A precise detection system for tuberculosis based on ultrasonic intervention

Abstract

The application belongs to the medical field and provides a tuberculosis precision detection system based on ultrasonic intervention, which comprises an ultrasonic probe for acquiring ultrasonic images of a target region; a three-dimensional modeling unit for constructing a three-dimensional model of the target region based on the ultrasonic images and spatial position information of the ultrasonic probe; an ultrasonic guiding unit connected with the three-dimensional modeling unit, for calculating an optimal detection path based on the three-dimensional model and real-time position of the ultrasonic probe, and guiding an operator to acquire second ultrasonic images through intercostal spaces; and a detection module connected with the ultrasonic guiding unit, for performing tuberculosis precision detection based on the second ultrasonic images.

Description

Technical Field

[0001] This invention belongs to the medical field, and specifically relates to a precision tuberculosis detection system based on ultrasound intervention. Background Technology

[0002] Tuberculosis (TB) is a chronic infectious disease caused by Mycobacterium tuberculosis, primarily affecting lung tissue. Early detection and accurate diagnosis are crucial for controlling disease transmission and improving treatment outcomes. Imaging examinations are important tools for TB diagnosis, with chest X-rays and computed tomography (CT) scans being commonly used clinical imaging techniques. However, these methods have limitations such as radiation risks, high costs, and strong equipment dependence, making them unsuitable for pregnant women, children, and patients requiring frequent follow-up examinations. In contrast, ultrasound imaging, due to its advantages of being radiation-free, providing real-time imaging, and being portable, has become a highly regarded non-invasive auxiliary detection method for TB diagnosis in recent years.

[0003] However, the propagation characteristics of ultrasound in different tissues are mainly affected by acoustic impedance. The acoustic impedance of bone tissue is much higher than that of soft tissue. When ultrasound waves encounter ribs, sternum, and other bony structures of the thoracic cage, most of the sound energy is reflected or absorbed, with only a very small amount of ultrasound waves able to penetrate the bone tissue and reach the deeper lung tissue. This phenomenon results in acoustic shadowing behind the bones, appearing as a black, signal-free area on the ultrasound image. This makes it difficult to observe important anatomical structures behind the bones, such as the lung tissue, mediastinum, and pleura, severely affecting the application of ultrasound in the detection of pulmonary tuberculosis.

[0004] In clinical practice, the ultrasound probe can be tilted or rotated to allow ultrasound waves to propagate along the intercostal spaces, avoiding the high acoustic impedance interfaces of the ribs, thereby entering the lung tissue for imaging. Alternatively, a suitable detection area can be selected, such as the intercostal space, subxiphoid process, or supraclavicular fossa, to obtain the best possible lung ultrasound echo signal.

[0005] However, the above methods are highly dependent on the doctor's experience and technique. Different doctors have varying degrees of mastery over probe angle, incident path, and parameter adjustments, leading to inconsistencies in ultrasound image quality and test results. Furthermore, individual differences exist in the thoracic anatomy of different patients, and the penetrability of ultrasound waves varies among individuals, further increasing the difficulty of ultrasound detection of pulmonary tuberculosis. Therefore, relying solely on the doctor's experience to optimize the ultrasound detection path and angle makes it difficult to achieve standardized, automated, and highly repeatable ultrasound detection of tuberculosis in clinical practice. Summary of the Invention

[0006] To address the problems in the prior art, this invention provides a precise tuberculosis detection system based on ultrasound intervention, comprising the following modules:

[0007] An ultrasonic probe is used to acquire ultrasonic images of a target area; the ultrasonic probe also includes an inertial measurement unit, which measures the spatial position information of the ultrasonic probe.

[0008] A 3D modeling unit is used to construct a 3D model of the target area based on the ultrasound image and the spatial position information of the ultrasound probe. The 3D modeling unit includes:

[0009] The data acquisition unit is used to receive multi-angle ultrasound image data acquired by the ultrasound probe and simultaneously record the spatial position information of the ultrasound probe.

[0010] The point cloud generation unit is used to calculate the point cloud data of each tissue layer based on the depth information of the ultrasound image, and generate a preliminary three-dimensional point cloud model of the target area by fusing multiple frames of images.

[0011] The skeletal structure recognition unit is used to identify the skeletal boundaries within the target area based on the point cloud model, so as to determine the three-dimensional anatomical structure of the ribs, sternum and hard tissues.

[0012] The soft tissue modeling unit is used to construct a soft tissue model within the target area based on the ultrasound echo characteristics.

[0013] An ultrasound guiding unit, connected to the three-dimensional modeling unit, is used to calculate the optimal detection path based on the three-dimensional model and the real-time position of the ultrasound probe, and guide the operator to acquire a second ultrasound image through the intercostal space. The ultrasound guiding unit includes:

[0014] The path planning unit is used to calculate the optimal ultrasound detection path based on the three-dimensional model and the anatomical structure of the target area to avoid areas obscured by bones.

[0015] The probe angle optimization unit is used to calculate the recommended probe tilt angle based on the current position of the ultrasound probe and the positional relationship of the target area.

[0016] The force feedback unit is used to monitor the pressure of the ultrasonic probe during operation and provide tactile feedback to determine the optimal position and angle of the ultrasonic probe.

[0017] The detection module, connected to the ultrasound guide, is used for accurate tuberculosis detection based on the second ultrasound image.

[0018] This invention provides a precise tuberculosis detection system based on ultrasound intervention, which has the following advantages compared to existing technologies:

[0019] This invention utilizes intelligent ultrasound guidance, probe angle optimization, and a real-time force feedback mechanism to effectively avoid obstructions from the ribs and sternum, optimizing the ultrasound propagation path, improving the penetration ability of the ultrasound signal and the imaging quality of the lesion area, and reducing detection errors caused by differences in physician experience. This enhances the accuracy and consistency of ultrasound detection for pulmonary tuberculosis. Employing automated ultrasound detection path planning and intelligent probe angle recommendation algorithms, it dynamically guides physicians to adjust the probe position and angle, reducing the high dependence of ultrasound detection on physician experience. This allows non-specialist ultrasound physicians to quickly master the optimal detection operation, improving the standardization of ultrasound detection, reducing repetitive scans, increasing detection efficiency, and optimizing clinical application value. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a system diagram of the method of the present invention. Detailed Implementation

[0022] The invention will now be described in preferred form with reference to the accompanying drawings and specific embodiments.

[0023] This embodiment solves the above problems through the following steps:

[0024] In one embodiment, reference Figure 1 This invention provides a precision tuberculosis detection system based on ultrasound intervention. The system combines ultrasound imaging technology, spatial positioning technology, and intelligent guidance algorithm. It constructs a three-dimensional anatomical model by acquiring ultrasound data of the target area in real time, and uses an optimized ultrasound detection path and force feedback mechanism to guide the operator to accurately acquire high-quality ultrasound images, thereby improving the detection accuracy and diagnostic precision of tuberculosis lesions.

[0025] The following will describe each module of the system in detail.

[0026] An ultrasonic probe is used to acquire ultrasonic images of a target area; the ultrasonic probe also includes an inertial measurement unit, which measures the spatial position information of the ultrasonic probe.

[0027] The ultrasound probe is used to acquire ultrasound images of the target area. The ultrasound probe is a medical ultrasound device with high-resolution imaging capabilities, including an ultrasound transducer unit and a signal processing unit. The ultrasound transducer unit is used to emit ultrasound signals and receive echo signals from the target area. The signal processing unit is used to analyze the echo signals in real time to generate ultrasound image data.

[0028] The ultrasonic probe also includes an inertial measurement unit (IMU), which measures the spatial position information of the ultrasonic probe, including its displacement, rotation angle, and motion trajectory. The IMU consists of an accelerometer, a gyroscope, and a magnetometer. Through multi-sensor data fusion technology, it performs high-precision measurement of the motion state of the ultrasonic probe, ensuring the accuracy of the probe's position information during scanning.

[0029] By integrating an inertial measurement unit (IMU) into the ultrasound probe, this system can accurately record the probe's spatial position information while performing ultrasound imaging. This results in highly spatially traceable ultrasound image data, enabling precise matching of multi-angle ultrasound images and improving the accuracy of 3D modeling. Simultaneously, the IMU can monitor the ultrasound probe's operational status in real time, optimizing the operator's path, reducing imaging errors, and improving the stability and repeatability of the detection, providing more efficient image data support for the accurate detection of tuberculosis.

[0030] The 3D modeling unit is used to construct a 3D model of the target area based on the ultrasound image and the spatial position information of the ultrasound probe.

[0031] The three-dimensional modeling unit is a computing module used to process ultrasound data and generate visualized anatomical structures. It includes a data acquisition unit, a point cloud generation unit, a bone structure recognition unit, and a soft tissue modeling unit to achieve accurate reconstruction of the anatomical structure of the target area.

[0032] The three-dimensional model refers to a three-dimensional visualization structure generated based on ultrasound imaging data and the spatial position information of the ultrasound probe. This three-dimensional model includes the tissue boundary, anatomical structure features, and spatial distribution information of the lesion area of ​​the target region. It can provide complete anatomical information outside the ultrasound imaging plane, improving the comprehensiveness and accuracy of the detection.

[0033] The 3D modeling unit acquires multi-angle ultrasound image data, combines it with the spatial position information of the ultrasound probe, and uses point cloud computing, depth information fusion, and image registration algorithms to generate a 3D model of the target area. This 3D model accurately reflects the anatomical morphology of the target area and is dynamically updated to adapt to real-time changes during probe movement.

[0034] By constructing a 3D modeling unit, the system overcomes the limitations of traditional 2D ultrasound imaging, achieving multi-angle and multi-level information fusion and improving the accuracy of lesion location. Simultaneously, this 3D model can be used for navigation of the ultrasound-guided unit, assisting operators in optimizing the probe path and improving the repeatability and consistency of detection. Furthermore, this technology reduces the influence of operator experience on detection results, making interventional ultrasound detection more intelligent, precise, and standardized, thereby enhancing the early screening and accurate diagnosis of tuberculosis.

[0035] The data acquisition unit is used to receive multi-angle ultrasound image data acquired by the ultrasound probe and simultaneously record the spatial position information of the ultrasound probe.

[0036] The data acquisition unit is used to receive multi-angle ultrasound image data acquired by the ultrasound probe and simultaneously record the spatial position information of the ultrasound probe. The data acquisition unit is the core module for performing ultrasound image data reception, position information storage and synchronous processing, and includes an image signal receiving module, a position information storage module and a data synchronization processing module to ensure the integrity and spatial consistency of ultrasound data.

[0037] The multi-angle ultrasound imaging data refers to ultrasound imaging data acquired by the ultrasound probe at different spatial orientations and angles. This data reflects the tissue characteristics and lesion information of the target area at different imaging angles and can be used for subsequent three-dimensional reconstruction and lesion analysis.

[0038] The spatial position information refers to the real-time displacement, rotation angle, and trajectory information of the ultrasound probe in three-dimensional space during the scanning process. This information is acquired by an inertial measurement unit (IMU), optical tracking system, or magnetic positioning system and is used to map the real position of the ultrasound image data to ensure the spatial consistency of the imaging data at different scanning angles.

[0039] The data acquisition unit receives ultrasound image data in real time and stores the spatial motion information of the probe simultaneously. It uses a high-precision time synchronization algorithm to ensure that the image data and the probe position can correspond one-to-one, thus guaranteeing the traceability of the image data.

[0040] This unit effectively improves the accuracy and spatial matching of ultrasound imaging data, enabling multi-frame fusion, depth calculation, and 3D reconstruction of images during ultrasound detection, thereby enhancing the imaging quality of the target area. Furthermore, this data acquisition unit reduces image deviations caused by probe movement, resulting in more accurate lesion localization. It provides high-precision, structured image data for subsequent ultrasound-guided lesion analysis, improving the intelligence and standardization of tuberculosis detection.

[0041] The point cloud generation unit is used to calculate the point cloud data of each tissue layer based on the depth information of the ultrasound image, and generate a preliminary three-dimensional point cloud model of the target area by fusing multiple frames of images.

[0042] The point cloud generation unit is used to calculate point cloud data of each tissue layer based on the depth information of the ultrasound image, and generate a preliminary three-dimensional point cloud model of the target area by fusing multiple frames of images. The point cloud generation unit includes a depth information calculation module, a point cloud data construction module and a multi-frame fusion module to achieve high-precision three-dimensional modeling of the target area.

[0043] The depth information refers to the echo signal time difference, attenuation characteristics, and sound propagation path of each tissue structure in the target area obtained by the ultrasound probe at different scanning angles. This information can be used to calculate the relative depth of the tissue structure in the ultrasound imaging plane to establish the positional relationship in three-dimensional space.

[0044] The point cloud data refers to a spatial data set composed of multiple three-dimensional coordinate points. Each point is determined by the intensity and location information of the reflected signal from ultrasound imaging. This point cloud data can reflect the morphological characteristics of tissue structures and provide a basis for subsequent tissue identification and model reconstruction.

[0045] The multi-frame image fusion refers to registering, matching, and fusing ultrasound image data acquired by the ultrasound probe at different angles and times to enhance the stability of ultrasound imaging, reduce noise or artifacts that may exist in a single frame image, and improve the integrity and accuracy of the three-dimensional point cloud model.

[0046] The point cloud generation unit can establish a three-dimensional point cloud model using the following methods:

[0047] The depth information of the target tissue is calculated using the propagation time of ultrasound and mapped to a three-dimensional coordinate space to construct initial point cloud data.

[0048] By combining the phase characteristics of ultrasound, the location of tissue boundary points can be inferred by calculating the phase difference of different ultrasound beams, thereby improving the accuracy of point cloud data.

[0049] Based on the motion trajectory of the ultrasound probe, the propagation path of ultrasound waves in tissues is calculated, and point cloud data is generated using the changes in echo intensity.

[0050] Through the point cloud generation unit, this system can effectively improve the three-dimensional visualization capability of ultrasound image data, enhance the spatial resolution capability of lesion areas, provide high-precision structural information support for ultrasound interventional detection, and make tuberculosis screening and diagnosis more intelligent and accurate.

[0051] The skeletal structure recognition unit is used to identify the skeletal boundaries within the target area based on the point cloud model, so as to determine the three-dimensional anatomical structure of the ribs, sternum and hard tissues.

[0052] The skeletal structure recognition unit is used to identify the skeletal boundaries within the target area based on the point cloud model, so as to determine the three-dimensional anatomical structure of the ribs, sternum and hard tissues. The skeletal structure recognition unit includes a skeletal boundary detection module, a skeletal point cloud extraction module and a three-dimensional anatomical modeling module, which are used to accurately construct a spatial distribution model of the bones and provide stable anatomical reference information for ultrasound interventional detection.

[0053] The bone boundary refers to the reflective interface formed by the high acoustic impedance of bone tissue in ultrasound images. This boundary can be used to distinguish bone from soft tissue and serves as a spatial reference for ultrasound guidance.

[0054] Three-dimensional anatomical structure refers to the skeletal spatial model within the target area reconstructed based on point cloud data, including the spatial arrangement information of ribs, sternum and other hard tissues. This structure can be used to optimize the imaging path of ultrasound probes and improve the accuracy of interventional detection.

[0055] The bone boundary detection can employ ultrasonic echo signal feature analysis to detect regions of abrupt changes in acoustic impedance between tissues, thereby extracting the bone contour. Furthermore, it can be combined with gradient-enhanced edge detection algorithms (such as the Sobe or Canny operators) to accurately locate bone boundaries and improve recognition accuracy.

[0056] Hard tissues such as ribs and sternum can obstruct the propagation of ultrasound waves, limiting imaging in certain areas and affecting the integrity of ultrasound detection. By establishing an accurate skeletal structure model, the optimal ultrasound incident angle can be determined, the ultrasound propagation path can be optimized, and artifact interference can be reduced.

[0057] The skeletal structure recognition unit, combined with point cloud data analysis, ultrasound boundary detection, and 3D modeling algorithms, can accurately identify the 3D anatomical structure of ribs, sternum, and other hard tissues, improve ultrasound imaging quality, optimize ultrasound guidance paths, enhance the accuracy of tuberculosis detection, and provide key technical support for intelligent ultrasound navigation.

[0058] The soft tissue modeling unit is used to construct a soft tissue model within a target area based on ultrasound echo characteristics.

[0059] The soft tissue modeling unit is used to construct a soft tissue model within a target area based on ultrasound echo characteristics. The soft tissue modeling unit includes an echo characteristic analysis module, a soft tissue point cloud generation module, and a three-dimensional reconstruction module, which are used to accurately extract the echo information of soft tissue and reconstruct the soft tissue anatomical structure of the target area to improve the accuracy and visualization effect of ultrasound detection.

[0060] The three-dimensional representation of tissue structures within the target area reconstructed from ultrasound echo data, including morphological information of lung tissue, pleura, lymph nodes, muscle tissue, and other non-bone tissues, can be used for lesion detection, tissue segmentation, and surgical planning. The soft tissue modeling unit uses ultrasound echo feature analysis, such as time gain compensation technology, to correct the attenuation effect of ultrasound waves at different depths and improve image contrast. It then uses the echo time calculation method to calculate the relative depth information of the soft tissue layers based on the round-trip time of ultrasound waves. Finally, it employs an iterative nearest-point algorithm for multi-frame data registration to improve the spatial stability of the soft tissue point cloud.

[0061] In ultrasound-guided precision tuberculosis detection, accurate soft tissue models help doctors optimize puncture paths, improving the safety and accuracy of interventional procedures. Intelligent modeling reduces noise interference, improves the clarity of soft tissue boundaries, makes lesion areas more prominent, and enhances the readability of ultrasound images.

[0062] The ultrasound guidance unit, connected to the three-dimensional modeling unit, is used to calculate the optimal detection path based on the three-dimensional model and the real-time position of the ultrasound probe, and guide the operator to obtain a second ultrasound image through the intercostal space.

[0063] The ultrasound guidance unit includes a path planning unit, a probe angle optimization unit, and a force feedback unit, which are used to optimize the ultrasound imaging process and improve the accuracy and repeatability of the detection.

[0064] The path planning unit is used to calculate the optimal ultrasound detection path based on the three-dimensional model and the anatomical structure of the target area to avoid areas obstructed by bone. The path planning unit includes an anatomical information parsing module, a path optimization calculation module, and a dynamic adjustment module to ensure smooth ultrasound signal propagation, improve ultrasound imaging quality, and optimize the ultrasound detection operation process.

[0065] The anatomical information parsing module is used to analyze the three-dimensional anatomical structure of the target region, extract key anatomical features, and establish an anatomical dataset that can be used for path planning. Further, this module includes:

[0066] The ultrasound image segmentation unit employs deep learning-based image segmentation algorithms (such as U-Net or DeepLabV3+) to automatically identify bones, blood vessels, soft tissues, and lesion areas in ultrasound images. Combined with an adaptive region growing algorithm, it refines tissues with blurred boundaries, improving segmentation accuracy.

[0067] The skeletal shielding region identification unit uses the Hough transform algorithm to detect rib boundaries in ultrasound images and establish a skeletal shielding region model. Combined with skeletal point cloud data, it calculates the location of interrib cages to determine the optimal area where ultrasound can penetrate.

[0068] The target tissue modeling unit employs a voxel reconstruction algorithm to transform ultrasound images into a three-dimensional anatomical model. Combined with Bayesian inference, it optimizes anatomical structure prediction and reduces errors caused by missing data.

[0069] The path optimization calculation module is used to calculate the optimal ultrasound detection path based on anatomical information, ensuring smooth ultrasound signal propagation and improving image quality. This module employs the following algorithm for path optimization:

[0070] The goal of this algorithm is to find a suitable starting point for ultrasound detection, enabling ultrasound waves to propagate in a straight line to the target area while avoiding areas obstructed by bones, thus ensuring the imaging quality and effectiveness of ultrasound detection.

[0071] The following constraints must be followed during the route planning process:

[0072] The target point is fixed and cannot be changed; that is, the target detection position.

[0073] The detection starting point is adjustable; the optimal detection position needs to be found so that the ultrasonic signal can propagate to the target point in a straight line.

[0074] The detection path must be a straight line; the ultrasonic wave cannot bend.

[0075] The path cannot pass through the bone region; it must spread within the soft tissue region.

[0076] The model is divided into a regular grid, with each grid representing an optional detection starting point.

[0077] Assign properties to each grid cell:

[0078] Can it be used as the starting point for ultrasonic detection?

[0079] Is the area a soft tissue region that allows ultrasound waves to penetrate?

[0080] Does it fall within an area obstructed by bone, preventing ultrasound waves from passing through?

[0081] The current position of the ultrasound probe is selected as the starting point. If the point does not meet the path requirements, the algorithm will find the optimal detection starting point.

[0082] Based on the current locations of the detection starting point and the target point, calculate the straight-line path between them. Discretize this straight line into multiple discrete points to perform path feasibility detection within a grid.

[0083] Examine each grid point on the straight path to determine if it passes through a skeletal region: if all path points are within soft tissue regions, the detection starting point is usable, and the search terminates. If the path contains skeletal regions, the current detection starting point is not feasible, and the detection starting point needs to be adjusted.

[0084] During adjustments, new detection points are searched within the candidate detection point area. Using the target point as a reference, the search is expanded outwards to find possible detection points. The search range is set by the physician and typically includes a certain area in front of the target point to cover all possible ultrasound incidence angles.

[0085] Calculate a straight path from the candidate probe starting point to the target point, checking if it completely avoids the skeletal region. If the path is feasible, set the probe starting point as the final probe starting point and terminate the search. If the path is still not feasible, continue searching for other candidate points until the optimal probe starting point is found.

[0086] Prioritize selecting the feasible detection starting point closest to the target point to ensure the shortest ultrasound detection path and improve signal quality. If multiple feasible detection starting points exist, select the one with the optimal ultrasound angle, i.e., the point where the ultrasound incident angle is close to perpendicular to the lesion area, to obtain the best echo signal.

[0087] The dynamic adjustment module is used to adjust the path based on real-time feedback during the detection process to improve detection accuracy. This module includes:

[0088] The real-time feedback adjustment unit, combined with ultrasound echo quality analysis, monitors the intensity and distribution of the ultrasound signal to determine the current imaging quality. Furthermore, fuzzy logic control can be employed to adjust the probe position and optimize the path based on signal quality.

[0089] The target area tracking unit employs an optical flow tracking algorithm to monitor the movement of target tissue in ultrasound images, ensuring that lesions do not deviate from the field of view during scanning. Furthermore, particle filtering can be incorporated to improve tracking stability and reduce errors.

[0090] The intelligent path adaptive unit dynamically adjusts the path during scanning through neural network optimization, improving adaptability. Furthermore, it can be combined with reinforcement learning algorithms to continuously optimize the probe path, enhancing long-term detection performance.

[0091] The path planning unit optimizes the path, allowing the ultrasound probe to avoid the rib area, minimizing the bone shielding effect, improving deep tissue imaging capabilities, ensuring uniform ultrasound signal coverage of the target area, and improving the accuracy of lesion identification.

[0092] The probe angle optimization unit is used to calculate the recommended probe tilt angle based on the current position of the ultrasound probe and the positional relationship of the target area.

[0093] The probe angle optimization unit is used to comprehensively analyze the ultrasonic wave propagation path, tissue acoustic characteristics and imaging requirements, calculate the optimal probe angle, and provide dynamic adjustment suggestions to improve detection accuracy and operational consistency.

[0094] When ultrasound waves propagate between different tissues, they are affected by reflection, refraction, scattering, and absorption. The intensity and penetration depth of the tissue echo also vary depending on the incident angle of the ultrasound wave. An optimal probe angle can effectively reduce ultrasound signal loss and improve imaging contrast.

[0095] Based on the depth of the target area and the current position of the ultrasound probe, the optimal incident angle of the ultrasound waves is calculated to maximize signal penetration and reduce reflection interference at tissue interfaces. An ultrasonic acoustic model is used to analyze the propagation path of ultrasound waves between different tissues, and the optimal detection angle is determined by combining this with the theory of oblique incidence wave propagation.

[0096] Calculate the normal direction from the ultrasound probe to the target area, and adjust the probe angle based on the law of incidence of sound waves so that the incident angle is as close as possible to the tissue surface perpendicular to the lesion area, in order to reduce the reflection loss of ultrasound signals.

[0097] The force feedback unit is used to monitor the pressure, position, and angle of the ultrasonic probe during operation and to provide tactile feedback to determine the optimal position and angle of the ultrasonic probe.

[0098] The force feedback unit includes a pressure sensing module, a position detection module, an angle calculation module, a feedback control module, and a real-time adjustment module. These modules work together to ensure that the ultrasonic probe is in the optimal operating state, thereby improving the stability and accuracy of ultrasonic imaging.

[0099] The pressure sensing module is located at the bottom of the ultrasound probe. It monitors the pressure between the probe and the skin in real time using a piezoresistive sensor, strain gauge sensor, or capacitive sensor, and transmits the pressure data to the feedback control module to prevent tissue deformation due to excessive pressure or ultrasound signal attenuation due to insufficient pressure. The position detection module, based on an inertial measurement unit (IMU), magnetic positioning system, or optical tracking system, acquires the spatial coordinates of the ultrasound probe in real time and calculates the relative positional deviation between the probe and the target area, ensuring the probe is always on the optimal detection path. The angle calculation module, using a gyroscope, accelerometer, and magnetometer, measures the tilt angle of the ultrasound probe and, combined with the anatomical features of the target area, calculates the optimal incident angle of the ultrasound waves to reduce reflection loss and improve echo signal quality.

[0100] The feedback control module receives pressure, position, and angle data and analyzes the probe's operating status. When the probe pressure exceeds the preset range, the position deviates from the target area, or the angle does not reach the optimal incident angle, the feedback control module provides adjustment suggestions to the operator through tactile feedback. This tactile feedback can be provided by a vibration motor built into the probe handle, allowing the operator to perceive probe adjustment needs through changes in vibration intensity during the testing process. The real-time adjustment module, based on the instructions from the feedback control module, dynamically optimizes the probe's operating strategy, enabling the probe to automatically adapt to individual differences among patients during testing, thus improving the stability and consistency of the testing.

[0101] The detection module, connected to the ultrasound guide, is used for accurate tuberculosis detection based on the second ultrasound image.

[0102] The detection module includes an ultrasound image preprocessing unit, a feature extraction unit, a lesion identification unit, and a classification and diagnosis unit, to analyze and process ultrasound image data of the target area and automatically identify tuberculosis-related lesions, thereby improving the accuracy and reliability of detection.

[0103] The ultrasound image preprocessing unit receives the second ultrasound image acquired by the ultrasound guide unit and performs noise removal, contrast enhancement, and edge optimization on the image to ensure that the image quality reaches the optimal analysis conditions. This unit uses an adaptive filtering algorithm to remove high-frequency noise from the ultrasound image, while combining histogram equalization to improve the visualization of the lesion area. It also uses an image segmentation algorithm to extract possible lesion areas to reduce background interference and improve the accuracy of subsequent detection.

[0104] The feature extraction unit extracts tuberculosis-related image features based on the texture, morphology, and echo characteristics of ultrasound images. This unit employs methods such as Gray-Level Co-occurrence Matrix Analysis (GLCM), Local Binary Pattern (LBP), and wavelet transform to extract multi-scale features from lesion areas in ultrasound images. Furthermore, it combines these features with a deep learning feature extraction network to further enhance the ability to distinguish lesion areas from normal tissue.

[0105] The lesion identification unit, based on feature extraction results, employs an improved U-Net deep learning network to perform semantic segmentation of the target region, accurately delineating the tuberculosis lesion area. It then combines this with a classification model based on ultrasound echo intensity to determine the extent, location, and type of the lesion. Furthermore, this unit integrates multi-frame ultrasound image fusion to perform time-series analysis of the ultrasound images of the target region, thereby improving the stability and accuracy of lesion detection.

[0106] The classification and diagnosis unit, based on lesion identification results, employs machine learning algorithms such as Support Vector Machine (SVM), Random Forest (RF), or Deep Neural Network (DNN) to classify lesion areas. It also combines this classification with clinical tuberculosis cases from the image database to assess the malignancy probability, development trend, and pathological characteristics of the lesions. This unit can also be integrated with expert systems or telemedicine platforms to provide automated tuberculosis diagnosis assistance suggestions, improving diagnostic efficiency.

[0107] In summary, the detection module, through ultrasound image preprocessing, feature extraction, lesion identification, and intelligent classification, can accurately analyze the second ultrasound image, realize the automatic detection and classification of tuberculosis lesion areas, improve the level of intelligence in detection, reduce human error, and provide efficient and stable technical support for clinical diagnosis.

[0108] For any module structures not specifically defined in this invention, the existing technical descriptions shall prevail. The prior art mentioned in the foregoing background and specific embodiments sections can be considered part of this invention and used to understand the meaning of certain technical features or parameters.

Claims

1. A precision tuberculosis detection system based on ultrasound intervention, characterized in that, The system includes: An ultrasonic probe is used to acquire ultrasonic images of a target area; the ultrasonic probe also includes an inertial measurement unit, which measures the spatial position information of the ultrasonic probe. A 3D modeling unit is used to construct a 3D model of the target area based on the ultrasound image and the spatial position information of the ultrasound probe. The 3D modeling unit includes: The data acquisition unit is used to receive multi-angle ultrasound image data acquired by the ultrasound probe and simultaneously record the spatial position information of the ultrasound probe. The point cloud generation unit is used to calculate the depth information of the target tissue using the propagation time of ultrasound and map it to a three-dimensional coordinate space to construct initial point cloud data. Combining the phase characteristics of ultrasound, the position of tissue boundary points is inferred by calculating the phase difference of different ultrasound beams. Based on the motion trajectory of the ultrasound probe, the propagation path of ultrasound in the tissue is calculated, and point cloud data is generated using the change in echo intensity. Finally, a preliminary three-dimensional point cloud model of the target area is generated by fusing multiple frames of images. The skeletal structure recognition unit is used to identify the skeletal boundaries within the target area based on the point cloud model, so as to determine the three-dimensional anatomical structure of the ribs, sternum and hard tissues. The soft tissue modeling unit is used to construct a soft tissue model within the target area based on the ultrasound echo characteristics. An ultrasound guiding unit, connected to the three-dimensional modeling unit, is used to calculate the optimal detection path based on the three-dimensional model and the real-time position of the ultrasound probe, and guide the operator to acquire a second ultrasound image through the intercostal space. The ultrasound guiding unit includes: The path planning unit is used to calculate the optimal ultrasound detection path based on the three-dimensional model and the anatomical structure of the target area to avoid areas obscured by bones. The probe angle optimization unit is used to calculate the recommended probe tilt angle based on the current position of the ultrasound probe and the positional relationship of the target area. The force feedback unit is used to monitor the pressure of the ultrasonic probe during operation and provide tactile feedback to determine the optimal position and angle of the ultrasonic probe. The detection module, connected to the ultrasound guide, is used for accurate tuberculosis detection based on the second ultrasound image.

2. The tuberculosis precision detection system based on ultrasound intervention according to claim 1, characterized in that, The skeletal structure recognition unit uses ultrasonic echo signal feature analysis to detect regions of abrupt changes in acoustic impedance between tissues in order to extract the skeletal contour.

3. The tuberculosis precision detection system based on ultrasound intervention according to claim 1, characterized in that, The soft tissue modeling unit uses time gain compensation technology to correct the attenuation effect of ultrasound at different depths and improve image contrast; then it uses echo time calculation method to calculate the relative depth information of soft tissue layers based on the round-trip time of ultrasound; finally, it uses iterative nearest point algorithm to perform multi-frame data registration.

4. The tuberculosis precision detection system based on ultrasound intervention according to claim 1, characterized in that, The path planning unit includes an anatomical information parsing module, a path optimization calculation module, and a dynamic adjustment module.

5. The tuberculosis precision detection system based on ultrasound intervention according to claim 4, characterized in that, The anatomical information parsing module includes: The ultrasound image segmentation unit uses a deep learning-based image segmentation algorithm to automatically identify bones, blood vessels, soft tissues, and lesion areas in ultrasound images. The bone shielding region identification unit detects rib boundaries in ultrasound images using the Hough transform algorithm and establishes a bone shielding region model. The target tissue modeling unit uses a voxel reconstruction algorithm to convert ultrasound images into a three-dimensional anatomical model.

6. The tuberculosis precision detection system based on ultrasound intervention according to claim 5, characterized in that, The path optimization calculation module uses the following algorithm for path optimization: The model is divided into a regular grid, with each grid representing an optional detection starting point; Assign characteristics to each grid cell: whether it can be used as a starting point for ultrasound detection, whether it is a soft tissue area, or whether it belongs to a bone-covered area; The current ultrasonic probe position is selected as the starting point. If the starting point does not meet the path requirements, the algorithm will find the optimal detection starting point. Calculate the straight-line path between the current detection starting point and the target point; Check all grids on the straight path one by one to determine if it passes through the skeletal region: if all path points are in the soft tissue region, the detection starting point is usable and the search is terminated; If there are skeletal regions in the path, the current detection starting point is not feasible and needs to be adjusted. During adjustment, new detection starting points are searched within the candidate detection starting point area. Based on the target point, the search is expanded outward to find possible detection starting points. Calculate a straight path from the candidate detection starting point to the target point, and check whether it completely avoids the skeletal region. If the path is feasible, set the detection starting point as the final detection starting point and terminate the search. If the path is still not feasible, continue searching for other candidate points until the best detection starting point is found.

7. The tuberculosis precision detection system based on ultrasound intervention according to claim 6, characterized in that, The path optimization includes: Prioritize selecting the feasible detection starting point closest to the target point to ensure the shortest ultrasonic detection path; If multiple feasible detection points exist, select the detection point with the optimal ultrasound angle, that is, the point where the ultrasound incident angle is close to perpendicular to the lesion area, in order to obtain the best echo signal.

8. The tuberculosis precision detection system based on ultrasound intervention according to claim 1, characterized in that, The force feedback unit includes a pressure sensing module, a position detection module, an angle calculation module, a feedback control module, and a real-time adjustment module.

9. The tuberculosis precision detection system based on ultrasound intervention according to claim 1, characterized in that, The detection module includes an ultrasound image preprocessing unit, a feature extraction unit, a lesion identification unit, and a classification and diagnosis unit.

Images