Systems and methods for ultrasound perfusion imaging

Abstract

A perfusion imaging processing method includes collecting multiple digital images containing continuous B-mode ultrasound reflectance data, calculating a decorrelation trend of autocorrelation data to determine blood flow and perfusion levels, using the decorrelation trend to reduce noise content in the data, and / or forming a visual representation based on the decorrelation trend. This system and method provide an ultrasound imaging system and method that provides perfusion data without the need for injected contrast agents.

Description

Technical Field

[0001] This invention relates to perfusion imaging of in vivo tissues using ultrasound. Specifically, this invention relates to a method and system for non-invasive perfusion imaging of living tissues using ultrasound and noise reduction of ultrasound images. Background Technology

[0002] Dynamic imaging of physiological data is used in medicine to detect abnormalities in tissue perfusion, such as diagnosing acute stroke, brain tumors, or for tumor detection and classification. Ultrasound imaging or ultrasound examination is used to image living tissue in animals and humans. Diagnostic ultrasound, also known as ultrasound examination, is an ultrasound-based diagnostic imaging technique used to visualize a patient's subcutaneous body structures, such as tendons, muscles, joints, blood vessels, and internal organs, to understand possible conditions or lesions. Ultrasound images, or sonographs, are created by sending ultrasound pulses into tissues using an ultrasound transducer or probe. Sound reflects off parts of the tissue and produces echoes, which are recorded and displayed as images to the operator of the medical imaging system. Typically, in sonograph images, denser tissues appear as bright areas, and less dense tissues appear as darker areas. The most common form of ultrasound image is the B-mode image (luminance mode), which shows changes in acoustic impedance across a two-dimensional or three-dimensional cross-section of the tissue.

[0003] Doppler ultrasound is used to study blood flow, including flow direction and velocity, as well as muscle movement. In Doppler images, different detected flow velocities and movement speeds are typically represented by color for easier interpretation. For example, in a leaky heart valve, the leak appears as a flash of a unique color. Ultrasound contrast agents containing encapsulated gaseous microbubbles can be used to enhance echogenicity in tissues, or to improve the ability to generate echoes. Intravenous injection of the contrast agent introduces the gaseous microbubble contrast agent into the systemic circulation, and the enhanced echogenicity provides enhanced images for blood perfusion imaging of organs. Perfusion imaging provides information about physiological tissue behavior, such as blood volume, blood flow, mean transit time (MTT), and peak time (TTP) in vascularized tissue. Although the minimum flow velocity and vessel size for energy Doppler are slightly lower than for color Doppler, blood flow in tissues is difficult or likely to be undetectable without contrast agents unless the vessels exceed a certain minimum size, such as in the heart, where the vessels are advantageously oriented towards the imaging plane and have a relatively high flow velocity. Unlike color flow imaging, which measures blood flow rate, it is preferable to measure perfusion levels within biological tissues.

[0004] In perfusion imaging, a contrast agent is typically injected, and subsequent images covering the target volume are repeatedly acquired after its distribution. The contrast agent acts as a blood tracer and provides signal changes to indicate blood flow. Depending on the actual physiological process, the short-term distribution of blood flow (perfusion) (less than 1 minute) or the long-term (greater than 1 minute) diffusion process of tracer particles in the microvascular membrane (tissue dynamics) is encoded in the changing signals of the image voxels. The time-intensity curve of each extracted voxel is typically converted into a relative concentration-time curve. Conventional ultrasound resolution limits the minimum side length of voxels to greater than 150 micrometers.

[0005] High-frequency ultrasound (also known as micro-ultrasound) is becoming a valuable diagnostic technique due to the development of high-frequency ultrasound array transducers. In a micro-ultrasound system, the transducer generates sound waves in the range of 15 to 80 MHz, which then propagate through living tissue, reflect these sound waves, and then return to the transducer. The sound waves are then converted into two-dimensional or three-dimensional images. A benefit of high-frequency ultrasound is its ability to image small voxels, which improves image resolution.

[0006] In one example of ultrasound imaging, U.S. Patent 9,955,941 to Rafter et al. describes an ultrasound diagnostic imaging system that scans multiple planar slices within a volumetric region containing tissue that has been perfused with contrast agent. After detecting the image data, the slice data is combined by projecting the data along an elevation dimension to produce an elevation-combined slice image, and the image data is combined using an average or maximum intensity detection or weighting process or by ray projection along the elevation dimension during volumetric rendering. The combined slice image provides a measure of perfusion.

[0007] There is still a need for an ultrasound imaging system and method that can provide perfusion data without the need for contrast agents.

[0008] This background information is provided to disclose information that the applicant believes may be relevant to the present invention. It is not intended to be an admission, nor should it be construed as, that any of the foregoing information constitutes prior art to the present invention. Summary of the Invention

[0009] One object of the present invention is to provide an ultrasound imaging system and method that provides perfusion data without the need for contrast agents.

[0010] In one aspect, an imaging processing method is provided, comprising: processing multiple digital images containing continuous B-mode ultrasound reflectance data, and calculating decorrelation trends of autocorrelation data to determine flow and perfusion levels.

[0011] In one embodiment, the method uses a decorrelation trend to reduce noise content in ultrasound data. In another embodiment, the method forms a visual representation of perfusion levels based on the decorrelation trend. In another embodiment, the method uses a decorrelation trend to reduce noise content in autocorrelation data and forms a visual representation of perfusion levels based on the decorrelation trend. In another embodiment, the method has an image capture rate of 20 frames / second or higher. In another embodiment, the image data is normalized before autocorrelation. In another embodiment, the autocorrelation data is normalized. In another embodiment, the decorrelation trend of the autocorrelation data is determined by linear regression, mean difference, or change in overall amplitude. In another embodiment, autocorrelation data with a nonlinear trend is processed to determine one or more of the following: decorrelation period, exponential decay rate, time to local minimum, or other measures. In another embodiment, Spearman correlation, Pearson correlation, or Fourier transform is used to calculate the autocorrelation. In another embodiment, the method further includes a smoothing step before the visual representation of the signal. In another embodiment, the method further includes applying a threshold and rescaling to the decorrelation trend before the visual representation. In another embodiment, the method further includes a logarithmic transformation of the decorrelation trend. In another embodiment of the method, decorrelation trends are mapped to different color or grayscale values ​​for representation in the image. In another embodiment, the method further includes aligning multiple digital images to correct for movement by matching local signal patterns between frames. In another embodiment, the method includes at least five consecutive frames of mode B ultrasound reflectance data. In another embodiment of the method, a subset of the image field is processed from each image to reduce processing time. In another embodiment of the method, the images are downsampled to a lower resolution to reduce processing time. In another embodiment of the method, the ultrasound data is high-frequency ultrasound with a frequency range greater than 15 MHz.

[0012] On the other hand, a perfusion imaging system is provided, comprising a high-frequency ultrasound transducer for capturing and / or collecting multiple digital images containing continuous B-mode ultrasound reflectance data; and a signal processing unit operatively connected to the transducer, the signal processing unit being configured to calculate the decorrelation trend of autocorrelation data to determine blood flow and perfusion levels.

[0013] On the other hand, a computer-readable storage medium is provided, comprising executable instructions that, when executed by a process, cause a processor to: process multiple digital images containing sequential B-mode ultrasound reflectance data obtained using microsound, and calculate decorrelation trends of autocorrelation data to determine blood flow and perfusion levels. Attached Figure Description

[0014] To better understand the invention and other aspects and further features, reference is made to the following description used in conjunction with the accompanying drawings, wherein:

[0015] Figure 1 A high-frequency medical imaging system is shown;

[0016] Figure 2A and 2B These are example ultrasound images before and after denoising by eliminating low-correlation pixels;

[0017] Figure 3 An example is shown comparing the perfusion signal from the blood vessel with the signal from a noisy region in the image;

[0018] Figure 4 Examples of signals from misaligned and aligned images are shown;

[0019] Figure 5A -C is an example of forearm muscle activity before and after; and

[0020] Figure 6 This is a flowchart illustrating an example method of ultrasound perfusion imaging. Detailed Implementation

[0021] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0022] As used in the specification and claims, the singular forms “a,” “an,” and “the” include plural references unless the context clearly specifies otherwise.

[0023] As used herein, the term “comprising” will be understood to mean that what follows is not exhaustive and may include or exclude any other suitable additional items, such as one or more additional features / components and / or elements (as the case may be).

[0024] This article provides an ultrasound modality for tissue perfusion imaging. Specifically, the system and method provide an ultrasound imaging system and method that delivers perfusion data without the need for injected contrast agents. By utilizing certain features of the autocorrelation sequences in the collected ultrasound images, blood perfusion can be detected and distinguished from noise in the ultrasound signal. In this way, perfusion within biological tissues can be measured without the need for contrast agents.

[0025] Blood perfusion in tissue can be detected and imaged by examining time-decorrelated signals at the pixel level in micro-ultrasound images configured for very high frame rates. This is used, for example, in techniques such as plane wave imaging. At the high frame rates and small voxel sizes provided by micro-ultrasound, the movement of each reflector in and out of the sensitive detectable region causes slow changes over time. The detection of these changes over time, as a result of blood perfusion in the tissue, contrasts with noise that is completely uncorrelated over time and solid tissue that is correlated over time. Perfusion may be more difficult to measure in this way with conventional ultrasound because the sensitive region (voxel) in the tissue is too large, and too many reflectors move in and out simultaneously, causing intensity changes to be averaged out. Furthermore, conventional ultrasound is insensitive to individual reflectors in the blood, while high-frequency micro-ultrasound is sensitive to them. In another embodiment, perfusion can be predicted to be measured using conventional ultrasound by using another reflector. For example, large proteins or macrophages can serve as reflectors for conventional ultrasound.

[0026] Unbound by theory, we assume that cells in the bloodstream act as endogenous contrast agents and can be detected using appropriate signal processing as described herein. Specifically, red blood cells (erythrocytes) are the most abundant cells in the blood, comprising approximately 40% to 45% of its volume. Red blood cells have a diameter of approximately 6–8 micrometers, within the same size range as ultrasound microbubble contrast agents, and possess sufficient echogenicity to be detected by micro-ultrasound. Capillary blood flow rates are approximately less than 1 mm / s, or about 0.03 cm / s, and perfusion at these rates can be detected using high-frequency micro-ultrasound with the imaging modalities currently described. This is in comparison to the much faster arterial blood flow of tens of millimeters per second, which can be imaged using conventional low-frequency ultrasound (e.g., echocardiography). Capillaries have a diameter of approximately 5–10 micrometers (μm), so small that red blood cells typically pass through them in a single file. The average density of capillaries in human tissue is approximately 600 / mm³, meaning the average spacing between adjacent capillaries is approximately 40 micrometers. The use of contrast agents has limitations, such as the need for injection, potential censorship in all jurisdictions, potential censorship for all uses, and the possibility that contrast agents may break down or dissipate over time. Contrast agents work in ultrasound because they are filled with gas and produce a non-linear response due to their oscillating manner.

[0027] If the frame rate used for image acquisition is not fast enough, the time between frames becomes too long, resulting in too many reflectors entering and exiting voxels simultaneously. In this case, averaging occurs, which confuses the correlation of the signal. Therefore, the frame rate must be at least as fast as the reflector movement to ensure that blood perfusion of tissue can be captured. In one embodiment, the frame rate is approximately at least greater than or equal to (≥) 30 s. -1 The frame rate. In another embodiment, the frame rate is approximately at least greater than or equal to (>) 20s. -1The frame rate also requires motion compensation.

[0028] Furthermore, the size of the voxels or sensitive regions imaged in the tissue must be small enough to observe and detect one or more reflectors as they are perfused within the voxel (i.e., such that portions of the reflectors are present in the same voxel across frames). In one embodiment, the voxel size is approximately 70 micrometers. A voxel size of approximately 70 micrometers corresponds to high-frequency ultrasound greater than or equal to (≥) 15 MHz. Those skilled in the art understand that voxel size varies with ultrasound frequency.

[0029] Increased blood flow in tissues can occur during exercise, and increased angiogenesis due to angiogenesis is characteristic of many cancers and tumors. Under conventional ultrasound, blood appears dark, while under high-frequency ultrasound, without signal decorrelation, blood appears as noise. It has been found that increased capillary angiogenesis can be visualized by altering micro-ultrasound imaging settings, allowing for sufficiently rapid image acquisition to monitor changes in the ultrasound signal by limiting the number of scan lines and the focus area, and by isolating the decorrelation signal characteristics of capillary blood flow. Specifically, it has been found that signals previously observed as noise from vascularized tissue can be identified as echo-reflected blood by observing how noise in high-frequency ultrasound changes over time at specific locations. Therefore, this technique enables the identification and imaging of perfusion areas within tissues.

[0030] Unlike Doppler ultrasound, which requires blood vessels or capillaries to flow directly across the imaging plane to visualize blood flow, the detection of echogenicity in each measurement voxel of tissue imaged using the currently described high-frequency ultrasound method is agnostic to the direction of capillaries. By using only amplitude information in each voxel without phase or spatial information, the current imaging modality is completely agnostic to flow direction. This means that if a bundle of capillaries exists in the same voxel, with each capillary flowing in different and / or opposite directions, the current imaging modality is still able to determine the total flow velocity across all capillaries. With conventional Doppler ultrasound, even if two vessels in the same voxel are aligned with the imaging plane, no signal will be seen if their flow directions are opposite, as they will cancel each other out.

[0031] In B-mode (brightness mode) ultrasound (also known as 2D ultrasound), a linear array of transducers simultaneously scans a plane of the body, which can be viewed as a two-dimensional image on a screen. B-mode is an ultrasound imaging modality that digitally highlights moving reflectors, likely primarily red blood cells, while suppressing signals from surrounding stationary tissue. Therefore, B-mode can simultaneously visualize flowing blood and surrounding stationary tissue. Each pixel in a B-mode image represents a tissue volume 70 micrometers wide. Typical capillaries are less than 70 micrometers in diameter, so the collected image focuses on the capillaries within the tissue volume, providing brightness associated with increased capillary flow towards the perfused area.

[0032] Data at a resolution provided by B-mode microsound produces perfusion images of approximately 5 megapixels per image, where areas of increased blood perfusion are shown as bright areas in the image. Using B-mode, images are collected at a high frame rate (at least 20 frames per second) sufficient for comparison and data autocorrelation, for example, at least 5 frames.

[0033] Image processing of micro-ultrasound data uses the time constant of brightness variation in each pixel and examines the statistical rate of change between time points and over time. Optionally, the image can be processed to reduce processing time, for example, by selecting a subset of pixels or by creating a lower-resolution downsampled version of the image. Optionally, the image can also be spatially or temporally processed to suppress noise, smooth the image, and improve fidelity. For example, in one embodiment, a low-pass or filter may be applied to the image first.

[0034] The signal is the slope of the correlation trend, also known as the decorrelation rate. The decorrelation slope provides information about the fluid flow velocity in the injection region.

[0035] To derive the signal from the image, the data in each pixel across all frames is first statistically normalized using the following equation:

[0036]

[0037] Then use x 基准 Autocorrelation is used as a basis to calculate the repetition of the normalized signal at each pixel. Autocorrelation examines the similarity between the signal at a pixel and the signal at the same pixel after a certain period of time, such as after 1 to 4 frames (0.03 to 0.13 seconds at 30 frames per second).

[0038] The autocorrelation y can be calculated from standardized image data using the following equation:

[0039]

[0040] in:

[0041] i and j represent the 2D coordinates of the pixel.

[0042] t is the time index of an image frame being processed, and

[0043] Lag is the amount of time between one frame and another, measured in frames.

[0044] The autocorrelation results can be normalized using the following equation:

[0045]

[0046] in:

[0047] n is the number of time points (i.e. frames) collected for analysis.

[0048] Other methods that can be used to calculate the autocorrelation of data include Spearman correlation, Pearson correlation, or Fourier transform. Alternatively, a Gaussian kernel function can be used, for example, to smooth the calculated autocorrelation.

[0049] The decorrelation rate is the trend of normalized autocorrelation of a standardized B-mode signal over time. The decorrelation rate can be used to measure tissue perfusion, including flow velocity, and to differentiate tissue types. Typically, a timescale in which the signal changes linearly over time can be chosen, and therefore the decorrelation rate can be quantitatively determined by applying linear regression to fit a straight line, the mean difference (by averaging the differences between each point at t and t+1), or the overall amplitude variation (max(y) - min(y)). Other calculations can be applied to fit more complex data and provide additional information about flow rate; these calculations may include determining the period, exponential decay rate, or time to a local minimum.

[0050] Voxels with very low or negligible decorrelation rates represent noise because noise is an uncorrelated process. Therefore, in one embodiment, voxels with decorrelation rates below a specified threshold (empirically chosen based on noise levels) are set to 0 in the B-mode image before display, reducing the noise content of the image and improving the signal-to-noise ratio and contrast. In another embodiment, the brightness of pixels in the B-mode image is reduced based on the decorrelation rate when it falls below a certain threshold, as shown in the following equation. Those skilled in the art will understand that parameters can be set empirically to optimize image quality for specific situations, and other similar functions can be used to achieve the same effect.

[0051]

[0052] Creating a visual representation of decorrelation rates is useful, as it represents various aspects of perfusion. In a simple map, decorrelation values ​​are assigned a brightness (0-255), which can be displayed as an image on the screen. A threshold can be useful, below which these values ​​are not used (e.g., slope less than 10 = brightness value 0). Logarithmic maps can also be used to “compress” the signal, similar to the compression performed on ultrasound B-mode imaging data, where brightness equals log(slope + 1) to show smaller differences between higher values ​​and highlight differences between lower values. Other transformations of the data, such as flow rate, can be applied if determined to be useful for highlighting target aspects. In some representations, different colors can be used to further distinguish target aspects.

[0053] In one embodiment, perfusion imaging and denoising are based on voxel decorrelation rates. Perfusion imaging and denoising can be applied independently of each other. Alternatively, perfusion imaging and denoising can be used together.

[0054] An ultrasound system suitable for acquiring data using this imaging method should be adapted to receive ultrasound signals with a frequency of at least 15 MHz and a frame rate of at least 20 frames per second (fps). High-frequency (HF) imaging (above 15 MHz) improves spatial resolution compared to conventional ultrasound imaging, which typically uses frequencies from 2 to 15 MHz. The signal processing described herein requires a higher frame rate to generate ultrasound images from the acquired ultrasound signals. In another embodiment, the signal processing may work with conventional ultrasound frequencies that have strong reflectors (e.g., microbubble contrast agents and / or nanoparticle contrast agents).

[0055] The transducer used for image acquisition can be a linear array transducer, a phased array transducer, a two-dimensional (2-D) array transducer, or a curved array transducer. The center emission frequency of the transducer used is preferably equal to or greater than 15 MHz. For example, the center emission frequency can be approximately 15 MHz, 20 MHz, 30 MHz, 40 MHz, 50 MHz, 55 MHz, or higher. In some examples, the ultrasonic transducer can transmit ultrasound to the object at a center frequency in the range of approximately 15 MHz to approximately 80 MHz. Preferably, the transducer comprises a high-frequency linear array having an imaging system in the range of 20-50 MHz.

[0056] Figure 1A high-frequency medical imaging system 100 for use with this method is shown. The system includes an ultrasound transducer 102 with transducer elements 104, an ultrasound transducer interface 106, a spatial sensor 108, and a server 110. The ultrasound transducer 102 is configured to: (A) convert received echo signals (by the ultrasound transducer 102) into ultrasound information; and (B) transmit ultrasound information (through an output port). The ultrasound transducer 102 is also referred to as an ultrasound probe. The ultrasound transducer 102 has transducer elements 104 arranged in an array; for example, the transducer elements 104 may be aligned one after another relative to each other along a row. The transducer elements 104 are configured to be activated (they may be selectively activated or deactivated). The transducer elements 104 are also referred to as transmitting and receiving elements because they transmit ultrasound pulses and receive reflections of ultrasound pulses. The collection of transducer elements 104 is also referred to as a transducer array. When the ultrasonic transducer 102 is configured to both transmit (output ultrasonic pulses) and receive (reflect ultrasonic pulses), the ultrasonic transducer 102 is also referred to as an ultrasonic transceiver. The medical imaging system 100 uses the ultrasonic transducer 102 on a principle similar to radar or sonar, wherein the medical imaging system 100 is configured to assess the properties of a target by interpreting the echoes (reflections) from the sound waves. The ultrasonic transducer 102 is configured to: (A) generate relatively high-frequency sound waves; and (B) receive echoes from the target. The medical imaging system 100 is configured to: (A) evaluate the ultrasonic information provided by the ultrasonic transducer 102; (B) calculate the time interval between transmitting the output signal (from the ultrasonic transducer 102) and receiving the echo; and (C) determine the distance to the target or object based on the calculated time interval. The ultrasonic transducer 102 is configured to generate sound waves in the ultrasonic range, typically above about 18 kHz, by converting electrical energy into sound; then, upon receiving an echo, the ultrasonic transducer 102 is configured to convert the reflected sound waves into electrical energy, which can be measured and displayed by the medical imaging system 100.

[0057] Ultrasound is an oscillating sound pressure wave with a frequency greater than the upper limit of human hearing. While this limit varies from person to person, it is approximately 20 kHz in healthy young adults. Some ultrasound devices operate at frequencies ranging from approximately 20 kHz to several gigahertz (GHz). The ultrasound transducer 102 is configured to emit a signal comprising short pulses of ultrasound energy. After each pulse, the ultrasound transducer 102 is configured to receive a return (reflected) signal within a small time window corresponding to the time it takes for the energy to pass through the patient's tissue; the signal received during this period is then eligible for additional signal processing by the medical imaging system 100. The ultrasound transducer 102 (medical ultrasound transducer or probe) can be configured to have various shapes and sizes for creating images of different parts of the body. The ultrasound transducer 102 can be inserted through the surface of the body (patient), via laparoscopy, or through an opening in the patient's body, such as the esophagus, rectum, or vagina. The ultrasound transducer 102 can be configured (by a clinician or operator performing an ultrasound-guided procedure) for use with a probe positioning system (not depicted but known) configured to hold and / or move the ultrasound transducer 102; the ultrasound transducer 102 includes an array of transducer elements 104. The rows of transducer elements 104 of the ultrasound transducer 102 can be aligned in a straight line or a curved line. Each transducer element 104 is configured to: (A) emit an incident sound signal toward a target; and (B) receive an echo signal representing sound reflected back from the target to the transducer element 104.

[0058] The ultrasonic transducer interface 106 is configured to control the operation of the ultrasonic transducer 102. The ultrasonic transducer interface 106 is... Figure 1 The processor component 120 is described as a software program (depending on the option). The processor component 120 controls the ultrasonic transducer 102 via the ultrasonic transducer interface 106. The ultrasonic transducer interface 106 is also referred to as a beamformer. According to one example, the ultrasonic transducer interface 106 may include server executable code (software program) tangibly stored in a non-transitory computer-readable medium 112 (hereinafter referred to as memory 112) of the server 110; according to another example, the ultrasonic transducer interface 106 includes a combination of electronic hardware components that cooperate with the server executable code. Generally, the ultrasonic transducer interface 106 is configured to: (A) be operatively connected to the ultrasonic transducer 102 (via the output port of the ultrasonic transducer 102); (B) control the shape of the incident acoustic signal to be transmitted by the transducer elements 104; (C) receive ultrasonic information from the ultrasonic transducer 102; and (D) provide scan lines mapped to the transducer elements 104, which are activated in such a way as to generate the scan lines to be provided (not all transducer elements 104 will be activated, therefore these unused instances of transducer elements 104 will be deactivated). The ultrasonic transducer interface 106 is a device configured to facilitate electronically controlled focusing of the ultrasonic energy emitted and / or received by the ultrasonic transducer 102.

[0059] Typically, the space sensor 108 is configured to: (A) detect spatial movement of the ultrasonic transducer 102; and (B) provide spatial information indicating the spatial movement of the ultrasonic transducer 102 when the ultrasonic transducer 102 transmits ultrasonic information to the ultrasonic transducer interface 106. The space sensor 108 may be attached to the ultrasonic transducer 102. Alternatively, the space sensor 108 may be integrated with the ultrasonic transducer 102.

[0060] Server 110 is also referred to as a computer, etc. Typically, server 110 is configured to: (A) connect to ultrasonic transducer interface 106; (B) connect to space sensor 108; and (C) have memory 112 that tangibly stores executable code 114 (also referred to as processor executable code, hereinafter referred to as program 114). Program 114 is a combination of operational tasks to be performed by server 110. Server 110 is a system composed of software and appropriate computer hardware. Server 110 may include a dedicated computer or a combination of computers. Server 110 may be configured for a client-server architecture (if desired).

[0061] Memory 112 can refer to a physical device or any equivalent thereof used to store computer-executable programs or processor-executable programs (sequences of instructions or operations) and / or data (e.g., program state information) on a temporary or permanent basis for use in server 110. Unlike secondary storage, which is used for information in a high-speed physical system (e.g., RAM or random access memory), secondary storage is a physical device used for program and data storage, with slower access speeds but larger storage capacities. Main memory stored on secondary storage is called "virtual memory." For example, memory 112 may include volatile and / or non-volatile memory. For example, memory 112 may include secondary storage such as magnetic tape, disks, and optical discs (CD-ROM or compact optical disc ROM, and DVD-ROM or digital video optical disc ROM).

[0062] Program 114 is constructed using well-known software tools known to those skilled in the art; computer program instructions are assembled in a high-level computer programming language and translated into executable code using a compiler and other tools. It should be understood that program 114 provides a sequence of methods or operations to be executed by processor component 120. Memory 112 includes (tangibly storing) executable code 114 (also referred to as program 114). Executable code 114 includes a combination of operational tasks to be performed by processor component 120. For example, executable code 114 is configured to instruct server 110 to receive ultrasound information associated with a scanline group having a finite number of selectable scanlines of ultrasound transducer 102. For example (and not limited to this), the scanline group may have a finite number of scanlines mapped to a finite set of transducer elements 104 of ultrasound transducer 102 (for generating selected scanlines of the scanline group), if desired.

[0063] It should be understood that, in view of the foregoing, generally, a method is provided for operating a medical imaging system 100 having an ultrasound transducer interface 106; the ultrasound transducer interface 106 is configured to be operatively connected to an ultrasound transducer 102; the ultrasound transducer 102 includes a transducer element 104; the medical imaging system 100 also has a space sensor 108 configured to provide spatial information indicative of spatial movement of the ultrasound transducer 102; the method includes receiving ultrasound information associated with a scan line group having a finite number of selectable scan lines of the ultrasound transducer 102. Furthermore, a server 110 is configured (programmed) to receive ultrasound information associated with a scan line group having a finite number of selectable scan lines of the ultrasound transducer 102. In addition, the non-transitory computer-readable medium 112 includes executable code 114 tangibly stored in the non-transitory computer-readable medium 112; the executable code 114 includes a combination of operational tasks that can be performed by the server 110; the executable code 114 is configured (programmed) to instruct the server 110 to receive ultrasound information associated with a scan line group having a limited number of optional scan lines of the ultrasound transducer 102.

[0064] Server 110 also includes a display component 116; an input / output interface module 118; a processor component 120; a database 122 tangibly stored in memory 112; ultrasonic data 123; and spatial data 124. Ultrasonic data 123 and spatial data 124 are stored in database 122 or in memory 112. Input / output interface module 118 is configured to operatively connect processor component 120 to display component 116, ultrasonic transducer interface 106 (and indirectly to ultrasonic transducer 102), and spatial sensor 108. In this way, processor component 120 can control the operation of display component 116, ultrasonic transducer interface 106, and spatial sensor 108, and can also control ultrasonic transducer 102 through direct control of ultrasonic transducer interface 106. Input / output interface module 118 is also configured to connect processor component 120 to user interface devices (e.g., keyboard, mouse, touchscreen display component, etc.).

[0065] Processor component 120 (also referred to as a central processing unit or CPU) is hardware within server 110 that executes the instructions described in program 114 by performing arithmetic, logic, and input / output operations. Processor component 120 may have one or more instances of CPUs. A CPU may include a microprocessor (meaning the CPU is contained on a single silicon chip). Some integrated circuits (ICs) may contain multiple CPUs on a single chip; these ICs are called multi-core processors. An IC containing a CPU may also contain peripherals and other components of the computer system; this is called a system-on-a-chip (SoC). The components of a CPU are an arithmetic logic unit (ALU) that performs arithmetic and logical operations, and a control unit that fetches instructions from memory, decodes and executes them, and invokes the ALU when necessary. Processor component 120 may include an array processor or vector processor with multiple parallel computing elements, where no single unit is considered a "center". In a distributed computing model, the problem is solved by a set of distributed interconnected processors. Images to be displayed by medical imaging system 100 may be displayed in real time and / or after an acquisition or processing delay (via display component 116).

[0066] Figure 2A and 2B This involves denoising ultrasound images before and after the process by eliminating pixels with very short decorrelation times. These pixels are often products of random noise in the received signal rather than actual biological effects, unlike pixels with measurable medium or long decorrelation times, which are associated with flow or static anatomical structures. By removing or reducing the magnification of these short decorrelation pixels, the overall signal-to-noise ratio of the image is improved. Figure 2A The image before correction is displayed. Figure 2BThe same image is shown after removing pixels with short decorrelation times (after correction).

[0067] In addition to brightening the perfused area to highlight the blood flow area, it can suppress the brightness signal of the surrounding quiescent tissue area to provide contrast between perfused and non-perfused tissues.

[0068] Figure 3 An example comparing perfusion signals from blood vessels with signals from noisy regions in an image is shown. The blood signal has a clear linear slope, while the noise signal appears random and lacks an overall trend.

[0069] Figure 4 Examples of signals from misaligned and aligned images are shown, where anomalous signals that moved during imaging were removed. Movement of the target region or within the target region can also occur during image acquisition. This could be due to patient movement or movement of tissues and organs in the body. Optionally, this movement can be corrected by aligning images through matching local signal patterns between frames. In one alignment method, the neighborhood of each pixel is identified, and temporal correlation cross-correlation of signals across a 2D grid of pixel neighborhoods is matched to remove anomalous signals. Faster imaging rates may help reduce alignment problems.

[0070] Blood flow to tissues may increase during exercise, and increased blood flow is also common in solid tumors and other cancerous growths. Figure 5A -C shows that the exercise-induced increase in blood flow occurs before and after forearm muscle activity. Figure 5A The anatomical structure of the wrist is shown in a B-mode micro-ultrasound image. Figure 5B An image of a stationary flexor muscle is shown. Figure 5C An image of the flexor muscles at rest after 1 minute of activity is shown.

[0071] Figure 6This is a flowchart illustrating an example method for ultrasound imaging modalities used for tissue imaging perfusion. First, multiple B-mode ultrasound image data are acquired using high-frequency ultrasound. In another embodiment, multiple B-mode high-frequency ultrasound image data (frames) 202 from memory 112 are processed. Then, a decorrelation trend of the autocorrelation image data is calculated to determine the flow and perfusion level between selected image frames for each pixel in the image frames 204. Then, one of three steps can be performed. In one embodiment, the noise content in the image data is reduced by using the decorrelation trend 206. Alternatively, a visual representation of the perfusion level is formed based on the decorrelation trend 208. Alternatively, the noise content in the image data is reduced by using the decorrelation trend, and a visual representation of the perfusion level is formed based on the decorrelation trend (as shown in 210 and 212). Step 210 is the same as 206, and step 212 is the same as 208. Steps 210 and 212 can occur in any order. Furthermore, the image data can be normalized (not shown) at each pixel of all images before autocorrelation.

[0072] The following terms are provided as an example of the device for further description. Any one or more of the following terms may be combined with any other one or more of the following terms and / or with any segment or one or more parts and / or clauses of any other terms. Any one of the following terms may be valid on its own merit without having to be combined with any other terms or any part of any other terms.

[0073] Clause 1: An imaging processing method comprising: processing a plurality of digital images containing continuous B-mode ultrasound reflectance data; calculating a decorrelation trend of autocorrelation data to determine flow and perfusion levels. Clause 2: A method according to any clause or any part thereof mentioned in this paragraph, further comprising: using the decorrelation trend to reduce noise content in the ultrasound data. Clause 3: A method according to any clause or any part thereof mentioned in this paragraph, further comprising: forming a visual representation of the perfusion level based on the decorrelation trend. Clause 4: A method according to any clause or any part thereof mentioned in this paragraph, further comprising: using the decorrelation trend to reduce noise content in the autocorrelation data; and forming a visual representation of the perfusion level based on the decorrelation trend. Clause 5: A method according to any clause or any part thereof mentioned in this paragraph, wherein the method has an image capture rate of 20 frames / second or higher. Clause 6: A method according to any clause or any part thereof mentioned in this paragraph, wherein the image data is normalized prior to autocorrelation. Clause 7: The method according to any clause or any part of any clause mentioned in this paragraph, wherein the autocorrelation data is normalized. Clause 8: The method according to any clause or any part of any clause mentioned in this paragraph, wherein the decorrelation trend of the autocorrelation data is determined by linear regression, mean difference, or population magnitude change. Clause 9: The method according to any clause or any part of any clause mentioned in this paragraph, wherein the autocorrelation data with a non-linear trend is processed to determine one or more of the following measures: decorrelation period, exponential decay rate, time to local minimum, or other measures. Clause 10: The method according to any clause or any part of any clause mentioned in this paragraph, wherein the autocorrelation is calculated using Spearman correlation, Pearson correlation, or Fourier transform. Clause 11: The method according to any clause or any part of any clause mentioned in this paragraph, further including a smoothing step prior to the visual representation of the signal. Clause 12: The method according to any clause or any part of any clause mentioned in this paragraph, further including applying a threshold and rescaling to the decorrelation trend prior to the visual representation. Clause 13: The method described according to any or any part of any clause mentioned in this paragraph further includes a logarithmic transformation of the decorrelation trend. Clause 14: The method described according to any or any part of any clause mentioned in this paragraph, wherein the decorrelation trend is mapped to different color or grayscale values ​​for representation in the image. Clause 15: The method described according to any or any part of any clause mentioned in this paragraph further includes aligning the plurality of digital images to correct for motion by matching local signal patterns between frames.Clause 16: A method according to any clause or any part of any clause mentioned in this paragraph, wherein the method comprises at least 5 consecutive frames of B-mode ultrasound reflectance data. Clause 17: A method according to any clause or any part of any clause mentioned in this paragraph, wherein a subset of the image field is processed from each image to reduce processing time. Clause 18: A method according to any clause or any part of any clause mentioned in this paragraph, wherein the image is downsampled to a lower resolution to reduce processing time. Clause 19: A method according to any clause or any part of any clause mentioned in this paragraph, wherein the ultrasound data is high-frequency ultrasound with a frequency range greater than 15 MHz. Clause 20: A perfusion imaging system comprising: a high-frequency ultrasound transducer for capturing and collecting multiple digital images containing consecutive B-mode ultrasound reflectance data; and a signal processing unit operatively connected to the transducer, the signal processing unit being configured to calculate decorrelation trends of autocorrelation data to determine flow and perfusion levels. Clause 21: A system according to any or any part of the clauses mentioned in this paragraph, wherein the signal processing unit is further configured to reduce the noise content in the ultrasound data using the decorrelation trend. Clause 22: A system according to any or any part of the clauses mentioned in this paragraph, wherein the signal processing unit is further configured to form a visual representation of the perfusion level based on the decorrelation trend. Clause 23: A system according to any or any part of the clauses mentioned in this paragraph, wherein the signal processing unit is further configured to reduce the noise content in the ultrasound data using the decorrelation trend and to form a visual representation of the perfusion level based on the decorrelation trend. Clause 24: A system according to any or any part of the clauses mentioned in this paragraph, wherein the system has an image capture rate of 20 frames per second or higher. Clause 25: A system according to any or any part of the clauses mentioned in this paragraph, wherein the signal processing unit is configured to normalize the image data before autocorrelation. Clause 26: A computer-readable storage medium comprising executable instructions that, when executed by a process, cause a processor to: process a plurality of digital images comprising sequential B-mode ultrasound reflectance data obtained using microsound; and calculate a decorrelation trend of autocorrelation data to determine flow and perfusion levels. Clause 27: A computer-readable storage medium according to any or any part of the clause mentioned in this paragraph, further comprising instructions to: reduce the noise content in the ultrasound data using the decorrelation trend. Clause 28: A computer-readable storage medium according to any or any part of the clause mentioned in this paragraph, further comprising instructions to: form a visual representation of the perfusion level based on the decorrelation trend.Clause 29: The computer-readable storage medium described under any or any part of the clause mentioned in this paragraph further includes the following instructions: using the decorrelation trend to reduce the noise content in the ultrasound data; and forming a visual representation of the perfusion level based on the decorrelation trend.

[0074] All publications, patents, and patent applications mentioned in this specification demonstrate the skill of a person skilled in the art to which this invention pertains and are incorporated herein by reference. It is obvious that the invention thus described can be varied in many ways. Such variations should not be considered as departing from the scope of the invention, and it will be apparent to those skilled in the art that all such modifications are intended to be included within the scope of the appended claims.

Claims

1. An imaging processing method, the imaging processing method comprising: Processing multiple digital images containing continuous B-mode ultrasound reflectance data; Calculate the decorrelation trend of autocorrelation data to determine flow and perfusion levels. Specifically, the decorrelation trend is used to reduce the noise content in the ultrasonic reflectivity data. The method has an image capture rate of 20 frames per second or higher. The ultrasonic reflectivity data refers to high-frequency ultrasound with a frequency range greater than 15 MHz.

2. The method according to claim 1, further comprising: A visual representation of the perfusion level is formed based on the decorrelation trend.

3. The method according to claim 1, further comprising: The decorrelation trend is used to reduce the noise content in the autocorrelation data; as well as A visual representation of the perfusion level is formed based on the decorrelation trend.

4. The method according to any one of claims 1 to 3, wherein the digital image is normalized prior to autocorrelation.

5. The method according to any one of claims 1 to 3, wherein the autocorrelation data is normalized.

6. The method according to any one of claims 1 to 3, wherein the decorrelation trend of the autocorrelation data is determined by linear regression, mean difference, or change in overall magnitude.

7. The method according to any one of claims 1 to 3, wherein the autocorrelation data having a nonlinear trend is processed to determine one or more of the following: the decorrelation period, the exponential decay rate, the time to a local minimum, or other measures.

8. The method according to any one of claims 1 to 3, wherein Spearman correlation, Pearson correlation or Fourier transform is used to calculate the autocorrelation.

9. The method according to any one of claims 1 to 3, further comprising a smoothing step prior to the visual representation of the signal.

10. The method according to any one of claims 1 to 3, further comprising applying a threshold and rescaling to the decorrelation trend prior to the visualization representation.

11. The method of claim 10, further comprising the logarithmic transformation of the decorrelation trend.

12. The method according to any one of claims 1 to 3, wherein the decorrelation trend is mapped to different color or grayscale values ​​to be represented in the image.

13. The method according to any one of claims 1 to 3, further comprising aligning the plurality of digital images to correct for movement by matching local signal patterns between frames.

14. The method according to any one of claims 1 to 3, wherein the method comprises at least 5 consecutive frames of B-mode ultrasound reflectivity data.

15. The method according to any one of claims 1 to 3, wherein a subset of the image field is processed from each image to reduce processing time.

16. The method according to any one of claims 1 to 3, wherein the image is downsampled to a lower resolution to reduce processing time.

17. A perfusion imaging system, the perfusion imaging system comprising: A high-frequency ultrasonic transducer used to capture and collect multiple digital images containing continuous B-mode ultrasonic reflectivity data. A signal processing unit, operatively connected to the transducer, is configured to calculate the decorrelation trend of autocorrelation data to determine flow and perfusion levels. The signal processing unit is further configured to use the decorrelation trend to reduce the noise content in the ultrasonic reflectivity data. The system has an image capture rate of 20 frames per second or higher. The ultrasonic reflectivity data refers to high-frequency ultrasound with a frequency range greater than 15 MHz.

18. The system of claim 17, wherein the signal processing unit is further configured to form a visual representation of the perfusion level based on the decorrelation trend.

19. The system of claim 17, wherein the signal processing unit is further configured to reduce the noise content in the ultrasonic reflectivity data using the decorrelation trend, and to form a visual representation of the perfusion level based on the decorrelation trend.

20. The system according to any one of claims 17 to 19, wherein the signal processing unit is configured to normalize the digital image prior to autocorrelation.

21. A computer-readable storage medium comprising executable instructions that, when executed by a process, cause a processor to: Processing multiple digital images containing continuous B-mode ultrasound reflectance data obtained using micro-ultrasound, wherein, The image capture rate is 20 frames per second or higher, and the ultrasonic reflectivity data is high-frequency ultrasound with a frequency range greater than 15 MHz; Calculate the decorrelation trend of autocorrelation data to determine flow and perfusion levels. The executable instructions, when executed by the process, further enable the processor to: The decorrelation trend is used to reduce the noise content in the ultrasonic reflectivity data.

22. The storage medium according to claim 21, wherein, The executable instructions, when executed by the process, further enable the processor to: A visual representation of the perfusion level is formed based on the decorrelation trend.

23. The storage medium according to claim 21, wherein, The executable instructions, when executed by the process, further enable the processor to: The decorrelation trend is used to reduce the noise content in the ultrasonic reflectivity data; and A visual representation of the perfusion level is formed based on the decorrelation trend.

Images