A method and system for jointly photoacoustic-ultrasound depth-resolved multilayered cranial brain microvessels

Abstract

The application discloses a method and system for jointly performing photoacoustic-ultrasound depth resolution multilayer craniocerebral microvessel imaging, and the method comprises the following steps: firstly, threshold processing is performed on the amplitude of collected photoacoustic-ultrasound data; secondly, a single pixel value on a scanning plane is cross-correlated with an ultrasonic echo signal of a reference glass to determine the outer surface of the skull, the process is repeated for all scanning points, and the inner surface of the skull is identified in combination with a photoacoustic image; the inner and outer surfaces of the skull in a selected region on the skull are manually segmented using the above information, the average skull thickness is calculated, then a skull angle correction uniform model is established, and the effective thickness of the skull is calculated; finally, the effective thickness of the skull is used to perform depth layer imaging on craniocerebral microvessels in a photoacoustic image. Based on the application, a fine craniocerebral microvessel network image can be obtained, which is helpful for the analysis of craniocerebral vascular structure and can be expected to be used in the field of medical image analysis.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of photoacoustic imaging and ultrasonic imaging, and particularly relates to a method and system for jointly photoacoustic ultrasonic depth-resolved multilayer craniocerebral microvessels. BACKGROUND

[0002] Photoacoustic imaging visualizes tissue chromophores by detecting the tiny ultrasonic vibrations caused by the absorption of short-time light pulses, which can be used to image and functionally analyze mouse brain vasculature at multiple spatial scales. However, a single imaging technique cannot comprehensively and in detail describe the biological tissue structure and function information, and photoacoustic imaging technology cannot distinguish craniocerebral microvessels and meningeal microvessels. Therefore, we propose a combined method of photoacoustic imaging and ultrasonic imaging to visualize and accurately locate the entire parietal bone and most of the frontal bone and the craniocerebral microvessel system between the parietal bones. Ultrasonic imaging reflects the acoustic impedance difference characteristics between biological tissues, and the acoustic impedance difference between the skull and soft tissue is strongly mismatched, so the region to which the skull belongs can be determined according to the ultrasonic image of the skull, and the region to which the craniocerebral microvessel network belongs is located in the photoacoustic image.

[0003] The craniocerebral vascular system plays a core role in many nervous system diseases, and the craniocerebral microvessel system is responsible for transporting oxygen and nutrients, and plays a crucial role in the maintenance, development and hematopoiesis of bone marrow. The morphology and function of craniocerebral vessels have important research significance. In the past, the craniocerebral vascular system was mainly studied by using young mice, but the degree of vascularization of the skull of young mice is low, so the skull and its microvessel system have been largely ignored in previous studies. Therefore, it is necessary to accurately locate and extract the craniocerebral microvessel network from the photoacoustic image.

[0004] Establishing a skull angle correction uniform model can calculate the effective thickness of the skull, which helps to accurately locate the outer surface and inner surface of the skull. The Chinese patent "A correction method for plane ultrasonic transcranial brain imaging" (Patent publication number: CN112927145A, application date 2021.06.08) discloses an imaging method based on ultrasonic plane wave, which compensates and corrects the ultrasonic distortion caused by the skull, but does not mention the influence of the position difference of the ultrasonic transducer relative to the skull on the skull response. This patent separates the position influence of the ultrasonic transducer by using the cross-correlation method of the ultrasonic image pixel value and the reference glass ultrasonic echo signal. The Chinese patent "Photoacoustic endoscopic quantitative tomographic imaging method and system based on ultrasonic structure layering guidance" (Patent publication number: CN111829956A, application date 2021.10.01) only mentions using the Sobel edge detection algorithm to identify the edge of the cavity wall blood vessels in ultrasonic endoscopic imaging, but does not mention the skull thickness calculation and its angle correction method. This patent proposes a skull angle correction uniform model, which effectively calculates the skull thickness, effectively segments the cranial brain structure, and accurately extracts the skull microvascular network and meningeal microvascular network. SUMMARY

[0005] The main purpose of the present application is to make up for the shortcomings of the existing technology, and provide a method and system for jointly photoacoustic ultrasonic depth resolution of multi-layer skull brain microvessels. The present application can effectively combine the ultrasonic image with the photoacoustic image to accurately locate the skull, thereby accurately extracting the skull microvascular network and providing technical support for the study of nervous system diseases.

[0006] In order to achieve the above purpose, the technical scheme adopted by the present application is as follows:

[0007] In a first aspect, the present application provides a method for jointly photoacoustic ultrasonic depth resolution of multi-layer skull brain microvessels, comprising the following steps:

[0008] (1) Collecting ultrasonic and photoacoustic original images of the same skull brain region;

[0009] (2) Image preprocessing of the collected ultrasonic and photoacoustic original images, setting threshold to exclude noise signal interference according to the overall characteristics of the ultrasonic original image and the overall characteristics of the photoacoustic original image;

[0010] (3) Cross-correlation of the ultrasonic image pixel value with the reference glass to obtain the cross-correlated ultrasonic echo signal, thereby eliminating the influence of the position difference of the ultrasonic transducer relative to the skull on the skull response;

[0011] (4) According to the difference in acoustic impedance characteristics between the skull and the soft tissue, a high-amplitude echo signal is generated on the outer surface of the skull, and the position matrix of the outer surface of the skull is obtained in the processed ultrasonic image;

[0012] (5) The photoacoustic image and the ultrasound image are combined to identify the inner surface of the skull, and whether there is a peak value factor of the skull blood vessel interfering with the identification of the inner surface of the skull is judged according to the photoacoustic signal, so as to obtain a position matrix of the inner surface of the skull;

[0013] (6) The inner surface of the skull has different tissue structures, and a roughness parameter of the inner surface of the skull is obtained by calculation, so as to adjust the accuracy of the position of the inner surface of the skull;

[0014] (7) A reliable point with clear ultrasound signal and photoacoustic signal characteristics in different brain regions is selected, and the obtained position matrix of the outer surface of the skull and the position matrix of the inner surface of the skull are used to calculate the average thickness of the skull;

[0015] (8) A least square fitting local plane is adopted to establish a skull angle correction uniform model, and the effective thickness of the skull is calculated;

[0016] (9) According to the relationship between the ultrasound signal and the photoacoustic signal in the time response, the depth position information of the skull microvessel network corresponding to the photoacoustic data is obtained through the given effective thickness of the skull in the vertical scanning axis direction;

[0017] (10) The skull microvessel image, the cerebral cortex shallow microvessel network and the cerebral cortex deep microvessel network are segmented according to the depth position information of the photoacoustic data, and maximum projection is performed respectively.

[0018] As a preferred technical scheme, in step (3), the ultrasound image pixel value is cross-correlated with the reference glass, specifically:

[0019] A glass reference with the same or similar thickness as the skull is selected, and the ultrasound echo signal of the reference glass is collected, for a single pixel on the scanning plane, the ultrasound echo signal on the skull is cross-correlated with the ultrasound echo signal of the reference glass, and the cross-correlation is realized based on MATLAB.

[0020] As a preferred technical scheme, in step (4), the position matrix of the outer surface of the skull is obtained, specifically:

[0021] The image pixel value is cross-correlated with the reference glass, the maximum value of the cross-correlation function is used to determine the position of the outer surface of the skull, and the remaining positive peaks in the cross-correlation function are stored;

[0022] This process is repeated for all scanning points to generate the skull outer surface echo signal and obtain the position matrix of the skull outer surface.

[0023] As a preferred technical scheme, in step (5), the photoacoustic image and the ultrasound image are combined to identify the inner surface of the skull, specifically:

[0024] The ultrasound image skull outer surface position matrix determines the region to which the skull microvessel network in the photoacoustic image belongs;

[0025] According to the photoacoustic signal, it is judged whether a peak corresponding to a skull blood vessel exists to interfere with the recognition of the inner surface of the skull. If the scanning point corresponds to a position where a skull microvessel exists at this time, the inner surface position is divided as the first point after the first peak of the photoacoustic image. If the scanning point corresponds to a position where a skull microvessel does not exist at this time, the inner surface position is divided as a point before the first peak of the photoacoustic image.

[0026] As a preferred technical solution, the skull microvessel network depth position information corresponding to the photoacoustic data is obtained, specifically:

[0027] The pulsed echo ultrasound realizes the recognition of the outer surface depth of the skull through the focused transmitting beam and receiving beam, and the photoacoustic signal is received through the ultrasonic transducer located around the tissue. In the same imaging area, the position index of the ultrasonic echo signal in the depth dimension and the position index of the photoacoustic signal exist an approximate two times relationship, and the starting position of the region to which the skull microvessel network belongs in the photoacoustic image is determined through such a quantity relationship.

[0028] As a preferred technical solution, in step (6), the evaluation of the skull inner surface roughness is specifically:

[0029] The skull inner surface roughness measures the heterogeneity inside the skull, which is reflected in the complex structure inside the skull, including blood vessels, bone marrow compartments and inner skull surfaces. The measurement of the skull inner surface roughness reflects the skull non-uniformity level of the local region of interest (ROI), and thus it is defined as the moving standard deviation of the second trailing peak in the ultrasonic pulse echo signal.

[0030] For the ultrasonic signal of each scanning point of the local region of interest (ROI), the distance between the second trailing peak and the first significant peak is first calculated. The standard deviation of these distances can obtain the skull inner surface roughness parameter, which is used to adjust the accuracy of the skull inner surface position estimation.

[0031] Skull inner surface roughness Wherein, i represents the peak serial number in each region of interest (ROI), n represents the number of peaks in the region of interest (ROI), z i is the distance from them to the outer surface of the skull, represents the average distance.

[0032] As a preferred technical solution, in step (7), the skull average thickness calculation specific steps are:

[0033] According to the combination of the ultrasonic image and the photoacoustic image, the ultrasonic image is used to determine the position of the outer surface of the skull, and the photoacoustic image is used to judge whether there is a peak corresponding to a blood vessel, so as to determine the position of the inner surface of the skull.

[0034] Select n trust points with clear ultrasonic signal and photoacoustic signal characteristics to guide manual segmentation of images, each trust point corresponds to a thickness d1, d2, …, dn, and the average thickness of the skull is calculated

[0035] As a preferred technical solution, in step (8), the local plane is fitted by the least squares method, specifically:

[0036] By estimating the outer surface of the skull, the fitting of the local plane is converted into a least squares plane fitting estimation problem;

[0037] The least squares plane fitting estimation step is specifically:

[0038] Based on the nearest neighbor search, the nearest neighbor points are found;

[0039] A covariance matrix is created at the nearest neighbor of the query point, the covariance matrix is solved, and the covariance matrix eigenvectors and eigenvalues are analyzed, the covariance matrix is:

[0040]

[0041] Where k is p i The number of nearest neighbor points of the point, is the three-dimensional centroid of the nearest neighbor point, λ j is the jth eigenvalue of the covariance matrix, is the jth eigenvector;

[0042] The eigenvector corresponding to the smallest eigenvalue of the obtained covariance matrix is the vertex normal vector.

[0043] As a preferred technical solution, in step (8), the skull angle correction uniform model is specifically:

[0044] The skull thickness along the vertical scanning axis is defined by the outer surface of the skull, the inner surface and the surface angle, the outer surface of the skull is outlined by identifying the first significant amplitude peak in the ultrasonic signal, the normal vector is obtained by fitting the local plane according to the estimated outer surface of the skull, and the included angle θ i , i = 1, 2, …, n, the surface angle of each scanning position is obtained, and the effective thickness of the skull is calculated according to the surface angle;

[0045] The skull effective thickness calculation step is:

[0046]

[0047] Where b i is the effective thickness of the skull corresponding to each scanning point, d is the average thickness of the skull, and θ i is the included angle between the normal vector and the vertical scanning axis.

[0048] In a second aspect, the present application provides a system for a method of jointly photoacoustic and ultrasound depth-resolved multilayered cerebral microvessel imaging, comprising an image acquisition module, an image processing module, an intracranial and extracranial surface recognition module, a photoacoustic image structure tomography module;

[0049] The image acquisition module is configured to acquire photoacoustic and ultrasound images of the same cerebral region.

[0050] The image processing module is configured to separate additional scattering signals in the cerebral space by cross-correlating the ultrasound image pixel values with reference glass.

[0051] The intracranial and extracranial surface recognition module is configured to recognize the intracranial surface of the skull, the roughness evaluation process of the intracranial surface of the skull, and the average thickness calculation of the skull by combining photoacoustic images with ultrasound images.

[0052] The photoacoustic image structure tomography module is configured to segment and extract the multilayered microvessel network of the brain along the axial depth dimension in the photoacoustic image, and perform maximum projection.

[0053] Compared with the prior art, the present application has the following advantages and beneficial effects:

[0054] (1) The present application uses the method of cross-correlating the ultrasound image pixel values with the reference glass ultrasound echo signal image pixel values to separate the influence of the ultrasound transducer in the skull response, while ignoring the additional ultrasound scattering signals generated by the skull microvessel system and the skull marrow compartment, which helps to accurately identify the intracranial and extracranial surfaces of the skull.

[0055] (2) The present application establishes a skull angle correction uniform model to correct the thickness variation of different skull regions and the curvature variation of the skull.

[0056] (3) The present application accurately identifies the intracranial and extracranial surface positions of the skull by combining photoacoustic imaging and ultrasound imaging, accurately calculates the effective thickness of the skull, and can extract the cerebral microvessel network layer by layer, providing an effective technical means for the study of the cerebral vascular network. BRIEF DESCRIPTION OF DRAWINGS

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0058] Figure 1 A flowchart of a method of jointly photoacoustic and ultrasound depth-resolved multilayered cerebral microvessel imaging is provided for the embodiments of the present application.

[0059] Figure 2 This is a schematic diagram of the location of the selected point on the inner surface of the skull provided in an embodiment of the present invention, using a combination of photoacoustic ultrasound and manual segmentation.

[0060] Figure 3 This is a projection diagram of the maximum value of the cranial microvascular network provided in an embodiment of the present invention;

[0061] Figure 4 This is a maximum projection map of the superficial microvascular network of the cerebral cortex provided in an embodiment of the present invention;

[0062] Figure 5 This is a maximum projection map of the deep microvascular network in the cerebral cortex provided in an embodiment of the present invention. Detailed Implementation

[0063] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative effort are within the scope of protection of the present application.

[0064] Example

[0065] like Figure 1 As shown in this embodiment, a method for depth-resolved multilayer cranial microvascular network combining photoacoustic and ultrasound imaging is presented. Based on this method, the process employs cross-correlation between image pixel values ​​and a reference glass, uses photoacoustic and ultrasound to identify the inner surface of the skull, assesses the roughness of the inner surface, and establishes a uniform model for skull angle correction. This method performs layered extraction of the cranial microvascular network in the depth dimension, which is helpful for the research and analysis of the cranial microvascular network structure. The method of this invention specifically includes the following steps:

[0066] (1) Acquire 500*2000 ultrasound images and 500*500 photoacoustic images of the same cranial region;

[0067] (2) Image preprocessing: The acquired image dataset is first bandpass filtered between 1MHz and 100MHz using a first-order Butterworth filter, and the dataset is mapped onto a regular-interval transverse scanning network with nearest-neighbor interpolation. The preprocessed dataset is then stored.

[0068] (3) Cross-correlation is performed between the pixel values ​​of the ultrasound image and the pixel values ​​of the ultrasound echo signal from the reference glass;

[0069] Further, the shape and amplitude of the ultrasound echo signal depends not only on the structure of the skull and its acoustic impedance characteristics, but also on the position of the ultrasound transducer relative to the skull. In addition, the skull microvascular system and skull marrow compartment generate additional ultrasound scattering signals, which are displayed as trailing peaks after the first significant peak in the ultrasound image. These effects can be eliminated by cross-correlation. The ultrasound echo signal of a glass reference with the same or similar thickness as the skull is collected. For a single pixel on the scanning plane, the glass reference with the same or similar thickness as the skull is selected for cross-correlation. The cross-correlation is realized based on MATLAB.

[0070] (4) According to the difference in acoustic impedance characteristics between the skull and soft tissue, a high-amplitude echo signal is generated on the outer surface of the skull, and the position matrix of the outer surface of the skull is obtained in the processed ultrasound image;

[0071] Further, the skull outer surface identification step in the ultrasound image is specifically:

[0072] (4-1) For a single scanning point, the position of the maximum value of the cross-correlation of the ultrasound image pixel value and the reference glass is found to determine the depth position of the outer surface of the skull, and the remaining positive peak values in the cross-correlation function are stored.

[0073] (4-2) For all scanning points, repeat the process to generate the skull outer surface echo signal, determine its position, and obtain the skull outer surface position matrix.

[0074] It can be understood that in the skull outer surface identification process in the ultrasound image, the index value at the maximum value of the cross-correlation function of all scanning points is obtained to obtain the index value of the skull outer surface.

[0075] (5) As shown in Figure 2 , the skull inner surface is identified by combining the photoacoustic signal and the ultrasound signal. It is observed from the photoacoustic signal image whether there is a peak value corresponding to the blood vessel interfering with the identification of the skull inner surface.

[0076] Further, step (5) is specifically:

[0077] (5-1) The skull microvascular network region in the photoacoustic image is determined according to the skull outer surface position matrix in the ultrasound image.

[0078] Further, the skull microvascular network region in the photoacoustic image is determined, specifically:

[0079] (5-1-1) According to the 500*500 photoacoustic image and the 500*2000 ultrasound image, the collection points of the photoacoustic image and the ultrasound image are at the same position. The starting position of the skull microvascular network in the photoacoustic image is determined by the skull outer surface position index value in the ultrasound image, which can be calculated by the following formula:

[0080] (5-1-2) Set the index value of the initial position of the skull microvascular network in the photoacoustic image as p, the index value of the position of the outer surface of the skull in the ultrasound image as u, and the initial position as n. The calculation formula is

[0081] (5-2) According to the photoacoustic signal, it is judged whether the peak value corresponding to the skull blood vessel and other factors interfere with the recognition of the inner surface of the skull. If the scanning point corresponds to the position of the skull microvessel at this time, the inner surface position is divided into the first point after the first peak value in the photoacoustic image. If the scanning point does not correspond to the position of the skull microvessel at this time, the inner surface position is divided into a point before the first peak value in the photoacoustic image.

[0082] (6) There are different tissue structures in the inner surface of the skull. The roughness parameter of the inner surface of the skull is calculated to adjust the accuracy of the position of the inner surface of the skull.

[0083] Further, the calculation of the roughness of the inner surface of the skull is as follows:

[0084] (6-1) The roughness of the inner surface of the skull measures the heterogeneity inside the skull, which is reflected in the complex structure inside the skull, including blood vessels, bone marrow compartments and the inner surface of the skull. The roughness parameter of the inner surface of the skull reflects the non-uniformity level of the region of interest (ROI) (100x100 μm 2 ) in the ultrasound pulse echo signal, and is defined as the moving standard deviation of the second trailing peak.

[0085] (6-2) Select the left and right parietal bone regions of the skull as the region of interest (ROI). For the ultrasound signal of each scanning point in the region of interest, first calculate the distance between the second trailing peak and the first significant peak. The standard deviation of these distances can obtain the roughness parameter of the inner surface of the skull, which is used to adjust the accuracy of the estimation of the position of the inner surface of the skull.

[0086] (6-3) The calculation formula of the roughness of the inner surface of the skull is:

[0087] Where i represents the number of peak value serial numbers in the region of interest (ROI), n represents the number of peak values in the region of interest (ROI), z i is the distance from them to the outer surface of the skull, represents the average distance.

[0088] (7) Select the reliable points with clear ultrasound signal and photoacoustic signal characteristics in different brain regions. Using the obtained outer surface position matrix of the skull and the inner surface position matrix of the skull, the average thickness of the skull is calculated.

[0089] Further, the average thickness of the skull is calculated as follows:

[0090] (7-1) The skull average thickness calculation considers the thickness variation of different skull regions covering the whole brain. The skull average thickness calculation is performed in the frontal bone, anterior fontanel and plate bone regions. Three different scanning positions are selected in each different skull region for robust estimation of the average thickness;

[0091] (7-2) Taking the frontal bone region calculation as an example, three reliable points are selected on the frontal bone. The time dimension positions of the outer surface and the inner surface of the skull corresponding to each scanning point are determined by combining the ultrasound image and the photoacoustic image. The skull thickness is calculated by substituting the skull sound speed of 3695 m / s. The three reliable points can obtain three skull thicknesses, d1, d2 and d3, and the average thickness is

[0092] (7-3) The skull thickness calculation of the remaining regions repeats the above process, and then the average skull thickness of each region is added to obtain the average value of the overall skull region;

[0093] (8) For skull angle correction, the skull outer surface fitting local plane is converted into a least squares plane fitting estimation problem; a skull angle correction uniform model is established to calculate the effective thickness of the skull;

[0094] Further, the least squares plane fitting estimation step specifically comprises:

[0095] (8-1-1) Based on the nearest neighbor search, the nearest neighbor points are found;

[0096] (8-1-2) A covariance matrix is created at the nearest neighbor of the query point, and the covariance matrix is solved. The covariance matrix eigenvalues and eigenvectors are analyzed. The covariance matrix is:

[0097]

[0098] Where k is the number of p i nearest neighbor points of the point, is the three-dimensional centroid of the nearest neighbor point, λ j is the jth eigenvalue of the covariance matrix, is the jth eigenvector;

[0099] (8-1-3) The eigenvector corresponding to the smallest eigenvalue of the obtained covariance matrix is the vertex normal vector;

[0100] Further, a skull angle correction uniform model is established, specifically:

[0101] (8-2-1) The skull angle correction uniform model defines the skull thickness along the vertical scanning axis by the outer surface, the inner surface and the surface angle of the skull. By identifying the first significant amplitude peak in the ultrasound signal, the outer surface of the skull can be outlined;

[0102] (8-2-2) According to the estimated outer surface of the skull, a local plane is fitted, a normal vector of the obtained local plane is calculated, and an included angle θ between the normal vector and a vertical scanning axis direction vector is calculated i (i = 1, 2, …, n), the surface angle of each scanning position is obtained, and the effective thickness of the skull is calculated according to the surface angle;

[0103] The effective thickness calculation formula of the skull is:

[0104] Where b i is the effective thickness of the skull corresponding to each scanning point, d is the average thickness of the skull, and θ i is the included angle between the normal vector and the vertical scanning axis.

[0105] (9) According to the relationship between the ultrasonic signal and the photoacoustic signal in the time response, the depth position information of the skull microvessel network corresponding to the photoacoustic data is obtained through the obtained effective thickness of the skull along the vertical scanning axis direction.

[0106] (10) As shown in Figures 3-5 , it is the maximum value projection diagram of the imaging effect of the skull angle corrected skull brain microvessel network, the skull microvessel image, the cerebral cortex shallow microvessel network and the cerebral cortex deep microvessel network are segmented out according to the depth position information of the photoacoustic data, and maximum value projection is carried out respectively.

[0107] The embodiment of the application also provides a system for jointly imaging the photoacoustic ultrasonic depth-resolved multilayer skull brain microvessel, which specifically comprises: an image acquisition module, an image processing module, a skull inner and outer surface identification module, and a photoacoustic image structure tomography module.

[0108] The image acquisition module acquires photoacoustic and ultrasonic images of the same skull brain region.

[0109] The image processing module uses the cross correlation of the ultrasonic image pixel value and the reference glass to separate the additional scattering signal in the skull brain space.

[0110] The skull inner and outer surface identification module comprises a skull inner surface identification module, a skull inner surface roughness evaluation process, and a skull average thickness calculation module.

[0111] The photoacoustic image structure tomography module separates and extracts the multilayer microvessel network of the skull brain along the axial depth dimension in the photoacoustic image, enhances the image contrast using contrast limited adaptive histogram equalization (CLAHE), and performs maximum value projection.

[0112] It has to be remarked that in this specification, terms such as "comprise", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that processes, methods, articles, or apparatuses that comprise a list of elements are not limited to those elements, but can include other elements not expressly listed or inherent to such processes, methods, articles, or apparatuses. Without further limitation, an element preceded by "comprises a" does not, without more constraints, foreclose the existence of additional identical elements in the process, method, article, or apparatus that comprises the recited element.

[0113] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "an example", "a specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any appropriate manner in one or more embodiments or examples. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the present specification and the features of the different embodiments or examples without contradiction.

[0114] Although the embodiments of the present application have been shown and described above, it is to be understood that the above-described embodiments are exemplary, and are not to be construed as limiting the present application, and the person skilled in the art can make changes, modifications, replacements and variations to the above-described embodiments within the scope of the present application.

Claims

1. A method for depth-resolution of multilayer cranial microvascular bundles using a combination of photoacoustic and ultrasound methods, characterized in that, Includes the following steps: (1) Acquire raw ultrasound and photoacoustic images of the same cranial region; (2) Perform image preprocessing on the acquired ultrasound and photoacoustic raw images, and set thresholds to exclude noise signal interference based on the overall characteristics of the ultrasound raw images and the overall characteristics of the photoacoustic raw images; (3) Cross-correlate the pixel values ​​of the ultrasound image with the reference glass to obtain the cross-correlated ultrasound echo signal, thereby eliminating the influence of the different positions of the ultrasound transducer relative to the skull on the skull response. (4) Based on the difference in acoustic impedance characteristics between the skull and soft tissue, a high-amplitude echo signal is generated on the outer surface of the skull, and the position matrix of the outer surface of the skull is obtained in the processed ultrasound image. (5) The inner surface of the skull is identified by combining photoacoustic images and ultrasound images. The presence of peak factors corresponding to cranial blood vessels that interfere with the identification of the inner surface of the skull is determined based on the photoacoustic signals, and the position matrix of the inner surface of the skull is obtained. (6) Different tissue structures exist on the inner surface of the skull. The roughness parameters of the inner surface of the skull are calculated to adjust the accuracy of the position of the inner surface of the skull. (7) Select trustworthy points with clear ultrasound and photoacoustic signal characteristics in different brain regions, and use the obtained skull outer surface position matrix and skull inner surface position matrix to calculate the average thickness of the skull. (8) The least squares method was used to fit the local plane, a uniform model for skull angle correction was established, and the effective thickness of the skull was calculated. (9) Based on the given effective skull thickness along the vertical scanning axis, and according to the relationship between ultrasound and photoacoustic signals in time response, obtain the depth location information of the skull microvascular network corresponding to the photoacoustic data; (10) Based on the depth location information of the photoacoustic data, the skull microvascular image, the superficial microvascular network of the cerebral cortex and the deep microvascular network of the cerebral cortex are segmented and the maximum value projection is performed respectively.

2. The method for combined photoacoustic and ultrasound depth-resolved multilayer cranial microvascular system according to claim 1, characterized in that, In step (3), the cross-correlation between the ultrasonic image pixel values ​​and the reference glass is specifically as follows: A glass reference with the same or similar thickness as the skull is selected, and the ultrasonic echo signal of the reference glass is acquired. For a single pixel on the scanning plane, the ultrasonic echo signal on the skull and the ultrasonic echo signal of the reference glass are cross-correlated. The cross-correlation is implemented based on MATLAB.

3. The method for combined photoacoustic and ultrasound depth-resolved multilayer cranial microvascular system according to claim 1, characterized in that, In step (4), obtaining the position matrix of the outer surface of the skull specifically involves: Image pixel values ​​are cross-correlated with a reference glass. The maximum value of the cross-correlation function is used to determine the location of the outer surface of the skull, and the remaining positive peak values ​​in the cross-correlation function are stored. Repeat this process for all scanning points to generate echo signals from the outer surface of the skull and obtain the position matrix of the outer surface of the skull.

4. The method for combined photoacoustic and ultrasound depth-resolved multilayer cranial microvascular system according to claim 1, characterized in that, In step (5), the combination of photoacoustic and ultrasound images to identify the inner surface of the skull specifically involves: The position matrix of the outer surface of the skull in ultrasound images determines the region to which the skull microvascular network belongs in photoacoustic images; The photoacoustic signal is used to determine whether there are peak factors corresponding to cranial blood vessels that interfere with the identification of the inner surface of the skull. If there are cranial microvessels at the corresponding position of the scanning point, the inner surface position is defined as the first point after the first peak of the photoacoustic image. If there are no cranial microvessels at the corresponding position of the scanning point, the inner surface position is defined as the point before the first peak of the photoacoustic image.

5. The method for combined photoacoustic and ultrasound depth-resolved multilayer cranial microvascular system according to claim 1, characterized in that, The acquisition of the depth location information of the cranial microvascular network corresponding to the photoacoustic data specifically involves: Pulse echo ultrasound identifies the depth of the outer surface of the skull by focusing the transmitted and received beams. The photoacoustic signal is received by an ultrasound transducer located around the tissue. In the same imaging area, the location index of the ultrasound echo signal in the depth dimension is approximately twice that of the location index of the photoacoustic signal. This quantitative relationship is used to determine the starting position of the area to which the skull microvascular network belongs in the photoacoustic image.

6. The method for combined photoacoustic and ultrasound depth-resolution of multilayer cranial microvascular structures according to claim 1, characterized in that, In step (6), the assessment of the roughness of the inner surface of the skull is specifically as follows: The intracranial surface roughness measure reflects the heterogeneity within the skull, which is manifested in the complex structures within the skull, including blood vessels, bone marrow compartments, and the inner skull surface. The measurement of intracranial surface roughness reflects the level of non-uniformity of the skull in the region of interest (ROI), and is therefore defined as the moving standard deviation of the second trailing peak in the ultrasound pulse echo signal. For the ultrasound signal at each scan point of the region of interest (ROI), the distance between the second trailing peak and the first significant peak is first calculated. The standard deviation of these distances can be used to obtain the roughness parameter of the inner surface of the skull, which is used to adjust the accuracy of the position estimation of the inner surface of the skull. Skull inner surface roughness Where i represents the peak index within each region of interest (ROI). This represents the number of peak values ​​within the region of interest (ROI). The distance from the region of interest (ROI) to the outer surface of the skull. This indicates the average distance.

7. The method for combined photoacoustic and ultrasound depth-resolved multilayer cranial microvascular system according to claim 1, characterized in that, In step (7), the specific steps for calculating the average skull thickness are as follows: By combining ultrasound images and photoacoustic images, ultrasound images are used to determine the location of the outer surface of the skull, while photoacoustic images are used to determine whether there are peaks corresponding to blood vessels, thereby determining the location of the inner surface of the skull. Select A set of trust points with clear ultrasonic and photoacoustic signal characteristics are used to guide manual image segmentation. Each trust point corresponds to a thickness of [missing information]. Calculate the average thickness of the skull. .

8. The method for combined photoacoustic and ultrasound depth-resolved multilayer cranial microvascular system according to claim 1, characterized in that, In step (8), the least squares method is used to fit the local plane, specifically as follows: By estimating the obtained outer surface of the skull, fitting the local plane is transformed into a least squares plane fitting estimation problem. The least squares plane fitting estimation steps are as follows: Find nearest neighbor points based on nearest neighbor search; Create a covariance matrix in the nearest neighbors of the query point, calculate the covariance matrix, and analyze the eigenvectors and eigenvalues ​​of the covariance matrix. The covariance matrix is: in for The number of nearest neighbors of a point The three-dimensional centroid of the nearest neighbor point, The first covariance matrix is ​​the first... 1 eigenvalue, For the first 1 eigenvector; The eigenvector corresponding to the smallest eigenvalue of the obtained covariance matrix is ​​the vertex normal vector.

9. The method for combined photoacoustic ultrasound depth-resolution of multilayer cranial microvascular structures according to claim 1, characterized in that, In step (8), the uniform model for skull angle correction specifically refers to: The skull thickness along the vertical scanning axis is defined by the outer surface, inner surface, and surface angle of the skull. The outer surface of the skull is delineated by identifying the first significant amplitude peak in the ultrasound signal. Based on the estimated outer surface of the skull, its normal vector is obtained by fitting a local plane, and the angle between the normal vector and the direction vector of the vertical scanning axis is calculated. , The surface angle at each scanning position is obtained, and the effective thickness of the skull is calculated based on the surface angle. The steps for calculating the effective thickness of the skull are as follows: in The effective thickness of the skull corresponding to each scan point. The average thickness of the skull. The angle between the normal vector and the vertical scan axis.

10. A system based on the method of combined photoacoustic ultrasound depth-resolved multilayer cranial microvascular system according to any one of claims 1-9, characterized in that, It includes an image acquisition module, an image processing module, a skull inner and outer surface recognition module, and a photoacoustic image structure tomography module; The image acquisition module is used to acquire photoacoustic and ultrasound images of the same cranial region; The image processing module is used to cross-correlate the pixel values ​​of the ultrasound image with the reference glass to separate additional scattered signals in the cranial space. The skull inner and outer surface recognition module is used for recognizing the inner surface of the skull by combining photoacoustic images with ultrasound images, evaluating the roughness of the inner surface of the skull, and calculating the average thickness of the skull. The photoacoustic image structure tomography module is used to segment and extract the multilayer microvascular network of the brain along the axial depth dimension in the photoacoustic image, and perform maximum value projection.

Images