An inertial navigation gastric tube intubation system and method

Abstract

The present application relates to the field of medical devices, in particular to a kind of inertial navigation gastric tube intubation system and method.The system includes mainframe workstation, miniature camera guiding operation tube and conventional gastric tube.Inertial navigation positioning module of mainframe workstation is matched with navigation track by real-time analysis operation tube attitude data and sensor probe;Data processing module receives contrast image generation three-dimensional dissection path model, and is converted into navigation reference track;Display screen dynamic display dissection path and real-time navigation track superimposed view and cavity image.Minisize camera guiding operation tube integrates navigation sensor, minisize camera and cold light illumination unit, and is worn in the inner cavity of conventional gastric tube.The present application solves the problem of traditional gastric tube intubation blind insertion, improves success rate and safety, shortens operation time, and reduces the requirement for patient cooperation.

Description

Technical Field

[0001] This invention relates to the field of medical devices, specifically to an inertial navigation gastric tube insertion system and method. Background Technology

[0002] A nasogastric tube is a commonly used medical device, required in current gastrointestinal and esophageal surgeries. Inserted through the nose, it passes through the oropharynx, esophagus, and reaches the stomach, facilitating the cleaning or aspiration of gastric fluids for gastrointestinal decompression and nutrient infusion. However, current nasogastric tube technology lacks visualization during insertion, anatomical guidance, and is hampered by esophageal strictures or obstructions caused by esophageal lesions. This often results in blind insertion by physicians relying on clinical experience, frequently leading to failure to insert the tube, mucosal damage and bleeding at the insertion site, and mechanical damage to surrounding tissues. Insertion may fail or result in unsatisfactory placement or depth. Because the insertion and anatomical location cannot be directly observed, the tube often kinks and clumps within the mouth and esophagus, leading to insertion failure. Common complications include airway obstruction, aspiration, nasopharyngeal mucosal damage and bleeding, infection, and airway compression. The success rate of blind nasogastric tube placement is even lower in critically ill patients, children, some elderly patients, and certain special populations who cannot cooperate with clinical procedures. Therefore, a visual nasogastric tube placement guidance device is needed to address these shortcomings. Summary of the Invention

[0003] The purpose of this invention is to provide an inertial navigation gastric tube insertion system and method to solve the problems of insertion failure, poor positioning, injury and bleeding caused by blind insertion based on clinical experience in the prior art.

[0004] To achieve the above objectives, the following technical solution is adopted.

[0005] An inertial navigation gastric tube insertion system, comprising: Main unit workbench, miniature camera-guided operating tube, and standard gastric tube; The host workbench includes an inertial navigation and positioning module, a data processing module, and a display screen, wherein... The inertial navigation and positioning module is used to analyze motion attitude data in real time. Based on the real-time attitude data collected by the navigation sensor probe, it calculates the current spatial coordinates of the operating tube and performs a matching degree analysis with the navigation reference trajectory. The data processing module is used to receive the patient's upper gastrointestinal contrast imaging data and generate a three-dimensional anatomical path model. At the same time, it receives the navigation data of the miniature camera-guided operation tube in real time and converts the three-dimensional anatomical path model into a navigation reference trajectory. The display screen dynamically shows an overlay view of the three-dimensional anatomical path model and the real-time navigation trajectory, and simultaneously displays real-time cavity images captured by the miniature digital camera and overlaid navigation deviation indicator marks; The miniature camera-guided operation tube includes a flexible tube body. The front end of the flexible tube body integrates a navigation sensor probe, a miniature digital camera, and a cold light illumination unit, while the rear end is connected to the host workbench via a data transmission line. The navigation sensor probe is communicatively connected to an inertial navigation and positioning module to collect real-time motion attitude data of the flexible tube body. The miniature digital camera is set along the axial direction of the flexible tube body, and its shooting angle covers the forward direction of the flexible tube body. The miniature camera-guided operating tube is detachably inserted into the working channel of the conventional gastric tube.

[0006] Optionally, the navigation sensor probe includes: A triaxial MEMS accelerometer is fixed to the front end of the flexible tube and is used to detect the linear acceleration of the tube in three-dimensional space. A three-axis MEMS gyroscope is coaxially mounted with the three-axis MEMS accelerometer to collect angular velocity data of the flexible tube about three orthogonal axes. The signal conditioning circuit amplifies and filters the analog signals output by the three-axis MEMS accelerometer and the three-axis MEMS gyroscope. The SPI communication interface transmits the processed digital signals to the inertial navigation and positioning module.

[0007] Optionally, the inertial navigation and positioning module includes a navigation and positioning algorithm unit, which is implemented using an extended Kalman filter and specifically includes: The motion modeling subunit is used to establish a kinematic model of the tube based on the angular velocity and acceleration data output by the navigation sensor probe; The data fusion subunit is used to fuse the output of the tube kinematic model with the visual feature point data collected by the miniature digital camera to calculate the attitude correction parameters. The trajectory prediction subunit is used to build an autoregressive model based on historical motion data using attitude correction parameters to predict the trajectory of the tube within a preset time period.

[0008] Optionally, the data processing module includes: The image receiving unit is used to receive DICOM format upper gastrointestinal contrast CT image sequences, the image sequences containing axial continuous slice data from the esophagus to the stomach; A three-dimensional reconstruction unit is used to perform three-dimensional voxel reconstruction on the CT image sequence based on the improved Marching Cubes algorithm to generate an anatomical model of the esophageal-cardia-gastric junction. A path planning unit is used to calculate the optimal cannulation path in the anatomical model using an improved A algorithm. The data fusion unit is used to receive six-degree-of-freedom pose data output by the inertial navigation and positioning module in real time; and to simultaneously acquire the coordinates of cavity feature points transmitted by the miniature digital camera. Inertial navigation data and visual odometry data are spatiotemporally aligned and fused using an extended Kalman filter to generate a calibrated three-dimensional spatial coordinate sequence. The navigation reference generation unit converts the three-dimensional spatial coordinate sequence into a navigation reference trajectory and optimizes the trajectory continuity through B-spline curve interpolation. The error correction unit constructs an error prediction model based on a historical operation database. The error prediction model takes into account tube contact pressure, blood oxygen saturation change rate, and real-time trajectory offset as inputs and outputs compensation and correction parameters. The optimal intubation path output by the path planning unit and the navigation reference trajectory output by the navigation reference generation unit form a closed-loop iterative optimization mechanism.

[0009] Optionally, the inner wall of the working channel is provided with a guide groove that slides into contact with the protrusion on the outer wall of the miniature camera guiding operation tube. The guide groove has a T-shaped cross-section and includes: The main direction extends along the working channel axis; Limiting flanges are symmetrically arranged on both sides of the main direction part.

[0010] Optionally, a security monitoring module may also be included, the security monitoring module comprising: The pressure sensing unit, located at the front end of the working channel, is used to detect the contact pressure between the conventional gastric tube and the tissue; The photoplethysmography unit analyzes changes in blood color in mucosal images captured by a miniature digital camera to calculate local blood oxygen saturation in real time. When the contact pressure exceeds the preset threshold or the rate of decrease in blood oxygen saturation is greater than 5% / second, an audible and visual alarm is triggered and navigation command output is paused.

[0011] Optionally, the cold light illumination unit employs a multispectral LED array, including: The main illumination group emits a blue cold light source with a wavelength of 400-450nm; Auxiliary lighting group, emitting green light source with wavelength of 520-550nm; The blue cold light source and the green light source flash alternately with a 1:3 duty cycle, and the miniature digital camera synchronously switches the corresponding filter mode to enhance the contrast of blood vessels.

[0012] A medical inertial navigation-guided visual intubation method includes the following steps: S1. Obtain the patient's upper gastrointestinal contrast imaging data, extract the anatomical structure from the esophagus to the stomach based on the image segmentation algorithm, and generate a three-dimensional anatomical path model containing a spatial coordinate sequence; S2. Plan the optimal cannulation path in the anatomical structure model and convert the path data into a navigation reference trajectory including a sequence of position coordinates; S3. Insert the miniature camera-guided operating tube and the conventional gastric tube through the mouth or nose, and activate the inertial navigation and positioning module to collect the motion posture data of the flexible tube in real time; S4. The data processing module matches the real-time motion attitude data with the navigation reference trajectory and calculates the pose deviation vector; S5. Generate navigation correction instructions based on the pose deviation vector, the instructions including suggested values ​​for steering angle and propulsion speed; S6. A miniature digital camera captures images of the cavity in real time, and combined with multispectral illumination from a cold light illumination unit, identifies the distribution characteristics of mucosal blood vessels; S7. When the tip of the conventional gastric tube is detected to have reached the cardia, the locking mechanism between the miniature camera-guided operating tube and the conventional gastric tube is released, and the miniature camera-guided operating tube is withdrawn separately.

[0013] Optionally, the method for calculating the pose deviation vector in step S4 includes: S41. Establish the transformation matrix for the tube's motion coordinate system:

[0014] Where φ, θ, and ψ are the roll angle, pitch angle, and yaw angle, respectively; S42. Calculate the Euclidean distance Δd between the current coordinate point P(x,y,z) and the nearest point P'(x',y',z') on the navigation reference trajectory; S43. Calculate the angle deviation Δα between the trajectory tangent direction and the pipe axis based on the directional derivative; S44. Combine Δd and Δα to generate a two-dimensional pose deviation vector [Δd, Δα].

[0015] Optionally, the mucosal vessel recognition in step S6 includes: S61. Extract the G channel component from the image under blue light source to obtain the initial contour of the blood vessels; S62. Perform RB channel differential operation on the image under green light source to enhance the contrast of veins; S63. Morphological closing operations are used to perform connectivity restoration on vascular images; S64. Based on convolutional neural networks, feature matching is performed on blood vessel bifurcation points to establish a local blood vessel topology map; S65. When an abnormal vascular distribution pattern is detected, the high-risk area is automatically marked and the cannulation operation is paused.

[0016] Compared with the prior art, the present invention has the following beneficial effects: The inertial navigation gastric tube insertion system and method of the present invention effectively solves many problems existing in the gastric tube insertion technology through innovative system design and method steps, and significantly improves the safety, accuracy and efficiency of insertion.

[0017] First, the system achieves visualization and precise navigation of the gastric tube insertion process through the organic integration of the main workbench, the miniature camera-guided operating tube, and the conventional gastric tube. The inertial navigation and positioning module of the main workbench can analyze motion posture data in real time and perform matching degree analysis with the navigation reference trajectory, thereby providing accurate spatial positioning and path guidance for the insertion operation. The data processing module further generates a three-dimensional anatomical path model by receiving the patient's upper gastrointestinal contrast imaging data and converts it into a navigation reference trajectory, making the insertion path more scientific and reasonable. The display screen dynamically displays an overlay view of the three-dimensional anatomical path model and the real-time navigation trajectory, as well as real-time cavity images captured by the miniature digital camera and navigation deviation indicator marks, providing the operator with an intuitive operation reference. This effectively solves the problems of insertion failure, mucosal bleeding, and tissue damage caused by blind insertion based on clinical experience in existing technologies, improving the success rate and safety of insertion, shortening operation time, and reducing the requirements for patient cooperation, making it particularly suitable for critically ill patients, elderly patients, and patients in special populations.

[0018] Secondly, the system's functions and technical effects were further refined and improved. The navigation sensor probes, through the coordinated operation of a three-axis MEMS accelerometer and a three-axis MEMS gyroscope, achieved high-precision acquisition of the flexible tube's motion attitude data. Signal processing and transmission were performed via signal conditioning circuitry and an SPI communication interface, providing an accurate and reliable data source for the inertial navigation and positioning module, further improving navigation accuracy and reliability. The inertial navigation and positioning module employs a navigation and positioning algorithm unit implemented with an extended Kalman filter. Through the coordinated operation of sub-units such as motion modeling, data fusion, and trajectory prediction, it can effectively correct attitude errors and predict future motion trajectories, further enhancing navigation accuracy and stability. The data processing module not only achieved efficient reception and 3D reconstruction of upper gastrointestinal contrast CT image sequences, but also optimized the generation and calibration of the navigation reference trajectory through improved A algorithm for path planning, and spatiotemporal alignment fusion and error correction mechanisms implemented with an extended Kalman filter, improving the scientific rigor and safety of the intubation path. The sliding fit design between the guide groove on the inner wall of the working channel and the protrusion on the outer wall of the miniature camera-guided operating tube effectively improves the stability and guiding accuracy of the operating tube within a conventional gastric tube. The safety monitoring module, through pressure sensing and photoplethysmography units, monitors the contact pressure between the conventional gastric tube and tissue, as well as local blood oxygen saturation, in real time. It can promptly trigger audible and visual alarms and pause navigation command output when abnormalities are detected, further ensuring the safety of the intubation procedure. The cold light illumination unit employs a multispectral LED array design. Through alternating flashing of blue and green light sources and switching of filter modes by the miniature digital camera, it effectively enhances vascular contrast, providing better visual conditions for identifying mucosal vascular distribution characteristics. The medical inertial navigation visualization intubation method, through a series of scientifically sound steps, achieves full-process visualization and precise navigation from acquiring upper gastrointestinal contrast imaging data to completing the intubation procedure, further improving the standardization and safety of the intubation operation.

[0019] In summary, the inertial navigation gastric tube insertion system and method of the present invention, through innovative technical solutions and system design, effectively solves many problems existing in the gastric tube insertion technology, significantly improves the safety, accuracy and efficiency of insertion, and has broad application prospects and important clinical significance. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of a module according to an embodiment of an inertial navigation gastric tube insertion system based on the present invention.

[0021] Figure 2 This is a schematic diagram of an embodiment of an inertial navigation gastric tube insertion system according to the present invention.

[0022] Figure 3This is a schematic flowchart illustrating the steps of an embodiment of an inertial navigation gastric tube insertion method according to the present invention.

[0023] The components include: 1. Main unit workbench; 2. Miniature camera-guided operation tube; 3. Standard gastric tube; 4. Display screen; 5. Data transmission cable. Detailed Implementation

[0024] The present invention will now be described in detail with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other.

[0025] The following detailed description is exemplary and intended to provide further detailed explanation of the invention. Unless otherwise specified, all technical terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains. The terminology used in this invention is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention.

[0026] Example 1 like Figure 1 and Figure 2 As shown, the inertial navigation gastric tube insertion system of the present invention is an innovative medical device designed to solve the problems of blindness, operational complexity, and potential damage to patients in traditional gastric tube insertion processes through visualization and precise navigation technology. The system mainly consists of three parts: a main unit workbench 1, a miniature camera-guided operating tube 2, and a conventional gastric tube 3. These parts work together to achieve precise guidance for gastric tube insertion.

[0027] The main workbench 1 is the system's control center, comprising an inertial navigation and positioning module, a data processing module, and a display screen 4. The inertial navigation and positioning module analyzes the motion attitude in real time using attitude data collected by navigation sensor probes, calculates the current spatial coordinates of the operating tube, and performs a matching degree analysis with the navigation reference trajectory. The core of this module lies in its navigation and positioning algorithm unit, implemented using an extended Kalman filter, which effectively integrates multiple data sources to improve navigation accuracy. Specifically, the motion modeling subunit establishes a tube kinematic model based on the angular velocity and acceleration data output by the navigation sensor probes. The data fusion subunit fuses the output of the tube kinematic model with visual feature point data collected by a miniature digital camera to calculate attitude correction parameters. The trajectory prediction subunit establishes an autoregressive model based on the attitude correction parameters and historical motion data to predict the tube's trajectory within a preset timeframe, thereby achieving real-time guidance and correction for the insertion operation.

[0028] The data processing module is responsible for receiving and processing the patient's upper gastrointestinal contrast imaging data, generating a three-dimensional anatomical path model, and converting it into a navigation reference trajectory. This module receives DICOM format CT image sequences through the image receiving unit, performs three-dimensional reconstruction using an improved Marching Cubes algorithm, and calculates the optimal intubation path using an improved A algorithm. Specifically, the image receiving unit receives CT image sequences containing axial continuous slice data from the esophagus to the stomach, and the three-dimensional reconstruction unit generates an anatomical model of the esophageal-cardia-stomach junction based on this data. The path planning unit calculates the optimal intubation path within the anatomical model, and the data fusion unit receives six-DOF pose data output from the inertial navigation positioning module in real time and simultaneously acquires the coordinates of cavity feature points transmitted by a miniature digital camera. An extended Kalman filter is used to spatiotemporally align and fuse the inertial navigation data with visual odometry data to generate a calibrated three-dimensional spatial coordinate sequence. The navigation reference generation unit converts the three-dimensional spatial coordinate sequence into a navigation reference trajectory and optimizes trajectory continuity using B-spline curve interpolation. The error correction unit constructs an error prediction model based on a historical operation database. Inputs include tube contact pressure, blood oxygen saturation change rate, and real-time trajectory deviation; outputs compensation and correction parameters. The optimal cannulation path output by the path planning unit and the navigation reference trajectory output by the navigation reference generation unit form a closed-loop iterative optimization mechanism, further improving navigation accuracy.

[0029] The four displays dynamically show an overlay view of the three-dimensional anatomical path model and the real-time navigation trajectory, while simultaneously displaying real-time cavity images captured by a miniature digital camera and navigation deviation indicators, providing the operator with intuitive operational references. This visualization design allows the operator to observe the intubation path and surrounding anatomical structures in real time, avoiding accidental entry into the airway or other non-target areas, significantly improving the safety and success rate of intubation.

[0030] The miniature camera-guided operating tube 2 is a key component of the system, comprising a flexible tube body, a navigation sensor probe, a miniature digital camera, and a cold light illumination unit. The navigation sensor probe integrated at the front end of the flexible tube body consists of a three-axis MEMS accelerometer and a three-axis MEMS gyroscope, capable of acquiring real-time motion attitude data of the flexible tube body in three-dimensional space. A signal conditioning circuit amplifies and filters the acquired analog signals to ensure signal stability and accuracy. The SPI communication interface is responsible for transmitting the processed digital signals to the inertial navigation and positioning module. The miniature digital camera is positioned along the axial direction of the flexible tube body, with a shooting angle covering the direction of movement of the flexible tube body, enabling real-time capture of image information within the cavity. The cold light illumination unit uses a multispectral LED array, including a main illumination group and an auxiliary illumination group. Through alternating flashing and switching of the filter modes of the miniature digital camera, it enhances the contrast of blood vessels, providing better visual conditions for the identification of mucosal vascular distribution characteristics.

[0031] In practical applications, the accuracy of the navigation sensor probe can be adjusted according to specific application scenarios to meet the varying precision requirements of intubation procedures. For example, for surgical scenarios requiring extremely high precision, higher-precision MEMS sensors can be used, and the filtering algorithm of the signal conditioning circuit can be optimized to further improve signal quality. The 3D reconstruction algorithm and path planning algorithm in the data processing module can also be optimized based on individual patient differences. For instance, for patients with esophageal stenosis or lesions, the parameters of the 3D reconstruction algorithm can be adjusted to more accurately reconstruct the anatomical structure of the lesion area, and the path planning algorithm can be optimized to avoid stenosis or lesion sites and select a safer intubation path. Furthermore, the system can integrate more sensors and monitoring functions, such as temperature and humidity sensors, to further enrich the monitoring information during intubation and improve the safety and reliability of the operation.

[0032] The conventional gastric tube 3 serves as the carrier for the intubation procedure, while the miniature camera-guided intubation tube 2 is detachably inserted into its inner working channel. The inner wall of the working channel has a guide groove that slides into contact with a protrusion on the outer wall of the miniature camera-guided intubation tube 2. The guide groove has a T-shaped cross-section, including a main guide portion extending axially along the working channel and symmetrically positioned limiting flanges on either side of the main guide portion. This design effectively improves the stability and guiding accuracy of the intubation tube within the conventional gastric tube 3, preventing deviation or rotation of the intubation tube during insertion and ensuring a smooth intubation procedure.

[0033] In addition, the system includes a safety monitoring module for real-time monitoring of safety during intubation. A pressure sensing unit, located at the front end of the working channel, detects the contact pressure between the conventional gastric tube 3 and the tissue; the photoplethysmography unit analyzes changes in blood color in mucosal images captured by a miniature digital camera to calculate local blood oxygen saturation in real time. When the contact pressure exceeds a preset threshold or the rate of decrease in blood oxygen saturation exceeds 5% / second, the system triggers an audible and visual alarm and pauses navigation command output, effectively preventing injury to the patient. This safety monitoring mechanism provides additional safety assurance for intubation procedures, especially suitable for critically ill patients and elderly patients, who have high safety requirements.

[0034] In practical applications, the parameters of the safety monitoring module can be adjusted according to individual patient differences. For example, for patients with low blood oxygen saturation, the threshold for the rate of decrease in blood oxygen saturation can be appropriately lowered to trigger the alarm mechanism earlier.

[0035] In addition, the system can be combined with other physiological monitoring data, such as heart rate and blood pressure, to further improve the comprehensiveness and accuracy of safety monitoring.

[0036] Example 2 like Figure 3As shown, the medical inertial navigation visualization intubation method of the present invention is based on the above-mentioned system and achieves precise guidance for gastric tube intubation through a series of steps. The method includes the following steps: First, upper gastrointestinal contrast imaging data of the patient is acquired. An image segmentation algorithm is used to extract the anatomical structures from the esophagus to the stomach, generating a three-dimensional anatomical path model containing spatial coordinate sequences. This is completed by the data processing module of the main unit 1. The image receiving unit receives CT image sequences in DICOM format, and the three-dimensional reconstruction unit uses an improved Marching Cubes algorithm to perform three-dimensional voxel reconstruction of the image sequences, generating an anatomical structure model of the esophageal-cardia-stomach junction. The path planning unit then uses an improved A algorithm to calculate the optimal intubation path within the anatomical structure model and converts the path data into a navigation reference trajectory including positional coordinate sequences.

[0037] In practical applications, image segmentation algorithms can be optimized based on individual patient differences. For example, for patients with esophageal strictures or lesions, more refined segmentation algorithms can be used to more accurately extract the anatomical structures of the lesion area. 3D reconstruction algorithms can also be adjusted according to specific needs, such as increasing voxel resolution or optimizing reconstruction algorithm parameters to improve the accuracy and detail of the anatomical structure model. Furthermore, pathway planning algorithms can be optimized based on the patient's anatomy and clinical needs, such as considering avoiding strictures or lesions during pathway planning to select a safer intubation route.

[0038] Next, the miniature camera-guided insertion tube 2 and the conventional gastric tube 3 are inserted into the patient's body through the mouth or nose. The inertial navigation and positioning module is activated to collect real-time motion posture data of the flexible tube. The data processing module matches the real-time motion posture data with the navigation reference trajectory to calculate the pose deviation vector. During this process, the navigation and positioning algorithm unit of the inertial navigation and positioning module works collaboratively through the motion modeling subunit, data fusion subunit, and trajectory prediction subunit. Based on the angular velocity and acceleration data output by the navigation sensor probe, it establishes a kinematic model of the tube and fuses the model output with the visual feature point data collected by the miniature digital camera to calculate the posture correction parameters. The trajectory prediction subunit then establishes an autoregressive model based on the posture correction parameters and historical motion data to predict the tube's motion trajectory within a preset time period, thereby achieving real-time guidance and correction of the insertion operation.

[0039] In practical applications, navigation and positioning algorithms can be optimized according to specific needs. For example, for cannulation operations requiring higher precision, the parameters of the extended Kalman filter can be further optimized to improve the accuracy and stability of data fusion. The motion modeling subunit can be adjusted based on the physical characteristics of the flexible tube, such as considering the influence of the elastic deformation of the flexible tube on the motion posture, thereby establishing a more accurate kinematic model. The data fusion subunit can incorporate more data sources, such as pressure sensor data or blood oxygen saturation data, to further improve the accuracy and reliability of posture correction.

[0040] Based on the calculated pose deviation vector, navigation correction commands are generated, including suggested values ​​for steering angle and propulsion speed. A miniature digital camera captures images of the cavity in real time, and combined with multispectral illumination from a cold light illumination unit, identifies the distribution characteristics of mucosal vessels. The main illumination group of the cold light illumination unit emits a blue cold light source with a wavelength of 400-450nm, while the auxiliary illumination group emits a green light source with a wavelength of 520-550nm. The blue and green light sources flash alternately with a duty cycle of 1:3, and the miniature digital camera synchronously switches the corresponding filter mode, thereby enhancing vessel contrast and providing better visual conditions for identifying the distribution characteristics of mucosal vessels.

[0041] In practical applications, the spectral parameters of the cold light illumination unit can be adjusted according to different tissue characteristics. For example, for certain specific lesions, the wavelength range or duty cycle of the light source can be adjusted to better highlight the characteristics of the lesion area. The filtering mode of the miniature digital camera can also be optimized according to specific needs, such as introducing more advanced image processing algorithms to further improve blood vessel contrast and image clarity.

[0042] When the tip of the conventional gastric tube 3 is detected to have reached the cardia, the locking mechanism between the miniature camera-guided operating tube 2 and the conventional gastric tube 3 is released, and the miniature camera-guided operating tube 2 is withdrawn independently, completing the intubation operation. During this process, the system monitors the position of the gastric tube in real time through a miniature digital camera to ensure that it accurately reaches the target position. At the same time, the safety monitoring module monitors the safety status during the intubation process in real time. When the contact pressure exceeds a preset threshold or the rate of decrease in blood oxygen saturation is greater than 5% / second, the system will trigger an audible and visual alarm and suspend the output of navigation commands, thereby effectively avoiding injury to the patient.

[0043] In the calculation of the pose deviation vector, the transformation matrix of the tube motion coordinate system is first established, where φ, θ, and ψ represent the roll angle, pitch angle, and yaw angle, respectively. Then, the Euclidean distance Δd between the current coordinate point P(x,y,z) and the nearest point P'(x',y',z') on the navigation reference trajectory is calculated, and the angular deviation Δα between the trajectory tangent direction and the tube axis is calculated based on the directional derivative. Finally, Δd and Δα are combined to generate a two-dimensional pose deviation vector [Δd, Δα], providing a basis for the generation of navigation correction commands.

[0044] In practical applications, the method for calculating the pose deviation vector can be optimized according to specific needs. For example, for cannulation operations requiring higher precision, more error correction parameters can be introduced, such as the zero-bias error of the accelerometer and the drift error of the gyroscope, to further improve the calculation accuracy of the pose deviation vector. Simultaneously, the calculation method for the coordinate system transformation matrix can be optimized, for example, by introducing more efficient numerical calculation algorithms to improve computational efficiency and accuracy.

[0045] During mucosal vessel identification, G-channel components are extracted from images under blue light to obtain the initial vessel contour; RB-channel differential operations are performed on images under green light to enhance vein contrast; morphological closing operations are used to repair connectivity in the vessel images; and feature matching of vessel bifurcation points is performed based on convolutional neural networks to establish a local vessel topology map. When abnormal vessel distribution patterns are detected, high-risk areas are automatically marked and cannulation is paused, effectively avoiding damage to the patient.

[0046] In practical applications, mucosal vessel recognition algorithms can be optimized according to specific needs. For example, for certain special lesions, the parameters of the image processing algorithm can be adjusted to better identify the vascular features of the lesion area. The structure and training data of the convolutional neural network can also be adjusted according to specific needs, such as introducing more vascular feature data to improve the network's recognition accuracy and robustness.

[0047] As is known from common technical knowledge, this invention can be implemented through other embodiments that do not depart from its spirit or essential characteristics. Therefore, the disclosed embodiments described above are merely illustrative in all respects and are not the only ones. All modifications within the scope of this invention or equivalent to the scope of this invention are included in this invention.

Claims

1. An inertial navigation gastric tube insertion system, characterized in that, include: The host workbench (1), the miniature camera-guided operation tube (2), and the conventional gastric tube (3); The host workbench (1) includes an inertial navigation and positioning module, a data processing module, and a display screen (4), wherein, The inertial navigation and positioning module is used to analyze motion attitude data in real time. Based on the real-time attitude data collected by the navigation sensor probe, it calculates the current spatial coordinates of the operating tube and performs a matching degree analysis with the navigation reference trajectory. The data processing module is used to receive the patient's upper gastrointestinal contrast imaging data and generate a three-dimensional anatomical path model. At the same time, it receives the navigation data of the miniature camera-guided operation tube (2) in real time and converts the three-dimensional anatomical path model into a navigation reference trajectory. The display screen (4) dynamically displays the superimposed view of the three-dimensional anatomical path model and the real-time navigation trajectory, and synchronously displays the real-time cavity images captured by the miniature digital camera and the superimposed navigation deviation indicator marks; The miniature camera-guided operation tube (2) includes a flexible tube body. The front end of the flexible tube body integrates a navigation sensor probe, a miniature digital camera, and a cold light illumination unit. The rear end is connected to the host workbench (1) via a data transmission line (5). The navigation sensor probe is connected to the inertial navigation positioning module for real-time acquisition of the motion attitude data of the flexible tube body. The miniature digital camera is set along the axial direction of the flexible tube body, and the shooting angle covers the forward direction of the flexible tube body. The miniature camera-guided operating tube (2) is detachably inserted into the working channel of the conventional gastric tube (3).

2. The inertial navigation gastric tube insertion system according to claim 1, characterized in that, The navigation sensor probe includes: A triaxial MEMS accelerometer is fixed to the front end of the flexible tube and is used to detect the linear acceleration of the tube in three-dimensional space. A three-axis MEMS gyroscope is coaxially mounted with the three-axis MEMS accelerometer to collect angular velocity data of the flexible tube about three orthogonal axes. The signal conditioning circuit amplifies and filters the analog signals output by the three-axis MEMS accelerometer and the three-axis MEMS gyroscope. The SPI communication interface transmits the processed digital signals to the inertial navigation and positioning module.

3. The inertial navigation gastric tube insertion system according to claim 1, characterized in that, The inertial navigation and positioning module includes a navigation and positioning algorithm unit, which is implemented using an extended Kalman filter and specifically includes: The motion modeling subunit is used to establish a kinematic model of the tube based on the angular velocity and acceleration data output by the navigation sensor probe; The data fusion subunit is used to fuse the output of the tube kinematic model with the visual feature point data collected by the miniature digital camera to calculate the attitude correction parameters. The trajectory prediction subunit is used to build an autoregressive model based on historical motion data using attitude correction parameters to predict the trajectory of the tube within a preset time period.

4. The inertial navigation gastric tube insertion system according to claim 1, characterized in that, The data processing module includes: The image receiving unit is used to receive DICOM format upper gastrointestinal contrast CT image sequences, the image sequences containing axial continuous slice data from the esophagus to the stomach; A three-dimensional reconstruction unit is used to perform three-dimensional voxel reconstruction on the CT image sequence based on the improved Marching Cubes algorithm to generate an anatomical model of the esophageal-cardia-gastric junction. A path planning unit is used to calculate the optimal cannulation path in the anatomical model using an improved A algorithm. The data fusion unit is used to receive six-degree-of-freedom pose data output by the inertial navigation and positioning module in real time; and to simultaneously acquire the coordinates of cavity feature points transmitted by the miniature digital camera. Inertial navigation data and visual odometry data are spatiotemporally aligned and fused using an extended Kalman filter to generate a calibrated three-dimensional spatial coordinate sequence. The navigation reference generation unit converts the three-dimensional spatial coordinate sequence into a navigation reference trajectory and optimizes the trajectory continuity through B-spline curve interpolation. The error correction unit constructs an error prediction model based on a historical operation database. The error prediction model takes into account tube contact pressure, blood oxygen saturation change rate, and real-time trajectory offset as inputs and outputs compensation and correction parameters. The optimal intubation path output by the path planning unit and the navigation reference trajectory output by the navigation reference generation unit form a closed-loop iterative optimization mechanism.

5. The inertial navigation gastric tube insertion system according to claim 1, characterized in that, The inner wall of the working channel is provided with a guide groove that slides into contact with the outer wall protrusion of the miniature camera guide operation tube (2). The cross-section of the guide groove is T-shaped and includes: The main direction extends along the working channel axis; Limiting flanges are symmetrically arranged on both sides of the main direction part.

6. The inertial navigation gastric tube insertion system according to claim 1, characterized in that, It also includes a security monitoring module, which includes: The pressure sensing unit is located at the front end of the working channel and is used to detect the contact pressure between the conventional gastric tube (3) and the tissue; The photoplethysmography unit analyzes changes in blood color in mucosal images captured by a miniature digital camera to calculate local blood oxygen saturation in real time. When the contact pressure exceeds the preset threshold or the rate of decrease in blood oxygen saturation is greater than 5% / second, an audible and visual alarm is triggered and navigation command output is paused.

7. The inertial navigation gastric tube insertion system according to claim 1, characterized in that, The cold light illumination unit employs a multispectral LED array, including: The main illumination group emits a blue cold light source with a wavelength of 400-450nm; Auxiliary lighting group, emitting green light source with wavelength of 520-550nm; The blue cold light source and the green light source flash alternately with a 1:3 duty cycle, and the miniature digital camera synchronously switches the corresponding filter mode to enhance the contrast of blood vessels.

8. A medical inertial navigation visualization intubation method, based on the medical inertial navigation visualization intubation system according to any one of claims 1-7, characterized in that, Includes the following steps: S1. Obtain the patient's upper gastrointestinal contrast imaging data, extract the anatomical structure from the esophagus to the stomach based on the image segmentation algorithm, and generate a three-dimensional anatomical path model containing a spatial coordinate sequence; S2. Plan the optimal cannulation path in the anatomical structure model and convert the path data into a navigation reference trajectory including a sequence of position coordinates; S3. Insert the miniature camera-guided operation tube (2) and the conventional gastric tube (3) through the mouth or nose, and start the inertial navigation and positioning module to collect the motion posture data of the flexible tube in real time; S4. The data processing module matches the real-time motion attitude data with the navigation reference trajectory and calculates the pose deviation vector; S5. Generate navigation correction instructions based on the pose deviation vector, the instructions including suggested values ​​for steering angle and propulsion speed; S6. A miniature digital camera captures images of the cavity in real time, and combined with multispectral illumination from a cold light illumination unit, identifies the distribution characteristics of mucosal blood vessels; S7. When the tip of the conventional gastric tube (3) is detected to have reached the cardia, the locking mechanism between the miniature camera-guided operation tube (2) and the conventional gastric tube (3) is released, and the miniature camera-guided operation tube (2) is withdrawn separately.

9. A medical inertial navigation visualization intubation method according to claim 8, characterized in that, The method for calculating the pose deviation vector in step S4 includes: S41. Establish the transformation matrix for the tube's motion coordinate system: Where φ, θ, and ψ are the roll angle, pitch angle, and yaw angle, respectively; S42. Calculate the Euclidean distance Δd between the current coordinate point P(x,y,z) and the nearest point P'(x',y',z') on the navigation reference trajectory; S43. Calculate the angle deviation Δα between the trajectory tangent direction and the pipe axis based on the directional derivative; S44. Combine Δd and Δα to generate a two-dimensional pose deviation vector [Δd, Δα].

10. A medical inertial navigation visualization intubation method according to claim 8, characterized in that, The mucosal vessel identification in step S6 includes: S61. Extract the G channel component from the image under blue light source to obtain the initial contour of the blood vessels; S62. Perform RB channel differential operation on the image under green light source to enhance the contrast of veins; S63. Morphological closing operations are used to perform connectivity restoration on vascular images; S64. Based on convolutional neural networks, feature matching is performed on blood vessel bifurcation points to establish a local blood vessel topology map; S65. When an abnormal vascular distribution pattern is detected, the high-risk area is automatically marked and the cannulation operation is paused.

Images