Intra-abdominal flexible tissue respiratory real-time image correction method and device

Abstract

The application discloses an intra-abdominal flexible tissue respiratory real-time image correction method and device, which can correct respiration in a three-dimensional space through a two-dimensional ultrasonic image in surgery, dynamically corrects a preoperative plan according to respiration, and corrects a static interventional path of the preoperative plan into a planned path at a current respiration node, so that a doctor does not need to perform additional operations in surgery, and the application has very high practicability. The method comprises the following steps: (1) initializing image parameters, performing pose registration and target selection; (2) preoperative motion analysis, acquiring a spatial position of the target; (3) extracting a respiration signal rule: tracking a position of the target in a three-dimensional space in real time, synchronously tracking a spatial position of a body surface sensor, and fitting a mapping relationship between the two signals; and (4) respiration motion correction: predicting a spatial position of the target under the influence of respiration by using the spatial position of the body surface sensor at a current time in surgery, and correcting a static interventional path planned in advance at each time.

Description

Technical Field

[0001] This invention relates to the technical field of surgical navigation, and more particularly to a method for real-time image correction of respiration in flexible tissues within the abdominal cavity, and a device for real-time image correction of respiration in flexible tissues within the abdominal cavity. Background Technology

[0002] With the rapid development of minimally invasive intervention and precision medicine, image-based surgical navigation systems have gradually become the mainstream in interventional surgery. They mainly use various medical image information (US, CT, MRI and PET multimodal data) to navigate surgical instruments, providing doctors with as much information as possible about the vicinity of the surgical site, building a bridge between two-dimensional medical images and the actual site, thereby reducing surgical trauma, shortening operation time and improving surgical quality.

[0003] Ultrasound imaging involves emitting ultrasound waves into the body through an ultrasound probe, scanning linearly or in a fan-shaped pattern. When encountering tissues with different acoustic impedances, reflected sound waves of varying amplitudes are generated. These reflected waves are received by the probe, processed, and displayed on a screen. Due to its advantages such as low cost, real-time operation, non-invasiveness, and no radiation, ultrasound is often the preferred method for observing intra-abdominal organs. Surgical navigation systems applied to the abdominal cavity also frequently utilize ultrasound image-assisted guidance.

[0004] Because respiratory movements significantly affect the abdominal cavity, they can cause a deviation between the patient's spatial position during surgery and the preoperative planned position, impacting the accuracy of surgical navigation. During respiration, lung contraction causes the diaphragm to move the liver and surrounding areas in a quasi-periodic motion, making it difficult for surgeons to execute static preoperative plans on moving targets during interventional procedures. Currently, ultrasound-guided surgical navigation systems lack a systematic solution for respiratory motion, although some researchers have attempted to address this. In 2019, Huang applied a two-dimensional ultrasound-based target tracking method to a motion monitoring system in abdominal radiotherapy. Principal component analysis and slow feature extraction were used to extract image features, and the K-nearest neighbor algorithm was used to find index frames to estimate the target position. However, the lack of three-dimensional information meant that respiratory motion could only be corrected in a plane. Since targets in the two-dimensional ultrasound plane are not fixed to the same cross-section during surgery, relying solely on two-dimensional image information is insufficient for clinical correction of intraoperative respiratory motion. Therefore, real-time three-dimensional correction during surgery is needed to meet the needs of surgeons in ultrasound-guided surgical navigation. Summary of the Invention

[0005] To overcome the shortcomings of existing technologies, the technical problem to be solved by this invention is to provide a real-time image correction method for breathing in flexible intraperitoneal tissue. This method can perform breathing correction in three-dimensional space through intraoperative two-dimensional ultrasound imaging, and dynamically correct the preoperative plan with breathing movements, so that the preoperative static intervention path is corrected to the planned path under the current breathing node. No additional operation by the doctor is required during the operation, which has very high practicality.

[0006] The technical solution of this invention is: a real-time image correction method for respiration in flexible intraperitoneal tissue, comprising the following steps:

[0007] (1) Initialize image parameters, perform pose registration and target selection;

[0008] (2) Preoperative motion analysis to obtain the spatial location of the target;

[0009] (3) Extracting respiratory signal patterns: Real-time tracking of the target's position in three-dimensional space, synchronous tracking of the spatial position of the surface sensor, and fitting the mapping relationship between the two signals;

[0010] (4) Respiratory movement correction: During the operation, the spatial position of the target under the influence of breathing is predicted by the spatial position of the surface sensor at the current moment, and the pre-planned static intervention path is corrected at each moment.

[0011] This invention tracks the target position in a two-dimensional ultrasound plane in real time during surgery, matches the local image with the preoperative three-dimensional image to obtain the angle and position of the ultrasound plane in three-dimensional space, and applies the corrected three-dimensional displacement vector to the preoperatively planned treatment plan. The preoperative plan is dynamically corrected with respiratory movements, so that the preoperatively planned static interventional path is corrected to the planned path under the current respiratory node. No additional operation by the doctor is required during the operation, which has a very high degree of practicality.

[0012] It also provides a real-time image correction device for respiration in flexible tissues within the abdominal cavity, which includes:

[0013] The initialization module is configured to initialize image parameters, perform pose registration, and select targets.

[0014] The analysis module is configured for preoperative motion analysis to obtain the spatial location of the target.

[0015] The extraction module is configured to extract respiratory signal patterns: it tracks the target's position in three-dimensional space in real time, simultaneously tracks the spatial position of the surface sensor, and fits the mapping relationship between the two signals.

[0016] The correction module is configured for respiratory motion correction: during the operation, the spatial position of the target under the influence of breathing is predicted by the spatial position of the surface sensors at the current moment, and the pre-planned static intervention path is corrected at each moment. Attached Figure Description

[0017] Figure 1 This is a flowchart of the method for real-time image correction of respiration in flexible tissues within the abdominal cavity according to the present invention.

[0018] Figure 2This is a schematic diagram of obtaining the target spatial location provided by the present invention. Detailed Implementation

[0019] like Figure 1 As shown, this method for real-time image correction of respiration in flexible intraperitoneal tissue includes the following steps:

[0020] (1) Initialize image parameters, perform pose registration and target selection;

[0021] (2) Preoperative motion analysis to obtain the spatial location of the target;

[0022] (3) Extracting respiratory signal patterns: Real-time tracking of the target's position in three-dimensional space, synchronous tracking of the spatial position of the surface sensor, and fitting the mapping relationship between the two signals;

[0023] (4) Respiratory movement correction: During the operation, the spatial position of the target under the influence of breathing is predicted by the spatial position of the surface sensor at the current moment, and the pre-planned static intervention path is corrected at each moment.

[0024] This invention tracks the target position in a two-dimensional ultrasound plane in real time during surgery, matches the local image with the preoperative three-dimensional image to obtain the angle and position of the ultrasound plane in three-dimensional space, and applies the corrected three-dimensional displacement vector to the preoperatively planned treatment plan. The preoperative plan is dynamically corrected with respiratory movements, so that the preoperatively planned static interventional path is corrected to the planned path under the current respiratory node. No additional operation by the doctor is required during the operation, which has a very high degree of practicality.

[0025] Preferably, in step (1), the spatial position of the surface sensor is tracked in the navigation system and its pose is registered with the preoperative image. The tracking target is selected in the two-dimensional ultrasound image. When the respiratory node of the target is closest to that of the preoperative image, the spatial position P0 of the target is recorded, and the position of the ultrasound plane (θ0) is recorded. r0), where θ0 is the angle between the ultrasonic plane and the y=0 plane. θ is the angle between the ultrasonic plane and the z=0 plane, and r0 is the intercept of the ultrasonic plane with the z-axis.

[0026] Preferably, in step (2), in the preoperative image, the target block B is extracted with P0 as the center, and spatial sampling is performed on B to obtain several two-dimensional image slices; when the patient breathes freely, tracking is performed on the two-dimensional ultrasound plane to obtain a local two-dimensional ultrasound image U centered on the target. t The spatial location P of the target t , will U t Match with several two-dimensional slices to find the most similar slice whose planar position (θ) t , rt ) for U t At the current position, calculate P. t In (θ0, The projection points of the r0 plane are used to obtain the true spatial position of the current target.

[0027] Furthermore, in step (2), B is sampled in spherical coordinate space with P0 as the center, resulting in N two-dimensional cross-sections (θ). i , r i ), where i = 1, 2, ..., N; at time t, the spatial position of the target in two-dimensional ultrasound is P. t , with P t Extract a local image containing the complete target from the center U t , will U t With each two-dimensional cut surface (θ) i , r i Perform a similarity metric based on local structural gradients to find the optimal matching cross section (θ). t , r t At this time, P t For the tangent (θ) t , r t Points on the plane (θ0, ...) are used to establish the plane (θ0, θ0). r0) and plane (θ) t , r t Transformation relation T t , will T t Acting on P t The true spatial location P of the current target can then be obtained. t→0 .

[0028] Preferably, in step (3), a polynomial fitting method is used to establish the mapping relationship between the position of the body surface sensor and the target position.

[0029] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium. When executed, the program includes the steps of the methods of the above embodiments. The storage medium can be ROM / RAM, magnetic disk, optical disk, memory card, etc. Therefore, corresponding to the method of the present invention, the present invention also includes a real-time image correction device for respiration in flexible intraperitoneal tissue. This device is typically represented in the form of functional modules corresponding to the steps of the method. Figure 2 As shown, the device includes:

[0030] The initialization module is configured to initialize image parameters, perform pose registration, and select targets.

[0031] The analysis module is configured for preoperative motion analysis to obtain the spatial location of the target.

[0032] The extraction module is configured to extract respiratory signal patterns: it tracks the target's position in three-dimensional space in real time, simultaneously tracks the spatial position of the surface sensor, and fits the mapping relationship between the two signals.

[0033] The correction module is configured for respiratory motion correction: during the operation, the spatial position of the target under the influence of breathing is predicted by the spatial position of the surface sensors at the current moment, and the pre-planned static intervention path is corrected at each moment.

[0034] Preferably, the initialization module tracks the spatial position of the surface sensor in the navigation system, performs pose registration with the preoperative image, selects the tracking target in the two-dimensional ultrasound image, and records the target spatial position P0 when the respiratory node of the target is closest to that in the preoperative image, and records the ultrasound plane position (θ0). r0).

[0035] Preferably, the analysis module, in the preoperative image, extracts a target block B centered on P0, and performs spatial sampling on B to obtain several two-dimensional image slices; when the patient breathes freely, it tracks the target on a two-dimensional ultrasound plane to obtain a local two-dimensional ultrasound image U centered on the target. t The spatial location P of the target t , will U t Match with several two-dimensional slices to find the most similar slice whose planar position (θ) t , r t ) for U t At the current position, calculate P. t In (θ0, The projection points of the r0 plane are used to obtain the true spatial position of the current target.

[0036] Furthermore, the analysis module performs spherical coordinate space sampling on B with P0 as the center, obtaining N two-dimensional cross-sections (θ). i , r i ), where i = 1, 2, ..., N; at time t, the spatial position of the target in two-dimensional ultrasound is P. t , with P t Extract a local image containing the complete target from the center U t , will U t With each two-dimensional cut surface (θ) i , r iPerform a similarity metric based on local structural gradients to find the optimal matching cross section (θ). t , r t At this time, P t For the tangent (θ) t , r t Points on the plane (θ0, ...) are used to establish the plane (θ0, θ0). r0) and plane (θ) t , r t Transformation relation T t , will T t Acting on P t The true spatial location P of the current target can then be obtained. t→0 .

[0037] Preferably, the extraction module uses a polynomial fitting method to establish a mapping relationship between the body surface sensor position and the target position.

[0038] Compared with existing breathing correction methods, the advantages of this invention are:

[0039] 1. Only two-dimensional ultrasound is needed before surgery to obtain the three-dimensional respiratory motion trajectory of the target, which can realize the correction of respiratory motion in space.

[0040] 2. It overcomes the reliance on intraoperative guiding images and establishes a mapping relationship between surface sensors and respiratory motion. The spatial position of the target under the influence of respiration can be obtained solely by relying on the position information of the surface sensors.

[0041] 3. It can correct the effects of intra-abdominal respiratory movements in real time, presenting doctors with the best preoperative planning scheme, meeting real-time requirements, and has low requirements for computing resource allocation, making it easy to promote.

[0042] 4. Respiratory movement correction can be completed automatically without any external intervention during the operation, without increasing the difficulty of the operation or changing the surgeon's operating habits, thus having higher applicability and practicality.

[0043] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention shall still fall within the protection scope of the present invention.

Claims

1. An apparatus for rectifying a flexible tissue respiration real-time image in an abdominal cavity, characterized in that: It includes: An initialization module configured to initialize image parameters, perform pose registration and target selection; An analysis module configured to perform preoperative motion analysis and obtain the spatial position of the target; An extraction module configured to extract the breathing signal rule: real-time tracking of the position of the target in three-dimensional space, synchronous tracking of the spatial position of the body surface sensor, and fitting of the mapping relationship between the two signals; a correction module configured to correct respiratory motion: predicting the spatial position of the target under the influence of respiration at the current time using the spatial position of the body surface sensor at the current time, correcting the static interventional path planned preoperatively at each time; the initialization module tracking the spatial position of the body surface sensor in the navigation system, performing pose registration with the preoperative image, selecting the tracking target in the two-dimensional ultrasound image, and recording the spatial position of the target when the respiratory node at which the target is located is closest to the preoperative image , recording the position of the ultrasound plane ; The analysis module, in the preoperative image, takes a target block centered at ;​​​​​​​​​​​​​​​​​​​​​ The extraction module uses a polynomial fitting method to establish the mapping relationship between the position of the body surface sensor and the position of the target.

Images