Intelligent navigation system and device for gastrointestinal endoscope

Abstract

This invention provides an intelligent navigation system and device for gastrointestinal endoscopy. The intelligent navigation system for gastrointestinal endoscopy includes: a multi-source sensing and acquisition module, a 3D modeling and registration module, an intelligent path planning module, a real-time pose calculation module, a visual navigation interaction module, and a closed-loop control and execution module. This invention provides an intelligent navigation system for gastrointestinal endoscopy that, by employing multi-source data fusion positioning technology, combined with visual 3D reconstruction and inertial pose calculation, overcomes the shortcomings of single positioning methods, achieving sub-millimeter-level positioning accuracy. It effectively eliminates positioning deviations caused by image interference, inertial cumulative errors, and body position shifts, significantly improving navigation stability and accuracy. Based on an improved algorithm, it achieves adaptive path planning, automatically identifying intestinal obstacles and blind spots, dynamically optimizing the diagnostic and treatment path, and achieving full coverage of the entire digestive tract without blind spots. This significantly reduces the probability of missed diagnoses and misdiagnoses, while also avoiding the risk of intestinal compression damage.

Description

Technical Field

[0001] This invention relates to the field of medical endoscopic navigation technology, and in particular to an intelligent navigation system and device for gastrointestinal endoscopy. Background Technology

[0002] Gastrointestinal diseases are common digestive system diseases in clinical practice. The screening and treatment of diseases such as gastritis, gastric ulcer, enteritis, intestinal polyps, and gastrointestinal tumors are highly dependent on endoscopic equipment. At present, the clinical methods of gastrointestinal endoscopic diagnosis and treatment mainly rely on medical staff to manually operate the endoscope by hand. They determine the position and treatment area by observing the two-dimensional images transmitted by the endoscope in real time, and complete the examination, sampling, minimally invasive surgery and other operations.

[0003] Current gastrointestinal endoscopic diagnostic and treatment techniques suffer from several drawbacks. The reliance on manual operation is heavily dependent on the clinical experience of medical personnel. Novice operators are prone to operational errors, leading to endoscope deviations, missed diagnoses in blind spots, and poor repeatability. Relying solely on two-dimensional images cannot accurately determine the endoscopic's three-dimensional pose and intestinal depth information, resulting in low positioning accuracy and a risk of the endoscope touching the intestinal wall or causing mucosal damage due to excessive pressure. Furthermore, the numerous folds, bends, and narrow areas within the intestine make it difficult to anticipate obstacles and plan the optimal treatment path, resulting in low efficiency and numerous blind spots. Finally, the endoscopic pose adjustment during movement is subject to human error, preventing real-time dynamic correction and significantly increasing the risk of diagnosis and treatment in complex intestinal structures.

[0004] Existing endoscopic navigation technologies rely solely on visual or inertial positioning, which suffers from problems such as visual positioning being affected by intestinal fluid accumulation and shadows, and inertial positioning having large cumulative errors. These technologies cannot achieve high-precision and high-stability intelligent navigation and lack path adaptive planning and closed-loop attitude control functions, making it difficult to meet the clinical needs for precise, safe, and efficient diagnosis and treatment.

[0005] Therefore, it is necessary to provide an intelligent navigation system and device for gastrointestinal endoscopy to solve the above-mentioned technical problems. Summary of the Invention

[0006] This invention provides an intelligent navigation system and device for gastrointestinal endoscopy, which solves the problems of existing endoscopic navigation technologies that only use single visual positioning or inertial positioning, resulting in visual positioning being affected by intestinal fluid accumulation and shadow interference, and inertial positioning having large cumulative errors.

[0007] To address the aforementioned technical problems, this invention provides an intelligent navigation system for gastrointestinal endoscopy, comprising: a multi-source sensing and acquisition module, a 3D modeling and registration module, an intelligent path planning module, a real-time pose calculation module, a visual navigation interaction module, and a closed-loop control execution module;

[0008] The multi-source sensing and acquisition module is used to collect visual image data, inertial pose data, intestinal pressure data and human body posture data in real time during the endoscope's movement, and to complete the synchronous acquisition and preliminary noise reduction preprocessing of multi-source raw data.

[0009] The 3D modeling and registration module is communicatively connected to the multi-source sensing and acquisition module. It is used to construct a real-time 3D scene model of the digestive tract based on the preprocessed visual image data, and to accurately register the real-time 3D scene model with the preset standard digestive tract anatomical model to achieve the fusion and matching of real-time scene and anatomical prior information.

[0010] The intelligent path planning module is communicatively connected to the 3D modeling and registration module, and is used to adaptively plan the optimal endoscopic treatment path based on the registered 3D model, the location of suspected lesions in the digestive tract, intestinal physiological structure parameters, and preset obstacle avoidance rules and treatment priority strategies.

[0011] The real-time pose calculation module is communicatively connected to the multi-source sensing acquisition module and the intelligent path planning module, respectively, and is used to fuse inertial pose data and visual positioning data to calculate the three-dimensional coordinates, deflection angle, travel speed and attitude parameters of the endoscope probe in real time, and generate real-time deviation data by comparing with the optimal travel path.

[0012] The visualization navigation interaction module is communicatively connected to the real-time pose calculation module and the 3D modeling and registration module, and is used to render and display the 3D scene of the digestive tract, the real-time pose of the endoscope, the planned navigation path, the lesion marking points and deviation warning information in real time, and to provide a human-computer interaction control interface.

[0013] The closed-loop control execution module is communicatively connected to the real-time pose calculation module, the intelligent path planning module, and the visual navigation interaction module, respectively. It is used to generate attitude adjustment and movement control commands based on real-time deviation data, drive the endoscope to complete adaptive attitude correction, speed adjustment, and path tracking, and realize closed-loop intelligent navigation control.

[0014] Preferably, the multi-source sensing and acquisition module includes a visual acquisition unit, an inertial sensing unit, a pressure sensing unit, and a body position acquisition unit.

[0015] The visual acquisition unit is mounted on the inner lens end and includes a high-definition miniature camera and a supplementary light source, used to continuously acquire high-definition time-series image data of the inner wall of the digestive tract;

[0016] The inertial sensing unit has a built-in high-precision inertial measurement unit (IMU) for real-time acquisition of the endoscope's three-axis acceleration, three-axis angular velocity, and attitude angle data.

[0017] The pressure sensing unit includes a miniature flexible pressure sensor for collecting pressure data between the endoscope and the intestinal wall, thus avoiding rigid compression that could damage intestinal tissue.

[0018] The body position acquisition unit includes a body surface posture sensing component, which is used to collect data on changes in the patient's body position during the diagnosis and treatment process, and to compensate for positioning errors caused by body position deviation.

[0019] Preferably, the 3D modeling and registration module is specifically used for: extracting feature points, matching frames, and reconstructing 3D from time-series image data to generate a real-time local 3D point cloud model and mesh model of the digestive tract;

[0020] The iterative nearest point algorithm is used to achieve accurate registration between real-time 3D models and standard anatomical models, with matching accuracy error controlled within 0.5mm.

[0021] By combining patient position data, the model is dynamically corrected to eliminate model matching deviations caused by patient positional movement and intestinal deformation, thus achieving real-time dynamic updates of the entire digestive tract scenario.

[0022] Preferably, the intelligent path planning module includes an obstacle avoidance unit and a path optimization unit;

[0023] The obstacle avoidance unit is used to identify obstacles such as narrow areas of the intestine, folds and protrusions, fluid accumulation blind spots, and suspected lesion protrusions, and generates an obstacle bounding box to avoid dangerous travel areas;

[0024] The path optimization unit is based on an improved algorithm, with the optimization goals of full coverage of diagnosis and treatment, shortest travel distance, and lowest risk of intestinal damage. Combined with the flexible motion constraints of the endoscope, it dynamically updates the optimal navigation path and automatically generates deceleration and fine-tuning detour paths for intestinal bends and folds.

[0025] Preferably, the real-time pose calculation module adopts a visual-inertial fusion positioning algorithm, which corrects the cumulative error of inertial measurement through visual positioning results and compensates for the positioning delay between image frames through inertial data, thereby achieving real-time 10ms-level updates of endoscope pose and sub-millimeter-level positioning accuracy.

[0026] Preferably, the visual navigation interaction module supports the synchronous overlay display of two-dimensional endoscopic real-scene images and three-dimensional navigation scenes, and has functions such as lesion location marking, path deviation warning, and abnormal status sound and light alarm. It also supports one-click switching between manual intervention and intelligent automatic navigation mode.

[0027] A digestive tract endoscope intelligent navigation device is used in a digestive tract endoscope intelligent navigation system. The digestive tract endoscope intelligent navigation device includes: an endoscope body, a multi-dimensional sensing component, a micro-drive adjustment component, a data transmission component, an external navigation terminal, an operating device, and an insertion tube.

[0028] The multi-dimensional sensing component is integrated into the tip of the endoscope body and includes a high-definition camera component, an inertial measurement sensor, and a flexible pressure sensor, used to achieve synchronous acquisition of multi-source data;

[0029] The micro-drive adjustment component is disposed in the flexible section of the endoscope body, including a micro stepper motor and a flexible adjustment joint, and is used to receive control commands from the closed-loop control execution module to complete the adaptive adjustment of the endoscope's up-down, left-right deflection and travel speed.

[0030] The data transmission component employs a wireless low-latency transmission module to enable real-time bidirectional transmission of multi-source data and control commands between the endoscope body and the external navigation terminal. The insertion tube is connected to one side of the data transmission component pipe, and the operating device is located at one end of the insertion tube.

[0031] The external navigation terminal has a built-in data processing chip, display screen, and human-computer interaction buttons, which are used to complete data calculation, three-dimensional visualization display, and navigation control operations.

[0032] Preferably, the endoscope body is a flexible and bendable structure with an outer diameter of no more than 5 mm at the tip. The flexible section can achieve 360° all-round fine adjustment and deflection to adapt to the narrow, curved and complex physiological structure of the digestive tract.

[0033] Preferably, the micro-drive adjustment component has micron-level fine-tuning precision, can correct the endoscope posture in real time according to path deviation data, has a maximum deflection adjustment angle of ±90°, and can adaptively and continuously adjust the travel speed in the range of 0-5mm / s.

[0034] Preferably, the external navigation terminal has a built-in data storage module for real-time storage of image data, navigation path data, and pose data during the diagnosis and treatment process, supporting the retrospective analysis and review of diagnosis and treatment data.

[0035] Compared with related technologies, the intelligent navigation system for gastrointestinal endoscopy provided by this invention has the following advantages:

[0036] This invention provides an intelligent navigation system for gastrointestinal endoscopy. By employing multi-source data fusion positioning technology, combined with visual 3D reconstruction and inertial pose calculation, it overcomes the shortcomings of single positioning methods, achieving sub-millimeter positioning accuracy. It effectively eliminates positioning deviations caused by image interference, inertial cumulative errors, and body position shifts, significantly improving navigation stability and accuracy. Based on an improved algorithm, it achieves adaptive path planning, automatically identifying intestinal obstacles and blind spots, dynamically optimizing the treatment path, and achieving full coverage of the digestive tract without blind spots. This significantly reduces the probability of missed diagnoses and misdiagnoses, while avoiding the risk of intestinal compression damage. Employing a closed-loop intelligent control mode, it can monitor endoscopic deviations in real time and automatically adjust posture and speed, eliminating the need for continuous and precise manual control. This significantly reduces the workload of medical staff, lessens reliance on the operator's clinical experience, and improves the standardization of diagnosis and treatment. It achieves simultaneous visualization of 2D real-world scenes and 3D navigation scenarios, supports flexible human-computer interaction, and allows free switching between automatic navigation and manual control modes. Balancing intelligence and clinical operational flexibility, it is adaptable to various gastrointestinal endoscopic diagnosis and treatment scenarios, making it widely applicable. Attached Figure Description

[0037] Figure 1 A schematic diagram of the structure of an intelligent navigation system for gastrointestinal endoscopy provided by the present invention;

[0038] Figure 2 for Figure 1 The diagram shows a multi-source sensing and acquisition module.

[0039] Figure 3 for Figure 1 The diagram shows the module diagram of the intelligent path planning module;

[0040] Figure 4 This is a schematic diagram of the structure of an intelligent navigation system for digestive tract endoscopy provided by the present invention.

[0041] The following components are labeled in the diagram: 1. Endoscope body; 2. Multi-dimensional sensing component; 3. Micro-drive adjustment component; 4. Data transmission component; 5. External navigation terminal; 6. Operating device; 7. Insertion tube. Detailed Implementation

[0042] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0043] First Embodiment

[0044] Please refer to the following: Figure 1 , Figure 2 and Figure 3 ,in, Figure 1 A schematic diagram of the structure of an intelligent navigation system for gastrointestinal endoscopy provided by the present invention; Figure 2 for Figure 1 The diagram shows a multi-source sensing and acquisition module. Figure 3 for Figure 1 The diagram shows the intelligent path planning module.

[0045] A digestive tract endoscopy intelligent navigation system includes: a multi-source sensing and acquisition module, a three-dimensional modeling and registration module, an intelligent path planning module, a real-time pose calculation module, a visual navigation interaction module, and a closed-loop control and execution module;

[0046] The multi-source sensing and acquisition module is used to collect visual image data, inertial pose data, intestinal pressure data and human body posture data in real time during the endoscope's movement, and to complete the synchronous acquisition and preliminary noise reduction preprocessing of multi-source raw data.

[0047] The 3D modeling and registration module is communicatively connected to the multi-source sensing and acquisition module. It is used to construct a real-time 3D scene model of the digestive tract based on the preprocessed visual image data, and to accurately register the real-time 3D scene model with the preset standard digestive tract anatomical model to achieve the fusion and matching of real-time scene and anatomical prior information.

[0048] The intelligent path planning module is communicatively connected to the 3D modeling and registration module, and is used to adaptively plan the optimal endoscopic treatment path based on the registered 3D model, the location of suspected lesions in the digestive tract, intestinal physiological structure parameters, and preset obstacle avoidance rules and treatment priority strategies.

[0049] The real-time pose calculation module is communicatively connected to the multi-source sensing acquisition module and the intelligent path planning module, respectively, and is used to fuse inertial pose data and visual positioning data to calculate the three-dimensional coordinates, deflection angle, travel speed and attitude parameters of the endoscope probe in real time, and generate real-time deviation data by comparing with the optimal travel path.

[0050] The visualization navigation interaction module is communicatively connected to the real-time pose calculation module and the 3D modeling and registration module, and is used to render and display the 3D scene of the digestive tract, the real-time pose of the endoscope, the planned navigation path, the lesion marking points and deviation warning information in real time, and to provide a human-computer interaction control interface.

[0051] The closed-loop control execution module is communicatively connected to the real-time pose calculation module, the intelligent path planning module, and the visual navigation interaction module, respectively. It is used to generate attitude adjustment and movement control commands based on real-time deviation data, drive the endoscope to complete adaptive attitude correction, speed adjustment, and path tracking, and realize closed-loop intelligent navigation control.

[0052] The multi-source sensing and acquisition module includes a visual acquisition unit, an inertial sensing unit, a pressure sensing unit, and a body position acquisition unit.

[0053] The visual acquisition unit is mounted on the inner lens end and includes a high-definition miniature camera and a supplementary light source, used to continuously acquire high-definition time-series image data of the inner wall of the digestive tract;

[0054] The inertial sensing unit has a built-in high-precision inertial measurement unit (IMU) for real-time acquisition of the endoscope's three-axis acceleration, three-axis angular velocity, and attitude angle data.

[0055] The pressure sensing unit includes a miniature flexible pressure sensor for collecting pressure data between the endoscope and the intestinal wall, thus avoiding rigid compression that could damage intestinal tissue.

[0056] The body position acquisition unit includes a body surface posture sensing component, which is used to collect data on changes in the patient's body position during the diagnosis and treatment process, and to compensate for positioning errors caused by body position deviation.

[0057] The 3D modeling and registration module is specifically used to: extract feature points, match frames, and reconstruct 3D from time-series image data to generate a real-time local 3D point cloud model and mesh model of the digestive tract.

[0058] The iterative nearest point algorithm is used to achieve accurate registration between real-time 3D models and standard anatomical models, with matching accuracy error controlled within 0.5mm.

[0059] By combining patient position data, the model is dynamically corrected to eliminate model matching deviations caused by patient positional movement and intestinal deformation, thus achieving real-time dynamic updates of the entire digestive tract scenario.

[0060] The intelligent path planning module includes an obstacle avoidance unit and a path optimization unit;

[0061] The obstacle avoidance unit is used to identify obstacles such as narrow areas of the intestine, folds and protrusions, fluid accumulation blind spots, and suspected lesion protrusions, and generates an obstacle bounding box to avoid dangerous travel areas;

[0062] The path optimization unit is based on an improved algorithm, with the optimization goals of full coverage of diagnosis and treatment, shortest travel distance, and lowest risk of intestinal damage. Combined with the flexible motion constraints of the endoscope, it dynamically updates the optimal navigation path and automatically generates deceleration and fine-tuning detour paths for intestinal bends and folds.

[0063] The pose real-time calculation module adopts a visual-inertial fusion positioning algorithm, which corrects the cumulative error of inertial measurement through visual positioning results and compensates for the positioning delay between image frames through inertial data, so as to achieve real-time update of endoscope pose at the 10ms level and achieve a positioning accuracy of sub-millimeter level.

[0064] The visualization navigation interaction module supports the synchronous overlay display of two-dimensional endoscopic real-scene images and three-dimensional navigation scenes. It has functions such as lesion location marking, path deviation warning, and abnormal status sound and light alarm. It also supports one-click switching between manual intervention and intelligent automatic navigation mode.

[0065] The working principle of the intelligent navigation system for gastrointestinal endoscopy provided by this invention is as follows:

[0066] During operation, the multi-source sensing and acquisition module first synchronously acquires multi-dimensional data such as visual images, inertial pose, intestinal contact pressure, and patient body posture after the endoscope enters the digestive tract, and performs time-stamp alignment and preliminary filtering and noise reduction on the raw signals. Subsequently, the 3D modeling and registration module constructs a local 3D point cloud and mesh model based on high-quality visual sequences, and registers it with the standard anatomical model with high precision through the Iterative Closest Point (ICP) algorithm, with the registration error controlled within 0.5mm. At the same time, it integrates body position sensing data to dynamically correct model deformation caused by patient movement or intestinal peristalsis.

[0067] Based on this, the intelligent path planning module combines the identified suspected lesion areas, intestinal anatomical features and real-time obstacle information, and uses an improved path optimization algorithm to generate the optimal travel trajectory that takes into account the integrity, safety and efficiency of diagnosis and treatment, and automatically inserts deceleration and fine-tuning detour strategies for curved sections or folded areas.

[0068] Meanwhile, the real-time pose calculation module continuously integrates visual and inertial data, outputting the 3D coordinates, deflection angle, and motion status of the endoscope at a 10ms update frequency, and comparing it with the planned path to generate a deviation. This deviation information is sent to the visualization navigation interaction module, which simultaneously presents the 3D scene, endoscope pose, navigation path, lesion markings, and warning prompts on the external terminal, supporting real-time monitoring and mode switching by doctors. On the other hand, it is transmitted to the closed-loop control execution module, driving the micro-stepping motor and flexible adjustment joint to make micron-level adaptive adjustments to the endoscope's posture and travel speed, ensuring that the endoscope strictly tracks the planned path. The entire process includes a closed-loop control link of perception, modeling, planning, calculation, display, and control, effectively overcoming the limitations of a single sensor and significantly improving navigation accuracy, stability, and clinical applicability.

[0069] Compared with related technologies, the intelligent navigation system for gastrointestinal endoscopy provided by this invention has the following advantages:

[0070] This invention provides an intelligent navigation system for gastrointestinal endoscopy. By employing multi-source data fusion positioning technology, combined with visual 3D reconstruction and inertial pose calculation, it overcomes the shortcomings of single positioning methods, achieving sub-millimeter positioning accuracy. It effectively eliminates positioning deviations caused by image interference, inertial cumulative errors, and body position shifts, significantly improving navigation stability and accuracy. Based on an improved algorithm, it achieves adaptive path planning, automatically identifying intestinal obstacles and blind spots, dynamically optimizing the treatment path, and achieving full coverage of the digestive tract without blind spots. This significantly reduces the probability of missed diagnoses and misdiagnoses, while avoiding the risk of intestinal compression damage. Employing a closed-loop intelligent control mode, it can monitor endoscopic deviations in real time and automatically adjust posture and speed, eliminating the need for continuous and precise manual control. This significantly reduces the workload of medical staff, lessens reliance on the operator's clinical experience, and improves the standardization of diagnosis and treatment. It achieves simultaneous visualization of 2D real-world scenes and 3D navigation scenarios, supports flexible human-computer interaction, and allows free switching between automatic navigation and manual control modes. Balancing intelligence and clinical operational flexibility, it is adaptable to various gastrointestinal endoscopic diagnosis and treatment scenarios, making it widely applicable.

[0071] Second Embodiment

[0072] Please refer to the following: Figure 4 Based on the intelligent navigation system for gastrointestinal endoscopy provided in the first embodiment of this application, the second embodiment of this application proposes another intelligent navigation device for gastrointestinal endoscopy. The second embodiment is merely a preferred embodiment of the first embodiment, and the implementation of the second embodiment will not affect the separate implementation of the first embodiment.

[0073] Specifically, the second embodiment of this application provides a digestive tract endoscope intelligent navigation device that is different in that it is used in a digestive tract endoscope intelligent navigation system. The digestive tract endoscope intelligent navigation device includes: an endoscope body 1, a multi-dimensional sensing component 2, a micro-drive adjustment component 3, a data transmission component 4, an external navigation terminal 5, an operating device 6, and an insertion tube 7.

[0074] The multi-dimensional sensing component 2 is integrated into the tip of the endoscope body 1 and includes a high-definition camera component, an inertial measurement sensor, and a flexible pressure sensor, used to achieve synchronous acquisition of multi-source data;

[0075] The micro-drive adjustment component 3 is disposed on the flexible section of the endoscope body 1, including a micro stepper motor and a flexible adjustment joint, and is used to receive control commands from the closed-loop control execution module to complete the adaptive adjustment of the endoscope's up-down, left-right deflection and travel speed.

[0076] The data transmission component 4 adopts a wireless low-latency transmission module to realize real-time bidirectional transmission of multi-source data and control commands between the endoscope body 1 and the external navigation terminal 5. The insertion tube 7 is connected to one side of the pipeline of the data transmission component 4, and the operating device 6 is set at one end of the insertion tube 7.

[0077] The external navigation terminal 5 has a built-in data processing chip, display screen and human-computer interaction buttons, which are used to complete data calculation, three-dimensional visualization display and navigation control operations.

[0078] The endoscope body 1 is a flexible and bendable structure with an outer diameter of no more than 5mm at the tip. The flexible section can achieve 360° all-round fine adjustment and deflection, adapting to the narrow, curved and complex physiological structure of the digestive tract.

[0079] The micro-drive adjustment component 3 has micron-level fine-tuning precision, can correct the endoscope posture in real time according to path deviation data, has a maximum deflection adjustment angle of ±90°, and can adaptively and continuously adjust the travel speed in the range of 0-5mm / s.

[0080] The external navigation terminal 5 has a built-in data storage module for real-time storage of image data, navigation path data, and pose data during the diagnosis and treatment process, supporting the retrospective analysis and review of diagnosis and treatment data.

[0081] The device employs a miniaturized multimodal sensor array at the head end, which simultaneously acquires visual, inertial, pressure, and body position information while ensuring an outer diameter of ≤5mm, effectively improving the spatiotemporal consistency of data. The micro-drive adjustment component 3 built into the flexible segment has micron-level response accuracy and a large angle deflection capability of ±90°. Combined with sub-millimeter-level pose calculation results, it can achieve highly compliant tracking of complex intestinal curvature.

[0082] The wireless low-latency data transmission mechanism of data transmission component 4 ensures real-time closed-loop interaction between control commands and sensor data, meeting the needs of dynamic clinical regulation; the external navigation terminal 5 integrates computing, display, storage and interaction, not only supporting full-element recording and retrospection of the diagnosis and treatment process, but also reducing the cognitive load of operation through a two-dimensional / three-dimensional fusion visualization interface.

[0083] The insertion tube 7 and the data transmission component 4 are connected by a pipe. The insertion tube 7 is also connected to the operating device 6, which is a control handle device and is based on existing endoscopic technology.

[0084] The working principle of the intelligent navigation device for gastrointestinal endoscopy provided by this invention is as follows:

[0085] During operation, the external navigation terminal 5 first loads a standard three-dimensional anatomical model of the human digestive tract, completes the calibration of equipment sensors and the debugging of data transmission, and establishes a unified three-dimensional world coordinate system.

[0086] After the endoscope body 1 enters the human digestive tract, the visual acquisition unit continuously acquires time-series images of the inner wall, the inertial sensing unit outputs the endoscope's position and posture parameters in real time, the pressure sensing unit monitors the intestinal contact pressure in real time, and the body position acquisition unit captures changes in the patient's body position in real time. All data are synchronously transmitted to the terminal and preprocessed by filtering and noise reduction.

[0087] The external navigation terminal 5 completes the reconstruction of local three-dimensional point cloud of the digestive tract based on time-series images, corrects model deformation by combining patient position data, and accurately registers the real-time three-dimensional model with the standard anatomical model through the iterative nearest point algorithm to construct a complete real-time three-dimensional navigation scene of the digestive tract.

[0088] The system automatically identifies obstacles such as intestinal folds, narrowings, and fluid accumulation blind spots. Combined with the preset full-area diagnosis and treatment coverage requirements, it plans the optimal travel path through improved algorithms, automatically generates deceleration and detour trajectories for curved areas, and marks key observation points for suspected lesion areas.

[0089] By integrating visual positioning and inertial positioning data, the three-dimensional coordinates, deflection angle, and travel speed of the endoscope body 1 are calculated in real time, and the real-time position deviation and attitude deviation are calculated by comparing with the planned path.

[0090] The system generates precise adjustment commands based on deviation data, driving the micro-drive adjustment component 3 to complete the fine-tuning of the endoscope body 1's posture and adaptive adjustment of its travel speed, ensuring that the endoscope body 1 travels smoothly along the optimal path; when the intestinal pressure is detected to exceed the threshold, it automatically decelerates and retracts to avoid it, protecting the intestinal mucosa.

[0091] The external navigation terminal 5 simultaneously displays two-dimensional endoscopic real-scene images and three-dimensional navigation paths, and displays the endoscopic posture, obstacle positions, and lesion markers in real time. It automatically issues audio-visual warnings when path deviation or abnormal pressure occurs. Medical staff can switch to manual control mode at any time to intervene and adjust the diagnosis and treatment operation.

[0092] During the diagnosis and treatment process, the system stores all image data, navigation path data, and pose data. After the diagnosis and treatment is completed, it can automatically generate a diagnosis and treatment report, support data retrospection and review analysis, and provide data support for subsequent diagnosis and treatment.

[0093] Compared with related technologies, the intelligent navigation device for gastrointestinal endoscopy provided by the present invention has the following beneficial effects:

[0094] This invention provides an intelligent navigation device for gastrointestinal endoscopy. By integrating multi-dimensional perception, intelligent decision-making, and precise execution into an integrated device structure, it achieves deep coupling between the navigation system and the endoscope hardware. Compared with traditional endoscope devices, it improves navigation autonomy, operational safety, and diagnostic traceability, and is especially suitable for precise exploration and intervention in traditional blind spots such as the deep colon and small intestine.

[0095] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A smart navigation system for gastrointestinal endoscopy, characterized in that, include: Multi-source sensing and acquisition module, 3D modeling and registration module, intelligent path planning module, real-time pose calculation module, visual navigation interaction module, and closed-loop control and execution module; The multi-source sensing and acquisition module is used to collect visual image data, inertial pose data, intestinal pressure data and human body posture data in real time during the endoscope's movement, and to complete the synchronous acquisition and preliminary noise reduction preprocessing of multi-source raw data. The 3D modeling and registration module is communicatively connected to the multi-source sensing and acquisition module. It is used to construct a real-time 3D scene model of the digestive tract based on the preprocessed visual image data, and to accurately register the real-time 3D scene model with the preset standard digestive tract anatomical model to achieve the fusion and matching of real-time scene and anatomical prior information. The intelligent path planning module is communicatively connected to the 3D modeling and registration module, and is used to adaptively plan the optimal endoscopic treatment path based on the registered 3D model, the location of suspected lesions in the digestive tract, intestinal physiological structure parameters, and preset obstacle avoidance rules and treatment priority strategies. The real-time pose calculation module is communicatively connected to the multi-source sensing acquisition module and the intelligent path planning module, respectively, and is used to fuse inertial pose data and visual positioning data to calculate the three-dimensional coordinates, deflection angle, travel speed and attitude parameters of the endoscope probe in real time, and generate real-time deviation data by comparing with the optimal travel path. The visualization navigation interaction module is communicatively connected to the real-time pose calculation module and the 3D modeling and registration module, and is used to render and display the 3D scene of the digestive tract, the real-time pose of the endoscope, the planned navigation path, the lesion marking points and deviation warning information in real time, and to provide a human-computer interaction control interface. The closed-loop control execution module is communicatively connected to the real-time pose calculation module, the intelligent path planning module, and the visual navigation interaction module, respectively. It is used to generate attitude adjustment and movement control commands based on real-time deviation data, drive the endoscope to complete adaptive attitude correction, speed adjustment, and path tracking, and realize closed-loop intelligent navigation control.

2. The intelligent navigation system for gastrointestinal endoscopy according to claim 1, characterized in that, The multi-source sensing and acquisition module includes a visual acquisition unit, an inertial sensing unit, a pressure sensing unit, and a body position acquisition unit. The visual acquisition unit is mounted on the inner lens end and includes a high-definition miniature camera and a supplementary light source, used to continuously acquire high-definition time-series image data of the inner wall of the digestive tract; The inertial sensing unit has a built-in high-precision inertial measurement unit (IMU) for real-time acquisition of the endoscope's three-axis acceleration, three-axis angular velocity, and attitude angle data. The pressure sensing unit includes a miniature flexible pressure sensor for collecting pressure data between the endoscope and the intestinal wall, thus avoiding rigid compression that could damage intestinal tissue. The body position acquisition unit includes a body surface posture sensing component, which is used to collect data on changes in the patient's body position during the diagnosis and treatment process, and to compensate for positioning errors caused by body position deviation.

3. The intelligent navigation system for gastrointestinal endoscopy according to claim 1, characterized in that, The 3D modeling and registration module is specifically used to: extract feature points, match frames, and reconstruct 3D from time-series image data to generate a real-time local 3D point cloud model and mesh model of the digestive tract. The iterative nearest point algorithm is used to achieve accurate registration between real-time 3D models and standard anatomical models, with matching accuracy error controlled within 0.5mm. By combining patient position data, the model is dynamically corrected to eliminate model matching deviations caused by patient positional movement and intestinal deformation, thus achieving real-time dynamic updates of the entire digestive tract scenario.

4. The intelligent navigation system for gastrointestinal endoscopy according to claim 1, characterized in that, The intelligent path planning module includes an obstacle avoidance unit and a path optimization unit; The obstacle avoidance unit is used to identify obstacles such as narrow areas of the intestine, folds and protrusions, fluid accumulation blind spots, and suspected lesion protrusions, and generates an obstacle bounding box to avoid dangerous travel areas; The path optimization unit is based on an improved algorithm, with the optimization goals of full coverage of diagnosis and treatment, shortest travel distance, and lowest risk of intestinal damage. Combined with the flexible motion constraints of the endoscope, it dynamically updates the optimal navigation path and automatically generates deceleration and fine-tuning detour paths for intestinal bends and folds.

5. The intelligent navigation system for gastrointestinal endoscopy according to claim 2, characterized in that, The pose real-time calculation module adopts a visual-inertial fusion positioning algorithm, which corrects the cumulative error of inertial measurement through visual positioning results and compensates for the positioning delay between image frames through inertial data, so as to achieve real-time update of endoscope pose at the 10ms level and achieve a positioning accuracy of sub-millimeter level.

6. The intelligent navigation system for gastrointestinal endoscopy according to claim 1, characterized in that, The visualization navigation interaction module supports the synchronous overlay display of two-dimensional endoscopic real-scene images and three-dimensional navigation scenes. It has functions such as lesion location marking, path deviation warning, and abnormal status sound and light alarm. It also supports one-click switching between manual intervention and intelligent automatic navigation mode.

7. A digestive tract endoscopy intelligent navigation device, used in the digestive tract endoscopy intelligent navigation system as described in any one of claims 1-6, characterized in that, The intelligent navigation device for gastrointestinal endoscopy includes: an endoscope body, a multi-dimensional sensing component, a micro-drive adjustment component, a data transmission component, an external navigation terminal, an operating device, and an insertion tube; The multi-dimensional sensing component is integrated into the tip of the endoscope body and includes a high-definition camera component, an inertial measurement sensor, and a flexible pressure sensor, used to achieve synchronous acquisition of multi-source data; The micro-drive adjustment component is disposed in the flexible section of the endoscope body, including a micro stepper motor and a flexible adjustment joint, and is used to receive control commands from the closed-loop control execution module to complete the adaptive adjustment of the endoscope's up-down, left-right deflection and travel speed. The data transmission component employs a wireless low-latency transmission module to enable real-time bidirectional transmission of multi-source data and control commands between the endoscope body and the external navigation terminal. The insertion tube is connected to one side of the data transmission component pipe, and the operating device is located at one end of the insertion tube. The external navigation terminal has a built-in data processing chip, display screen, and human-computer interaction buttons, which are used to complete data calculation, three-dimensional visualization display, and navigation control operations.

8. The intelligent navigation device for gastrointestinal endoscopy according to claim 7, characterized in that, The endoscope body is a flexible and bendable structure with an outer diameter of no more than 5mm at the tip. The flexible section can achieve 360° all-round fine adjustment and deflection, adapting to the narrow, curved and complex physiological structure of the digestive tract.

9. The intelligent navigation device for gastrointestinal endoscopy according to claim 7, characterized in that, The micro-drive adjustment component has micron-level fine-tuning precision, can correct the endoscope posture in real time according to path deviation data, has a maximum deflection adjustment angle of ±90°, and can adaptively and continuously adjust the travel speed in the range of 0-5mm / s.

10. The intelligent navigation device for gastrointestinal endoscopy according to claim 7, characterized in that, The external navigation terminal has a built-in data storage module for real-time storage of image data, navigation path data, and pose data during the diagnosis and treatment process, supporting the retrospective analysis and review of diagnosis and treatment data.

Images