Method and system for photoacoustic microscopy imaging of brain electrode implantation sites and path planning

Abstract

The present invention provides a method and system for photoacoustic microscopy imaging of brain electrode implantation sites and path planning. The method comprises: an image extraction step: using a cerebral vascular enhancement extraction algorithm to extract input photoacoustic microscopy scanning imaging results to obtain cerebral vascular information; the cerebral vascular information includes two-dimensional image information and three-dimensional reconstruction results of the cerebral vessels; and an implantation site and path acquisition step: using a brain electrode implantation site path planning algorithm to process the cerebral vascular information to obtain the implantation site on the two-dimensional image and the implantation path in the three-dimensional structure. By utilizing the cerebral vascular enhancement extraction algorithm, the present invention solves the problems of high noise and poor imaging quality in photoacoustic scanning imaging, achieving precise extraction and three-dimensional reconstruction of cerebral vessels.

Description

Technical Field

[0001] The present invention relates to the field of medical image processing technology, and in particular to a method and system for photoacoustic microscopy imaging of brain electrode implantation sites and path planning thereof. Background Art

[0002] Electrode implantation techniques include single-needle electrode implantation and multi-needle electrode implantation (for example, two electrodes or four electrodes are implanted at a time). Compared with single-needle electrodes, multi-needle electrodes can usually obtain higher quality signals and are currently the most commonly used electrode implantation mode in animal experiments. In practical applications, before performing the electrode implantation operation, it is often necessary to reasonably plan the electrode implantation position (or implantation site). Taking brain electrode implantation as an example, compared with scalp electrodes, neural electrodes implanted in the cerebral cortex can record neuroelectrophysiological signals with a higher signal-to-noise ratio, which can help researchers understand changes in brain activity more effectively. The cerebral cortex has a large number of blood vessels. If the implantation position planning is not accurate enough, the electrode may be implanted into the vascular area, causing bleeding risks. Therefore, it is very important to ensure the accuracy of the planned electrode implantation site.

[0003] Photoacoustic microscopy (PAM) is a preoperative imaging technique based on photoacoustic microscopy (PAM). With a resolution of tens to dozens of microns, it can image both surface and deep vessels. It is a novel biomedical imaging technique commonly used for imaging blood vessels and other tissues. Under 532nm laser excitation, cerebral vessels exhibit strong contrast with surrounding brain tissue, enabling high-contrast, high signal-to-noise ratio imaging of brain vessels.

[0004] However, photoacoustic scanning imaging has high noise and unsatisfactory imaging quality, which may cause bleeding during surgery and even damage brain function.

[0005] In a Chinese patent document with publication number CN116196097A, a method and apparatus for electrode implantation site planning, a readable storage medium, and a terminal are disclosed. The method includes: determining an original image collected for an electrode implantation object; performing image segmentation on the original image to determine a non-implantation area and an implantable area; determining multiple first candidate implantation sites within the implantable area based on a first preset electrode distance; for each first candidate implantation site, determining a preliminary pairing point for the first candidate implantation site based on the first preset electrode distance; and screening each first candidate implantation site and its preliminary pairing point based on a second preset electrode distance to determine an electrode implantation site planning result. This patent document uses traditional imaging methods with low spatial resolution, which cannot capture all vascular information. It also only considers site planning on a two-dimensional image plane, and does not consider spatial path planning on a three-dimensional plane. Summary of the Invention

[0006] In view of the defects in the prior art, the purpose of the present invention is to provide a method and system for planning the implantation site and path of brain electrodes for photoacoustic microscopy.

[0007] According to the present invention, a method for photoacoustic microscopy brain electrode implantation site and path planning thereof includes:

[0008] Image extraction step: using a cerebral vascular enhancement extraction algorithm to extract the input photoacoustic microscopy scanning imaging results to obtain cerebral vascular information; the cerebral vascular information includes two-dimensional image information and three-dimensional reconstruction results of the cerebral vessels;

[0009] Implantation site and path acquisition steps: Use the brain electrode implantation site path planning algorithm to process cerebral vascular information to obtain the implantation site on the two-dimensional image and the implantation path in the three-dimensional structure.

[0010] Preferably, the image extraction step includes:

[0011] Step S1.1: Use the trained denoising model to denoise the acquired photoacoustic microscopy image to improve the signal-to-noise ratio;

[0012] Step S1.2: Enhance the photoacoustic microscopy image using the trained image enhancement model to correct blurred areas, areas where the resolution does not meet the preset value, and areas where the image is deformed, thereby improving the resolution of the photoacoustic microscopy image.

[0013] Step S1.3: Use the trained blood vessel segmentation model to segment the blood vessels in the photoacoustic microscopy image, remove background and other tissue structure information, and obtain two-dimensional image information of the brain blood vessels;

[0014] Step S1.4: The depth information of the photoacoustic microscopy imaging is combined with the two-dimensional image information of the cerebral blood vessels, and a back-projection calculation is performed to obtain a three-dimensional reconstruction result of the cerebral blood vessels.

[0015] Preferably, the denoising model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output the signal-to-noise ratio photoacoustic microscopy image with noise removed; the training of the deep neural network is based on supervised learning, and the supervised information is the denoising and noising process.

[0016] Preferably, the image enhancement model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output the photoacoustic microscopy image with enhanced resolution; the training of the deep neural network is based on supervised learning, and the supervised information is the restored difference between the original high-resolution image and the downsampled image.

[0017] Preferably, the blood vessel segmentation model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output a mask image corresponding to the blood vessel area; the training of the deep neural network is based on supervised learning, and the supervised information is a coarse segmentation annotation of the blood vessel.

[0018] Preferably, the implantation point and path acquisition step includes:

[0019] Step S2.1: Calculate a two-dimensional cerebral vascular probability density map to obtain implantation sites; select starting points for a given number of implantation sites in the two-dimensional image plane such that the distance between any two of them is greater than or equal to the specified shortest distance between adjacent sites. The optimization goal is to maximize the sum of the cumulative probability values ​​of all implantation sites, and the optimization condition is to maintain the distance between any two points greater than or equal to the shortest distance between adjacent sites.

[0020] Step S2.2: Calculate the three-dimensional cerebral vascular probability density map to obtain the implantation path; select the implantation direction starting from the determined implantation site in three-dimensional space. Under the condition of a specified implantation depth, the optimization goal is to maximize the sum of the cumulative probability values ​​along all implantation paths.

[0021] According to the present invention, a system for photoacoustic microscopy imaging of brain electrode implantation sites and path planning thereof includes:

[0022] Image extraction module: uses a cerebral vascular enhancement extraction algorithm to extract the input photoacoustic microscopy scanning imaging results to obtain cerebral vascular information; the cerebral vascular information includes 2D image information and 3D reconstruction results of cerebral vessels;

[0023] Implantation site and path acquisition module: Use the brain electrode implantation site path planning algorithm to process cerebral vascular information to obtain the implantation site on the two-dimensional image and the implantation path in the three-dimensional structure.

[0024] Preferably, the image extraction module includes:

[0025] Module M1.1: Use the trained denoising model to denoise the acquired photoacoustic microscopy images to improve the signal-to-noise ratio;

[0026] Module M1.2: Enhance photoacoustic microscopy images using a trained image enhancement model, correcting blurred areas, areas where the resolution does not meet the preset value, and areas with image deformation, thereby improving the resolution of photoacoustic microscopy images.

[0027] Module M1.3: Use the trained vascular segmentation model to segment blood vessels in photoacoustic microscopy images, remove background and other tissue structure information, and obtain two-dimensional image information of cerebral blood vessels;

[0028] Module M1.4: Combine the depth information of photoacoustic microscopy with the two-dimensional image information of cerebral blood vessels, and perform back-projection calculation to obtain the three-dimensional reconstruction results of cerebral blood vessels.

[0029] Preferably, the denoising model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output the signal-to-noise ratio photoacoustic microscopy image with noise removed; the training of the deep neural network is based on supervised learning, and the supervised information is the denoising and noising process.

[0030] Preferably, the image enhancement model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output the photoacoustic microscopy image with enhanced resolution; the training of the deep neural network is based on supervised learning, and the supervised information is the restored difference between the original high-resolution image and the downsampled image.

[0031] Preferably, the blood vessel segmentation model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output a mask image corresponding to the blood vessel area; the training of the deep neural network is based on supervised learning, and the supervised information is a coarse segmentation annotation of the blood vessel.

[0032] Preferably, the implantation point and path acquisition module includes:

[0033] Module M2.1: Calculate the two-dimensional cerebral vascular probability density map to obtain implantation sites; select the starting points for the specified number of implantation sites in the two-dimensional image plane so that the distance between any two of them is greater than or equal to the specified minimum distance between adjacent sites. The optimization goal is to maximize the sum of the cumulative probability values ​​of all implantation sites, and the optimization condition is to keep the distance between any two points greater than or equal to the minimum distance between adjacent sites.

[0034] Module M2.2: Calculate the three-dimensional cerebral vascular probability density map to obtain the implantation path; select the implantation direction in three-dimensional space starting from the determined implantation site. Under the condition of specifying the implantation depth, the optimization goal is to maximize the sum of the cumulative probability values ​​along all implantation paths.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] 1. The present invention solves the problems of high noise and poor imaging quality in photoacoustic scanning by adopting a brain blood vessel enhancement extraction algorithm, and realizes the accurate extraction and three-dimensional reconstruction of brain blood vessels.

[0037] 2. Based on the extracted cerebral blood vessels and three-dimensional reconstruction, the present invention adopts a brain electrode implantation site path planning algorithm to avoid blood vessels during the brain electrode implantation process, reducing bleeding during the operation and potential damage to brain function, thereby improving the safety and success rate of the operation and providing great help to doctors.

[0038] Other beneficial effects of the present invention will be explained through the introduction of specific technical features and technical solutions in the specific implementation methods. Those skilled in the art should be able to understand the beneficial technical effects brought about by the introduction of these technical features and technical solutions. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Other features, objects and advantages of the present invention will become more apparent upon reading the detailed description of non-limiting embodiments with reference to the following drawings:

[0040] Figure 1 Flow chart of the method of the present invention.

[0041] Figure 2 This is a flow chart of the cerebral blood vessel enhancement extraction algorithm of the present invention.

[0042] Figure 3 This is a flow chart of the brain electrode implantation site path planning algorithm of the present invention. DETAILED DESCRIPTION

[0043] The present invention will be described in detail below with reference to specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those skilled in the art, several changes and improvements can be made without departing from the scope of the present invention. These all fall within the scope of protection of the present invention.

[0044] Reference Figure 1 As shown, a method for photoacoustic microscopy imaging of brain electrode implantation sites and path planning thereof includes:

[0045] Image extraction step: using a cerebral vascular enhancement extraction algorithm to extract the input photoacoustic microscopy scanning imaging results to obtain cerebral vascular information; the cerebral vascular information includes two-dimensional image information and three-dimensional reconstruction results of the cerebral vessels;

[0046] Specifically:

[0047] Step S1.1: Use the trained denoising model to denoise the acquired photoacoustic microscopy image to improve the signal-to-noise ratio;

[0048] Step S1.2: Enhance the photoacoustic microscopy image using the trained image enhancement model to correct blurred areas, areas where the resolution does not meet the preset value, and areas where the image is deformed, thereby improving the resolution of the photoacoustic microscopy image.

[0049] Step S1.3: Use the trained blood vessel segmentation model to segment the blood vessels in the photoacoustic microscopy image, remove background and other tissue structure information, and obtain two-dimensional image information of the brain blood vessels;

[0050] Step S1.4: The depth information of the photoacoustic microscopy imaging is combined with the two-dimensional image information of the cerebral blood vessels, and a back-projection calculation is performed to obtain a three-dimensional reconstruction result of the cerebral blood vessels.

[0051] Among them, the denoising model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output the signal-to-noise ratio photoacoustic microscopy image with noise removed; the training of the deep neural network is based on supervised learning, and the supervised information is the denoising and noisy process.

[0052] The image enhancement model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output the photoacoustic microscopy image with enhanced resolution; the training of the deep neural network is based on supervised learning, and the supervision information is the restored difference between the original high-resolution image and the downsampled image.

[0053] The vascular segmentation model includes a deep neural network that inputs the original photoacoustic microscopy image and outputs a mask image corresponding to the vascular area; the training of the deep neural network is based on supervised learning, and the supervised information is the coarse segmentation annotation of the blood vessels.

[0054] Implantation site and path acquisition steps: Use the brain electrode implantation site path planning algorithm to process cerebral vascular information to obtain the implantation site on the two-dimensional image and the implantation path in the three-dimensional structure.

[0055] Specifically:

[0056] Step S2.1: Calculate a two-dimensional cerebral vascular probability density map to obtain implantation sites; select starting points for a given number of implantation sites in the two-dimensional image plane such that the distance between any two of them is greater than or equal to the specified shortest distance between adjacent sites. The optimization goal is to maximize the sum of the cumulative probability values ​​of all implantation sites, and the optimization condition is to maintain the distance between any two points greater than or equal to the shortest distance between adjacent sites.

[0057] Step S2.2: Calculate the three-dimensional cerebral vascular probability density map to obtain the implantation path; select the implantation direction starting from the determined implantation site in three-dimensional space. Under the condition of a specified implantation depth, the optimization goal is to maximize the sum of the cumulative probability values ​​along all implantation paths.

[0058] This invention addresses the issues of high noise and poor imaging quality associated with photoacoustic scanning by employing an enhanced cerebral vascular extraction algorithm, enabling precise extraction and three-dimensional reconstruction of cerebral vessels. Based on the extracted cerebral vessels and 3D reconstruction, a path planning algorithm for the electrode implantation site is employed to avoid these vessels during the implantation process, reducing bleeding and potential damage to brain function during the procedure, thereby improving surgical safety and success rates.

[0059] The above is a basic embodiment of the present invention. The technical solution of the present invention is further described below through a preferred embodiment.

[0060] Example 1

[0061] Reference Figure 1 As shown, a method for photoacoustic microscopy imaging of brain electrode implantation sites and path planning thereof includes:

[0062] Step S1: executing a cerebral blood vessel enhancement extraction algorithm on the input photoacoustic microscopy scanning imaging result to extract and obtain two-dimensional image information and three-dimensional reconstruction results of the cerebral blood vessels.

[0063] Reference Figure 2 As shown in the figure, the process of the cerebral vascular enhancement extraction algorithm is as follows: First, a denoising operation is performed on the original acquired photoacoustic microscopy image. The trained denoising model is used to remove noise caused by background interference, instrument electromagnetic radiation, and imaging tissue movement, thereby improving the signal-to-noise ratio of the photoacoustic microscopy image. Next, the trained image enhancement model is used to enhance the photoacoustic microscopy image, correcting blurred imaging areas, areas with low resolution, and areas of imaging deformation to improve the resolution of the photoacoustic microscopy image. Furthermore, the trained vascular segmentation model is used to segment the blood vessels in the photoacoustic microscopy image, removing background and other tissue structure information, and obtaining two-dimensional image information of the cerebral blood vessels. Finally, the depth information of the photoacoustic microscopy imaging is combined with the two-dimensional cerebral blood vessel image information obtained in the previous step, and the back-projection calculation is performed to obtain the three-dimensional reconstruction result of the cerebral blood vessels.

[0064] The denoising model is implemented using a deep neural network. Its function is to input the original photoacoustic microscopy image and output a high signal-to-noise ratio photoacoustic microscopy image with noise removed. The training of the deep neural network is based on supervised learning, with the supervisory information being the denoising and noising processes.

[0065] The image enhancement model is based on a deep neural network. It takes as input the original photoacoustic microscopy image and outputs a photoacoustic microscopy image with enhanced resolution. The deep neural network is trained using supervised learning, where the supervisory information is the difference between the original high-resolution image and the downsampled image.

[0066] The vessel segmentation model is based on a deep neural network. It takes as input the original photoacoustic microscopy image and outputs a mask image corresponding to the vessel region. The deep neural network is trained using supervised learning, with the supervised information being the rough segmentation annotations of the vessels.

[0067] Step S2: For the extracted cerebral vascular information, a brain electrode implantation site path planning algorithm is executed to obtain the implantation site on the two-dimensional image and the implantation path in the three-dimensional structure.

[0068] Implantation sites are selected based on a two-dimensional cerebral vascular probability density map. The primary calculation method involves selecting starting points for a specific number of implantation sites within the two-dimensional image plane, ensuring that the distance between any two sites is greater than or equal to the specified minimum distance between adjacent sites. The optimization objective is to maximize the sum of the cumulative probability values ​​for all implantation sites, and the optimization condition is to maintain a distance between any two sites greater than or equal to the minimum distance between adjacent sites.

[0069] The implantation path is selected based on a three-dimensional cerebral vascular probability density map. The main calculation method involves selecting an implantation direction in three-dimensional space, starting from the designated implantation site. Under a specified implantation depth, the optimization goal is to maximize the sum of the cumulative probability values ​​along all implantation paths.

[0070] Reference Figure 3 As shown, the path planning algorithm for brain electrode implantation sites includes implantation site planning based on two-dimensional images and path planning based on three-dimensional reconstruction results. Based on the two-dimensional image information of cerebral blood vessels, the probability density distribution map of the implantation site is calculated. The calculation is mainly based on the image distance between each pixel in the image and the nearest blood vessel. In the two-dimensional coordinate system of the image, the closer the distance to the blood vessel, the smaller the probability of the pixel corresponding to the implantation site. The implantation site probability distribution map obtained in this way reflects the distance between each implantation site and the nearest blood vessel. The greater the corresponding probability, the farther away from the blood vessel, the safer it is during the implantation process, and the lower the probability of penetrating the cerebral blood vessels. After artificially setting the number of implantation sites and the shortest distance between adjacent sites, the implantation sites that meet the requirements are selected based on the implantation site probability density map to realize implantation site planning based on two-dimensional images. The main calculation formula is as follows:

[0071] Based on the two-dimensional image information of cerebral blood vessels, the probability density distribution map M is first calculated. For each pixel point (u, v) in the distribution map, its probability density value is defined as proportional to the distance d(u, v) to the nearest blood vessel, that is,

[0072] M(u,v)=kd(u,v)

[0073] And normalize the image, the normalized image is Normalization is calculated by the following formula:

[0074]

[0075] After calculating the corresponding normalized two-dimensional probability density distribution map, N implantation sites are randomly generated in the area where the initial threshold is set to be greater than 0.5, where N represents the number of implantation sites set manually, and the corresponding implantation site position is defined as At the same time, the shortest distance between the two implantation sites is set to s, and the implantation site P is optimized based on the above conditions:

[0076]

[0077] Finally, the optimal implantation site planning is obtained through optimization iteration

[0078] Similarly, based on the 3D reconstruction of cerebral vasculature, a probability density distribution map of the implantation path is calculated. This calculation is primarily based on the spatial distance between each voxel point and the nearest blood vessel in the 3D voxel space. Combining this with the completed 2D implantation site planning, after manually determining the implantation depth, a 3D implantation path that meets the requirements is calculated, preserving the avoidance of blood vessels. This allows for path planning based on the 3D reconstruction results.

[0079] In an in vivo animal experiment on brain electrode implantation, the method provided by the present invention achieved a bleeding site of less than 150 microns for each electrode implantation, bringing the surgical success rate to over 80%.

[0080] The present invention also provides a system for photoacoustic microscopy imaging of brain electrode implantation sites and their path planning. The system for photoacoustic microscopy imaging of brain electrode implantation sites and their path planning can be implemented by executing the process steps of the method for photoacoustic microscopy imaging of brain electrode implantation sites and their path planning. That is, those skilled in the art can understand the method for photoacoustic microscopy imaging of brain electrode implantation sites and their path planning as a preferred embodiment of the system for photoacoustic microscopy imaging of brain electrode implantation sites and their path planning.

[0081] Specifically, a system for photoacoustic microscopy imaging of brain electrode implantation sites and path planning thereof includes:

[0082] Image extraction module: uses a cerebral vascular enhancement extraction algorithm to extract the input photoacoustic microscopy scanning imaging results to obtain cerebral vascular information; the cerebral vascular information includes 2D image information and 3D reconstruction results of cerebral vessels;

[0083] Implantation site and path acquisition module: Use the brain electrode implantation site path planning algorithm to process cerebral vascular information to obtain the implantation site on the two-dimensional image and the implantation path in the three-dimensional structure.

[0084] The image extraction module includes:

[0085] Module M1.1: Use the trained denoising model to denoise the acquired photoacoustic microscopy images to improve the signal-to-noise ratio;

[0086] Module M1.2: Enhance photoacoustic microscopy images using a trained image enhancement model, correcting blurred areas, areas where the resolution does not meet the preset value, and areas with image deformation, thereby improving the resolution of photoacoustic microscopy images.

[0087] Module M1.3: Use the trained vascular segmentation model to segment blood vessels in photoacoustic microscopy images, remove background and other tissue structure information, and obtain two-dimensional image information of cerebral blood vessels;

[0088] Module M1.4: Combine the depth information of photoacoustic microscopy with the two-dimensional image information of cerebral blood vessels, and perform back-projection calculation to obtain the three-dimensional reconstruction results of cerebral blood vessels.

[0089] The denoising model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output the signal-to-noise ratio photoacoustic microscopy image with noise removed; the training of the deep neural network is based on supervised learning, and the supervisory information is the denoising and noising process.

[0090] The image enhancement model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output the photoacoustic microscopy image with enhanced resolution; the training of the deep neural network is based on supervised learning, and the supervision information is the restored difference between the original high-resolution image and the downsampled image.

[0091] The blood vessel segmentation model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output a mask image corresponding to the blood vessel area; the training of the deep neural network is based on supervised learning, and the supervised information is the coarse segmentation annotation of the blood vessel.

[0092] The implantation point and path acquisition module includes:

[0093] Module M2.1: Calculate the two-dimensional cerebral vascular probability density map to obtain implantation sites; select the starting points for the specified number of implantation sites in the two-dimensional image plane so that the distance between any two of them is greater than or equal to the specified minimum distance between adjacent sites. The optimization goal is to maximize the sum of the cumulative probability values ​​of all implantation sites, and the optimization condition is to keep the distance between any two points greater than or equal to the minimum distance between adjacent sites.

[0094] Module M2.2: Calculate the three-dimensional cerebral vascular probability density map to obtain the implantation path; select the implantation direction in three-dimensional space starting from the determined implantation site. Under the condition of specifying the implantation depth, the optimization goal is to maximize the sum of the cumulative probability values ​​along all implantation paths.

[0095] Those skilled in the art will appreciate that, in addition to implementing the system and its various devices, modules, and units provided by the present invention in purely computer-readable program code, it is entirely possible to implement the same functions of the system and its various devices, modules, and units provided by the present invention in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered a hardware component, and the devices, modules, and units included therein for implementing various functions can also be considered as structures within the hardware component; the devices, modules, and units for implementing various functions can also be considered as both software modules implementing the method and structures within the hardware component.

[0096] The above describes specific embodiments of the present invention. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention. The embodiments of this application and the features in the embodiments may be combined with each other in any manner unless there is a conflict.

Claims

1. A method for photoacoustic microscopy imaging of brain electrode implantation sites and path planning, characterized in that: include: Image extraction step: using a cerebral vascular enhancement extraction algorithm to extract the input photoacoustic microscopy scanning imaging results to obtain cerebral vascular information; The cerebral blood vessel information includes two-dimensional image information and three-dimensional reconstruction results of the cerebral blood vessels; Implantation site and path acquisition steps: Use the brain electrode implantation site path planning algorithm to process cerebral vascular information to obtain the implantation site on the two-dimensional image and the implantation path in the three-dimensional structure; The image extraction step comprises: Step S1.1: Use the trained denoising model to denoise the acquired photoacoustic microscopy image to improve the signal-to-noise ratio; Step S1.2: Enhance the photoacoustic microscopy image using the trained image enhancement model to correct blurred areas, areas where the resolution does not meet the preset value, and areas where the image is deformed, thereby improving the resolution of the photoacoustic microscopy image. Step S1.3: Use the trained blood vessel segmentation model to segment the blood vessels in the photoacoustic microscopy image, remove background and other tissue structure information, and obtain two-dimensional image information of the brain blood vessels; Step S1.4: combining the depth information of the photoacoustic microscopy imaging with the two-dimensional image information of the cerebral blood vessels, and performing back-projection calculation to obtain a three-dimensional reconstruction result of the cerebral blood vessels; The implantation site and path acquisition step includes: Step S2.1: Calculate a two-dimensional cerebral vascular probability density map to obtain implantation sites; select starting points for a given number of implantation sites in the two-dimensional image plane so that the distance between any two of them is greater than or equal to the specified shortest distance between adjacent sites. The optimization goal is to maximize the sum of the cumulative probability values ​​of all implantation sites, and the optimization condition is to maintain the distance between any two points greater than or equal to the shortest distance between adjacent sites. Step S2.2: Calculate the three-dimensional cerebral vascular probability density map to obtain the implantation path; select the implantation direction starting from the determined implantation site in three-dimensional space. Under the condition of a specified implantation depth, the optimization goal is to maximize the sum of the cumulative probability values ​​along all implantation paths.

2. The method for photoacoustic microscopy brain electrode implantation site and path planning according to claim 1, characterized in that: The denoising model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output the signal-to-noise ratio photoacoustic microscopy image with noise removed; the training of the deep neural network is based on supervised learning, and the supervisory information is the denoising and noising process.

3. The method for photoacoustic microscopy brain electrode implantation site and path planning according to claim 1, characterized in that: The image enhancement model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output the photoacoustic microscopy image with enhanced resolution; the training of the deep neural network is based on supervised learning, and the supervision information is the restored difference between the original high-resolution image and the downsampled image.

4. The method for photoacoustic microscopy brain electrode implantation site and path planning according to claim 1, characterized in that: The blood vessel segmentation model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output a mask image corresponding to the blood vessel area; the training of the deep neural network is based on supervised learning, and the supervised information is the coarse segmentation annotation of the blood vessel.

5. A system for photoacoustic microscopy of brain electrode implantation sites and path planning, characterized in that: include: Image extraction module: uses the cerebral vascular enhancement extraction algorithm to extract the input photoacoustic microscopy scanning imaging results to obtain cerebral vascular information; The cerebral blood vessel information includes two-dimensional image information and three-dimensional reconstruction results of the cerebral blood vessels; Implantation site and path acquisition module: uses the brain electrode implantation site path planning algorithm to process cerebral vascular information to obtain the implantation site on the two-dimensional image and the implantation path in the three-dimensional structure; The image extraction module includes: Module M1.1: Use the trained denoising model to denoise the acquired photoacoustic microscopy images to improve the signal-to-noise ratio; Module M1.2: Enhance photoacoustic microscopy images using a trained image enhancement model, correcting blurred areas, areas where the resolution does not meet the preset value, and areas with image deformation, thereby improving the resolution of photoacoustic microscopy images. Module M1.3: Use the trained vascular segmentation model to segment blood vessels in photoacoustic microscopy images, remove background and other tissue structure information, and obtain two-dimensional image information of cerebral blood vessels; Module M1.4: Combine the depth information of photoacoustic microscopy with the two-dimensional image information of cerebral blood vessels, and perform back-projection calculation to obtain the three-dimensional reconstruction results of cerebral blood vessels; The implantation site and path acquisition module includes: Module M2.1: Calculate the two-dimensional cerebral vascular probability density map to obtain implantation sites; select the starting points for the specified number of implantation sites in the two-dimensional image plane so that the distance between any two points is greater than or equal to the specified minimum distance between adjacent sites. The optimization goal is to maximize the sum of the cumulative probability values ​​of all implantation sites, and the optimization condition is to maintain the distance between any two points greater than or equal to the minimum distance between adjacent sites. Module M2.2: Calculate the three-dimensional cerebral vascular probability density map to obtain the implantation path; select the implantation direction in three-dimensional space starting from the determined implantation site. Under the condition of specifying the implantation depth, the optimization goal is to maximize the sum of the cumulative probability values ​​along all implantation paths.

6. The system for photoacoustic microscopy brain electrode implantation site and path planning according to claim 5, characterized in that: The denoising model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output the signal-to-noise ratio photoacoustic microscopy image with noise removed; the training of the deep neural network is based on supervised learning, and the supervisory information is the denoising and noising process.

7. The system for photoacoustic microscopy brain electrode implantation site and path planning according to claim 5, characterized in that: The image enhancement model includes a deep neural network, which is used to input the original photoacoustic microscopy image and output the photoacoustic microscopy image with enhanced resolution; the training of the deep neural network is based on supervised learning, and the supervision information is the restored difference between the original high-resolution image and the downsampled image.

Images