Search

CN-121971116-A - Dental X-ray tomography system and method based on intelligent mechanical arm

CN121971116ACN 121971116 ACN121971116 ACN 121971116ACN-121971116-A

Abstract

The invention discloses a dental X-ray tomography system and a dental X-ray tomography method based on an intelligent mechanical arm, which relate to the technical field of medical imaging, wherein the method comprises the steps of obtaining a high-precision three-dimensional point cloud model of an oral cavity target area through three-dimensional scanning of structured light; the method comprises the steps of generating a collision-free optimal scanning path sequence and a theoretical pose based on the model, controlling a mechanical arm to carry an X-ray source to move along the optimal path and synchronously trigger exposure and image acquisition, monitoring the actual pose of the tail end of the mechanical arm and the quality of a projection image in real time to form a double feedback closed loop, calculating pose compensation instructions and dynamically adjusting exposure parameters, and executing a three-dimensional iterative reconstruction algorithm by using accurate geometric parameters to generate a tomographic image and three-dimensional volume data. The invention can realize self-adaptive planning of scanning track, dynamic correction of pose error, millisecond-level self-adaptive dosage regulation and end-to-end full-automatic workflow, remarkably improve the spatial resolution and geometric fidelity of reconstructed images and reduce the radiation dosage of patients.

Inventors

  • CHEN QIANGJIANG
  • YIN XUFENG

Assignees

  • 上海汉缔医疗设备有限公司

Dates

Publication Date
20260505
Application Date
20260123

Claims (10)

  1. 1. A dental X-ray tomography system based on an intelligent mechanical arm, which is characterized by comprising the following components: the multi-mode sensing module is used for acquiring anatomical structure information of a patient oral cavity area, pose information of an end effector of the mechanical arm and X-ray projection image information in real time; The intelligent decision and control module is in communication connection with the multi-mode sensing module and is used for constructing an imaging geometric and kinematic coupling model based on sensing information and generating a dynamically optimized scanning path, a real-time pose compensation instruction and self-adaptive X-ray exposure parameters; The high-precision execution module comprises at least one end effector of a six-degree-of-freedom cooperative mechanical arm, wherein a miniature X-ray source and an optical positioning target are integrated on the end effector, and the high-precision execution module is connected with the intelligent decision and control module and is used for receiving and executing a motion control instruction and an exposure triggering instruction issued by the high-precision execution module; The image acquisition and reconstruction module comprises at least one digital X-ray detector and is used for receiving X-rays penetrating through tissue of a patient and generating projection images, and the image acquisition and reconstruction module is connected with the intelligent decision and control module and the multi-mode sensing module and is used for receiving synchronous signals and executing acquisition of projection data and reconstruction of three-dimensional tomographic images.
  2. 2. The smart robotic arm-based dental X-ray tomography system of claim 1, wherein the multi-modality perception module comprises a structured light three-dimensional scanning unit, an optical tracking unit, and a projection image analysis unit; The structured light three-dimensional scanning unit is used for carrying out non-contact three-dimensional reconstruction on the oral cavity area of a patient before scanning to obtain an initial three-dimensional point cloud model containing tooth, gum and jaw bone contours; The optical tracking unit is used for tracking a passive reflective marker ball fixed on the end effector of the mechanical arm in real time and outputting six-degree-of-freedom pose data of the end effector under a global coordinate system; The projection image analysis unit is used for analyzing each acquired frame of X-ray projection image in real time and extracting the signal-to-noise ratio, contrast and edge definition index of the key anatomical structure of the image.
  3. 3. The dental X-ray tomography system based on the intelligent mechanical arm according to claim 1, wherein the intelligent decision and control module is characterized in that an imaging geometry and kinematics coupling model is used for carrying out unified mathematical expression on a kinematics forward solution model of the mechanical arm, an imaging geometry model of an X-ray source and a detector and a three-dimensional model of a patient oral cavity acquired by a structured light three-dimensional scanning unit, and the imaging geometry and kinematics coupling model is used for generating an optimal scanning track of the tail end of the collision-free mechanical arm meeting the complete sampling condition based on a multi-objective optimization function.
  4. 4. The dental X-ray tomography system based on the intelligent mechanical arm according to claim 1, wherein the intelligent decision and control module comprises a real-time closed-loop compensation unit, the real-time closed-loop compensation unit receives real-time pose feedback from the optical tracking unit and image quality feedback from the projection image analysis unit, when detecting that the pose deviation exceeds a preset threshold value, the real-time closed-loop compensation unit calculates joint angular velocity correction amount in the next control period based on the deviation amount and a jacobian matrix of the mechanical arm and sends the joint angular velocity correction amount to the high-precision execution module for dynamic compensation, and meanwhile, the real-time closed-loop compensation unit dynamically adjusts the product value of tube voltage and tube current of the next frame of X-ray exposure according to the estimated thickness of the penetrating tissue and the image signal-to-noise ratio under the current projection angle, and the adjustment strategy follows the principle that the signal intensity received by the detector is maintained in the preset optimal interval.
  5. 5. The dental X-ray tomography system based on the intelligent mechanical arm according to claim 1, wherein the image acquisition and reconstruction module adopts an iterative reconstruction algorithm based on GPU acceleration, the iterative reconstruction algorithm takes accurate scanning geometric parameters provided by the intelligent decision and control module as prior knowledge input, a regularization term based on total variation minimization is introduced in the reconstruction process, and the iterative reconstruction algorithm adopts an algorithm framework based on ordered subset maximum expected values.
  6. 6. A dental X-ray tomography method based on a smart robotic arm, the method comprising the steps of: Step S110, a high-precision three-dimensional point cloud model of a patient oral cavity target area is obtained through structured light three-dimensional scanning, and a volume of interest to be scanned is defined based on the model; step S120, based on a three-dimensional point cloud model, mechanical arm kinematics constraint and imaging geometry constraint, an intelligent decision system solves a multi-objective optimization problem and generates a collision-free optimal scanning path sequence from an initial swing point to a scanning end point and a corresponding theoretical pose; Step S130, controlling the mechanical arm to carry the X-ray source to move along an optimal scanning path, and synchronously triggering X-ray exposure and detector image acquisition at each preset projection angle position; Step S140, monitoring the actual pose of the tail end of the mechanical arm through an optical tracking system in real time in the scanning movement and image acquisition process, performing image analysis, and monitoring the quality of a projection image to form a double-feedback closed loop; step S150, based on double feedback closed loop information, calculating pose compensation instructions in real time and sending the pose compensation instructions to a mechanical arm controller, and dynamically adjusting X-ray exposure parameters of the next projection angle; step S160, after all projection data acquisition is completed, a three-dimensional iterative reconstruction algorithm is executed by using accurate scanning geometric parameters subjected to real-time compensation and correction, so as to generate a tomographic image and three-dimensional volume data of a patient oral cavity target area.
  7. 7. The dental X-ray tomography method as defined in claim 6, wherein the specific process of generating the optimal scan path sequence in step S120 includes: Secondly, taking the center of the volume of interest to be scanned as an origin, discretizing under the spherical coordinate system to generate a series of candidate X-ray source positions, and excluding the positions which collide with the body of the patient or a treatment chair and exceed the working space of the mechanical arm; calculating the optimal pose of the corresponding detector for each candidate X-ray source position so that the projection of the interested volume completely falls into the effective area of the detector; and (3) applying the multi-objective optimization function, and finding out a continuous track which minimizes the objective function value from all feasible source-detector pose pairs through a graph search algorithm or an optimization algorithm.
  8. 8. The dental X-ray tomography method based on the intelligent mechanical arm according to claim 6, wherein the dual feedback closed loop in the step S140 is specifically implemented as: the optical tracking system outputs pose data at 50Hz frequency, and compares the pose data with the theoretical pose at the current moment to obtain pose errors; Triggering pose compensation when the pose error exceeds a position error threshold or an angle error threshold, increasing the mAs value of the next frame according to a preset proportion when the overall signal-to-noise ratio is lower than a lower limit threshold, and decreasing the mAs value according to a proportion when the overall signal-to-noise ratio is higher than an upper limit threshold.
  9. 9. The smart robotic arm-based dental X-ray tomography method of claim 6, wherein the three-dimensional iterative reconstruction algorithm in step S160 employs an ordered subset-based maximum expected value algorithm framework.
  10. 10. The dental X-ray tomography method based on the intelligent mechanical arm according to claim 6, wherein in the step S160, the post-processing is automatically performed on the reconstructed three-dimensional volume data, the three-dimensional shapes of the individual teeth are automatically segmented based on the threshold segmentation and region growing algorithm, and the key anatomical structures of the pulp chamber, the root canal and the periodontal ligament gap are automatically identified and labeled by using the machine learning model.

Description

Dental X-ray tomography system and method based on intelligent mechanical arm Technical Field The invention relates to the technical field of medical imaging, in particular to a dental X-ray tomography system and method based on an intelligent mechanical arm. Background With the continuous progress of the stomatology imaging technology, dental X-ray tomography has become a central tool for accurate diagnosis of dental, periodontal and jaw lesions. According to the technology, the X-ray projection data are acquired at multiple angles, the three-dimensional tomographic image is reconstructed, the diagnostic blind area caused by structural overlapping of the traditional two-dimensional dental film can be effectively overcome, and the visualization capability of key anatomical details such as micro focus, hidden crack, root canal form and the like is remarkably improved. The current dental imaging system is widely integrated with a digital detector, a low-dose radiation source and a computer tomography reconstruction algorithm, and the clinical value of the dental imaging system is fully verified in the fields of planting planning, orthodontic evaluation, early caries detection and the like. However, the existing imaging procedure is highly dependent on manual positioning and parameter adjustment of an operating doctor, has extremely high requirements on technician experience, is difficult to ensure consistency and repeatability of scanning geometry, and restricts the development of cooperation of high-precision quantitative analysis and remote diagnosis and treatment. In which, the dental X-ray imaging technology based on mechanical arm assistance becomes a research hot spot in recent years, and aims to realize the spatial cooperative motion of an X-ray source and a detector by replacing manual operation through an automatic motion platform. Such systems typically employ six degrees of freedom industrial robotic arms carrying miniature X-ray tubes and cooperate with intra-or extra-oral detectors to accomplish circular or non-circular track scanning. The basic principle is that an end effector of a mechanical arm is controlled to move along a specific space curve according to a preset scanning path, X-ray pulses and image acquisition are triggered synchronously, and finally three-dimensional volume data is generated by utilizing a filtering back projection or iterative reconstruction algorithm. The technical route can theoretically improve the positioning precision, reduce the operation threshold and support the personalized scanning protocol, and provides a hardware basis for the real-time imaging and intelligent diagnosis beside the chair. The prior art still faces the contradiction of multiple structures in the process of introducing an intelligent mechanical arm into a dental X-ray tomography, namely, firstly, a mechanical arm kinematic model and an X-ray imaging geometry are not in deep coupling, path planning only depends on a preset track and lacks perception and response to real-time anatomical features of a patient, so that scanning center deviation, insufficient visual field coverage or undersampling of a critical area are caused, secondly, an open-loop control architecture is generally adopted in the prior system, pose errors caused by micro motion of the patient, flexible deformation or thermal drift of the mechanical arm cannot be dynamically compensated in the scanning process, projection data inconsistency is caused, reconstructed image quality is seriously deteriorated, and thirdly, an X-ray dose regulation strategy and a mechanical arm motion state are disjointed, millisecond-level self-adaptive exposure adjustment cannot be carried out according to local tissue density, scanning angle and reconstruction requirements, so that unnecessary radiation exposure risks exist, diagnosis reliability is possibly influenced due to insufficient signal-to-noise ratio, and finally, the whole imaging flow lacks end-to-end intelligent decision capability, manual intervention is required from initial positioning and parameter configuration to reconstruction optimization, and ' one-key ' full-automatic workflow ' cannot be realized. Disclosure of Invention The invention aims to provide an efficient scheduling method and system based on artificial intelligence, which are used for solving the problems of insufficient field coverage, undersampling of a key area and high radiation dosage of a patient in the prior art. In order to solve the technical problems, the invention provides the following technical scheme: The dental X-ray tomography system based on the intelligent mechanical arm comprises a multi-mode sensing module, a detection module and a detection module, wherein the multi-mode sensing module is used for acquiring anatomical structure information of a patient oral cavity area, pose information of an end effector of the mechanical arm and an X-ray projection image in real time; the intelligent decision an