Accurate intervention positioning auxiliary device and method based on AI visual identification

Abstract

The invention discloses a precise intervention positioning auxiliary device and method based on AI visual identification, and aims to solve the problems of low positioning precision, lack of dynamic compensation, complex operation and poor comfort of a to-be-positioned target. The device main body integrates a bed body adjusting, lifting and positioning auxiliary mechanism, an AI visual identification system and a touch screen. The bed body adjusting assembly is matched with a to-be-positioned target body position through a soft cushion, an adjustable foot plate and a sliding block. The lifting mechanism achieves stable lifting through a gravity base and the like. The positioning auxiliary mechanism drives a rotating tooth rack through a motor to drive a flexible fixing clamp to precisely fix an instrument; the AI visual recognition system collects images, the body position and breath of a target to be positioned are tracked in real time through algorithm analysis, and positioning parameters are dynamically adjusted. The positioning method comprises the steps of body position adjustment, height adaptation, image acquisition and data generation, initial positioning, dynamic adjustment and completion. The system is accurate in positioning, high in comfort degree of a to-be-positioned target, convenient to operate, stable in structure and suitable for accurate positioning intervention scenes, and has wide popularization value.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, and in particular to a precise interventional positioning auxiliary device and method based on AI visual recognition. Background Technology

[0002] In the field of medical device technology, precise positioning intervention is a key and common diagnostic and treatment method, widely used in various medical scenarios. The accuracy and stability of its operation are crucial for ensuring the effectiveness of precise positioning intervention, alleviating the suffering of the patient being located, and reducing medical costs.

[0003] Traditionally, precise positioning interventions rely heavily on the operator's personal experience and intuition, a method with significant limitations. First, due to the complexity of the human body and individual differences, the operator's judgment often lacks high precision, increasing the risk of interventional failure. Second, traditional methods lack dynamic compensation for changes in the target's position and respiratory movements, making it difficult to avoid positioning deviations caused by minute movements of the target during precise interventions. These problems can not only increase the suffering of the target but also potentially lead to complications.

[0004] With the continuous advancement of medical technology, a number of precision-guided interventional devices have emerged on the market, aiming to improve the accuracy and stability of precision-guided interventions. However, most of these existing devices have limited functionality and cannot fully meet the complex and ever-changing clinical needs. For example: The multifunctional precision positioning interventional biopsy bed disclosed in CN209437576U, while improving patient comfort and ease of observation through features such as an adjustable-height abdominal pillow and a reflector, still falls short in dynamic positioning and adjustment. It lacks a real-time monitoring and dynamic adjustment mechanism for patient respiratory movements and positional changes, making it difficult to maintain high precision in complex clinical scenarios.

[0005] CN208611230U discloses an assisted precision positioning interventional fixation bed: This device, through a posture adjustment mechanism and lateral fixation components, achieves patient positioning and bed posture adjustment, improving surgical success rates and patient comfort. However, its operation procedure is relatively complex, requiring considerable manual adjustment by the operator, increasing workload. Furthermore, there is still room for improvement in the device's ability to dynamically compensate for changes in patient position and respiratory movements.

[0006] CN120788523A discloses a patient vein precision positioning interventional auxiliary guidance device based on intelligent visual recognition. This device utilizes intelligent visual recognition technology to clearly present the direction and depth of blood vessels through top and side vascular imaging instruments, providing medical staff with intuitive and accurate visual guidance. However, this device is mainly designed for precise vein positioning interventions, and its application scenarios are relatively limited. It also lacks comprehensive dynamic compensation capabilities for changes in the patient's overall body position and respiratory movements.

[0007] Specifically, some existing precision positioning interventional devices may employ mechanical positioning structures. While these structures are stable, they lack flexibility and adaptability, making it difficult to address individual differences in the target and dynamic changes during the precision positioning intervention process. Other devices, although integrating simple electronic sensors to monitor certain physiological parameters of the target, often fail to provide sufficiently accurate and real-time positioning information due to limitations in data processing capabilities and algorithm precision.

[0008] In addition, different precise positioning intervention scenarios in clinical practice have different requirements for the positioning accuracy and structural adaptability of the device, while existing auxiliary devices are mostly designed for a single scenario, lacking versatility and failing to cover diverse clinical needs.

[0009] In summary, existing technologies still have significant shortcomings in the precision positioning of interventional procedures, failing to meet the urgent needs of modern medicine for efficient, accurate, and stable diagnostic and treatment methods. Therefore, developing an intelligent clinical precision interventional positioning auxiliary device capable of real-time monitoring of changes in the patient's position and respiratory movements, dynamically adjusting positioning parameters, and being easy to operate, is of significant practical importance. This invention addresses this technological background and market demand, aiming to provide an efficient, accurate, and stable auxiliary solution for clinical precision interventional procedures through innovative structural design and intelligent control technology. Summary of the Invention

[0010] The technical problem to be solved by this invention is to provide a precise interventional positioning auxiliary device and method based on AI visual recognition, which solves the problems of low positioning accuracy, lack of dynamic compensation, complex operation and poor comfort of the target in existing clinical precise positioning interventional positioning technology. Through the integration of innovative mechanical structure design and intelligent visual recognition technology, it realizes the precision and dynamic adjustment of precise positioning interventional positioning, simplifies the operation process, improves the comfort of the target in the precise positioning interventional process, and provides an efficient and stable auxiliary solution for clinical precise positioning interventional diagnosis and treatment.

[0011] To achieve the above technical objectives, the present invention adopts the following technical solution: A precise intervention and positioning assistance device and method based on AI visual recognition are provided, as detailed below: (I) Precision Intervention and Positioning Assistance Device Based on AI Visual Recognition The device includes a main body, which serves as the core mounting carrier. The main body integrates and assembles a bed adjustment assembly, a lifting mechanism, a positioning auxiliary mechanism, an AI visual recognition system, and a display screen. These components are mechanically assembled or electrically connected to form a complete functional system. Overall assembly relationship: The bed adjustment components, lifting mechanism, and positioning auxiliary mechanism are all installed on the corresponding mounting surfaces of the main body of the device by bolt fixing or guide rail assembly; the control module of the AI ​​vision recognition system is integrated into the internal cavity of the main body of the device, and the display screen is mounted in the human-machine interaction area of ​​the main body of the device, realizing the integration of information display and operation control.

[0012] The bed adjustment assembly includes a cushion, an adjustable footplate, and a slider. The cushion is attached to the support panel of the bed adjustment assembly using an adhesive and snap-fit ​​method, providing comfortable support for the target being positioned. The adjustable footplate is hinged to the rear end of the bed adjustment assembly; the fixed end of the hinge is bolted to the support panel, and the movable end is fixed to the adjustable footplate, allowing it to rotate around the hinge axis to adjust the angle and accommodate different leg positions of the target. The lower end of the slider is engaged with the built-in guide rail of the bed adjustment assembly, and the upper end is fixed to the sliding support part of the bed, enabling sliding adjustment of the bed structure to meet diverse positioning needs.

[0013] Lifting mechanism: includes gravity base, drive unit, rotating rod, sliding buckle, top plate, tension spring, guide block and support block. The gravity base is fixed to the bottom mounting area of ​​the main body of the device by expansion bolts or positioning pins, using its own weight to ensure the stability of the device. The upper surface of the base has pre-drilled screw holes for the drive component. The drive component is vertically fixed to the center of the upper surface of the gravity base by bolts. The output shaft is hinged to one end of the rotating rod via a shaft pin to achieve power output. The other end of the rotating rod is hinged to the side ear plate of the sliding buckle via a shaft pin to transmit driving force. The sliding buckle has a U-shaped snap-fit ​​structure, which snaps into the sliding groove at the bottom of the top plate to form a sliding fit, allowing horizontal sliding along the groove. The guide block and support block are fixed to the upper surface of the top plate by bolts and are spaced apart along the length of the top plate. Each end of the tension spring has a hook, one end connected to the hook hole of the guide block and the other end connected to the hook hole of the support block, forming an elastic support circuit to assist in smooth lifting and lowering. The upper surface of the top plate is bolted to the bed-bearing part of the main body of the device. The extension and retraction of the lifting mechanism drives the upper structure of the main body of the device to rise and fall, adapting to different height requirements in different operating scenarios.

[0014] Positioning auxiliary mechanism: includes a motor, rotating gear, rack, and fixing clamp. The motor is fixed to the side mounting bracket of the device body by a motor mounting bracket bolt. Its output shaft is connected to the rotating gear by a flat key and locked with a nut to prevent loosening, providing precise driving force. The rotating gear meshes with the rack to form a transmission engagement. The rack is limited by a guide rail (fixed to the device body) and can make linear reciprocating motion along the rail. The fixing clamp has a connecting seat at the bottom, which is fixed to the upper end face of the rack by bolts or buckles. It moves synchronously with the linear movement of the rack to achieve positioning. The fixing clamp is made of flexible material, which can firmly and comfortably fix the interventional instruments for precise positioning and avoid instrument damage.

[0015] AI Visual Recognition System and Display Screen: The camera of the AI ​​visual recognition system is fixed to the upper crossbeam of the main body of the device via an adjustable bracket. The bracket can be folded or extended to adjust the camera angle. The camera is electrically connected to the control module inside the main body of the device via a data cable to acquire image information of the target area of ​​the target to be located. The control module is electrically connected to the display screen via a data cable, and also to the motor of the positioning auxiliary mechanism via a control cable. The display screen is a touch screen, used to display the analysis results of the AI ​​visual recognition system, precise positioning intervention information, and the working status of the device. It also supports the operator to set parameters, control operations, and view relevant information. The AI ​​visual recognition system can perform preprocessing such as denoising, enhancement, and segmentation on the acquired image information. It uses deep learning algorithms to identify lesion features and construct a 3D model of the structural features of the target to be located. Combined with respiratory monitoring information, it generates precise positioning data and transmits it to the positioning auxiliary mechanism. At the same time, it can track the changes in body position and respiratory movements of the target to be located in real time, dynamically adjust the precise positioning intervention parameters, and ensure the accuracy of the precise positioning intervention.

[0016] The AI ​​visual recognition system includes an image acquisition module, an image preprocessing module, a 3D model construction module, a positioning data generation and adjustment module, and an operation control and display module, as detailed below: 1. Image Acquisition Module The image acquisition module mainly consists of a high-resolution camera and an adjustable stand. The camera is responsible for acquiring image information of the target area, while the adjustable stand is used to adjust the angle and position of the camera to ensure the accuracy and comprehensiveness of image acquisition.

[0017] The camera captures images of the target area through an optical lens and converts them into digital signals, which are then transmitted to the control module for processing. The adjustable bracket adjusts the position and angle of the camera according to the operator's needs, adapting to different body positions and precise positioning of the intervention site.

[0018] 2. Image preprocessing module The image preprocessing module incorporates denoising, enhancement, and segmentation algorithms. These algorithms are integrated into the control module and are responsible for preprocessing the acquired image information.

[0019] Denoising Algorithm: A non-local mean denoising technique is employed. First, for each pixel in the image, similar pixel blocks are searched throughout the entire image; these similar pixel blocks may be distributed in different locations within the image. Then, a weight value is calculated based on the similarity between pixel blocks; the higher the similarity, the greater the weight. Finally, these weight values ​​are used to calculate a weighted average of the pixel values ​​of all similar pixel blocks, thus obtaining the denoised value of that pixel. This method removes random noise from the image, improving image quality.

[0020] Enhancement Algorithm: Utilizing histogram equalization. First, the gray-level histogram of the original image is statistically analyzed to obtain the frequency of each gray level. Then, the cumulative distribution function is calculated, and the gray values ​​of the original image are remapped according to the cumulative distribution function, resulting in a more uniform distribution of the new gray-level histogram. This improves the image's contrast and makes details clearer.

[0021] Segmentation Algorithm: Deep learning models such as U-Net are employed. First, the acquired image data is preprocessed, including normalization and cropping, to meet the model's input requirements. Then, the preprocessed images are input into the U-Net model, which performs feature extraction and image segmentation through convolutional layers, pooling layers, and upsampling layers. During training, a large amount of labeled data is used to train the model, enabling it to learn the features of the target region. In the prediction phase, the model segments new images, automatically identifying and segmenting the target region, providing a foundation for subsequent 3D reconstruction and localization.

[0022] 3. 3D Model Building Module The 3D model construction module includes a 3D reconstruction algorithm and a respiratory monitoring module. The 3D reconstruction algorithm is responsible for constructing a 3D feature model of the target area based on the preprocessed image data; the respiratory monitoring module is used to monitor the respiratory movement status of the target area in real time.

[0023] 3D Reconstruction Algorithm: This algorithm achieves precise alignment between images through multi-view image registration and fusion, employing algorithms such as Iterative Closest Point (ICP). First, image data of the target region is acquired from different viewpoints and preprocessed to extract feature points. Then, using the ICP algorithm, images from one viewpoint are gradually adjusted to align with the reference image. During each iteration, corresponding points between images are calculated, and the position and pose of the images are optimized by minimizing the distance error between corresponding points. After multiple iterations, precise alignment of images from all viewpoints is achieved. Finally, the registered multi-view images are fused to generate a high-precision 3D model.

[0024] Respiratory monitoring module: This module monitors the respiratory motion status of the target area in real time using an additional respiratory sensor or an image-based respiratory motion analysis algorithm. If a respiratory sensor is used, it collects respiratory motion-related signals and transmits them to the control module. If an image-based respiratory motion analysis algorithm is employed, feature points of the target area are first extracted from the image. Then, the motion trajectories of these feature points are tracked in consecutive image frames, and the respiratory motion parameters of the target area are calculated by analyzing the displacement changes of the feature points. Finally, the monitored respiratory motion status data is fed back to the control module.

[0025] 4. Positioning data generation and adjustment module The positioning data generation and adjustment module includes a positioning algorithm, a motor, a rotating gear, a rack, and a fixing clamp. The positioning algorithm generates initial positioning data based on the 3D model and respiratory monitoring data; the motor, rotating gear, rack, and fixing clamp are responsible for accurately positioning, moving, and fixing the interventional device based on the positioning data.

[0026] Positioning Algorithm: Combining a 3D model and respiratory monitoring data, the initial positioning position of the precise positioning interventional device is calculated. First, the precise location information of the target area is obtained from the 3D model, while considering the influence of respiratory motion on the target position, and the target position is dynamically corrected based on respiratory monitoring data. Then, based on the corrected target position, combined with the size and motion parameters of the precise positioning interventional device, the initial positioning coordinates of the precise positioning interventional device are calculated.

[0027] The motor drives the rotating gear to rotate based on the positioning data: the control module converts the positioning data calculated by the positioning algorithm into a control signal for the motor, and the motor starts to rotate after receiving the control signal. The rotating gear is connected to the motor and rotates along with the motor.

[0028] The fixed clamp is moved to the designated position through the meshing transmission of the rotating gear and rack: when the rotating gear rotates, it meshes with the rack, converting the rotational motion of the gear into the linear motion of the rack. The fixed clamp is mounted on the rack and moves with it, eventually reaching the designated position. During the precise positioning intervention, the positioning algorithm continuously tracks minute changes in the target area and dynamically adjusts the positioning data. The control module drives the motor to rotate again based on the adjusted positioning data, and through the transmission of the rotating gear and rack, the fixed clamp moves accordingly, ensuring that the precise positioning interventional instrument is always aligned with the target position.

[0029] 5. Operation control and display module The operation control and display module mainly consists of a touch screen display and operation control software. The touch screen display shows the analysis results, positioning information, and operating status of the AI ​​visual recognition system; the operation control software allows the operator to set parameters, control operations, and view relevant information.

[0030] The operator views the analysis results and positioning information of the AI ​​visual recognition system on the touchscreen display and sets parameters and controls the operation as needed. The operation control software transmits the operator's instructions to the control module, which adjusts the device's working status accordingly. This includes controlling the camera angle and position of the image acquisition module, adjusting the algorithm parameters of the image preprocessing module, controlling the reconstruction process of the 3D model construction module, and adjusting the positioning data of the positioning data generation and adjustment module. Simultaneously, the control module feeds back the device's real-time working status, such as image acquisition progress, 3D reconstruction progress, and precise positioning of the interventional device, to the operation control software. The operation control software then displays this information on the touchscreen display for the operator to view.

[0031] (II) Precise Intervention and Positioning Methods Based on AI Visual Recognition Applying the aforementioned AI-based visual recognition-based precise intervention and positioning auxiliary device, this invention also provides an AI-based visual recognition-based precise intervention and positioning method, specifically including the following steps: Position adjustment: The target lies flat on the soft pad of the bed adjustment component. The angle of the footboard is adjusted to match the leg posture of the target. At the same time, the position of the slider is adjusted to make the bed structure slide, so that the target is in a comfortable position that is easy to accurately locate and intervene.

[0032] Height Adaptation: Activate the lifting mechanism, and the drive unit will work. The rotating rod will drive the sliding buckle to slide in the slide groove. In conjunction with the tension spring, guide block, and support block, the device will be raised and lowered smoothly. The device will be adjusted to a height suitable for the operator. The gravity base will ensure the stability of the device during the lifting process.

[0033] Image acquisition and positioning data generation: The AI ​​visual recognition system is activated, and the camera acquires image information of the target area of ​​the target to be located. After preprocessing, the lesion features are identified by deep learning algorithms and a three-dimensional model of the structural features of the target to be located is constructed. Combined with respiratory monitoring information, accurate positioning data is generated and transmitted to the positioning assistance institution.

[0034] Initial positioning: The motor of the positioning auxiliary mechanism starts, driving the rotating gear to rotate. Through the meshing transmission between the rotating gear and the rack, the fixing clamp is moved to the designated position, which firmly fixes the relevant instruments for precise positioning intervention, thus completing the initial positioning of the precise positioning intervention.

[0035] Dynamic adjustment: During the precise positioning intervention, the AI ​​visual recognition system tracks the changes in the body position and respiratory movements of the target in real time, dynamically adjusts the positioning parameters of the precise positioning intervention, and the positioning auxiliary mechanism responds and adjusts synchronously to ensure that the precise positioning interventional instrument is always accurately aimed at the lesion.

[0036] Operation control and completion of precise positioning intervention: The operator views the image analysis results, precise positioning intervention location information and device working status on the display screen, performs necessary parameter settings and operation control, and finally completes the precise positioning intervention operation.

[0037] The precise intervention positioning assistance device and method based on AI visual recognition provided by this invention have the following beneficial effects: 1. This invention effectively solves the problems of insufficient positioning accuracy, poor operational stability, and lack of dynamic compensation for changes in the body position and respiratory movements of the target in the field of clinical precision positioning intervention technology. It significantly reduces the risk of precision positioning intervention failure, reduces the pain of the target, and provides a solid and reliable guarantee for clinical diagnosis and treatment.

[0038] 2. By integrating an AI visual recognition system, this invention achieves accurate acquisition and in-depth analysis of image information of the target to be located, and can construct an accurate three-dimensional structural feature model, providing unprecedented high-precision data support for precise positioning and intervention positioning.

[0039] 3. The AI ​​visual recognition system of the present invention has real-time tracking capability, which can dynamically capture changes in the body position and respiratory movements of the target to be located, and adjust the positioning parameters for precise positioning intervention in real time accordingly, ensuring the continuous accuracy of the precise positioning intervention process and effectively avoiding positioning deviations caused by the movement of the target to be located.

[0040] 4. While improving positioning accuracy, this invention also significantly optimizes the comfort of the target being located. The soft padding design of the bed adjustment components provides soft and comfortable support for the target, effectively alleviating the discomfort of prolonged bed rest.

[0041] 5. The bed adjustment assembly of the present invention also includes an adjustable footboard with a flexible angle, which can be adjusted in a personalized manner according to the leg posture of different targets to be positioned, further improving the comfort of the target to be positioned during the precise positioning intervention process.

[0042] 6. Through the slider design, the bed structure of the present invention realizes the sliding adjustment function, which can meet the needs of different target positions and provide more flexible and diverse position selection for precise clinical interventional operations.

[0043] 7. The lifting mechanism of this invention is ingeniously designed and can flexibly adjust the height of the device according to different clinical operation scenarios, ensuring that the operator can obtain the best operating view and comfort in different working environments.

[0044] 8. The introduction of a touchscreen display greatly simplifies the operation process of this invention. Operators can intuitively view the analysis results of the AI ​​visual recognition system, precise positioning information, and device operating status through the touchscreen, easily completing parameter settings and operation control.

[0045] 9. The positioning auxiliary mechanism of this invention uses a motor-driven rotating gear that meshes with a rack to achieve precise movement and positioning of the fixed clamp. This design not only improves the accuracy of positioning but also ensures the stability of precise positioning intervention operations.

[0046] 10. The fixing clip of this invention is made of flexible material, which can firmly fix and accurately position interventional instruments without damaging them. This design ensures the smooth execution of precise positioning interventional procedures and extends the service life of the instruments.

[0047] 11. The structure of this invention is stable and reliable, and the gravity base design of the lifting mechanism ensures the overall stability of the device. Meanwhile, the cooperation between the tension spring, guide block, and support block achieves smooth lifting, further improving the operational stability of the device.

[0048] 12. The highly intelligent design of this invention allows operators to focus more on the precise positioning and intervention operation itself, without being distracted by complex parameter settings and adjustments. The AI ​​visual recognition system can automatically complete tasks such as image analysis and positioning parameter adjustment, greatly reducing the operator's workload.

[0049] 13. The present invention has strong adaptability to various scenarios and is applicable to a variety of clinical precision positioning intervention scenarios. It can adapt to different operational needs and environmental conditions by adjusting the lifting mechanism and bed adjustment components.

[0050] 14. Through actual clinical application and testing, this invention has achieved remarkable results in improving the accuracy of precise positioning intervention, operational stability, and comfort of the target being located, bringing revolutionary innovation and practical value to the field of clinical precise positioning intervention. Attached Figure Description

[0051] The present invention will be further described below with reference to the accompanying drawings and embodiments: Figure 1 This is a schematic diagram of the overall structure of the device of the present invention; Figure 2 This is a top view of the device of the present invention; Figure 3 This is a side view of the device of the present invention; Figure 4 For the present invention Figure 1 Enlarged structural diagram at point A; Figure 5 This is a schematic diagram of the lifting assembly of the present invention; Figure 6 This is a schematic diagram of the positioning mechanism of the present invention; In the diagram: 1. Main body of the device; 2. Bed adjustment assembly; 3. Lifting mechanism; 4. Positioning auxiliary mechanism; 21. Soft pad; 22. Adjustable foot plate; 23. Slider; 31. Gravity base; 32. Drive component; 33. Rotating rod; 34. Sliding buckle; 35. Top plate; 36. Tension spring; 37. Guide block; 38. Support block; 305. Slide groove; 41. Motor; 42. Rotating gear; 43. Rack; 44. Fixing clamp; 45. Display screen. Detailed Implementation

[0052] The technical solutions of the present invention will be further described below with reference to the embodiments and accompanying drawings: Example 1 like Figures 1 to 6 As shown, this embodiment provides a precise intervention positioning assistance device based on AI visual recognition, as detailed below: like Figure 1 As shown, the AI ​​visual recognition-based precision intervention positioning assistance device of this embodiment includes a device body 1, a bed adjustment component 2, a lifting mechanism 3, a positioning assistance mechanism 4, an AI visual recognition system, and a display screen 45. The components are mechanically assembled to form a complete functional device, which can be directly applied to the precision positioning intervention scenario of the first target area.

[0053] like Figure 1 and Figure 4As shown, the bed adjustment assembly 2 is laid on the upper bearing area of ​​the main body 1 of the device. The soft pad 21 covers the bed bearing panel by adhesive and snap-fit, providing support for the back of the target to be positioned. The adjustment foot plate 22 is hinged to the rear end face of the bed. The fixed end of the hinge is bolted to the bearing panel, and the movable end is integrally fixed to the adjustment foot plate 22, allowing it to rotate around the hinge axis to adapt to the knee-bent or straight posture of the target. The slider 23 is snapped onto the built-in guide rail of the bed, and its upper end is fixed to the slidable bearing part of the bed. Figure 4 As shown in the enlarged structure at point A, slider 23 can slide horizontally along the guide rail to adjust the front and rear positions of the bed.

[0054] like Figure 1 and Figure 5 As shown, the lifting mechanism 3 is installed at the four corners of the bottom of the main body 1. The gravity base 31 is fixed to the bottom mounting surface of the main body 1 by expansion bolts, and its own weight ensures the stability of the device. The drive component 32 is fixed to the center of the upper surface of the gravity base 31 by vertical bolts. Its output shaft is hinged to one end of the rotating rod 33 by a shaft pin. The other end of the rotating rod 33 is connected to the side ear plate of the sliding buckle 34 by a shaft pin. The sliding buckle 34 has a U-shaped structure and is inserted into the sliding groove 305 at the bottom of the top plate 35 to form a sliding fit. The guide block 37 and the support block 38 are fixed to the upper surface of the top plate 35 by bolts. The two are arranged at intervals along the length of the top plate 35. The hooks at both ends of the tension spring 36 are respectively hooked into the hanging holes of the guide block 37 and the support block 38 to form an elastic buffer structure to ensure that the lifting process is smooth and without shaking.

[0055] like Figure 1 and Figure 6 As shown, the positioning auxiliary mechanism 4 is mounted on the side bracket of the main body 1 of the device. The motor 41 is fixed to the side bracket by mounting bolts. The output shaft of the motor 41 is connected to the rotating gear 42 by a flat key and locked by a nut. The rotating gear 42 meshes with the rack 43 to form a transmission structure. The rack 43 is limited by the guide rail on the side bracket and can move back and forth linearly along the rail. The fixing clamp 44 is fixed to the upper end face of the rack 43 by the bottom connecting seat buckle. It is made of medical flexible silicone material and has anti-slip texture on the inside, which can firmly clamp the precise positioning intervention needle without damaging the needle body.

[0056] like Figure 1 and Figure 2As shown, the display screen 45 is embedded in the human-computer interaction area at the front of the main body 1 of the device. It is a touch screen structure and is electrically connected to the AI ​​visual recognition system control module inside the main body 1 via a data cable. The camera of the AI ​​visual recognition system is fixed on the upper crossbeam of the main body 1 via a foldable bracket. The bracket can be extended and adjusted in height and angle. The camera lens faces the bed support area to collect image information of the target area of ​​the target to be located. The control module is integrated in the cavity inside the main body 1 and is electrically connected to the motor 41 and the drive component 32 via control lines to realize signal transmission and control.

[0057] The AI ​​visual recognition system includes an image acquisition module, an image preprocessing module, a 3D model construction module, a positioning data generation and adjustment module, and an operation control and display module, as detailed below: 1. Image Acquisition Module The image acquisition module mainly consists of a high-resolution camera and an adjustable stand. The camera is responsible for acquiring image information of the target area, while the adjustable stand is used to adjust the angle and position of the camera to ensure the accuracy and comprehensiveness of image acquisition.

[0058] The camera captures images of the target area through an optical lens and converts them into digital signals, which are then transmitted to the control module for processing. The adjustable bracket adjusts the position and angle of the camera according to the operator's needs, adapting to different body positions and precise positioning of the intervention site.

[0059] 2. Image preprocessing module The image preprocessing module incorporates denoising, enhancement, and segmentation algorithms. These algorithms are integrated into the control module and are responsible for preprocessing the acquired image information.

[0060] Denoising Algorithm: A non-local mean denoising technique is employed. First, for each pixel in the image, similar pixel blocks are searched throughout the entire image; these similar pixel blocks may be distributed in different locations within the image. Then, a weight value is calculated based on the similarity between pixel blocks; the higher the similarity, the greater the weight. Finally, these weight values ​​are used to calculate a weighted average of the pixel values ​​of all similar pixel blocks, thus obtaining the denoised value of that pixel. This method removes random noise from the image, improving image quality.

[0061] Enhancement Algorithm: Utilizing histogram equalization. First, the gray-level histogram of the original image is statistically analyzed to obtain the frequency of each gray level. Then, the cumulative distribution function is calculated, and the gray values ​​of the original image are remapped according to the cumulative distribution function, resulting in a more uniform distribution of the new gray-level histogram. This improves the image's contrast and makes details clearer.

[0062] Segmentation Algorithm: Deep learning models such as U-Net are employed. First, the acquired image data is preprocessed, including normalization and cropping, to meet the model's input requirements. Then, the preprocessed images are input into the U-Net model, which performs feature extraction and image segmentation through convolutional layers, pooling layers, and upsampling layers. During training, a large amount of labeled data is used to train the model, enabling it to learn the features of the target region. In the prediction phase, the model segments new images, automatically identifying and segmenting the target region, providing a foundation for subsequent 3D reconstruction and localization.

[0063] 3. 3D Model Building Module The 3D model construction module includes a 3D reconstruction algorithm and a respiratory monitoring module. The 3D reconstruction algorithm is responsible for constructing a 3D structural feature model of the target area based on the preprocessed image data; the respiratory monitoring module is used to monitor the respiratory motion status of the target area in real time.

[0064] 3D Reconstruction Algorithm: This algorithm achieves precise alignment between images through multi-view image registration and fusion, employing algorithms such as Iterative Closest Point (ICP). First, image data of the target region is acquired from different viewpoints and preprocessed to extract feature points. Then, using the ICP algorithm, images from one viewpoint are gradually adjusted to align with the reference image. During each iteration, corresponding points between images are calculated, and the position and pose of the images are optimized by minimizing the distance error between corresponding points. After multiple iterations, precise alignment of images from all viewpoints is achieved. Finally, the registered multi-view images are fused to generate a high-precision 3D model.

[0065] Respiratory monitoring module: This module monitors the respiratory motion status of the target area in real time using an additional respiratory sensor or an image-based respiratory motion analysis algorithm. If a respiratory sensor is used, it collects respiratory motion-related signals and transmits them to the control module. If an image-based respiratory motion analysis algorithm is employed, feature points of the target area are first extracted from the image. Then, the motion trajectories of these feature points are tracked in consecutive image frames, and the respiratory motion parameters of the target area are calculated by analyzing the displacement changes of the feature points. Finally, the monitored respiratory motion status data is fed back to the control module.

[0066] 4. Positioning data generation and adjustment module The positioning data generation and adjustment module includes a positioning algorithm, a motor, a rotating gear, a rack, and a fixing clamp. The positioning algorithm generates initial positioning data based on the 3D model and respiratory monitoring data; the motor, rotating gear, rack, and fixing clamp are responsible for accurately positioning, moving, and fixing the interventional device based on the positioning data.

[0067] Positioning Algorithm: Combining a 3D model and respiratory monitoring data, the initial positioning position of the precise positioning interventional device is calculated. First, the precise location information of the target area is obtained from the 3D model, while considering the influence of respiratory motion on the target position, and the target position is dynamically corrected based on respiratory monitoring data. Then, based on the corrected target position, combined with the size and motion parameters of the precise positioning interventional device, the initial positioning coordinates of the precise positioning interventional device are calculated.

[0068] The motor drives the rotating gear to rotate based on the positioning data: the control module converts the positioning data calculated by the positioning algorithm into a control signal for the motor, and the motor starts to rotate after receiving the control signal. The rotating gear is connected to the motor and rotates along with the motor.

[0069] The fixed clamp is moved to the designated position through the meshing transmission of the rotating gear and rack: when the rotating gear rotates, it meshes with the rack, converting the rotational motion of the gear into the linear motion of the rack. The fixed clamp is mounted on the rack and moves with it, eventually reaching the designated position. During the precise positioning intervention, the positioning algorithm continuously tracks minute changes in the target area and dynamically adjusts the positioning data. The control module drives the motor to rotate again based on the adjusted positioning data, and through the transmission of the rotating gear and rack, the fixed clamp moves accordingly, ensuring that the precise positioning interventional instrument is always aligned with the target position.

[0070] 5. Operation control and display module The operation control and display module mainly consists of a touch screen display 45 and operation control software. The touch screen display 45 is used to display the analysis results, positioning information, and working status of the AI ​​visual recognition system; the operation control software supports the operator in setting parameters, controlling operations, and viewing relevant information.

[0071] The operator views the analysis results and positioning information of the AI ​​visual recognition system on the touchscreen display 45, and sets parameters and controls the operation as needed. The operation control software transmits the operator's instructions to the control module, which adjusts the device's working status accordingly. This includes controlling the camera angle and position of the image acquisition module, adjusting the algorithm parameters of the image preprocessing module, controlling the reconstruction process of the 3D model construction module, and adjusting the positioning data of the positioning data generation and adjustment module. Simultaneously, the control module feeds back the device's real-time working status, such as image acquisition progress, 3D reconstruction progress, and precise positioning of the interventional device, to the operation control software. The operation control software then displays this information on the touchscreen display for the operator to view.

[0072] Example 2 In another preferred embodiment, based on Embodiment 1, this embodiment provides a precise intervention positioning method based on AI visual recognition. This method is a precise intervention positioning method for the first target area of ​​the precise intervention positioning auxiliary device based on AI visual recognition described in Embodiment 1. The specific steps are as follows: like Figure 1 and Figure 4 As shown, the first step is to adjust the body position: the target to be positioned lies flat on the soft pad 21 of the bed adjustment component 2. The operator rotates the adjusting foot plate 22 to adjust it to a suitable angle around the hinge axis according to the comfort of the target's legs. Then, the operator pushes the sliding load-bearing part of the bed and slides the slider 23 along the built-in guide rail to adjust the front and back position of the bed so that the area where the first part of the target is located is directly facing the camera acquisition range on the upper part of the device body 1, ensuring that the target is in a comfortable posture that is easy for precise positioning intervention.

[0073] like Figure 1 and Figure 5 As shown, height adaptation is performed: The lifting mechanism 3 is activated, and the drive component 32 is powered on. Its output shaft extends and retracts, driving the rotating rod 33 to rotate. The rotating rod 33 pulls the sliding buckle 34 to slide horizontally along the sliding groove 305 at the bottom of the top plate 35. Simultaneously, the tension spring 36 elastically extends and retracts between the guide block 37 and the support block 38, assisting the top plate 35 in driving the upper structure of the device body 1 to rise and fall smoothly. The operator can observe the height from the side (e.g., ...). Figure 3 As shown, the main body 1 of the device is adjusted to a position that matches its own operating height, and the gravity base 31 always ensures that the main body 1 of the device does not shift or shake during the lifting and lowering process.

[0074] like Figure 1 As shown, the AI ​​visual recognition system is activated: the angle of the camera on the upper crossbeam of the main body 1 is adjusted so that it is precisely aligned with the precise positioning intervention area of ​​the target to be located. The camera collects image information of the first part of the target area of ​​the target to be located and transmits it to the control module inside the main body 1 through the data cable. After the control module performs noise reduction, enhancement and segmentation preprocessing on the image, it identifies the features of the first part through deep learning algorithm, constructs a three-dimensional model of the structural features of the first part of the target to be located, and generates precise positioning intervention positioning data by combining the respiratory signals collected by the respiratory monitoring module.

[0075] like Figure 1 and Figure 6As shown, the initial positioning is completed: the control module transmits the positioning data to the motor 41 of the positioning auxiliary mechanism 4. The motor 41 starts and drives the rotating gear 42 to rotate. The rotating gear 42 meshes with the rack 43 and drives the rack 43 to move linearly along the guide rail. The fixing clamp 44 moves synchronously with the rack 43 to directly above the precise positioning intervention point. The operator puts the precise positioning intervention needle into the fixing clamp 44. The fixing clamp 44 firmly fixes the precise positioning intervention needle with the elastic clamping force of the flexible silicone material, thus completing the initial positioning of the precise positioning intervention.

[0076] like Figure 1 and Figure 2 As shown, dynamic adjustment and precise positioning intervention are performed: During the precise positioning intervention, the camera continuously acquires images of the target to be positioned, and the AI ​​visual recognition system tracks the changes in the target's position and breathing movements in real time, dynamically adjusts the positioning parameters and feeds them back to the motor 41. The motor 41 drives the rack 43 to fine-tune the position of the fixing clamp 44 to ensure that the precise positioning intervention needle is always aligned with the target area of ​​the first part. The operator can view the image analysis results, precise positioning intervention positioning coordinates and device working status in real time through the display screen 45 on the front of the device body 1. If manual adjustment is required, parameters can be directly input on the display screen 45 to control the motor 41 or the drive component 32 to complete the precise positioning intervention operation of the target area of ​​the first part.

[0077] Example 3 In another preferred embodiment, based on Embodiment 1, this embodiment provides a precise intervention positioning assistance device based on AI visual recognition, as detailed below: like Figure 1 As shown, the AI ​​visual recognition-based precision intervention positioning assistance device of this embodiment includes a device body 1, a bed adjustment component 2, a lifting mechanism 3, a positioning assistance mechanism 4, an AI visual recognition system, and a display screen 45. The components are fixed by bolts and assembled with guide rails to form an integrated device, which is specially adapted to the precise positioning intervention scenario of the target area of ​​the second clinical site. It can accurately locate the position of the second site and assist in intervention operation.

[0078] like Figure 1 and Figure 4 As shown, the bed adjustment assembly 2 is installed on the upper bearing surface of the main body 1. The soft pad 21 is made of medical breathable material and covers the bed bearing panel by adhesive and buckle fixing. It has an arc-shaped recess at the waist position corresponding to the target to be positioned, which is suitable for side-lying or semi-recumbent position support during the precise positioning intervention of the second target area. The adjustable foot plate 22 is hinged to the rear of the bed and can be flipped upward to form an angle of 30° to 60° with the bed, which is suitable for the leg bending support needs of the target to be positioned when lying on the side. The slider 23 is snapped onto the built-in guide rail of the bed, such as... Figure 4As shown in the enlarged structure at point A, slider 23 can move the sliding part of the bed body left and right to adjust the lateral position of the target body to be positioned, so that the target area of ​​the second part is aligned with the precise positioning intervention channel.

[0079] like Figure 1 and Figure 5 As shown, the lifting mechanism 3 is symmetrically installed on both sides of the bottom of the device body 1. The gravity base 31 is fixed to the bottom mounting surface of the device body 1 by a positioning pin to ensure that the device does not shift during the precise positioning intervention process. The drive component 32 is vertically fixed to the upper surface of the gravity base 31. Its output shaft is hinged to the rotating rod 33 by a shaft pin. The rotating rod 33 pulls the sliding buckle 34 to slide along the sliding groove 305 at the bottom of the top plate 35. The guide block 37 and the support block 38 at the upper end of the top plate 35 cooperate with the tension spring 36 to realize the smooth lifting and lowering of the upper structure of the device body 1, which is suitable for the operator's standing or sitting posture operation needs when the target area of ​​the second part is precisely positioned and intervened.

[0080] like Figure 1 and Figure 6 As shown, the positioning auxiliary mechanism 4 is mounted on the side crossbeam of the main body 1 of the device. The motor 41 is fixed on the crossbeam support by the mounting base. The output shaft of the motor 41 is keyed to the rotating gear 42. The rotating gear 42 meshes with the rack 43 for transmission. The rack 43 moves laterally along the guide rail, driving the top fixing clip 44 to align with the second part area of ​​the target to be positioned. The fixing clip 44 is made of flexible medical rubber material and has an arc-shaped groove on the inner side, which can be used to fix the intervention tube or the precise positioning intervention needle to avoid the displacement of the instrument during the precise positioning intervention.

[0081] like Figure 1 and Figure 2 As shown, the display screen 45 is embedded in the front operation panel of the main body 1 of the device and is electrically connected to the control module of the AI ​​visual recognition system via a data cable. The camera of the AI ​​visual recognition system is fixed to the upper crossbeam of the main body 1 of the device via a telescopic bracket. It can be adjusted downward to aim at the second part area of ​​the target to be located and collect image information of the second part area. The control module is integrated inside the main body 1 of the device and is electrically connected to the motor 41 and the drive component 32 to realize positioning data transmission and motion control.

[0082] Example 4 In another preferred embodiment, based on embodiments 1 to 3, this embodiment provides a precise intervention positioning method based on AI visual recognition. The positioning method for the second target area precise positioning intervention device of the precise intervention positioning auxiliary device based on AI visual recognition described in embodiment 3 includes the following specific steps: like Figure 1 and Figure 4As shown, the first step is to adjust the body position: the target is placed in a right lateral decubitus position on the soft pad 21 of the bed adjustment component 2. The operator pushes the sliding part of the bed according to the height of the target, and slides the slider 23 laterally along the built-in guide rail to adjust the left and right position of the target's body so that the area where the second part of the target is located is directly facing the camera's acquisition range. At the same time, the adjusting footplate 22 is rotated to flip it upward to a 45° angle to provide support for the target's right leg. The arc-shaped concave of the soft pad 21 fits the waist of the target, reducing the risk of body position changes.

[0083] like Figure 1 and Figure 5 As shown, height adaptation is performed: The lifting mechanism 3 is activated, and after the drive component 32 is powered on, the output shaft extends and retracts, causing the rotating rod 33 to rotate. The rotating rod 33 pulls the sliding buckle 34 to slide horizontally along the sliding groove 305 at the bottom of the top plate 35. The tension spring 36 elastically extends and retracts between the guide block 37 and the support block 38, assisting the top plate 35 in driving the upper structure of the main body 1 of the device to rise and fall; the operator can then... Figure 3 As shown in the side view, the device is adjusted to a height of 80cm (suitable for standing precise positioning intervention operation). The gravity base 31 ensures that the device is stable and does not shake during the lifting and lowering process.

[0084] like Figure 1 As shown, the AI ​​visual recognition system is activated: the camera angle on the upper crossbeam of the main body 1 is adjusted so that it is vertically aligned with the second part of the target area to be located. The camera collects tomographic image information of the second part and transmits it to the control module via a data cable. The control module performs noise reduction and enhancement preprocessing on the image, identifies the boundary of the target area of ​​the second part and the intervention target point through a deep learning algorithm, constructs a three-dimensional model of the structural features of the second part, and generates accurate positioning intervention positioning data by combining respiratory monitoring signals, which is displayed on the display screen 45 in real time.

[0085] like Figure 1 and Figure 6 As shown, initial positioning is completed: the control module sends the positioning data to the motor 41 of the positioning auxiliary mechanism 4. The motor 41 starts to drive the rotating gear 42 to rotate. The rotating gear 42 meshes with the rack 43 to drive the fixed clamp 44 to move laterally to directly above the intervention target point. The operator puts the precise positioning intervention needle into the arc-shaped groove of the fixed clamp 44. The fixed clamp 44 clamps the needle body with the elastic force of the flexible material to complete the initial positioning.

[0086] like Figure 1 and Figure 2As shown, dynamic adjustment and interventional operation are performed: During the precise positioning intervention, the AI ​​visual recognition system tracks the respiratory movement and slight changes in body position of the target in real time, dynamically adjusts the positioning parameters and feeds them back to the motor 41. The motor 41 drives the rack 43 to fine-tune the position of the fixing clamp 44 to ensure that the interventional needle is always aligned with the tumor target point. The operator can view the real-time positioning coordinates, image analysis results and device working status through the display screen 45. If it is necessary to adjust the precise positioning intervention depth, the operator can input parameters through the display screen 45 to control the drive component 32 to fine-tune the height, and finally complete the precise positioning interventional operation of the second target area.

[0087] Example 5 In another preferred embodiment, based on Embodiment 1, this embodiment provides a precise intervention positioning assistance device based on AI visual recognition, as detailed below: like Figure 1 As shown, the AI ​​visual recognition-based precision intervention positioning assistance device of this embodiment includes a device body 1, a bed adjustment component 2, a lifting mechanism 3, a positioning assistance mechanism 4, an AI visual recognition system, and a display screen 45. The components are mechanically assembled to form a complete device, which is suitable for precise positioning intervention scenarios in the third clinical location and can meet the precision positioning requirements for prone position precision positioning intervention.

[0088] like Figure 1 and Figure 4 As shown, the bed adjustment assembly 2 is laid on the upper part of the main body 1 of the device. The soft pad 21 adopts a segmented design and is fixed to the bed support panel by adhesive and buckle. Support protrusions are provided for the first part, second part and leg area of ​​the target to be positioned, which are adapted to the body force distribution during the precise positioning intervention in the prone position. The adjustable foot plate 22 is hinged to the rear of the bed and can be flipped down to a 180° horizontal position with the bed, which is adapted to the leg extension support when the target is prone. The slider 23 is snapped onto the built-in guide rail of the bed, such as Figure 4 As shown in the enlarged structure at point A, slider 23 can move the sliding part of the bed back and forth to adjust the position of the third part of the target to be positioned, so that the target area of ​​the third part is aligned with the precise positioning intervention channel.

[0089] like Figure 1 and Figure 5 As shown, the lifting mechanism 3 is installed at the four ends of the bottom of the device body 1. The gravity base 31 is fixed to the bottom of the device body 1 by expansion bolts to ensure that the device does not tilt when the target to be positioned is lying down in the prone position. The drive component 32 is vertically fixed to the upper surface of the gravity base 31. The output shaft is hinged to the rotating rod 33 by a shaft pin. The rotating rod 33 pulls the sliding buckle 34 to slide along the sliding groove 305 at the bottom of the top plate 35. The guide block 37 and the support block 38 on the top plate 35 cooperate with the tension spring 36 to realize the smooth lifting and lowering of the device body 1, which is adapted to the bending height of the operator when the target area of ​​the third part is accurately positioned.

[0090] like Figure 1 and Figure 6 As shown, the positioning auxiliary mechanism 4 is mounted on the rear bracket of the main body 1 of the device. The motor 41 is fixed on the bracket by the mounting seat. The output shaft of the motor 41 is locked and fixed with the rotating gear 42. The rotating gear 42 meshes with the rack 43 for transmission. The rack 43 moves longitudinally along the guide slide rail, driving the fixing clamp 44 to align with the target area of ​​the third part of the target to be positioned. The fixing clamp 44 is made of flexible silicone material and has anti-slip texture on the inner side, which can firmly fix the target area of ​​the third part and accurately position the intervention needle, avoiding the needle body shaking during the intervention process.

[0091] like Figure 1 and Figure 2 As shown, the display screen 45 is installed on the side of the operating area of ​​the main body 1 of the device and is electrically connected to the control module of the AI ​​visual recognition system via a data cable; the camera of the AI ​​visual recognition system is fixed to the upper rear crossbeam of the main body 1 of the device via a foldable bracket, and can be adjusted downward to aim at the third part of the target to be located, and collect image information of the target area of ​​the third part. The control module is integrated inside the main body 1 of the device and is electrically connected to the motor 41 and the drive component 32 to realize positioning control and information transmission.

[0092] Example 6 In another preferred embodiment, based on embodiments 1, 2, and 5, this embodiment provides a precise intervention positioning method based on AI visual recognition. The positioning method for the third target area precise positioning intervention device of the precise intervention positioning auxiliary device based on AI visual recognition described in embodiment 5 includes the following specific steps: like Figure 1 and Figure 4 As shown, the first step is to adjust the body position: the target lies prone on the soft pad 21 of the bed adjustment component 2, with the support protrusions of the first and second parts conforming to the body to reduce pressure; the operator flips down to adjust the footboard 22 to a horizontal position to accommodate the straight leg posture of the target; the sliding part of the bed is pushed, and the slider 23 moves back and forth along the built-in guide rail to adjust the position of the third part of the target, so that the area where the third part is located is directly facing the camera acquisition range on the upper rear side of the main body 1, ensuring that the precise positioning intervention channel is unobstructed.

[0093] like Figure 1 and Figure 5 As shown, height adaptation is performed: The lifting mechanism 3 is activated, the drive unit 32 is powered on, the output shaft extends and retracts, driving the rotating rod 33 to rotate. The rotating rod 33 pulls the sliding buckle 34 to slide along the sliding groove 305 at the bottom of the top plate 35. The tension spring 36 provides elastic buffering between the guide block 37 and the support block 38, assisting the top plate 35 in driving the upper structure of the main body 1 of the device to rise and fall; the operator can then... Figure 3As shown in the side view, the device is adjusted to a height of 70cm to accommodate bending over operation. The gravity base 31 ensures that the device remains stable when the target to be positioned turns over or makes minor adjustments to its position.

[0094] like Figure 1 As shown, the AI ​​visual recognition system is activated: the angle of the camera on the upper rear side of the main body 1 is adjusted so that it is vertically aligned with the target area of ​​the third part of the target to be located. The camera collects the tomographic image information of the third part and transmits it to the control module through the data cable. The control module performs segmentation and enhancement preprocessing on the image, and identifies the cortical and medullary boundaries and intervention target points of the target area of ​​the third part through deep learning algorithms. A three-dimensional model of the structural features of the third part is constructed, and precise positioning data is generated by combining respiratory monitoring signals and displayed on the display screen 45 in real time.

[0095] like Figure 1 and Figure 6 As shown, the initial positioning is completed: the control module sends the positioning data to the motor 41 of the positioning auxiliary mechanism 4. The motor 41 starts to drive the rotating gear 42 to rotate. The rotating gear 42 meshes with the rack 43 to drive the fixed clamp 44 to move longitudinally to directly above the intervention target point. The operator places the intervention needle for precise positioning in the third target area into the fixed clamp 44. The fixed clamp 44 fixes the needle body with the clamping force of the flexible material, thus completing the initial positioning.

[0096] like Figure 1 and Figure 2 As shown, dynamic adjustment and intervention are performed: During the precise positioning intervention, the AI ​​visual recognition system tracks the respiratory movements and minor changes in body position of the target in real time, dynamically adjusts the positioning parameters and feeds them back to the motor 41. The motor 41 drives the rack 43 to fine-tune the position of the fixing clamp 44 to ensure that the precise positioning intervention needle is always aligned with the intervention target point. The operator can view the real-time positioning information, image analysis results and device working status through the display screen 45. The operator can also input parameters through the display screen 45 to adjust the precise positioning intervention angle, and finally complete the precise positioning intervention operation of the third target area.

[0097] In the preferred embodiment, the bed adjustment component 2, lifting mechanism 3, and positioning auxiliary mechanism 4 are all installed on the corresponding mounting surfaces of the main body 1 by bolt fixing or guide rail assembly. The control module of the AI ​​visual recognition system is integrated into the internal cavity of the main body 1, and the display screen 45 is mounted in the human-machine interaction area of ​​the main body 1. This arrangement makes the overall structure of the device compact and stable, and facilitates the installation and disassembly of each component, enabling future maintenance and upgrades. Integrating the control module into the internal cavity of the main body effectively protects it from external interference and damage. The display screen 45, mounted in the human-machine interaction area, facilitates operation and information viewing for the operator. The rational layout of each component improves the reliability and stability of the device, ensures the accurate operation of the AI ​​visual recognition system, and thus improves the accuracy and efficiency of precise positioning intervention.

[0098] In a preferred embodiment, the bed adjustment assembly 2 includes a soft pad 21, an adjustable footplate 22, and a slider 23. The soft pad 21 is laid on the support panel of the bed adjustment assembly 2 by adhesive bonding and snap-fit ​​fixing, providing comfortable support for the target to be positioned. The adjustable footplate 22 is hinged to the rear end face of the bed adjustment assembly 2, with the fixed end of the hinge bolted to the support panel and the movable end fixed to the adjustable footplate 22, allowing for angle adjustment to accommodate different leg postures of the target to be positioned. The lower end face of the slider 23 is engaged with the built-in guide rail of the bed adjustment assembly 2, and the upper end face is fixed to the slidable support part of the bed, enabling sliding adjustment of the bed structure. These features allow for comprehensive adaptation to the body needs of the target to be positioned, improving comfort and operational adaptability. The stable placement of the soft pad 21 ensures the target lies comfortably, reducing swaying during precise positioning intervention. The adjustable footplate 22 allows for flexible angle adjustment to accommodate various leg postures, preventing improper posture from affecting precise positioning intervention. Slider 23 enables sliding adjustment of the bed structure, precisely adapting to the overall position of the target body and facilitating operator operation. The coordinated action of all components effectively improves the accuracy and stability of the precise positioning intervention, reduces the patient's discomfort, and ensures the smooth progress of the precise positioning intervention.

[0099] In a preferred embodiment, the lifting mechanism 3 includes a gravity base 31, which is fixed to the bottom mounting area of ​​the main body 1 by expansion bolts or positioning pins. A mounting screw hole for the drive component 32 is pre-drilled on the upper surface. The drive component 32 is vertically fixed to the center of the upper surface of the gravity base 31 by bolts, and its output shaft is hinged to one end of a rotating rod 33 via a shaft pin. The other end of the rotating rod 33 is hinged to the side ear plate of a sliding buckle 34 via a shaft pin. The sliding buckle 34 has a U-shaped snap-fit ​​structure, which engages with a sliding groove 305 at the bottom of the top plate 35 to form a sliding fit. Guide blocks 37 and support blocks 38 are fixed to the upper surface of the top plate 35 by bolts and are spaced apart along the length of the top plate 35. Tension springs 36 have hooks at both ends, one end connected to the hook hole of the guide block 37 and the other end connected to the hook hole of the support block 38, thus achieving the lifting function. The upper surface of the top plate 35 is bolted to the bed-bearing portion of the main body 1. These features ensure the smoothness and precision of the lifting process. The gravity base 31 is securely fixed, providing solid support for the entire mechanism and preventing swaying. The drive component 32 is centrally mounted, outputting power evenly and ensuring coordinated lifting and lowering movements. The rotating rod 33 is hinged to the sliding buckle 34, and works with the sliding groove 305 to achieve smooth sliding. The guide block 37 and support block 38 are arranged at intervals, working with the tension spring 36 to effectively buffer the impact of lifting and lowering, making lifting and lowering more stable. The top plate 35 is connected to the bed frame's load-bearing part, allowing for precise adjustment of the bed frame height to meet the needs of different target scenarios, improving the accuracy and safety of precise positioning intervention.

[0100] In a preferred embodiment, the positioning auxiliary mechanism 4 includes a motor 41, which is fixed to a side mounting bracket of the device body 1 by a motor mounting seat bolt. The output shaft of the motor 41 is connected to a rotating gear 42 via a flat key and locked with a nut to prevent loosening. The rotating gear 42 meshes with a rack 43 to form a transmission engagement. The rack 43 is limited by a guide rail, which is fixed to the device body 1. A fixing clamp 44 is mounted on the rack 43, with a connecting seat at its bottom. It is fixed to the upper surface of the rack 43 by bolts or buckles. The motor 41 drives the fixing clamp 44 to achieve precise movement and positioning. The fixing clamp 44 is made of flexible material to securely and comfortably fix interventional instruments for precise positioning. The above configuration greatly improves the accuracy and ease of operation of precise positioning interventional positioning. The motor 41 is stably fixed, and the output power is precisely transmitted through the rotating gear 42 and the rack 43. The guide rail effectively limits the movement of the fixing clamp 44, ensuring smooth movement and accurate positioning. The flexible clamp 44 can not only firmly fix the precise positioning interventional device and prevent it from shaking and affecting the precise positioning intervention, but also conform to the shape of the device, increase the comfort of the target to be positioned, and reduce the additional pain caused by improper device fixation, thus providing a reliable guarantee for precise positioning intervention in clinical practice.

[0101] In the preferred embodiment, the display screen 45 is a touchscreen, used to display the analysis results of the AI ​​visual recognition system, precise positioning intervention information, and the working status of the device. It also supports the operator in setting parameters, controlling operations, and viewing relevant information. These features greatly optimize the operator's experience and the efficiency of the precise positioning intervention process. The touchscreen 45 intuitively presents key information, allowing the operator to quickly obtain AI analysis results and positioning data without additional complex operations. The integrated parameter setting and operation control functions facilitate flexible adjustments to the device based on the target being located, improving the targeted nature of the precise positioning intervention. Convenient information viewing helps the operator fully understand the device's status and promptly handle any abnormalities. Overall, this improves the accuracy, safety, and convenience of precise positioning intervention, providing strong support for the efficient completion of clinical precise positioning intervention work.

[0102] In the preferred embodiment, in step 1, the angle of the adjustable footboard 22 and the position of the slider 23 of the bed adjustment component 2 are adjusted to adapt to the leg posture and overall body position requirements of the target to be positioned, while the soft pad 21 provides support and cushioning for the target. These settings significantly improve the comfort of the target and the suitability for precise positioning intervention. The adjustable footboard 22 can flexibly adjust its angle to precisely fit different leg postures of the target, avoiding muscle tension or discomfort caused by improper posture. The slider 23 moves as needed to precisely adapt to the overall body position of the target, ensuring the body is in the optimal state for precise positioning intervention. The support and cushioning effect of the soft pad 21 can distribute body pressure, reduce local pressure, and allow the target to remain relaxed during precise positioning intervention, thereby reducing the deviation in precise positioning intervention caused by the movement of the target and ensuring the accuracy and safety of precise positioning intervention.

[0103] In the preferred embodiment, in step 2, the driving component 32 of the lifting mechanism 3 drives the sliding buckle 34 to slide within the slide groove 305 via the rotating rod 33. This, combined with the tension spring 36, guide block 37, and support block 38, achieves smooth lifting. The gravity base 31 ensures the stability of the device during lifting. These features ensure accurate and safe lifting. The driving component 32, via the rotating rod 33, drives the sliding buckle 34, which, in conjunction with the tension spring 36 and other components, makes the lifting action smooth and fluid, avoiding jamming or shaking, greatly improving the stability and reliability of the operation. The gravity base 31 further enhances the overall stability of the device, effectively preventing tipping during lifting and providing a stable platform for precise positioning interventional operations. This not only ensures the safety of the target to be positioned but also facilitates the operator's precise control of the precise positioning intervention position and force, improving the success rate of precise positioning interventions and reducing medical risks.

[0104] In the preferred embodiment, in step 3, the AI ​​visual recognition system acquires image information through a camera. After noise reduction, enhancement, and segmentation preprocessing, it identifies lesion features using a deep learning algorithm and constructs a 3D model of the target structure to be located. This model is then combined with respiratory monitoring information to generate positioning data. These features ensure the accuracy and safety of the device's lifting and lowering process. The drive component 32, via the rotating rod 33, drives the sliding buckle 34 to slide, working in conjunction with components such as the tension spring 36 to ensure smooth and stable lifting and lowering movements, preventing jamming or shaking and greatly improving operational stability and reliability. The gravity base 31 further enhances the overall stability of the device, effectively preventing tipping during lifting and lowering, and providing a stable platform for precise positioning interventional procedures. This not only ensures the safety of the target to be located but also facilitates precise control of the positioning intervention position and force by the operator, improving the success rate of precise positioning interventions and reducing medical risks.

[0105] In summary, this invention proposes a precise interventional positioning assistance device and method based on AI visual recognition, effectively solving the problems of insufficient positioning accuracy, poor operational stability, and lack of dynamic compensation for changes in the body position and respiratory movements of the target patient in the field of clinical precise positioning interventional technology. Traditional precise positioning interventional methods mainly rely on the operator's experience and feel, which limits positioning accuracy and cannot adapt to subtle changes in the body position and breathing of the target patient in real time, increasing the risk of precise positioning interventional failure and the patient's discomfort. While existing precise positioning interventional assistance devices on the market have improved the accuracy of precise positioning intervention to some extent, most have limited functions and cannot fully meet the complex and ever-changing clinical needs. This invention aims to overcome the above-mentioned limitations of the prior art and improve the overall effect of precise positioning intervention.

[0106] This device applies an AI visual recognition system to precise clinical interventional positioning. By acquiring and analyzing the image information of the target to be positioned, it provides precise positioning data to the positioning assistance mechanism 4, achieving precise positioning for interventional procedures. The AI ​​visual recognition system in the device not only acquires image information but also tracks changes in the target's position and respiratory movements in real time, dynamically adjusting the positioning parameters to ensure the accuracy of the intervention. The positioning assistance mechanism 4 achieves precise positioning based on the positioning data from the AI ​​visual recognition system, and the fixing clip 44 is made of flexible material, providing a secure and comfortable fixation for interventional instruments. The bed adjustment assembly 2 includes a soft pad 21, an adjustable footboard 22, and a slider 23. The soft pad 21 is installed using adhesive and snap-fit ​​methods, the adjustable footboard 22 is angle-adjustable, and the slider 23 allows for sliding adjustment of the bed structure, better adapting to the target's leg posture and overall positional needs. The lifting mechanism 3 employs components such as a gravity base 31, a drive component 32, a rotating rod 33, a sliding buckle 34, a top plate 35, a tension spring 36, a guide block 37, and a support block 38 to achieve smooth lifting. The gravity base 31 ensures the stability of the device during lifting. The display screen 45 is a touchscreen, which not only displays the analysis results of the AI ​​visual recognition system, precise positioning intervention information, and the working status of the device, but also supports the operator in setting parameters, controlling operations, and viewing relevant information. In the precise positioning intervention method, the AI ​​visual recognition system acquires image information through a camera, performs noise reduction, enhancement, and segmentation preprocessing, and then uses deep learning algorithms to identify lesion features and construct a three-dimensional model of the structural features of the target to be located, combining respiratory monitoring information to generate positioning data.

[0107] This solution combines AI visual recognition technology with precise clinical interventional positioning. Utilizing the high-precision analysis capabilities of the AI ​​visual recognition system, it provides more accurate data support for precise interventional positioning, improving the success rate and safety of such interventions. The various components of the device are designed to work together as an organic whole. For example, the coordinated operation of the bed adjustment component 2, the lifting mechanism 3, and the positioning auxiliary mechanism 4 allows for flexible adjustments based on the actual situation of the target being positioned. The AI ​​visual recognition system tracks changes in the target's position and respiratory movements in real time and dynamically adjusts the precise interventional positioning parameters, solving the problem of inaccurate positioning caused by target movement in traditional precise interventional positioning. The fixing clamp 44 in the positioning auxiliary mechanism 4 is made of flexible material, ensuring the secure fixation of interventional instruments while improving the comfort of the target. The lifting mechanism 3 achieves smooth lifting through the cooperation of multiple components, and the gravity base 31 ensures the stability of the device, solving the swaying problem that may exist in traditional lifting mechanisms. In precise interventional localization methods, the AI ​​visual recognition system employs a complete and scientific processing flow for image information. Through deep learning algorithms, it constructs a 3D model of the structural features of the target to be located, providing more comprehensive and accurate information for precise interventional localization. The entire solution, from device design to the formulation of precise interventional localization methods, fully considers actual clinical needs, demonstrating high practicality and offering a novel solution for precise interventional localization in clinical practice.

Claims

1. A precise intervention and positioning assistance device based on AI visual recognition, characterized in that: The device includes a main body (1), which is equipped with a bed adjustment component (2), a lifting mechanism (3), a positioning auxiliary mechanism (4), an AI visual recognition system, and a display screen (45). The bed adjustment component (2) is used to adjust the position of the target to be positioned. The lifting mechanism (3) is used to adjust the overall height of the device to adapt to different scene requirements. The AI ​​visual recognition system is used to acquire and analyze the image information of the target to be positioned, provide accurate positioning data for the positioning auxiliary mechanism (4), and can track the position changes and breathing movements of the target to be positioned in real time and dynamically adjust the accurate positioning parameters. The positioning auxiliary mechanism (4) achieves accurate positioning based on the positioning data of the AI ​​visual recognition system.

2. The precise intervention positioning assistance device based on AI visual recognition according to claim 1, characterized in that: The bed adjustment assembly (2), lifting mechanism (3), and positioning auxiliary mechanism (4) are all installed on the corresponding mounting surface of the device body (1) by bolt fixing or guide rail assembly. The control module of the AI ​​visual recognition system is integrated into the internal cavity of the device body (1), and the display screen (45) is mounted in the human-computer interaction area of ​​the device body (1).

3. The precise intervention positioning assistance device based on AI visual recognition according to claim 1, characterized in that: The bed adjustment assembly (2) includes a soft pad (21), an adjusting foot plate (22), and a slider (23). The soft pad (21) is laid on the bearing panel of the bed adjustment assembly (2) by adhesive and snap fastening to provide comfortable support for the target to be positioned. The adjusting foot plate (22) is hinged to the tail end face of the bed adjustment assembly (2) by a hinge. The fixed end of the hinge is bolted to the bearing panel, and the movable end is fixed to the adjusting foot plate (22). The angle can be adjusted to adapt to different leg postures of the target to be positioned. The lower end face of the slider (23) is snapped into the built-in guide rail of the bed adjustment assembly (2), and the upper end face is fixed to the slidable bearing part of the bed to realize the sliding adjustment of the bed structure.

4. The precise intervention positioning assistance device based on AI visual recognition according to claim 1, characterized in that: The lifting mechanism (3) includes a gravity base (31), which is fixed to the bottom mounting area of ​​the main body (1) of the device by expansion bolts or positioning pins. The upper end face is reserved with mounting screw holes for the drive component (32). The drive component (32) is vertically fixed to the center position of the upper end face of the gravity base (31) by bolts. Its output shaft is hinged to one end of the rotating rod (33) by a shaft pin. The other end of the rotating rod (33) is hinged to the side ear plate of the sliding buckle (34) by a shaft pin. The sliding buckle (34) is a U-shaped snap-fit. The structure is fitted into the sliding groove (305) at the bottom of the top plate (35) to form a sliding fit; the guide block (37) and the support block (38) are fixed to the upper surface of the top plate (35) by bolts and are arranged at intervals along the length of the top plate (35); the tension spring (36) is provided with hooks at both ends, one end is connected to the hanging hole of the guide block (37) and the other end is connected to the hanging hole of the support block (38) to achieve the lifting function; the upper surface of the top plate (35) is bolted to the bed bearing part of the main body of the device (1).

5. The precise intervention positioning assistance device based on AI visual recognition according to claim 1, characterized in that: The positioning auxiliary mechanism (4) includes a motor (41), which is fixed to the side mounting bracket of the main body (1) of the device by a motor mounting seat bolt. The output shaft of the motor (41) is connected to the rotating gear (42) by a flat key and locked with a nut to prevent loosening. The rotating gear (42) meshes with the rack (43) to form a transmission engagement. The rack (43) is limited by the guide rail, which is fixed to the main body (1) of the device. The fixing clamp (44) is set on the rack (43) and has a connecting seat at its bottom. It is fixed to the upper end face of the rack (43) by bolts or buckles. The fixing clamp (44) is driven by the motor (41) to achieve precise movement and positioning. The fixing clamp (44) is made of flexible material and is used to securely and comfortably fix the interventional related instruments.

6. The precise intervention positioning assistance device based on AI visual recognition according to claim 1, characterized in that: The display screen (45) is a touch screen, used to display the analysis results, positioning information and working status of the AI ​​visual recognition system, and supports the operator to set parameters, control operations and view relevant information.

7. A precise intervention and positioning method based on AI visual recognition, characterized in that, The precise intervention positioning assistance device based on AI visual recognition as described in any one of claims 1 to 6 includes the following steps: Step 1: Adjust the position of the target body to be positioned using the bed adjustment component (2) so that the target body is in a comfortable posture that facilitates precise positioning and intervention; Step 2: Adjust the overall height of the device using the lifting mechanism (3) to suit the operator's needs; Step 3: Start the AI ​​visual recognition system, acquire image information of the target area of ​​the target to be located and analyze and process it, generate accurate positioning data and transmit it to the positioning assistance mechanism (4). Step 4: Positioning auxiliary mechanism (4) moves and fixes the relevant instruments for precise positioning intervention according to the positioning data to complete the initial positioning of precise positioning intervention; Step 5: During the precise positioning intervention, the AI ​​visual recognition system tracks the changes in the body position and respiratory movements of the target in real time, dynamically adjusts the precise positioning intervention positioning parameters, and the positioning auxiliary mechanism (4) responds and adjusts synchronously to ensure the accuracy of the precise positioning intervention. Step 6: The operator views relevant information on the display screen (45) and performs necessary operation controls to complete the precise positioning intervention operation.

8. The precise intervention positioning method based on AI visual recognition according to claim 7, characterized in that: In step 1, by adjusting the angle of the footboard (22) and the position of the slider (23) of the bed adjustment component (2), the leg posture and overall body position requirements of the target to be positioned are adapted, and the soft pad (21) provides support and cushioning for the target to be positioned.

9. The precise intervention positioning method based on AI visual recognition according to claim 7, characterized in that: In step 2, the driving component (32) of the lifting mechanism (3) drives the sliding buckle (34) to slide in the slide groove (305) through the rotating rod (33), and works with the tension spring (36), guide block (37) and support block (38) to achieve smooth lifting and lowering. The gravity base (31) ensures the stability of the device during the lifting process.

10. The precise intervention positioning method based on AI visual recognition according to claim 7, characterized in that: In step 3, the AI ​​visual recognition system collects image information through a camera, and after denoising, enhancement, and segmentation preprocessing, it identifies lesion features and constructs a three-dimensional model of the structural features of the target to be located through a deep learning algorithm, and generates location data by combining respiratory monitoring information.

Images