A visualized gland injection system

Abstract

The application relates to the technical field of medicine, and particularly discloses a visualized gland injection system, which comprises an ultrasonic detection module, a visualized construction module, an evaluation module and a planning module.The ultrasonic detection module comprises a detection ultrasonic probe, is used for scanning and two-dimensional imaging of a target region by using the detection ultrasonic probe, and moves the ultrasonic probe in different directions, and obtains a scanning direction in real time, and obtains a two-dimensional image for each scanning direction.The visualized construction module stacks the two-dimensional images according to the scanning directions in sequence to generate a three-dimensional sign model.The evaluation module evaluates the injection risk of the three-dimensional sign model to judge whether injection can be performed.The planning module plans an injection region, an injection starting point, an injection ending point and an initial injection direction according to the three-dimensional sign model.The application can provide comprehensive target region information and evaluate the injection risk through the synergistic effect of the ultrasonic detection, the visualized construction, the evaluation and the planning modules, so that the accuracy and safety of gland injection are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the medical technology field, in particular, it relates to a visual gland injection system. BACKGROUND

[0002] The submandibular gland is a pair of large salivary glands located below the mandible; the submandibular gland is one of the largest salivary glands in the human body, located on the inner side and below the mandible, at the angle of the mandible.

[0003] In the prior art, it is found that local injection of drugs into the submandibular gland can treat salivary cysts, initial bite syndrome, and can also be used to treat sialadenitis, however, the submandibular gland is located in the deep part of the face, surrounded by complex anatomical structures such as nerves, blood vessels and other important organs, the inserted needle can damage the blood vessel wall, and arterial blood or venous blood can enter the parotid gland tissue, causing hematoma, compression of the facial nerve or obstruction of saliva flow. Measured from the skin surface of the submandibular area, the depth of the facial blood vessels is between 10.0 and 26.0 millimeters, while the depth of the facial veins is between 10.9 and 21.0 millimeters, and if the drug is given at the intraluminal position, it can cause blood vessel damage or the product to be flushed into the blood, reducing the efficacy.

[0004] In addition, according to the study, the distance between the two glands varies greatly depending on gender, age and body mass index, the older the age, the higher the body mass index, the farther the distance, the distance of the parotid gland ranges from 1.0 to 10.1 millimeters, while the distance of the submandibular gland ranges from 1.0 to 17.3 millimeters, which shows that whether it can be safely injected depends more on the patient's personal data, however, there are important nerve structures near the submandibular gland area, and such individual characteristic directional injection has the potential risk of nerve damage.

[0005] Therefore, there is an urgent need to design a visual gland injection system to solve the technical problem of the prior art that the submandibular gland is located in the deep part of the face, and there are various injection risks. SUMMARY

[0006] In view of this, the present application provides a visual gland injection system, which aims to solve the technical problem in the prior art that the submandibular gland is located in the deep part of the face, and there are various injection risks.

[0007] The present application provides a visual gland injection system, comprising:

[0008] An ultrasonic detection module comprising a detection ultrasonic probe, the ultrasonic detection module is used to scan and two-dimensionally image a target region using the detection ultrasonic probe, and move the ultrasonic probe in different directions and acquire a scanning direction in real time, acquire a two-dimensional image for each scanning direction, the frequency of the detection ultrasonic probe is 7.5-12 MHz, and the target region comprises a corner of a mandible and a rear of the mandible;

[0009] A visualization constructing module, which stacks the two-dimensional images in sequence according to the scanning directions to generate a three-dimensional sign model;

[0010] An evaluation module, which evaluates an injection risk of the three-dimensional sign model to determine whether injection can be performed;

[0011] A planning module, which plans an injection region, an injection starting point, an injection ending point and an initial injection direction according to the three-dimensional sign model;

[0012] An injection module comprising a medical injector, the injection module is used to perform injection of a first depth using the medical injector according to the injection starting point and the initial injection direction;

[0013] An ultrasonic intervention module comprising an intervention ultrasonic probe, the ultrasonic intervention module is used to scan the injection region using the intervention ultrasonic probe to generate a guide image in real time, and the frequency of the intervention ultrasonic probe is 5-7.5 MHz;

[0014] A guide module, which continuously generates a planned avoidance angle according to the guide image until the injection is completed.

[0015] Preferably, the visualization constructing module performs data processing on each two-dimensional image, which comprises removing noise, enhancing contrast, edge detection, stacking each two-dimensional image in sequence according to its scanning direction, and converting into three-dimensional voxel data, each voxel representing image information of a specific position.

[0016] Preferably, the visualization constructing module stacks each two-dimensional image in sequence according to its scanning direction and converts into three-dimensional voxel data, which comprises:

[0017] Predefining a size and a resolution of a voxel, creating an empty three-dimensional array as three-dimensional sign data according to the size and the resolution of the voxel, traversing each pixel in the two-dimensional data, calculating a voxel coordinate of the pixel in the three-dimensional sign data according to the scanning direction and the position of the pixel, and mapping the pixel to the voxel coordinate.

[0018] Preferably, the visualization construction module pre-processes the three-dimensional sign data, extracts physiological features including shape, geometric properties, texture, color, encodes the physiological features using Bag-of-Features, analyzes and classifies the extracted physiological features using a machine learning algorithm, and generates the three-dimensional sign model, which is labeled with blood vessel distribution, soft tissue hierarchy, nerve tissue, submandibular gland, and position information and area information of each physiological feature.

[0019] Preferably, the evaluation module evaluates the injection risk of the three-dimensional sign model and determines whether injection can be performed, including:

[0020] The position information of the submandibular gland is recorded as a target position Y, the target position Y is compared with a pre-set safe injection area, and the injection risk of the submandibular gland is evaluated according to the comparison;

[0021] When the target position Y belongs to the safe injection area, the evaluation result is set as no injection risk, and it is determined that injection can be directly performed;

[0022] When the target position Y does not belong to the safe injection area, the evaluation result is set as injection risk, and it is determined that ultrasound intervention is needed for injection.

[0023] Preferably, the evaluation module evaluates the injection risk of the three-dimensional sign model and determines whether injection can be performed, including:

[0024] When the target position Y does not belong to the safe injection area, the area information of the submandibular gland is recorded as a submandibular gland area K, the submandibular gland area K is compared with a pre-set first submandibular gland area K1 and a second submandibular gland area K2, wherein K1

[0025] When K≤K1, it is determined that injection cannot be performed;

[0026] When K1

[0027] When K2

[0028] Preferably, the planning module plans an injection area, an injection starting point, an injection ending point, and an initial injection direction according to the three-dimensional sign model, including:

[0029] The three-dimensional sign model is inputted to avoid the blood vessel distribution, soft tissue level and nerve tissue, a preferred path with least angle change from the skin surface to the submandibular gland is obtained by using a heuristic search algorithm, the starting point of the preferred path is set as the injection starting point, the direction of the path is set as the initial injection direction, and the end point of the path is set as the injection end point.

[0030] Preferably, the planning module uses a heuristic search algorithm to obtain a preferred path with least angle change from the skin surface to the submandibular gland, which comprises the following steps:

[0031] The three-dimensional sign model is converted into a graph representation, and nodes and edges are defined, wherein the nodes are defined as positions in the model, and the edges are defined as connection relationships between the nodes;

[0032] The component defines a heuristic function to evaluate the priority of each node, calculates an expected cost estimate value of a node to a target position, the expected cost estimate value is used to represent the expected cost from the node to the target position, and the expected cost estimate value is generated based on the physiological characteristic factors, wherein a penalty term is added to the blood vessels, soft tissues and nerve tissues, so that the search algorithm selects a path that bypasses these structures;

[0033] A preset test injection starting point is set as a starting node, and a preset test injection end point is set as a target node;

[0034] The distance of the starting node is set to 0, and the priority of the starting node is set to the expected cost estimate value, a heuristic search algorithm is used for searching, the nodes are traversed in the order of high to low priority, and the distance and priority of the nodes are updated, the neighbor nodes of the nodes are traversed, the new distance value is calculated, and the priority of the nodes is updated, so as to obtain a preferred path with least angle change from the skin surface to the submandibular gland.

[0035] Preferably, when the first injection risk is determined, a straight line distance U1 between the injection starting point and the injection end point is obtained, a first injection depth proportion factor q1 is generated according to the target position Y, and the first depth is set as q1*U1, wherein 0

[0036] When the second injection risk is determined, a straight line distance U2 between the injection starting point and the injection end point is obtained, a second injection depth proportion factor q2 is generated according to the target position Y, and the first depth is set as q2*U2, wherein 0

[0037] Compared with the prior art, the present application has the following beneficial effects:

[0038] The two-dimensional image can be converted into a three-dimensional sign model through the visualization construction module, more intuitive injection area information is provided, the evaluation module can evaluate the injection risk through analysis of the submandibular gland position information and area information, and the safety of injection is improved, the planning module can plan an injection path according to the three-dimensional sign model, the influence on surrounding tissues is reduced, and the injection accuracy is improved, the ultrasound intervention module can generate a guide image in real time, visualized guidance is provided, and the operation accuracy is increased, and the guide module can continuously generate a planned avoidance angle, and the safety in the injection process is ensured.

[0039] In conclusion, the visualized gland injection system provided by the application constructs a three-dimensional sign model based on deep analysis of individual characteristics, and formulating an injection plan according to the three-dimensional sign model helps to improve injection accuracy and safety. BRIEF DESCRIPTION OF DRAWINGS

[0040] Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The accompanying drawings are included to provide a description of preferred embodiments, and are not meant to limit the present application. Furthermore, the same reference numerals in different drawings refer to the same or similar components. In the drawings:

[0041] Figure 1 A functional block diagram of the visualized gland injection system provided by the embodiment of the application is shown. DETAILED DESCRIPTION

[0042] Exemplary embodiments of the present disclosure will be described in detail with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments described herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be accurately conveyed to those skilled in the art. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict. The present application will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0043] Reference Figure 1 As shown in the drawings, the embodiment provides a visualized gland injection system, which comprises:

[0044] The ultrasound detection module comprises a detection ultrasound probe, and is used for scanning and two-dimensional imaging of a target area by using the detection ultrasound probe, moving the ultrasound probe in different directions, and acquiring a two-dimensional image for each scanning direction in real time. The frequency of the detection ultrasound probe is 7.5-12 MHz, and the target area comprises the corner of the mandible and the rear of the mandible.

[0045] a visualization construction module, which stacks two-dimensional images in sequence according to a scanning direction to generate a three-dimensional sign model;

[0046] an evaluation module, which evaluates an injection risk of the three-dimensional sign model to determine whether injection can be performed;

[0047] a planning module, which plans an injection area, an injection starting point, an injection ending point and an initial injection direction according to the three-dimensional sign model;

[0048] an injection module, which includes a medical injector, and is configured to perform injection of a first depth using the medical injector according to the injection starting point and the initial injection direction;

[0049] an ultrasound intervention module, which includes an intervention ultrasound probe, and is configured to scan the injection area using the intervention ultrasound probe to generate a guide image in real time, and the intervention ultrasound probe has a frequency of 5-7.5 MHz;

[0050] a guide module, which continuously generates a planned avoidance angle according to the guide image until the injection is completed.

[0051] Specifically, the medical injector of the embodiment is preferably a 1ml fine needle medical injector.

[0052] In some embodiments of the present application, the data processing of each two-dimensional image by the visualization construction module includes removing noise, enhancing contrast, and edge detection, and each two-dimensional image is stacked in sequence according to its scanning direction to be converted into three-dimensional voxel data, each voxel representing image information of a specific position.

[0053] In some embodiments of the present application, the visualization construction module stacks each two-dimensional image in sequence according to its scanning direction to be converted into three-dimensional voxel data, including:

[0054] The size and resolution of a preset voxel are determined, an empty three-dimensional array is created as three-dimensional sign data according to the size and resolution of the voxel, each pixel in the two-dimensional data is traversed, the voxel coordinates of each pixel in the three-dimensional sign data are calculated according to the position of the pixel and the scanning direction, and the pixel is mapped to each voxel coordinate.

[0055] In some embodiments of the present application, the visualization construction module pre-processes the three-dimensional sign data to extract physiological features therefrom, the physiological features include shape, geometric properties, texture, and color, Bag-of-Features is used to encode the physiological features, a machine learning algorithm is used to analyze and classify the extracted physiological features, a three-dimensional sign model is generated, and the three-dimensional sign model is labeled with position information and area information of blood vessel distribution, soft tissue hierarchy, nerve tissue, submandibular gland and each physiological feature.

[0056] Specifically, Bag-of-Features is a feature representation method used in the field of image and video processing, which is mainly used to extract and encode local features in images for tasks such as classification, retrieval and recognition, and is completed through the following steps:

[0057] First, local features are extracted from the image, and common features include SIFT, SURF, HOG, etc. These features can capture texture, shape, edge, etc. information in the image; then the extracted local features are clustered to obtain a set of "vocabulary". Common clustering algorithms include K-means clustering algorithm; for each image, match its local features with the "vocabulary" obtained by clustering to get the frequency statistics of each "vocabulary". Common encoding methods include Vector Quantization and Spatial Pyramid Encoding, etc.

[0058] It can be understood that the extracted physiological features are analyzed and classified using a machine learning algorithm, and the images are labeled according to the encoding, and the physiological feature information of each part is distinguished, and further integrated into an image information model.

[0059] In some embodiments of the present application, the evaluation module evaluates the injection risk of the three-dimensional sign model to determine whether injection can be performed, comprising:

[0060] The position information of the submandibular gland is recorded as the target position Y, and the target position Y is compared with the pre-set safe injection area, and the injection risk of the submandibular gland is evaluated according to the comparison;

[0061] When the target position Y belongs to the safe injection area, the evaluation result is set as no injection risk, and it is determined that injection can be directly performed;

[0062] When the target position Y does not belong to the safe injection area, the evaluation result is set as there is injection risk, and it is determined that ultrasound intervention is needed for injection.

[0063] In some embodiments of the present application, the evaluation module evaluates the injection risk of the three-dimensional sign model to determine whether injection can be performed, comprising:

[0064] When the target position Y does not belong to the safe injection area, the area information of the submandibular gland is recorded as the submandibular gland area K, and the submandibular gland area K is compared with the pre-set first submandibular gland area K1 and second submandibular gland area K2, wherein K1 < K2, and the injection risk of the submandibular gland is evaluated according to the comparison;

[0065] When K≤K1, it is determined that injection cannot be performed;

[0066] When K1

[0067] When K2

[0068] In some embodiments of the present application, the planning module plans the injection region, the injection starting point, the injection ending point and the initial injection direction according to the three-dimensional sign model, comprising:

[0069] In the three-dimensional sign model, the distribution of blood vessels to be avoided, the soft tissue level and the nerve tissue are inputted, a preferred path with the least change in angle from the skin surface to the submandibular gland is obtained by using a heuristic search algorithm, the starting point of the preferred path is taken as the injection starting point, the direction of the path is taken as the initial injection direction, and the ending point of the path is taken as the injection ending point.

[0070] In some embodiments of the present application, the planning module obtains a preferred path with the least change in angle from the skin surface to the submandibular gland by using a heuristic search algorithm, comprising:

[0071] The three-dimensional sign model is converted into a graph representation, and nodes and edges are defined, wherein the nodes are defined as positions in the model, and the edges are defined as connection relationships between the nodes;

[0072] The component defines a heuristic function to evaluate the priority of each node, calculates an expected cost estimate value of a node to a target position, the expected cost estimate value is used to represent the expected cost from the node to the target position, and the expected cost estimate value is generated based on physiological characteristic factors, wherein a penalty term is added for blood vessels, soft tissues and nerve tissues, so that the search algorithm selects a path that avoids these structures;

[0073] A preset test injection starting point is taken as a starting node, and a preset test injection ending point is taken as a target node;

[0074] The distance of the starting node is set to 0, and the priority of the starting node is set to the expected cost estimate value, a search is performed by using a heuristic search algorithm, the nodes are traversed in order from high to low according to the priority of the nodes, and the distance and the priority of the nodes are updated, the neighbor nodes of the nodes are traversed, a new distance value is calculated, and the priority of the nodes is updated, thereby obtaining a preferred path with the least change in angle from the skin surface to the submandibular gland.

[0075] Specifically, the three-dimensional sign model is converted into a graph representation, and nodes and edges are defined. In this way, the problem can be converted into a graph theory problem, facilitating path planning using a heuristic search algorithm; a heuristic function is used to evaluate the priority of each node, calculate the expected cost estimate value of the node to the target position, and this heuristic function is generated based on physiological characteristic factors, and a penalty term is added to the blood vessels, soft tissues and nerve tissues, so that the search algorithm selects a path that avoids these structures. This can ensure that the planned path is safer and more feasible; using a heuristic search algorithm, nodes are traversed in order of priority from high to low, and the distance and priority of the node are updated. By calculating the neighbor nodes of the node, a new distance value is calculated, and the priority of the node is updated. In this way, an optimal path from the starting point to the target node is gradually searched out; through the planning algorithm, an optimal path with the least change in angle from the skin surface to the submandibular gland is obtained. This means that during the injection process, the path will try to avoid changes in the angle in the injection path, reducing possible discomfort and risk.

[0076] In some embodiments of the present application, the injection module uses a medical syringe to perform injection at a first depth, comprising:

[0077] When judging the first injection risk, the straight-line distance U1 between the injection starting point and the injection ending point is obtained, and a first injection depth proportion factor q1 is generated according to the target position Y, and the first depth is set as q1*U1, wherein 0

[0078] When judging the second injection risk, the straight-line distance U2 between the injection starting point and the injection ending point is obtained, and a second injection depth proportion factor q2 is generated according to the target position Y, and the first depth is set as q2*U2, wherein 0

[0079] In summary, the technical effects of the present embodiment are:

[0080] The target area is scanned and imaged by the ultrasonic detection module, and the two-dimensional images are stacked according to the scanning direction to generate a three-dimensional sign model, providing real-time imaging and visual representation of the target area; the evaluation module analyzes the three-dimensional sign model, evaluates the injection risk according to the pre-set safe injection area and submandibular gland area parameters, and judges whether injection can be performed; according to the three-dimensional sign model, the injection area, the injection starting point, the injection ending point and the initial injection direction are planned, and a preferred path with the least angle change is obtained through a heuristic search algorithm, improving the accuracy and safety of injection; the guide image is generated by the ultrasonic intervention module, and the planned avoidance angle is continuously generated to guide the injection process, ensuring the accuracy and safety of the injection; through the guidance of the evaluation and planning module, combined with the real-time scanning and navigation of the ultrasonic intervention module, the accuracy and safety of the injection are improved, and the injection risk is reduced.

[0081] Those skilled in the art will appreciate that embodiments of the application can be readily used as a method, apparatus such as a system, or computer program product. Accordingly, the application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the application can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer readable program code.

[0082] The application is described in reference to the flowchart illustrations and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart Figure 1 one or more functions specified in the flowchart block or blocks. Figure 1 means for (a) or (b) as specified in the flowchart block or blocks.

[0083] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart block or blocks. Figure 1 one or more functions specified in the flowchart block or blocks. Figure 1 means for (a) or (b) as specified in the flowchart block or blocks.

[0084] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more functions specified in the flowchart block or blocks. Figure 1 means for (a) or (b) as specified in the flowchart block or blocks.

[0085] Finally, it should be noted that the above-mentioned embodiments are merely intended for describing and illustrating, not limiting, the technical solutions of the present application. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that the specific embodiments of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application, and any modification or equivalent replacement without departing from the spirit and scope of the present application should be covered in the protection scope of the claims of the present application.

Claims

1. A visualized gland injection system, characterized by, The method comprises the following steps: An ultrasonic detection module comprising a detection ultrasonic probe is used to scan and two-dimensionally image a target region using the detection ultrasonic probe, and the ultrasonic probe is moved in different directions and the scanning direction is acquired in real time, one two-dimensional image is acquired for each scanning direction, the frequency of the detection ultrasonic probe is 7.5-12 MHz, and the target region comprises the angle of the mandible and the rear of the mandible; A visualization construction module is used to sequentially stack the two-dimensional images according to the scanning direction to generate a three-dimensional sign model; An evaluation module is used to evaluate the injection risk of the three-dimensional sign model and determine whether injection can be performed; A planning module is used to plan an injection region, an injection starting point, an injection ending point and an initial injection direction according to the three-dimensional sign model; An injection module comprising a medical injector is used to perform first-depth injection using the medical injector according to the injection starting point and the initial injection direction; An ultrasonic intervention module comprising an intervention ultrasonic probe is used to scan the injection region using the intervention ultrasonic probe and generate a guide image in real time, the frequency of the intervention ultrasonic probe is 5-7.5 MHz; A guide module is used to continuously generate a planned avoidance angle according to the guide image until the injection is completed; The three-dimensional sign model is generated, and the three-dimensional sign model is labeled with blood vessel distribution, soft tissue hierarchy, nerve tissue, submandibular gland and position information and area information of each physiological feature; The evaluation module comprises the following steps: The position information of the submandibular gland is recorded as a target position Y, the target position Y is compared with a pre-set safe injection area, and the injection risk of the submandibular gland is evaluated according to the comparison; When the target position Y belongs to the safe injection area, the evaluation result is set as no injection risk, and it is determined that injection can be directly performed; When the target position Y does not belong to the safe injection area, the evaluation result is set as injection risk, and it is determined that ultrasonic intervention is needed for injection; The evaluation module comprises the following steps: When the target position Y does not belong to the safe injection area, the area information of the submandibular gland is recorded as a submandibular gland area K, the submandibular gland area K is compared with a pre-set first submandibular gland area K1 and a second submandibular gland area K2, K1 < K2, and the injection risk of the submandibular gland is evaluated according to the comparison; When K ≤ K1, it is determined that injection cannot be performed; When K1 < K ≤ K2, it is determined that there is a first-level injection risk; When K2 < K, it is determined that there is a second-level injection risk; The injection module comprises the following steps: When it is determined that there is the first-level injection risk, the straight-line distance U1 of the injection starting point and the injection ending point is acquired, a first injection depth proportion factor q1 is generated according to the target position Y, and the first depth is set as q1*U1, wherein 0 < q1 < 0.

2. When judging the risk of the secondary injection, a straight-line distance U2 between the injection starting point and the injection ending point is obtained, and a second injection depth scaling factor q2 is generated according to the target position Y, and the first depth is set as q2*U2, where 0 2. The visualized gland injection system of claim 1, wherein, The data processing of each two-dimensional image by the visualization construction module includes removing noise, enhancing contrast, and edge detection, and each two-dimensional image is stacked in sequence according to its scanning direction and converted into three-dimensional voxel data, each voxel representing image information of a specific position.

3. The visualized gland injection system of claim 2, wherein, The visualization construction module, which stacks each two-dimensional image in sequence according to its scanning direction and converts it into three-dimensional voxel data, includes: The size and resolution of the preset voxel are determined, an empty three-dimensional array is created as three-dimensional feature data according to the size and resolution of the voxel, each pixel in the two-dimensional data is traversed, the voxel coordinates of each pixel in the three-dimensional feature data are calculated according to the scanning direction and the point position of the pixel, and the pixel is assigned to each voxel coordinate.

4. The visualized gland injection system of claim 3, wherein, The visualization construction module pre-processes the three-dimensional feature data to extract physiological features, including shape, geometric properties, texture, and color, encodes the physiological features using Bag-of-Features, and analyzes and classifies the extracted physiological features using a machine learning algorithm.

5. The visualized gland injection system of claim 4, wherein, The planning module plans the injection area, injection starting point, injection ending point, and initial injection direction according to the three-dimensional feature model, including: The three-dimensional feature model is input to avoid the blood vessel distribution, soft tissue hierarchy, and neural tissue, a heuristic search algorithm is used to obtain an optimal path with the least change in angle from the skin surface to the submandibular gland, the starting point of the optimal path is set as the injection starting point, the direction of the path is set as the initial injection direction, and the ending point of the path is set as the injection ending point.

6. The visualized gland injection system of claim 5, wherein, The planning module uses a heuristic search algorithm to obtain an optimal path with the least change in angle from the skin surface to the submandibular gland, including: The three-dimensional feature model is converted into a graph representation to define nodes and edges, where nodes are defined as positions in the model and edges are defined as connection relationships between nodes; The heuristic function defines the priority of each node, calculates an expected cost estimate value of a node to the target position, and the expected cost estimate value is used to represent the expected cost from the node to the target position, and the expected cost estimate value is generated based on the physiological feature factors, where a penalty term is added for blood vessels, soft tissues, and neural tissues to make the search algorithm choose a path that avoids these structures; A test injection starting point is preset as the starting node, and a test injection ending point is preset as the target node; The distance of the start node is set to 0, and the priority of the start node is set to the expected cost estimate value, a heuristic search algorithm is used to search, nodes are traversed in order of priority from high to low, and the distance and priority of the nodes are updated, the neighbor nodes of the nodes are traversed, new distance values are calculated, and the priority of the nodes is updated, and a preferred path with the least change in angle from the skin surface to the submandibular gland is obtained.

Images