientations and diameters of the morphological features. Clause 15 includes the subject matter of any of Clauses 1-14, and further specifies that the absolute flow rates in the flow channels are based on velocities of the particles and the determined orientations.

[0094] Clause 16 includes the subject matter of any of Clauses 1-15, and further specifies that the morphological features are flow channel centerlines, flow channel dimensions, and flow channel orientations.

[0095] Clause 17 includes the subject matter of any of Clauses 1-16, and further includes determining absolute flow rate as a function of cross-sectional area for at least one of the plurality of flow channels.

[0096] Clause 18 includes the subject matter of any of Clauses 1-17, and further includes determining flow velocity as a function of cross-sectional area for at least one of the plurality of flow channels.

[0097] Clause 19 includes the subject matter of any of Clauses 1-18, and further specifies that the flow channels are blood vessels and the flow rates are absolute blood flow rates associated with the blood vessels.

[0098] Clause 20 includes the subject matter of any of Clauses 1-19, and further specifies that the morphological features include at least one of blood vessel centerlines, blood vessel dimensions, and blood vessel orientations.

[0099] Clause 21 includes the subject matter of any of Clauses 1-20, and further specifies that the 3D spatio-temporal representation includes a series of three-dimensional images of the sample volume as a sequence of optical coherence tomography (OCT) images, wherein each sequence includes two or more two-dimensional images separated by a time interval, each of the two- dimensional images include image data associated with a sample depth and location.

[0100] Clause 22 includes the subject matter of any of Clauses 1-21, and further specifies that each of the sequences includes at least two two-dimensional images that include image data associated with sample position in a fast scan direction.

[0101] Clause 23 is a system, including: an optical coherence tomography (OCT) system; a processor coupled to the OCT system and configured to acquire a 3D spatio-temporal representation of a sample by: directing the OCT system to acquire a sequence of scans along a first scan axis and process each scan to produce two-dimensional image data associated with a sample location along the first scan axis and a sample depth, wherein each sequence of scans is acquired at a respective location along a second scan axis; and processing the 3D spatio-temporal representation to determine velocities of particles in the sample. Clause 24 includes the subject matter of Clause 23, and further specifies that the sample includes at least one blood vessel, and multiple 3D spatio-temporal representation is acquired in sequence to fully sample a single cardiac cycle.

[0102] Clause 25 includes the subject matter of any of Clauses 23-24, and further specifies that each of the sequences includes a predetermined number of scans along the second direction.

[0103] Clause 26 includes the subject matter of any of Clauses 23-25, and further specifies that the predetermined number of scans is selected based on a duration of the single cardiac cycle.

[0104] Clause 27 includes the subject matter of any of Clauses 23-26, and further specifies that the processor is configured to identify streaks in the 3D spatio-temporal representation associated with particles, wherein the velocities of the particles are determined based on the streak orientation.

[0105] Clause 28 includes the subject matter of any of Clauses 23-27, and further specifies that an orientation of streaks is identified based on a three-dimensional Radon transform.

[0106] Clause 29 includes the subject matter of any of Clauses 23-28, and further specifies that the processor is configured to identify at least one morphological feature of the sample associated with the particles based on OCT angiography (OCTA) and determine the orientation of the at least one morphological feature.

[0107] Clause 30 includes the subject matter of any of Clauses 23-29, and further specifies that at least one morphological feature defines a flow channel associated with the particles, and further includes determining a flow rate in the flow channel based on the velocities of the particles and a flow channel orientation.

[0108] Clause 31 includes the subject matter of any of Clauses 23-30, and further specifies that the at least one morphological feature is a flow channel centerline, a flow channel dimension, and a flow channel orientation.

[0109] Clause 32 includes the subject matter of any of Clauses 23-31, and further specifies that the morphological feature is a blood vessel, the particle is a blood cell, and the flow rate is a blood flow rate associated with the blood vessel.

[0110] Clause 33 includes the subject matter of any of Clauses 23-32, and further specifies that the morphological feature is at least one of a blood vessel centerline, a blood vessel dimension, and a blood vessel orientation.

[0111] Clause 34 is a method, including: with an OCT system, for each of multiple positions along a second scan direction, scan an OCT optical beam across a sample along a first scan direction to obtain a sequence of a selected number of two-dimensional images that are functions of sample depth in a sample along an optical beam propagation direction and along an axis parallel to the first scan direction; and based on the sequences of two-dimensional images, determining at least one morphological feature of the sample associated with particles and velocities of the particles.

[0112] Clause 35 includes the subject matter of Clause 34, and further specifies that the particles are a plurality of blood cells, the morphological feature of the sample are a blood vessel and an orientation of the blood vessel, wherein the velocities of the particles are determined based on orientation of associated streaks.

[0113] Clause 36 includes the subject matter of any of Clauses 34-35, and further specifies that the orientations of the streaks are identified based on a modified three-dimensional Radon transform. Other aspects and advantages of the disclosed technology will become apparent from the following drawings, detailed description, and claims, all of which illustrate the principles of the disclosed technology, by way of example only.

Claims

We claim:

1. A method, comprising: receiving a 3D spatio-temporal representation of a sample volume; and based on the 3D spatio-temporal representation, determining absolute flow rates in a plurality of flow channels in the sample volume.

2. The method of claim 1 , wherein the flow channels of the plurality of flow channels are distributed throughout the sample volume in at least one spatial direction.

3. The method of claim 1, further comprising determining a flow velocity associated with at least one of the flow channels.

4. The method of claim 3, further comprising determining flow velocity associated with each of the plurality of flow channels.

5. The method of claim 1, wherein the flow channels of the plurality of flow channels have effective diameters between 10 pm and 150 pm.

6. The method of claim 5, wherein the flow channels are blood vessels.

7. The method of claim 6, wherein the determined absolute flow rates are associated with pulsatile blood flow in a single cardiac cycle.

8. The method of claim 1, further comprising determining absolute flow rate as a function of cross-section.

9. The method of claim 1, wherein a direction of flow discriminating arterial and venous flow in at least one of the plurality of flow channels is determined based on a streak angle associated with flow in the associated flow channel and a scan location.

10. The method of claim 1, wherein the 3D spatio-temporal representation of the sample volume is based on sequences of optical coherence tomography (OCT) scans in a first direction at a multiple of second scan positions.

11. The method of claim 10, wherein the OCT scans are obtained with an OCT system having a fast scan direction and a slow scan direction, wherein the first direction is the fast scan direction and the second scan positions correspond to positions along the slow scan direction.

12. The method of claim 1, wherein velocities of particles in the plurality of flow channels are determined by identifying associated streaks in the 3D spatio-temporal representation.

13. The method of claim 12, wherein orientations of a plurality of associated streaks are obtained based on a three-dimensional Radon transform.

14. The method of claim 1, further comprising: identifying morphological features in the sample volume associated with flow channels based on the 3D spatio-temporal representation; and determining orientations and diameters of the morphological features.

15. The method of claim 12, wherein the absolute flow rates in the flow channels are based on velocities of the particles and the determined orientations.

16. The method of claim 14, wherein the morphological features are flow channel centerlines, flow channel dimensions, and flow channel orientations.

17. The method of claim 14, further comprising determining absolute flow rate as a function of cross-sectional area for at least one of the plurality of flow channels.

18. The method of claim 14, further comprising determining flow velocity as a function of cross-sectional area for at least one of the plurality of flow channels.

19. The method of claim 14, wherein the flow channels are blood vessels and the flow rates are absolute blood flow rates associated with the blood vessels.

20. The method of claim 14, wherein the morphological features include at least one of blood vessel centerlines, blood vessel dimensions, and blood vessel orientations.

21. The method of claim 1, wherein the 3D spatio-temporal representation includes a series of three-dimensional images of the sample volume as a sequence of optical coherence tomography (OCT) images, wherein each sequence includes two or more two-dimensional images separated by a time interval, each of the two-dimensional images include image data associated with a sample depth and location.

22. The method of claim 10, wherein each of the sequences includes at least two two- dimensional images that include image data associated with sample position in a fast scan direction.

23. A system, comprising: an optical coherence tomography (OCT) system; a processor coupled to the OCT system and configured to acquire a 3D spatio-temporal representation of a sample by: directing the OCT system to acquire a sequence of scans along a first scan axis and process each scan to produce two-dimensional image data associated with a sample location along the first scan axis and a sample depth, wherein each sequence of scans is acquired at a respective location along a second scan axis; and processing the 3D spatio-temporal representation to determine velocities of particles in the sample.

24. The system of claim 23, wherein the sample includes at least one blood vessel, and multiple 3D spatio-temporal representation is acquired in sequence to fully sample a single cardiac cycle.

25. The system of claim 24, wherein each of the sequences includes a predetermined number of scans along the second direction.

26. The system of claim 25, wherein the predetermined number of scans is selected based on a duration of the single cardiac cycle.

27. The system of claim 23, wherein the processor is configured to identify streaks in the 3D spatio-temporal representation associated with particles, wherein the velocities of the particles are determined based on the streak orientation.

28. The system of claim 27, wherein an orientation of streaks is identified based on a three- dimensional Radon transform.

29. The system of claim 28, wherein the processor is configured to identify at least one morphological feature of the sample associated with the particles based on OCT angiography (OCTA) and determine the orientation of the at least one morphological feature.

30. The system of claim 25, wherein at least one morphological feature defines a flow channel associated with the particles, and further comprising determining a flow rate in the flow channel based on the velocities of the particles and a flow channel orientation.

1. The system of claim 30, wherein the at least one morphological feature is one or more of a flow channel centerline, a flow channel dimension, and a flow channel orientation.

32. The system of claim 30, wherein the morphological feature is a blood vessel, the particle is a blood cell, and the flow rate is a blood flow rate associated with the blood vessel.

33. The system of claim 32, wherein the morphological feature is at least one of a blood vessel centerline, a blood vessel dimension, and a blood vessel orientation.

34. A method, comprising: with an OCT system, for each of multiple positions along a second scan direction, scan an OCT optical beam across a sample along a first scan direction to obtain a sequence of a selected number of two-dimensional images that are functions of sample depth in a sample along an optical beam propagation direction and along an axis parallel to the first scan direction; and based on the sequences of two-dimensional images, determining at least one morphological feature of the sample associated with particles and velocities of the particles.

35. The method of claim 34, wherein the particles are a plurality of blood cells, the morphological feature of the sample are a blood vessel and an orientation of the blood vessel, wherein the velocities of the particles are determined based on orientation of associated streaks.

36. The method of claim 35, wherein the orientations of the streaks are identified based on a modified three-dimensional Radon transform.

Images