Puncture intervention system and spatial registration method

Abstract

The application relates to a puncture intervention system and a spatial registration method. The system comprises: a reference marker position is invariable, a reference space is determined; a tracking marker is fixed to a tissue part of an object; a medical imaging device is used for collecting a medical image; a camera is used for collecting positions of the reference marker and the tracking marker; a processor is used for acquiring a target position of the tracking marker in the reference space and a position of the tracking marker in an image space within at least one breathing cycle, and performing registration on the two to obtain a registration relationship at a corresponding moment; a registration error corresponding to the registration relationship is determined, a breathing phase corresponding to the registration relationship is determined based on the target position of the tracking marker in the reference space at the corresponding moment; a breathing phase corresponding to a registration relationship meeting a requirement in a registration error is selected as a target breathing phase; a surgery time reminder is performed based on the target breathing phase; and a puncture mechanical arm is used for performing a puncture operation at a surgery time. The system can decouple the dependence on the camera.

Description

Technical Field

[0001] This application relates to the field of intelligent medical technology, and in particular to a puncture intervention system and a spatial registration method. Background Technology

[0002] The surgical navigation and positioning system can achieve positioning through optical tracking. The system is equipped with a matching optical tracking device, which contains a reflective ball. The optical system tracks the reflective ball to achieve real-time tracking and positioning of the target. The surgical space needs to be registered before the surgical navigation and positioning system can be used.

[0003] In traditional soft tissue puncture surgical robots, surgical space registration involves recording the patient's respiratory movements using a respiratory acquisition device. To achieve a correspondence between the image space and the surgical space, the position and orientation of respiratory markers need to be determined in both the image and camera spaces. Surgical space registration is completed when the respiratory phase at the moment of image acquisition matches the respiratory phase identified in the camera space.

[0004] However, this requires the camera to remain absolutely still throughout the entire surgical procedure, which is too demanding for a real surgical environment. Summary of the Invention

[0005] Therefore, it is necessary to provide a puncture intervention system and spatial registration method that can decouple from the dependence on tracking devices to address the above-mentioned technical problems.

[0006] In a first aspect, this application provides a puncture intervention system, the system comprising:

[0007] A reference marker is used, and the position of the reference marker remains unchanged. A reference space is obtained based on the reference marker.

[0008] A tracking marker, which is fixed to the tissue portion of the object;

[0009] Medical imaging equipment, used to acquire medical images of the tissue sites of the object;

[0010] A tracking device for acquiring the positions of the reference marker and the tracking marker;

[0011] The processor is configured to, within at least one respiratory cycle, acquire the target position of the tracking marker in a reference space based on the acquired positions of the reference marker and the tracking marker, determine the position of the tracking marker in an image space based on the medical image, and register the target position of the tracking marker in the reference space at a corresponding time with the position of the tracking marker in the image space to obtain the registration relationship at the corresponding time.

[0012] Calculate the registration error corresponding to the registration relationship at each corresponding time, and determine the respiratory phase corresponding to the registration relationship based on the target position of the tracking marker in the reference space at the corresponding time; select the respiratory phase corresponding to the registration relationship whose registration error meets the requirements as the target respiratory phase; and provide a surgical time reminder based on the target respiratory phase.

[0013] A puncture robotic arm is used to perform the puncture procedure during the surgical time.

[0014] In one embodiment, the reference marker is mounted on an operating table, on a robotic arm trolley, or on the medical imaging device.

[0015] In one embodiment, the tracking markers include at least five optical markers or magnetic navigation positioning markers, which are sequentially connected to form the target geometry, and the distance between any two of the optical markers or magnetic navigation positioning markers is not the same;

[0016] The processor is further configured to identify the geometry of the tracking marker in the reference space and the image space at corresponding times; based on the geometry, determine the correspondence between the tracking marker in the reference space and the tracking marker in the image space; and based on the correspondence, register the target position of the tracking marker in the reference space with the position of the tracking marker in the image space to obtain the registration relationship at the corresponding time.

[0017] In one embodiment, the processor is further configured to convert the target position of the tracking marker in the reference space to the image space based on the registration relationship to obtain a reference position; calculate the difference between each reference position and the position of the tracking marker in the image space; and select the maximum difference as the registration error at the corresponding time.

[0018] In one embodiment, the tracking marker has a different structure from the reference marker; the processor is further configured to:

[0019] Based on the structure of the tracking marker and the reference marker, it is determined whether the tracking device has acquired the tracking marker and the position of the reference marker;

[0020] When the tracking device acquires the positions of the tracking marker and the reference marker, it determines the target position of the tracking marker in the reference space based on the positions of the tracking marker and the reference marker, and places the target position in the historical position sequence in chronological order;

[0021] When the tracking device cannot collect the positions of the tracking marker and the reference marker, it acquires the historical position sequence corresponding to the tracking marker, the historical position sequence including the positions of the tracking marker arranged in chronological order;

[0022] Based on the positions of the tracking markers arranged in chronological order, construct the current position vector and the initial historical position vector;

[0023] Based on the vector distance between each of the initial historical position vectors and the current position vector, a preset number of target historical position vectors are selected from the initial historical position vectors;

[0024] Based on the preset number of target historical location vectors, the location of the tracking marker is predicted to obtain the target location.

[0025] In one embodiment, the processor is further configured to: obtain the vector distance between each of the initial historical position vectors and the current position vector; select initial historical position vectors whose vector distance is less than a distance threshold, and count the number of selected initial historical position vectors; when the number is less than or equal to a preset number, use the selected initial historical position vectors as target historical position vectors; when the number is greater than the preset number, determine the preset number of target historical position vectors from the selected initial historical position vectors in ascending order of vector distance.

[0026] Based on the vector distance between each target historical position vector and the current position vector, the weight of each target historical position vector is determined; the historical position of the tracking marker corresponding to each target historical position vector is obtained; and the target position of the tracking marker is obtained according to the historical position of the tracking marker corresponding to each target historical position vector and the weight.

[0027] In one embodiment, the processor is further configured to obtain a sequence length, the sequence length being at least greater than the length of a respiratory cycle; when the current length of the historical position sequence is less than the sequence length, the positions of the tracking markers in the reference space are stored in the historical position sequence in chronological order; when the current length of the historical position sequence is greater than or equal to the sequence length, the positions of the tracking markers in the reference space that are earlier in the historical position sequence are deleted, and the current positions of the tracking markers in the reference space are stored in the historical position sequence in chronological order.

[0028] In one embodiment, the processor is further configured to determine the target centroid position based on the target position of the tracking marker; obtain a fluctuation range, and when the target centroid position is within the fluctuation range, normalize the target centroid position based on the fluctuation range to obtain the current respiratory phase; when the current respiratory phase matches the target respiratory phase, use the time corresponding to the current respiratory phase as the surgical operation execution time;

[0029] The system also includes:

[0030] The display module is used to output the execution time of the surgical operation.

[0031] In one embodiment, the processor is further configured to place the target location in a historical location sequence in chronological order, and generate a current breathing curve based on the target location of each tracking marker in the historical location sequence;

[0032] The display module is also used to display the current respiratory curve, and output the time corresponding to the current respiratory phase in the current respiratory curve as a prompt for the execution time of the surgical operation.

[0033] In one embodiment, the processor is further configured to acquire system lag time, determine prediction time based on system lag time, and predict the target position of the tracking marker in the reference space corresponding to the prediction time.

[0034] Secondly, this application also provides a space registration method, the method comprising:

[0035] During at least one respiratory cycle, acquire the target location of the tracking marker in the reference space;

[0036] During at least one respiratory cycle, the location of the tracking marker in the image space is determined;

[0037] The target position of the tracking marker in the reference space at the corresponding time is registered with the position of the tracking marker in the image space to obtain the registration relationship at the corresponding time.

[0038] Calculate the registration error corresponding to the registration relationship at each corresponding time, and determine the respiratory phase corresponding to the registration relationship based on the target position of the tracking marker in the reference space at the corresponding time; select the respiratory phase corresponding to the registration relationship whose registration error meets the requirements as the target respiratory phase for spatial registration, and provide surgical time reminders based on the target respiratory phase.

[0039] The aforementioned puncture intervention system and spatial registration method, within at least one respiratory cycle, obtains the target position of the tracking marker in a reference space based on the acquired positions of the reference marker and the tracking marker, and determines the position of the tracking marker in the image space based on the medical image. The target position of the tracking marker in the reference space at a corresponding time is registered with the position of the tracking marker in the image space to obtain the registration relationship for that time. The registration error corresponding to the registration relationship at each corresponding time is calculated, and the respiratory phase corresponding to the registration relationship is determined based on the target position of the tracking marker in the reference space at the corresponding time. The respiratory phase corresponding to the registration relationship whose registration error meets the requirements is selected as the target respiratory phase. Surgical time reminders are then provided based on the target respiratory phase. In this way, the target position of the tracking marker in the reference space is obtained based on the position of the reference marker, and the reference space is used instead of the reference space of the tracking device, decoupling the dependence on the position of the tracking device and making the position of the tracking device more flexible. Attached Figure Description

[0040] To more clearly illustrate the technical solutions in the embodiments or related technologies of this application, the accompanying drawings used in the description of the embodiments or related technologies will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0041] Figure 1 This is a scene diagram of surgical space registration in an open space, as shown in one embodiment.

[0042] Figure 2 A scene diagram illustrating surgical space registration in an open space, as shown in another embodiment;

[0043] Figure 3 This is a scene diagram of surgical space registration in an open space in another embodiment;

[0044] Figure 4 This is a scene diagram of surgical space registration in a confined space in one embodiment;

[0045] Figure 5 This is a scene diagram of surgical space registration in a confined space, as shown in another embodiment.

[0046] Figure 6 This is a schematic diagram of a tracking marker in one embodiment;

[0047] Figure 7 This is a schematic diagram of the various operational stages in one embodiment;

[0048] Figure 8This is a flowchart of the various operational stages in one embodiment;

[0049] Figure 9 Here is a flowchart of a spatial registration method in one embodiment;

[0050] Figure 10 This is a flowchart illustrating a method for obtaining the location of a tracking marker in one embodiment;

[0051] Figure 11 A flowchart of the steps for generating historical location sequences of tracking markers in one embodiment;

[0052] Figure 12 This is a schematic diagram of a historical location sequence in one embodiment;

[0053] Figure 13 This is a schematic diagram of the current position vector and the initial historical position vector in one embodiment;

[0054] Figure 14 This is a schematic diagram of the registration error calculation steps in one embodiment;

[0055] Figure 15 This is a schematic diagram of the fluctuation range acquisition step in one embodiment;

[0056] Figure 16 This is a schematic diagram of the motion of the center of mass in one embodiment;

[0057] Figure 17 This is a schematic diagram of the center of mass motion in another embodiment;

[0058] Figure 18 This is a flowchart illustrating a method for determining the execution time of a surgical procedure in one embodiment;

[0059] Figure 19 This is a schematic diagram of the dynamic registration process in one embodiment. Detailed Implementation

[0060] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0061] This application provides a puncture intervention system, which includes a reference marker, a tracking marker, a medical imaging device, a tracking device, a processor, and a puncture robotic arm. The reference marker remains stationary, and a reference space can be obtained based on it. The tracking marker is fixed to the target tissue site. The medical imaging device acquires medical images of the target tissue site, and the tracking device acquires the positions of the reference marker and the tracking marker. Within at least one respiratory cycle, the processor obtains the target position of the tracking marker in the reference space based on the acquired positions of the reference marker and the tracking marker, and determines the position of the tracking marker in the image space based on the medical images. The target position of the tracking marker in the reference space at a corresponding time is registered with the position of the tracking marker in the image space to obtain the registration relationship for that time. The registration error corresponding to the registration relationship at each corresponding time is calculated, and the respiratory phase corresponding to the registration relationship is determined based on the target position of the tracking marker in the reference space at the corresponding time. The respiratory phase corresponding to the registration relationship with the required registration error is selected as the target respiratory phase. The surgical time is then reminded based on the target respiratory phase. Finally, the puncture robotic arm performs the puncture operation during the surgical time.

[0062] The reference marker can provide a reference space, thereby transforming the position of the tracking marker to the reference space and reducing the influence of the camera. For this purpose, the position of the reference marker is fixed and can be installed on the operating table, on the robotic arm trolley, or on the medical imaging equipment.

[0063] For ease of understanding, combined with Figures 1 to 5 The figures illustrate different embodiments of the puncture intervention system of this application. The tracking marker can be an optical marker or a magnetic navigation marker, and it is fixed to the target tissue site, for example, pasted to the pre-puncture site of the lesion. The reference marker can be fixed to an operating table, robotic arm trolley, or medical imaging equipment, and this reference marker can be an optical marker or a magnetic navigation marker. Only [specific embodiments are shown in this application] Figures 1 to 5 The five embodiments described above can be combined with each other by those skilled in the art, without specific limitations.

[0064] in Figure 1This is a scene diagram illustrating surgical space registration in an open space in one embodiment. In this embodiment, the subject wears tracking markers to complete medical imaging and undergoes surgery on a support device. The surgical space is relatively open. The tracking markers are attached to the subject's tissue sites, such as around the pre-puncture skin of a lesion, and fluctuate with tissue movement, such as with the fluctuations of breathing. The tracking markers include at least five optical markers that can be recognized by the camera and imaged in the medical images. A reference marker is fixed to the CT operating table of the support device, maintaining a fixed pose with respect to the subject. The reference marker includes at least three optical markers that can be recognized by the camera. A robotic arm trolley is located near the subject and remains stationary. The camera can simultaneously recognize the reference markers and tracking markers; during the normal breathing phase of the surgery, the camera's position can be moved. Specifically, the camera acquires the position of the tracking marker and the pose of the reference marker, transferring the position of the tracking marker to the reference marker space. The algorithm acquires the position of the tracking marker in the medical image. At a certain moment, geometric feature clustering and SVD registration algorithms, or geometric feature clustering and quaternion registration algorithms, are used to establish the pose relationship between the tracking marker in the reference space and the tracking marker in the medical image space. Since the tracking marker in the reference space moves continuously with the subject's breathing fluctuations, the registration between the medical image space and the bedside reference space at a certain moment may not be optimal; that is, the breathing phase during medical image imaging may differ from the breathing phase of the subject at the time of registration. Within one respiratory cycle, registration is performed at a preset frequency much higher than the respiratory rate, resulting in a registration error array. The registration matrix and respiratory amplitude at the minimum registration error in the registration error array are selected as the target registration result, completing the registration between the medical image space and the reference space.

[0065] in Figure 2 This is a scene diagram of surgical space registration in an open space in another embodiment, in which... Figure 1 The difference in the embodiment is that the reference marker is fixed on the robotic arm trolley and remains fixed in position relative to the object. The reference marker includes at least three optical markers that can be identified by the camera. Thus, when the camera acquires the position of the tracking marker, the camera acquires the pose of the reference marker, and the position of the tracking marker is transferred to the reference marker space on the robotic arm trolley, subsequent operations are... Figure 1 The embodiments are the same as those in the previous examples, and will not be repeated here.

[0066] in Figure 3 This is a scene diagram of surgical space registration in an open space in another embodiment, in which... Figure 1The difference in the embodiments is that the tracking markers are attached to the tissue site of the object, such as around the pre-puncture skin of the lesion, and fluctuate with tissue movement, such as fluctuating with respiratory movements. The tracking markers include at least five magnetic sensors for magnetic navigation positioning, which can be identified by magnetic navigation devices and can also be imaged in medical imaging. In this way, the magnetic navigation device obtains the position of the tracking markers, obtains the position of the reference markers, and transfers the position of the tracking markers to the space of the reference markers.

[0067] in Figure 4 This is a scene diagram illustrating surgical space registration in a confined space in one embodiment. In this embodiment, the subject wears tracking markers to complete medical imaging and undergoes surgery on a support device. The surgical space is confined, and the subject is within the CT aperture of the medical imaging equipment. The tracking markers are attached to the subject's tissue sites, such as around the pre-puncture skin of a lesion, and fluctuate with tissue movement, such as with the fluctuations of breathing. The tracking markers include at least five optical markers that can be recognized by the camera and imaged in the medical imaging. A reference marker is fixed to the CT operating table of the support device, maintaining a fixed pose with respect to the subject. The reference marker includes at least three optical markers that can be recognized by the camera. A robotic arm trolley is located near the subject and remains stationary. The camera can simultaneously recognize the reference markers and tracking markers; its position can be moved during the normal breathing phase of the surgery. Specifically, the camera acquires the position of the tracking marker and the pose of the reference marker, transferring the position of the tracking marker to the reference marker space. The algorithm acquires the position of the tracking marker in the medical image. At a certain moment, geometric feature clustering and SVD registration algorithms, or set feature clustering and quaternion registration algorithms, are used to establish the pose relationship between the tracking marker in the reference space and the tracking marker in the medical image space. Since the tracking marker in the reference space moves continuously with the subject's breathing fluctuations, the registration between the medical image space and the bedside reference space at a certain moment may not be optimal; that is, the breathing phase during medical image imaging may differ from the breathing phase of the subject at the time of registration. Within one respiratory cycle, registration is performed at a preset frequency much higher than the respiratory rate. Simultaneously, due to camera field-of-view occlusion issues in the confined space, a respiratory motion model is used to predict the tracking marker data in the reference space for registration, resulting in a registration error array. The registration matrix and respiratory amplitude at the minimum registration error in the array are selected as the target registration result, completing the registration between the medical image space and the reference space.

[0068] in Figure 5 This is a scene diagram of surgical space registration in a confined space in another embodiment, in which... Figure 4The difference in the embodiments is that the reference markers are fixed on the CT medical imaging equipment and remain fixed in position relative to the patient. The reference markers include at least three optical markers that can be recognized by the camera. In this way, the camera acquires the position of the patient tracking markers, the camera acquires the pose of the reference markers on the CT equipment, and transfers the position of the tracking markers to the reference marker space.

[0069] Combination Figure 6 As shown, Figure 6 This is a schematic diagram of a tracking marker in one embodiment. In this embodiment, the tracking marker is affixed to the tissue site of the object, such as around the pre-puncture skin of a lesion, and fluctuates with tissue movement, such as fluctuating with respiratory movements, to monitor respiratory motion. The tracking marker is constructed with a specific geometry and consists of at least five optical markers, fixed to flexible connecting strips connected to each other. The tracking marker is adhered tightly to the skin, and each optical marker moves independently without affecting the others. The distance from the spatial centroid of the tracking marker to each individual optical marker is different for each pair of markers.

[0070] For ease of understanding, combined with Figure 7 As shown, Figure 7 This is a schematic diagram of the various operational stages in one embodiment, including a preoperative preparation stage, a respiratory data acquisition stage, and a respiratory normalization stage. The respiratory normalization stage includes a surgical space registration stage, a surgical operation execution time determination stage, and the surgical stage. This application mainly relates to the respiratory data acquisition stage and the surgical space registration stage and the surgical operation execution time determination stage within the respiratory normalization stage.

[0071] Among them, the combination Figure 8 As shown, Figure 8 This is a flowchart of the various operational stages in one embodiment. In this embodiment, the respiratory data acquisition stage involves the generation of historical position sequences, i.e., the generation of tissue motion models. Specifically, this includes acquiring the historical positions of tracking markers to generate historical position sequences for each tracking marker. These historical position sequences can be used to determine the positions of the tracking markers, thereby generating respiratory curves and registering the surgical space during the normal breathing phase. Furthermore, these historical position sequences can also be used to determine or predict the surgical operation execution time. When determining the surgical operation execution time, the historical position sequence is used to determine whether the current time is the surgical operation execution time. When predicting the surgical operation execution time, the historical position sequence can be used to predict whether a future time will be the surgical operation execution time.

[0072] In one exemplary embodiment, such as Figure 9 As shown, a spatial registration method is provided, including the following steps S902 to S910. Wherein:

[0073] S902: Acquire the target location of the tracking marker in the reference space within at least one respiratory cycle.

[0074] The respiratory cycle can be optionally used. During respiration, the posture of the respiratory tissues changes, thus requiring the acquisition of the target position of the tracking marker in the reference space. When the tracking marker is unobstructed, its position can be directly obtained using a camera or magnetic navigation device. When the tracking marker is obstructed, its position can be determined using the tracking marker position acquisition method described below. Here, obstruction refers to occlusion caused by the doctor's actions, etc.

[0075] In one embodiment, periodically acquiring the target position of the tracking marker in the reference space includes: determining whether the camera has acquired the position of the tracking marker; when the camera acquires the position of the tracking marker, determining the target position of the tracking marker in the reference space based on the positions of the tracking marker and the reference marker in the reference space, and placing the target position in the historical position sequence in chronological order; when the camera does not acquire the position of the tracking marker, determining the target position of the tracking marker in the reference space based on the tracking marker position acquisition method in any of the embodiments below.

[0076] One point to note is that in this embodiment, the position of the tracking marker is obtained directly when it is available; otherwise, it can be predicted based on the tracking marker position acquisition method described below. However, it should be noted that when the position of the tracking marker is predicted, it is not updated in the historical position sequence to avoid the historical position sequence being filled with predicted values ​​due to long-term prediction.

[0077] S904: Determine the location of a tracking marker in the image space within at least one respiratory cycle.

[0078] Specifically, the position of a tracking marker in the image space refers to the position of a tracking marker in the medical image of an object acquired by medical imaging equipment, such as the position of a tracking marker in the medical image of an object acquired by CT equipment.

[0079] S906: Register the target position of the tracking marker in the reference space with the position of the tracking marker in the image space at the corresponding time to obtain the registration relationship at the corresponding time.

[0080] Within a respiratory cycle, a preset frequency much higher than the respiratory cycle is used to register the target position of the tracking marker in the reference space with the position of the tracking marker in the image space at the corresponding time to obtain the registration relationship at the corresponding time.

[0081] S908: Calculate the registration error corresponding to the registration relationship at each corresponding time, and determine the respiratory phase corresponding to the registration relationship based on the target position of the tracking marker in the reference space at the corresponding time.

[0082] The registration error is the error of the registration relationship at each time point. Since the patient tracking marker in the bedside reference space moves continuously with the patient's breathing, the registration between the medical image space and the bedside reference space at a certain time may not be the optimal registration. That is, the breathing phase during medical image imaging may be different from the patient's breathing phase during registration. Therefore, it is necessary to calculate the registration error corresponding to each registration relationship.

[0083] The breathing phase is determined based on the target position of the tracking marker in the reference space at the corresponding time. For example, the centroid position is first determined based on the target position of the tracking marker in the reference space, and then the centroid position is normalized to obtain the breathing phase. Combined with the table below, the registration relationship, registration error and breathing phase at each time can be recorded:

[0084]

[0085] S910: Select the respiratory phase corresponding to the registration relationship that meets the registration error requirements as the target respiratory phase for spatial registration, and provide surgical time reminders based on the target respiratory phase.

[0086] In this embodiment, the registration error is considered to be the minimum. The moment when the dynamic registration error is at its minimum is the moment when the bedside reference space respiratory phase and the image space respiratory phase are consistent. The registration relationship and the target registration result are saved and recorded. In other embodiments, other registration errors can be selected.

[0087] In the above embodiments, predictive data is used for automatic registration in the limited surgical space, which overcomes the difficulties caused by space limitations. The registration relationship and target registration result are saved and recorded when the dynamic registration error reaches its minimum value, which is the moment when the respiratory phase of the bedside reference space and the respiratory phase of the imaging space are consistent.

[0088] In one optional embodiment, the method for obtaining the location of the tracking marker mentioned above includes the following steps 1002 to 1008. Wherein:

[0089] S1002: Obtain the historical position sequence corresponding to the tracking marker. The historical position sequence includes the positions of the tracking markers arranged in chronological order. The tracking markers are fixed at the tissue parts of the object.

[0090] The specific limitations of the tracking markers can be found above. In this embodiment, the tracking markers are attached to the lesion site to monitor the posture of the lesion site during respiration. The historical position sequence of the tracking markers refers to the positions of the tracking markers collected during the respiration sampling phase. Each tracking marker corresponds to a historical position sequence, and the historical positions in each historical position sequence are arranged in chronological order. It should be noted that the positions of the tracking markers in the historical position sequence are the positions of the tracking markers in the reference space, which reduces the amount of subsequent calculations.

[0091] Specifically, combining and Figure 11 and Figure 12 As shown, Figure 11 This is a flowchart of the steps for generating the historical location sequence of tracking markers in one embodiment. Figure 11 This is a schematic diagram of the historical position sequence in one embodiment. In this embodiment, the respiratory data acquisition phase involves converting the original three-dimensional coordinate data of the tracking markers to a reference space and then storing it in the historical position sequence, such as in a historical position array. The duration of the respiratory data acquisition phase includes at least two respiratory cycles. Here, a fixed duration of 10 seconds is used as an example. Other durations can be used in other embodiments, and no specific limitation is made here. The number of historical position sequences is the number of optical markers of the tracking markers, plus a historical position sequence of the tracking marker core. The tracking marker core is obtained by averaging the spatial coordinates of all optical markers of the patient's respiratory markers. The length of all historical position sequences is fixed, for example, 5000. When the data of the historical position sequence reaches the maximum value, the historical position sequence is dynamically updated, that is, old data is removed from the head of the array, and new data is added to the end of the array. The storage description of each historical position sequence is as follows. Taking respiratory marker optical marker 1 as an example, the position at time C-1 is the end of the sequence, and the position at time T is the head of the sequence. Each position stores the three-dimensional coordinates (x, y, z) of the tracking marker optical marker.

[0092] S1004: Construct the current position vector and the initial historical position vector based on the positions of the tracking markers arranged in chronological order.

[0093] The current position vector is generated based on the position of the tracking marker at the current time, such as the position of the tracking marker at the end of the historical position sequence. The initial historical position vector is generated based on the positions of other tracking markers in the historical position sequence. For ease of understanding, we can combine... Figure 13 As shown, Figure 13 This is a schematic diagram of the current position vector and the initial historical position vector in one embodiment. In this embodiment, the breathing state vector is constructed by using historical data from the current time T, historical time T-3, and historical time T-6, as shown below. Figure 12 As shown, Pn is the predicted value of the respiratory state vector Vn at the current historical moment; Note: T, T-3, T-6 can also be T, T-4, T-8, etc., and are not specifically limited here. In this embodiment, the current position vector Vc and the initial historical position vector Vn can be combined Figure 12 As shown.

[0094] S1006: Based on the vector distance between each initial historical position vector and the current position vector, select a preset number of target historical position vectors from the initial historical position vectors.

[0095] The vector distance can be calculated by comparing the initial historical position vector with the current position vector, and the calculation method can be as follows:

[0096]

[0097] Where dn is the vector distance. , , The element in the current position vector Vc , , It is an element in the nth initial historical position vector.

[0098] In this embodiment, a preset number of target historical position vectors are selected from the initial historical position vectors based on the vector distance between each initial historical position vector and the current position vector.

[0099] In one optional embodiment, a preset number of target historical position vectors are selected from the initial historical position vectors based on the vector distance between each initial historical position vector and the current position vector. This includes: obtaining the vector distance between each initial historical position vector and the current position vector; selecting initial historical position vectors whose vector distance is less than a distance threshold and counting the number of selected initial historical position vectors; when the number is less than or equal to the preset number, using the selected initial historical position vectors as target historical position vectors; when the number is greater than the preset number, determining the preset number of target historical position vectors from the selected initial historical position vectors in ascending order of vector distance.

[0100] Specifically, in this embodiment, the target historical position vector is obtained through a dual selection process using a threshold and K-nearest neighbors. First, a distance threshold is determined. Then, initial historical position vectors with a vector distance less than the threshold are selected. Since the length of the historical position sequence is fixed, the number of initial historical position vectors is also limited. Therefore, initial historical position vectors with a vector distance less than the threshold are selected, and the number of such vectors is counted. If this number is less than a preset number, for example, less than 10, all initial historical position vectors with a vector distance less than the threshold are obtained as the target historical position vector. Otherwise, a preset number of initial historical position vectors with a vector distance less than the threshold are selected in ascending order of vector distance as the target historical position vector. It should be noted that the above operation is performed on each group of historical position vectors to obtain the target historical position vector corresponding to each tracking marker. A similar operation is performed on the centroid historical position vector, which will not be elaborated here. In the above embodiment, the dual-layer selection improves the accuracy of the target historical position vector and reduces the number of target historical position vectors, thus improving computational efficiency.

[0101] S1008: Based on a preset number of historical target location vectors, predict the location of the tracking marker to obtain the target location.

[0102] Specifically, in this embodiment, the target position of the tracking marker is predicted based on a preset number of historical target position vectors. For example, the future position of the tracking marker is predicted based on the historical positions of the tracking markers corresponding to a preset number of historical target position vectors. For accuracy, this embodiment predicts the target position of the tracking marker based on at least one, and optionally multiple, preset number of historical target position vectors.

[0103] In one optional embodiment, the position of the tracking marker is predicted based on a preset number of target historical position vectors to obtain the target position, including: determining the weight of each target historical position vector based on the vector distance between each target historical position vector and the current position vector; obtaining the historical position of the tracking marker corresponding to each target historical position vector; and obtaining the target position of the tracking marker according to the historical position of the tracking marker corresponding to each target historical position vector and the weight.

[0104] The historical position of the tracking marker corresponding to the target's historical position vector can be combined with... Figure 12 As shown, Pn represents the historical position of the tracking marker corresponding to the target's historical position vector Vn at the current historical moment. The position between Pn and Vn can be represented as follows: Figure 13The position immediately following the target in the diagram represents the historical position of the tracking marker at the next moment. In other embodiments, position Pn can also be the historical position of the tracking marker at time t+n. This is not specifically limited here, where t can be determined based on system latency or actual prediction needs. Correspondingly, the target position of the tracking marker can be the position at the next moment or the target position at time t+c. This is not specifically limited here.

[0105] The weights of the target's historical position vector can be determined based on the vector distance between the target's historical position vector and its current position vector. Specifically,

[0106]

[0107] Among them W i It is the weight, d i It is the vector distance, and N is the preset number.

[0108] Thus tracking the target location of the marker. , where P i It is the historical position of the tracking marker corresponding to the target's historical position vector.

[0109] It should be noted that the above embodiments only provide a method for determining the target position of a tracking marker. The method can be used to determine the target position of other tracking markers and the target position of the centroid, and will not be elaborated here.

[0110] The above-described method for obtaining the location of tracking markers involves acquiring the historical location sequence corresponding to the tracking markers when the tracking system cannot obtain their location. Based on the chronologically ordered locations of the tracking markers, a current location vector and initial historical location vectors are constructed. Based on the vector distance between each initial historical location vector and the current location vector, a preset number of target historical location vectors are selected from the initial historical location vectors. Based on the preset number of target historical location vectors, the location of the tracking markers is predicted to obtain the target location. In this way, even if the tracking system cannot collect the target location of the tracking markers, the target location of the tracking markers can be estimated based on the historical location sequence corresponding to the tracking markers, thereby enabling registration of the surgical space when the field of view is limited.

[0111] In one embodiment, before obtaining the historical position sequence corresponding to the tracking marker, the process includes: collecting the positions of the tracking marker and the reference marker in the camera space, wherein the positions of the reference marker and the object are relatively fixed; obtaining the reference space corresponding to the reference marker; determining the position of the tracking marker in the reference space based on the positions of the tracking marker and the reference marker in the camera space; and storing the positions of the tracking marker in the reference space into the target sequence in chronological order to obtain the historical position sequence corresponding to the tracking marker.

[0112] Among them, combined Figures 1 to 5 As shown, the positions of tracking markers and reference markers in camera space can be acquired, for example, through a camera or magnetic navigation device. The positions of the reference markers and the object are relatively fixed. The reference space can be determined based on the reference markers. Therefore, there are at least three reference markers to uniquely determine the reference space. Since the reference markers are determined by a camera or magnetic navigation device, the transformation relationship between camera space and reference space can be determined based on the reference markers. Then, based on this transformation relationship, the positions of the tracking markers in camera space are transformed to the reference space and stored in the corresponding historical position sequence.

[0113] Each tracking marker corresponds to a historical position sequence. After obtaining the position of the tracking marker in the reference space, the corresponding historical position sequence is determined based on the identifier of the tracking marker, and the position of the tracking marker in the reference space is stored in the historical position sequence. In one optional embodiment, the storage of the tracking marker positions can be performed in parallel. That is, after obtaining the transformation relationship, the positions of each tracking marker in the camera space are processed in parallel based on the transformation relationship to obtain the positions of the tracking markers in the reference space, and stored in the corresponding historical position sequence. The number of parallel processing threads can be equal to the number of tracking markers or equal to the number of tracking markers plus one, that is, it also includes a centroid processing thread.

[0114] In the above embodiments, a bedside reference space is used instead of a camera reference space, decoupling the dependence on camera position and making the camera position more flexible. The bedside reference refers to an optical marker that can be recognized by the camera, and the bedside reference space refers to the space with the optical marker as the origin of the coordinate system. The relationship between the bedside reference and the object's position is defined, and the installation position is flexible; it can be installed on CT equipment, CT equipment support devices, robotic arm trolleys, etc. All spatial transformations during the surgical procedure are based on the bedside reference space. The tracking marker moves within the bedside reference space, and the robotic arm end-effector target is positioned within the bedside reference space.

[0115] In one embodiment, the positions of tracking markers in the reference space are stored in the target sequence in chronological order to obtain the historical position sequence corresponding to the tracking markers. This includes: obtaining the sequence length, which is at least greater than the length of one respiratory cycle; when the current length of the historical position sequence is less than the sequence length, storing the positions of tracking markers in the reference space in chronological order into the historical position sequence; when the current length of the historical position sequence is greater than or equal to the sequence length, deleting the positions of the tracking markers in the reference space that are earlier in the historical position sequence, and storing the current positions of the tracking markers in the reference space in chronological order into the historical position sequence.

[0116] The length of the historical position sequence is fixed, and the sequence length is at least greater than the length of one respiratory cycle, and optionally includes at least the length of two respiratory cycles. Therefore, the positions of the tracking markers in the historical position sequence change in real time. When the current length of the historical position sequence is less than the sequence length, the positions of the tracking markers in the reference space are stored in the historical position sequence in chronological order. If the current length of the historical position sequence is greater than or equal to the sequence length, the positions of the tracking markers in the historical position sequence are updated, for example, by deleting old data and adding new data. That is, the positions of the tracking markers in the reference space that are earlier in time in the historical position sequence are deleted, and the positions of the tracking markers in the current reference space are stored in the historical position sequence in chronological order.

[0117] In the above embodiments, the positions of tracking markers in the historical position sequence are dynamically adjusted to ensure the effectiveness of the tracking marker positions, thereby improving the accuracy of prediction. A historical position sequence is constructed by collecting and normalizing the raw motion data of the tracking markers. This historical position sequence can be updated in real time using a sliding window method, employing a dual screening method of threshold and K-nearest neighbors, and using a distance weighting method for prediction. The historical position sequence can predict the moment of camera field of view occlusion, thereby reducing the environmental requirements for the surgical operating space; the historical position sequence can predict the position of respiratory markers, making the dynamic registration algorithm more robust; and the establishment of the historical position sequence can predict the surgical operation execution time, reducing errors caused by system lag.

[0118] In one embodiment, the tracking markers include at least five optical markers or magnetic navigation positioning markers, which are sequentially connected to form the target geometry, and the distance between any two optical markers or magnetic navigation positioning markers is not the same. This gives the tracking markers a specific structure, thereby enabling registration between the tracking markers in the reference space and the markers in the image space through this specific structure. Specifically, registering the target position of the tracking marker in the reference space with the position of the tracking marker in the image space at a corresponding time to obtain the registration relationship at the corresponding time includes: identifying the geometric features, such as geometric shape, of the tracking markers in the reference space and the image space at the corresponding time; determining the correspondence between the tracking markers in the reference space and the tracking markers in the image space based on the geometric features; and registering the target position of the tracking marker in the reference space with the position of the tracking marker in the image space based on the correspondence to obtain the registration relationship at the corresponding time.

[0119] In one embodiment, identifying the geometric features of tracking markers in the reference space and the image space at corresponding time points includes: calculating the centroids of the tracking markers in the reference space and the image space at corresponding time points; sorting the tracking markers according to their distances from the corresponding centroids; and determining the correspondence between the tracking markers in the reference space and the tracking markers in the image space based on the geometric features, including: determining the correspondence between the tracking markers in the reference space and the tracking markers in the image space based on the sorted positions of the tracked markers in the reference space and the image space.

[0120] Geometric features are inherent characteristics of the tracking markers. For example, the distances from the spatial centroid of a tracking marker to each optical marker are different for each pair of markers. Therefore, in this embodiment, the centroid of the tracking marker in the reference space is first calculated based on its position in the reference space. Then, the distances from the centroid of the tracking marker in the reference space to each tracking marker are calculated, sorted from largest to smallest or smallest to largest. Similarly, the distances from the centroid of the tracking marker in the medical imaging space to each respiratory optical marker are calculated, sorted from largest to smallest or smallest to largest. Based on this sorted position, the tracking markers in the reference space are matched one-to-one with the tracking markers in the image space to determine the corresponding correspondence. Based on this correspondence, the target position of the tracking marker in the reference space is registered with the position of the tracking marker in the image space to obtain the registration relationship at the corresponding time. When the camera's field of view is obstructed, the target position of the tracking marker in the reference space is predicted using historical position sequences, and then registration is performed.

[0121] In the above embodiments, in order to complete the real-time registration of tracking markers, high requirements are placed on the real-time performance, robustness, and complexity of the registration algorithm. By designing a unique tracking marker geometry, a geometric feature clustering method is used to find the one-to-one correspondence between respiratory markers in the bedside reference space and the imaging space, thereby completing the registration in the surgical space and improving the efficiency and accuracy of registration.

[0122] In one embodiment, calculating the registration error corresponding to the registration relationship at each corresponding time point includes: converting the target position of the tracking marker in the reference space to the image space based on the registration relationship to obtain the reference position; calculating the difference between each reference position and the position of the tracking marker in the image space; and selecting the maximum difference as the registration error at the corresponding time point.

[0123] Among them, combined Figure 14 As shown, Figure 14This is a schematic diagram of the registration error calculation steps in one embodiment. After the registration relationship is calculated, the target position of the tracking marker in the reference space at the current time is converted to the image space through the registration relationship to obtain the reference position. The difference between each reference position and the position of the tracking marker in the image space is calculated. Based on the difference, the registration error is obtained. In this embodiment, the maximum difference is selected as the registration error at the corresponding time.

[0124] In other embodiments, the position of the tracking marker in the image space can be converted to the reference space based on the registration relationship, and the registration error can be determined based on the difference between the target position of the tracking marker in the reference space and the converted position.

[0125] In the above embodiments, since the tracking markers in the bedside reference space move continuously with the patient's breathing fluctuations, the registration between the medical image space and the bedside reference space at a certain moment may not be optimal. That is, the respiratory phase during medical imaging may differ from the respiratory phase of the patient during registration. Within a respiratory cycle, for example, when the centroid of the patient's respiratory marker changes from its maximum value to its minimum value and then back to its maximum value, registration is performed at a preset frequency much higher than the respiratory rate during this period. This yields a registration error array. The registration matrix and respiratory amplitude at the minimum registration error in the array are selected as the optimal registration result, thus completing the registration between the medical image space and the bedside reference space. The maximum error among the registration markers is used as the phase registration error.

[0126] In one embodiment, determining the respiratory phase corresponding to the registration relationship based on the target position of the tracking marker in the reference space at the corresponding time includes: obtaining the fluctuation range; determining the centroid position based on the target position of the tracking marker in the reference space at the corresponding time; when the centroid position is within the fluctuation range, normalizing the centroid position according to the fluctuation range to obtain the respiratory phase corresponding to the registration relationship; when the centroid position exceeds the fluctuation range, updating the fluctuation range based on the target position of the tracking marker.

[0127] The fluctuation range is predetermined, for example, determined during the respiratory data acquisition phase, and is adjusted in real time. For example, in one embodiment, obtaining the fluctuation range includes: obtaining the historical position sequence of the tracking markers; obtaining the historical position sequence of the centroid corresponding to each tracking marker based on the historical position sequence of each tracking marker; projecting each centroid position in the historical position sequence of the centroid onto a preset coordinate axis, and obtaining the coordinate axis with the largest projection range after projection as the principal axis, and obtaining the projection range of the centroid position on the principal axis as the fluctuation range.

[0128] Combination Figure 15 As shown, Figure 15This is a schematic diagram of the fluctuation range acquisition step in one embodiment. In this embodiment, the fluctuation range refers to the fluctuation range of the centroid. Since the centroid is determined based on the positions of all tracking markers, it is more accurate. Before determining the fluctuation range, the historical position sequence of the centroid corresponding to each tracking marker is obtained based on the historical position sequence of each tracking marker. The historical position of the centroid is obtained based on the historical position of the tracking marker at the corresponding time, for example, by calculating the average value. Figure 16 As shown, Figure 16 This is a schematic diagram of the center of mass motion in one embodiment. In this embodiment, each center of mass position in the historical position sequence is projected onto a preset coordinate axis, and the coordinate axis with the largest projection range is obtained as the principal axis. The projection range of the center of mass position on the principal axis is obtained as the fluctuation range. In the subsequent normal breathing phase, the center of mass position is normalized using this fluctuation range to obtain the breathing phase, such as... Figure 16 As shown, OB1 represents the three-dimensional motion of the centroid of the tracking marker during the respiratory acquisition phase. OA, OC, and OO1 represent the changes in the XYZ coordinate axes of OB1 projected onto the bedside reference space. Assuming OA is the direction of maximum fluctuation, the normalized respiratory amplitude is Xnorm. During normal breathing, if the normalization condition is not met, for example, when the centroid position exceeds the fluctuation range, the fluctuation range is updated based on the target position of the tracking marker, i.e., the direction of the main axis of data fluctuation is determined again to achieve adaptive data dimensionality reduction.

[0129] After determining the fluctuation range, the centroid position is normalized based on the fluctuation range to obtain the respiratory phase corresponding to the registration relationship. Since respiration is periodic, the three-dimensional motion of the respiratory marker centroid in the reference space is approximately linear and reciprocating, such as... Figure 17 As shown, P1P2 is projected onto the XYZ axes, and the axis with the largest fluctuation in the XYZ directions is calculated as the principal axis. Assuming that the Y-axis has the largest fluctuation, the respiratory phases Ymax-Ymin are recorded. The respiratory amplitude Vnorm is calculated using a normalization method.

[0130]

[0131] In the above embodiments, a respiratory data acquisition stage is set up during the surgical preparation process. During this stage, the direction of the maximum three-dimensional fluctuation of the respiratory marker data is calculated and used as the main axis direction. At the same time, in the subsequent data processing, if the normalization condition is not met, the main axis direction of data fluctuation is determined again to achieve adaptive data dimensionality reduction.

[0132] In one exemplary embodiment, such as Figure 18 As shown, step S910 above also includes providing surgical time reminders based on the target respiratory phase. This step S910, namely the step of determining the surgical operation execution time, may include:

[0133] S1802: Obtain the target respiratory phase of the pre-generated target registration result.

[0134] The calculation method for the target respiratory phase can be found in the target registration results above, and will not be repeated here.

[0135] S1804: Obtain the target position of the tracking marker in the reference space.

[0136] The method for obtaining the target position of the tracking marker can be found above and will not be repeated here. One point to note in this embodiment is that the target position of the tracking marker can be obtained in real time, or it can be a prediction of the target position of the tracking marker at the predicted time.

[0137] For real-time acquisition, the target position of the tracking marker can be obtained in real time. If there is no obstruction, the target position of the tracking marker can be obtained based on the camera or magnetic navigation device. If there is obstruction, the target position of the tracking marker can be obtained through the tracking marker position acquisition method mentioned above.

[0138] Due to the lag in the system, the respiratory phase calculated by tracking markers is delayed. The target position of the tracking markers can be predicted using the tracking marker position acquisition method described above.

[0139] S1806: Determine the centroid position of the target based on the target position of the tracking marker.

[0140] The target centroid position is calculated based on the positions of each tracking marker at the corresponding time, such as by taking the average value, without making specific limitations here.

[0141] S1808: Obtain the fluctuation range, and when the target centroid position is within the fluctuation range, normalize the target centroid position based on the fluctuation range to obtain the current breathing phase.

[0142] Specifically, the method for normalizing the target centroid position can be found in the section on centroid position normalization above, and will not be repeated here.

[0143] S1810: When the current respiratory phase matches the target respiratory phase, the time corresponding to the current respiratory phase is used as the surgical operation execution time.

[0144] The system compares the current respiratory phase with the target respiratory phase. If they match, the moment with the smallest matching error is reached, and this moment is taken as the surgical procedure execution time. This time can be output to notify the operator, such as a doctor. Optionally, the above-mentioned puncture intervention system also includes a display module for outputting the surgical procedure execution time.

[0145] In the above embodiments, the target respiratory phase at the optimal matching time is obtained through fully automated real-time registration; this respiratory phase is the respiratory phase at the puncture timing. During the surgical procedure, the constructed historical location sequence is used to predict the respiratory amplitude at the set time and compared with the respiratory phase at the puncture timing, thus allowing for advance perception of the surgical procedure time.

[0146] In one embodiment, after obtaining the fluctuation range and normalizing the target centroid position based on the fluctuation range to obtain the current respiratory phase, the method further includes: generating a current respiratory curve based on the current respiratory phase; and after using the time corresponding to the current respiratory phase as the surgical operation execution time, the method further includes: outputting a prompt in the current respiratory curve indicating that the time corresponding to the current respiratory phase is the surgical operation execution time.

[0147] To achieve visualization, a current respiratory curve can be generated based on the current respiratory phase. This curve displays the current respiratory phase in real time and provides a prompt at the corresponding time position on the current respiratory curve when the surgical procedure execution time is determined. The display module also displays the current respiratory curve, outputting the time corresponding to the current respiratory phase as a prompt for the surgical procedure execution time.

[0148] Specifically, in combination Figure 19 As shown, Figure 19 This is a schematic diagram of the dynamic registration process in one embodiment. In this embodiment, when the camera's field of view is unobstructed, the position of the tracking marker is acquired and then transferred to a reference space for registration with the position of the tracking marker in the medical image space. Since points in different spaces are not paired, point-to-point relationships are established through geometric feature clustering, and then registration is performed using SVD or quaternions. The registration error, registration relationship, and respiratory phase are recorded; the maximum value of the registration error is used. When the camera's field of view is obstructed, the position of the tracking marker in the reference space is predicted using the historical position sequence, and then registration is performed. After registration, the registration error, registration relationship, and respiratory phase are also recorded. It should be noted that in this embodiment, if the historical position sequence is used to generate the current fluctuation curve, the predicted position of the tracking marker is added to the historical position sequence to generate a smooth curve. If it is used for registration, the predicted position of the tracking marker is not added to the historical position sequence to avoid the historical position sequence containing only predicted values.

[0149] In one embodiment, obtaining the target position of the tracking marker in the reference space includes: obtaining the system lag time, determining the prediction time based on the system lag time; predicting the target position of the tracking marker in the reference space corresponding to the prediction time, and placing the target position in the historical position sequence in chronological order; using the time corresponding to the current respiratory phase as the surgical operation execution time includes: using the prediction time corresponding to the current respiratory phase as the surgical operation execution time.

[0150] The system lag time is obtained from actual system lag tests and experiments. Thus, during prediction, the prediction time is determined based on the system lag time. For example, if the prediction time is greater than or equal to the system lag time, the target position of the tracking marker in the reference space can be predicted in advance, and the current respiratory phase can be determined. When the current respiratory phase is equal to the target respiratory phase, the predicted time is determined as the surgical operation execution time and displayed in the current fluctuation amplitude curve to indicate the surgical operation execution time in advance.

[0151] In the above embodiments, a surgical operation prediction mechanism is provided to solve the system lag problem and make the timing of surgical operations more accurate.

[0152] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0153] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0154] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A puncture intervention system characterized by, The system includes: A reference marker is used, and the position of the reference marker remains unchanged. A reference space is obtained based on the reference marker. A tracking marker, which is fixed to the tissue portion of the object; Medical imaging equipment, used to acquire medical images of the tissue sites of the object; A tracking device for acquiring the positions of the reference marker and the tracking marker; The processor is configured to, within at least one respiratory cycle, acquire the target position of the tracking marker in a reference space based on the acquired positions of the reference marker and the tracking marker, and determine the position of the tracking marker in an image space based on the medical image, and register the target position of the tracking marker in the reference space at a corresponding time with the position of the tracking marker in the image space to obtain the registration relationship at the corresponding time. Calculate the registration error corresponding to the registration relationship at each corresponding time, and determine the respiratory phase corresponding to the registration relationship based on the target position of the tracking marker in the reference space at the corresponding time; select the respiratory phase corresponding to the registration relationship whose registration error meets the requirements as the target respiratory phase; and provide a surgical time reminder based on the target respiratory phase. A puncture robotic arm is used to perform the puncture procedure during the surgical time.

2. The puncture intervention system according to claim 1, characterized by, The reference marker is mounted on the operating table, or on the robotic arm trolley, or on the medical imaging equipment.

3. The puncture intervention system according to claim 1, characterized by, The tracking markers include at least five optical markers or magnetic navigation positioning markers, which are connected sequentially to form the target geometry, and the distance between any two of the optical markers or magnetic navigation positioning markers is not the same; The processor is also configured to identify the geometry of the tracking markers in the reference space and the image space at corresponding times; Based on the geometry, the correspondence between the tracking markers in the reference space and the tracking markers in the image space is determined; Based on the correspondence, the target position of the tracking marker in the reference space is registered with the position of the tracking marker in the image space to obtain the registration relationship at the corresponding time.

4. The puncture intervention system according to claim 1, characterized by, The processor is further configured to convert the target position of the tracking marker in the reference space to the image space based on the registration relationship to obtain the reference position; Calculate the difference between each of the reference positions and the position of the tracking marker in the image space; The maximum difference is selected as the registration error at the corresponding time point.

5. The puncture intervention system according to claim 1, characterized by, The tracking marker has a different structure from the reference marker; the processor is also used for: Based on the structure of the tracking marker and the reference marker, it is determined whether the tracking device has acquired the position of the tracking marker; When the tracking device acquires the position of the tracking marker, it determines the target position of the tracking marker in the reference space based on the positions of the tracking marker and the reference marker, and places the target position in the historical position sequence in chronological order; When the tracking device cannot acquire the location of the tracking marker, it acquires the historical location sequence corresponding to the tracking marker, the historical location sequence including the locations of the tracking marker arranged in chronological order; Based on the positions of the tracking markers arranged in chronological order, construct the current position vector and the initial historical position vector; Based on the vector distance between each of the initial historical position vectors and the current position vector, a preset number of target historical position vectors are selected from the initial historical position vectors; Based on the preset number of target historical location vectors, the location of the tracking marker is predicted to obtain the target location.

6. The puncture intervention system according to claim 5, characterized by, The processor is further configured to: obtain the vector distance between each of the initial historical position vectors and the current position vector; select initial historical position vectors whose vector distance is less than a distance threshold, and count the number of selected initial historical position vectors; when the number is less than or equal to a preset number, use the selected initial historical position vectors as target historical position vectors; when the number is greater than the preset number, determine the preset number of target historical position vectors from the selected initial historical position vectors in ascending order of vector distance; The weight of each target historical position vector is determined based on the vector distance between each target historical position vector and the current position vector. Obtain the historical position of the tracking marker corresponding to each of the target historical position vectors; based on the historical position of the tracking marker corresponding to each of the target historical position vectors and the weight, obtain the target position of the tracking marker.

7. The puncture intervention system according to claim 5, characterized by, The processor is also configured to acquire the sequence length, which is at least greater than the length of one respiratory cycle; when the current length of the historical position sequence is less than the sequence length, the positions of the tracking markers in the reference space are stored in the historical position sequence in chronological order. When the current length of the historical location sequence is greater than or equal to the sequence length, delete the position of the tracking marker in the reference space that is earlier in the historical location sequence, and store the current position of the tracking marker in the reference space into the historical location sequence in chronological order.

8. The puncture intervention system according to any one of claims 1 to 6, characterized in that, The processor is also configured to determine the target centroid position based on the target position of the tracking marker; obtain the fluctuation range; and when the target centroid position is within the fluctuation range, normalize the target centroid position based on the fluctuation range to obtain the current respiratory phase. When the current respiratory phase matches the target respiratory phase, the time corresponding to the current respiratory phase is taken as the surgical operation execution time; The system also includes: The display module is used to output the execution time of the surgical operation.

9. The puncture intervention system according to claim 8, characterized in that, The processor is also configured to place the target location in a historical location sequence in chronological order, and generate a current breathing curve based on the target location of each of the tracking markers in the historical location sequence; The display module is also used to display the current respiratory curve, and output the time corresponding to the current respiratory phase in the current respiratory curve as a prompt for the execution time of the surgical operation.

10. The puncture intervention system according to claim 8, characterized in that, The processor is also used to acquire the system lag time, determine the prediction time based on the system lag time, and predict the target position of the tracking marker in the reference space corresponding to the prediction time.

Images