Ureteral stent positioning method, device, equipment and robotic surgery system

Abstract

The present invention discloses a ureteral stent positioning method, device, equipment and robotic surgery system, which belong to the field of ureteral stent positioning. After obtaining the CT slice images and their sequence of the ureteral stent, a two-dimensional neural network is used to obtain the detection frame of the ureteral stent in each CT slice image; the main branch sequence of the ureteral stent is obtained according to the sequence of the CT slice images and the detection frame; the main branch sequence is supplemented according to the CT slice image to obtain a complete detection frame sequence, and the position of the ureteral stent is determined according to the complete detection frame sequence. The present application solution only uses a two-dimensional neural network to process the CT slice image, without the need to use a three-dimensional slice image, with a small amount of calculation and low cost; in addition, the main branch sequence is first obtained; then the main branch sequence is supplemented to obtain a complete detection frame sequence, and then the position of the ureteral stent is determined according to the complete detection frame sequence, without the need to map the two-dimensional image to the three-dimensional space, which greatly reduces the amount of calculation while ensuring that no information is missed.

Description

Technical Field

[0001] The present invention relates to the field of ureteral stent positioning, and in particular to a ureteral stent positioning method, device and equipment, and a robotic surgery system. Background Art

[0002] Ureteral stents are the most commonly used tools in the treatment of various benign and malignant urological diseases. Figure 1 However, stent use is often associated with complications such as crusting, infection, pain, discomfort after placement, and stent migration or failure. These complications significantly impact patient treatment outcomes and quality of life. Therefore, patients now need to confirm the position of the stent after placement.

[0003] Existing technology uses CT to locate the stent. CT (Computed Tomography) uses precisely collimated X-ray beams, gamma rays, ultrasound waves, and other technologies, along with highly sensitive detectors, to scan a specific part of the human body in sections one after another. CT can produce CT slice images of multiple sections of a specific part.

[0004] However, when positioning based on CT sequence images (i.e., continuous cross-sectional images of a certain part obtained in a certain order (generally from one end to the other)), if a two-dimensional neural network model is directly used, information is often missed. Constructing a three-dimensional neural network model for feature extraction is computationally intensive and costly. After the two-dimensional image is segmented, it is mapped back to three-dimensional space for processing. This solution also requires a large amount of computation and is costly. Summary of the Invention

[0005] In order to overcome the deficiencies of the prior art, the present invention provides a ureteral stent positioning method, device and equipment and a robotic surgery system to solve the problem of large computational complexity and high cost when ensuring that no information is missed when positioning based on CT sequence images.

[0006] The technical solution adopted by the present invention to solve its technical problem is:

[0007] Firstly,

[0008] A ureteral stent positioning method comprises the following steps:

[0009] Acquiring CT sequence image information of the ureteral stent, wherein the CT sequence image information includes each CT slice image and its sequence;

[0010] Inputting each CT slice image into a two-dimensional neural network to obtain a detection frame of the ureteral stent;

[0011] Obtaining a main branch sequence of the ureteral stent according to the order of the CT slice images and the detection frame;

[0012] Supplementing the main branch sequence according to the CT sequence image to obtain a complete detection frame sequence;

[0013] The position of the ureteral stent is determined according to the complete detection frame sequence.

[0014] Furthermore, obtaining the main branch sequence of the ureteral stent according to the order of the CT slice images and the detection frame includes:

[0015] Acquiring a first number of CT slice images in which the same detection frame appears continuously according to the order of the CT slice images;

[0016] When the first number is greater than a first preset number, the continuous identical detection frames are used as the main branch sequence of the ureteral stent.

[0017] Furthermore, the step of obtaining the first number of CT slice images in which the same detection frame appears continuously further includes:

[0018] acquiring, according to the order of the CT slice images, a second number of target CT slice images between two CT slice images in which the same detection frame appears;

[0019] When the second number is not greater than a second preset number, the same detection frame is added to the target CT slice image that does not have the same detection frame.

[0020] Furthermore, the method of supplementing the main branch sequence according to the CT sequence image to obtain a complete detection frame sequence includes:

[0021] Acquire, in the order of the CT slice images, first detection frames of a non-main branch sequence that appear continuously on the CT slice images, wherein the number of CT slice images in which the first detection frames appear continuously is greater than a third preset number;

[0022] According to the order of the CT slice images, the first detection frame intersecting with the main branch sequence is added to the main branch sequence.

[0023] Furthermore, it also includes:

[0024] Acquire, in an order reverse to the order of the CT slice images, second detection frames of the non-supplemented main branch sequence that appear continuously on the CT slice images, wherein the number of CT slice images in which the second detection frames appear continuously is greater than a fourth preset number;

[0025] In an order opposite to the order of the CT slice images, the second detection frame that intersects with the supplemented main branch sequence is supplemented to the main branch sequence to obtain a complete detection frame sequence.

[0026] Furthermore, when there are at least two main branch sequences, and any two of them intersect, the following steps are performed:

[0027] Get the target detection frame of the intersection part;

[0028] The target detection frame is taken as a detection frame belonging only to any one of the two intersecting main branch sequences; and the target detection frame is deleted from the detection frame belonging to the other main branch sequence.

[0029] Furthermore, inputting each CT slice image into a two-dimensional neural network to obtain a detection frame of the ureteral stent includes:

[0030] Inputting the CT slice image into a target detection network to extract features of the CT slice image;

[0031] The extracted features are used to generate regions of interest through RPN;

[0032] Mapping the coordinates of the region of interest back to the extracted features through the ROIAlign operation;

[0033] Then, the detection frame of the ureteral stent is obtained through the FC layer.

[0034] Secondly,

[0035] A ureteral stent positioning device, comprising:

[0036] An image information acquisition module, configured to acquire CT sequence image information of the ureteral stent, wherein the CT sequence image information includes each CT slice image and its sequence;

[0037] A detection frame acquisition module is used to input each CT slice image into a two-dimensional neural network to obtain a detection frame of the ureteral stent;

[0038] a main branch sequence acquisition module, configured to obtain a main branch sequence of the ureteral stent according to the order of the CT slice images and the detection frame;

[0039] A complete detection frame acquisition module, configured to supplement the main branch sequence according to the CT sequence image to obtain a complete detection frame sequence;

[0040] A position determination area is used to determine the position of the ureteral stent according to the complete detection frame sequence.

[0041] Thirdly,

[0042] A computer device comprising:

[0043] processor;

[0044] a memory for storing instructions executable by the processor;

[0045] The processor is configured to execute the method according to any one of claims 1 to 7.

[0046] Fourthly,

[0047] A robotic surgery system comprising:

[0048] a positioning module, configured to execute the method according to any one of claims 1 to 7 to obtain the position of the ureteral stent;

[0049] A surgery module is used to perform surgery according to the position obtained by the positioning module.

[0050] Beneficial effects:

[0051] The technical solution of the present application provides a ureteral stent positioning method, device, equipment and robotic surgery system. After obtaining the CT slice images and their sequence of the ureteral stent, a two-dimensional neural network is used to obtain the detection frame of the ureteral stent in each CT slice image; then, the main branch sequence of the ureteral stent is obtained according to the sequence of the CT slice images and the detection frame; then, the main branch sequence is supplemented according to the CT slice images to obtain a complete detection frame sequence, and finally, the position of the ureteral stent is determined according to the complete detection frame sequence. The present application solution only uses a two-dimensional neural network to process the CT slice images, without the need to use three-dimensional slice images, with low computational complexity and low cost; in addition, the main branch sequence is first obtained; then, the main branch sequence is supplemented to obtain a complete detection frame sequence, and then the position of the ureteral stent is determined according to the complete detection frame sequence, without the need to map the two-dimensional image to three-dimensional space, which greatly reduces the computational complexity while ensuring that no information is missed. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0053] Figure 1 It is a schematic diagram of the structure of a ureteral stent;

[0054] Figure 2 This is a flow chart of a ureteral stent positioning method provided by an embodiment of the present invention;

[0055] Figure 3 1 is a schematic diagram of a detection frame on a CT slice image provided by an embodiment of the present invention;

[0056] Figure 4 This is a flow chart of a specific ureteral stent positioning method provided by an embodiment of the present invention;

[0057] Figure 5 This is a structural diagram of a coarse detection neural network provided by an embodiment of the present invention;

[0058] Figure 6 This is a structural schematic diagram of a ureteral stent positioning device provided by an embodiment of the present invention;

[0059] Figure 7 This is a schematic diagram of the structure of a robotic surgery system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0060] To make the purpose, technical solutions, and advantages of this application more clear, the technical solutions of the present invention are described in detail below with reference to the accompanying drawings and examples. Obviously, the described embodiments are only some of the embodiments of this application, rather than all of them. Based on the embodiments in this application, all other implementation methods obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0061] Reference Figure 2 , an embodiment of the present invention provides a ureteral stent positioning method, comprising the following steps:

[0062] Acquire CT sequence image information of the ureteral stent, where the CT sequence image information includes each CT slice image and its sequence;

[0063] Each CT slice image is input into a two-dimensional neural network to obtain the detection frame of the ureteral stent;

[0064] The main branch sequence of the ureteral stent is obtained according to the order of CT slice images and the detection frame;

[0065] According to the CT sequence images, the main branch sequence is supplemented to obtain a complete detection frame sequence;

[0066] The position of the ureteral stent was determined based on the complete detection frame sequence.

[0067] The embodiment of the present invention provides a ureteral stent positioning method. After obtaining the CT slice images and their sequence of the ureteral stent, a two-dimensional neural network is used to obtain the detection frame of the ureteral stent in each CT slice image; then, the main branch sequence of the ureteral stent is obtained according to the sequence of the CT slice images and the detection frame; then, the main branch sequence is supplemented according to the CT slice images to obtain a complete detection frame sequence, and finally, the position of the ureteral stent is determined according to the complete detection frame sequence. The present application solution only uses a two-dimensional neural network to process the CT slice images, without the need to use three-dimensional slice images, with low computational complexity and low cost; in addition, the main branch sequence is first obtained; then, the main branch sequence is supplemented to obtain a complete detection frame sequence, and then the position of the ureteral stent is determined according to the complete detection frame sequence. There is no need to map the two-dimensional image to three-dimensional space, which greatly reduces the computational complexity while ensuring that no information is missed.

[0068] As a supplementary explanation to the above embodiment, obtaining the main branch sequence of the ureteral stent according to the order of the CT slice images and the detection frame includes: obtaining a first number of CT slice images in which the same detection frame appears continuously according to the order of the CT slice images; when the first number is greater than a first preset number, using the continuous same detection frame as the main branch sequence of the ureteral stent. Figure 3 As shown in the figure, of two adjacent slice images, the first image includes both detection frames A and B, while the next image only includes detection frame A. Detection frame B could be a mistakenly detected detection frame or the detection frames at either end of the stent. As long as detection frame A appears continuously on the CT slice image for a number of times greater than a first preset number, detection frame A is selected as the main branch sequence.

[0069] Optionally, when obtaining a first number of CT slice images in which the same detection frame appears consecutively, the method further includes: obtaining a second number of target CT slice images between two CT slice images in which the same detection frame appears, in the order of the CT slice images; when the second number is not greater than a second preset number, adding the same detection frame to the target CT slice images that do not have the same detection frame. It is understandable that in actual practice, a certain CT slice image should have a detection frame for the main branch, but for various reasons it is not displayed on the CT slice image. In this case, the number of consecutive appearances of the detection frame may not be greater than the first preset number; therefore, when the number of target CT slice images in which the detection frame is missing is small (not greater than the second preset number), the missing detection frame is directly added to the target CT slice image that does not have the detection frame, so that the number of consecutive appearances of the detection frame is greater than the first preset number.

[0070] The main branch sequence obtained in the above steps only includes the main trunk portion of the stent. The circled portions at both ends cannot be displayed in the main branch sequence. Therefore, the main branch sequence needs to be further supplemented. Supplementing the main branch sequence based on the CT sequence images to obtain a complete detection frame sequence includes: obtaining, in the order of the CT slice images, first detection frames that are not in the main branch sequence and appear continuously on the CT slice images, where the number of CT slice images in which the first detection frame appears continuously is greater than a third preset number; and supplementing, in the order of the CT slice images, the first detection frames that intersect with the main branch sequence to the main branch sequence.

[0071] The method further includes: obtaining, in an order reverse to the order of the CT slice images, second detection frames of the main branch sequence that are not supplemented and appear consecutively on the CT slice images, wherein the number of CT slice images in which the second detection frames appear consecutively is greater than a fourth preset number; and supplementing, in an order reverse to the order of the CT slice images, the second detection frames that intersect with the supplemented main branch sequence to the main branch sequence to obtain a complete detection frame sequence. The two supplementations are repeated in reverse order to obtain a detection frame sequence that is identical to the stent as much as possible.

[0072] In practice, there may be two ureteral stents, and the two end points of the ureteral stents are connected to the kidney at one end and the bladder at the other end. Generally, the positions of the two ureteral stents at the bladder intersect on the CT slice image. When actually mapped to the two-dimensional CT slice, there will be many complex small points, and it is impossible to determine which tube they belong to. As an optional implementation method of an embodiment of the present invention, when there are at least two main branch sequences and any two of them intersect, it includes: obtaining a target detection frame of the intersection part; using the target detection frame as a detection frame belonging only to any one of the two intersecting main branch sequences; and deleting the target detection frame from the detection frame belonging to the other main branch sequence.

[0073] Optionally, each CT slice image is input into a two-dimensional neural network to obtain a detection frame of the ureteral stent, including: inputting the CT slice image into a target detection network to extract features of the CT slice image; and generating a region of interest through an RPN based on the extracted features; mapping the coordinates of the region of interest back to the extracted features through an ROIAlign operation; and then obtaining a detection frame of the ureteral stent through an FC layer.

[0074] According to the obtained complete detection frame sequence, the coordinate information of the complete detection frame sequence is obtained; and the position of the bracket is accurately located according to the coordinate information.

[0075] In order to further illustrate the solution of this application, a specific method for obtaining a complete detection frame sequence is provided below. Figure 4 As shown:

[0076] 1. Use Figure 5The object detection network shown performs coarse detection of ureteral stents on a sequence of 2D slices of CT images.

[0077] 1) The processed image data is input into the target detection network, and the CT image features are first extracted through the backbone + FPN feature extraction module. The backbone network selected in the present invention is ResNet34, and the FPN layer obtains four layers of features of different scales.

[0078] 2) Next, the RPN generates a region of interest (ROI), and the coordinates of the ROI are mapped back to the original features through the ROIAlign operation. The features obtained in the ROI region are then connected to two FC layers to regress the detection box coordinates and predict the category.

[0079] 3) Because there is only one category of ureteral stents, only the image name and its corresponding detection box (confidence greater than 0.2) of each image are output.

[0080] 2. Based on the continuity of the CT image sequence, the rough detection results of the neural network are post-processed to obtain a complete ureteral stent detection sequence.

[0081] 1) Process the rough detection results output by the neural network, write the name of an image and all its detection boxes into a dictionary, and write all images of a case into a list, so as to establish case-based data division for subsequent continuity judgment.

[0082] 2) Determine whether the sequence is continuous based on whether the detection boxes of the two images before and after intersect. Use a list stack to store the continuous main branch sequence. Gaps in the continuous sequence with no more than three images will be filled, and sequences with a length of less than 30 in the stack will be deleted.

[0083] 3) For double ureteral stents, there will be two main branch sequences in the stack. To ensure the same subsequent processing as for single ureteral stents, the detection frames of the two main branches are merged in the order of image names to form a single main branch. Specifically, the intersection area of ​​the two ureteral stents in the bladder is merged into the first branch, and the detection frames of the intersection area in the second branch are deleted to ensure that the detection frame sequences of the two ureteral stents do not intersect after processing.

[0084] The main sequence is traversed twice more in the forward and reverse directions, and the continuous secondary branches that intersect with it are added, and finally a complete ureteral stent detection frame sequence is obtained.

[0085] The ureteral stent detection frame sequence provided by the embodiment of the present invention uses a deep neural network to perform coarse detection on the two-dimensional slices of the CT image, and then performs fine detection by screening and completing the target frame through continuity post-processing, and finally obtains a complete ureteral stent detection sequence. The target frame roughly detected by the neural network is judged for continuity, and both single and double ureteral stents are processed to complete the complex endpoint part, and similar features (such as extra-tubular stones) detected in the two-dimensional neural network are screened out. It does not require the use of a three-dimensional neural network to obtain global information on the ureteral stent, which increases the time and difficulty of training. It also solves the problems of false detection and missed detection caused by the lack of continuous information of the previous and next CT slices during two-dimensional image detection.

[0086] In one embodiment, the present invention provides a ureteral stent positioning device, such as Figure 6 Shown, including:

[0087] The image information acquisition module 61 is used to acquire CT sequence image information of the ureteral stent. The CT sequence image information includes each CT slice image and its sequence.

[0088] The detection frame acquisition module 62 is used to input each CT slice image into the two-dimensional neural network to obtain the detection frame of the ureteral stent; specifically, the CT slice image is input into the target detection network to extract the features of the CT slice image; and the extracted features are used to generate the region of interest through the RPN; the coordinates of the region of interest are mapped back to the extracted features through the ROIAlign operation; and the detection frame of the ureteral stent is then obtained through the FC layer.

[0089] The main branch sequence acquisition module 63 is used to obtain the main branch sequence of the ureteral stent based on the order of the CT slice images and the detection frame. Specifically, the main branch sequence acquisition module 63 obtains a first number of CT slice images in which the same detection frame appears continuously in the order of the CT slice images. When the first number is greater than a first preset number, the continuous same detection frame is used as the main branch sequence of the ureteral stent.

[0090] When obtaining a first number of CT slice images in which the same detection frame appears continuously, the method further includes: the main branch sequence acquisition module 63 obtains a second number of target CT slice images between two CT slice images in which the same detection frame appears, according to the order of the CT slice images; and when the second number is not greater than a second preset number, the same detection frame is added to the target CT slice images that do not have the same detection frame.

[0091] In some embodiments, when the main branch sequence acquisition module 63 obtains that there are at least two main branch sequences and any two of them intersect, the main branch sequence acquisition module 63 obtains the target detection frame of the intersection part; uses the target detection frame as the detection frame belonging only to any one of the two intersecting main branch sequences; and deletes the target detection frame from the detection frame belonging to the other main branch sequence.

[0092] The complete detection frame acquisition module 64 is configured to supplement the main branch sequence based on the CT sequence images to obtain a complete detection frame sequence. Specifically, the complete detection frame acquisition module 64 acquires first detection frames that are not part of the main branch sequence and appear continuously on the CT slice images in the order of the CT slice images, where the number of CT slice images in which the first detection frame appears continuously is greater than a third preset number. Furthermore, the complete detection frame acquisition module 64 acquires second detection frames that are not part of the main branch sequence and appear continuously on the CT slice images in the reverse order of the CT slice images, where the number of CT slice images in which the second detection frame appears continuously is greater than a fourth preset number. Furthermore, the complete detection frame acquisition module 64 acquires second detection frames that are ... second detection frame appears continuously on the main branch sequence and appear in the reverse order of the CT slice images, where the second detection frame appears continuously on the main branch sequence and appear in the reverse order of the CT slice images, where the second detection frame appears in the reverse order of the CT slice images, where the second detection frame appears in the reverse order of the CT slice images, where the second detection frame appears in the reverse order of the CT slice images, where the second detection frame appears in the reverse order of the main branch sequence, to obtain a complete detection frame sequence.

[0093] The position determination area 65 is used to determine the position of the ureteral stent according to the complete detection frame sequence.

[0094] An embodiment of the present invention provides a ureteral stent positioning device, wherein an image information acquisition module acquires each CT slice image of the ureteral stent and its sequence; a detection frame acquisition module inputs each CT slice image into a two-dimensional neural network to obtain a detection frame of the ureteral stent; a main branch sequence acquisition module obtains the main branch sequence of the ureteral stent based on the sequence of the CT slice images and the detection frame; a complete detection frame acquisition module supplements the main branch sequence based on the CT sequence images to obtain a complete detection frame sequence; and a position determination area determines the position of the ureteral stent based on the complete detection frame sequence. The device provided by the embodiment of the present invention only uses a two-dimensional neural network to process CT slice images, without the need for three-dimensional slice images, resulting in low computational complexity and low cost. In addition, the main branch sequence is first obtained; then the main branch sequence is supplemented to obtain a complete detection frame sequence, and then the position of the ureteral stent is determined based on the complete detection frame sequence, without the need to map the two-dimensional image to three-dimensional space, thus greatly reducing the computational complexity while ensuring that no information is missed.

[0095] In one embodiment, the present invention provides a computer device comprising:

[0096] processor;

[0097] a memory for storing processor-executable instructions;

[0098] The processor is configured to execute the method provided in the above embodiments.

[0099] The embodiment of the present invention stores executable instructions in a memory. When the processor executes the executable instructions, after obtaining the CT slice images and their sequence of the ureteral stent, it can use a two-dimensional neural network to obtain the detection frame of the ureteral stent in each CT slice image; obtain the main branch sequence of the ureteral stent according to the sequence of the CT slice images and the detection frame; supplement the main branch sequence according to the CT slice images to obtain a complete detection frame sequence, and determine the position of the ureteral stent according to the complete detection frame sequence. The embodiment of the present invention only uses a two-dimensional neural network to process the CT slice images, without the need to use three-dimensional slice images, with low computational complexity and low cost; in addition, the main branch sequence is first obtained; then the main branch sequence is supplemented to obtain a complete detection frame sequence, and then the position of the ureteral stent is determined according to the complete detection frame sequence, without the need to map the two-dimensional image to three-dimensional space, which greatly reduces the computational complexity while ensuring that no information is missed.

[0100] In one embodiment, the present invention provides a robotic surgery system, such as Figure 7 Shown, including:

[0101] A positioning module 71 is used to execute the method provided in the above embodiment to obtain the position of the ureteral stent;

[0102] The surgery module 72 is used to perform surgery based on the position obtained by the positioning module.

[0103] The robotic surgical system in the embodiment of the present invention obtains the position of the ureteral stent according to the positioning module, and then the surgical module performs surgery according to the position, which can accurately and quickly remove or correct the ureteral stent. The positioning module can use a two-dimensional neural network to obtain the detection frame of the ureteral stent in each CT slice image after obtaining the CT slice image and its sequence of the ureteral stent; obtain the main branch sequence of the ureteral stent according to the sequence and detection frame of the CT slice image; supplement the main branch sequence according to the CT slice image to obtain a complete detection frame sequence, and determine the position of the ureteral stent according to the complete detection frame sequence. The embodiment of the present invention only uses a two-dimensional neural network to process the CT slice image, without the need to use a three-dimensional slice image, with a small amount of calculation and low cost; in addition, the main branch sequence is first obtained; then the main branch sequence is supplemented to obtain a complete detection frame sequence, and then the position of the ureteral stent is determined according to the complete detection frame sequence, without the need to map the two-dimensional image to the three-dimensional space, which greatly reduces the amount of calculation while ensuring that no information is missed.

[0104] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0105] It should be noted that, in the description of this application, the terms "first", "second", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance. In addition, in the description of this application, unless otherwise specified, the meaning of "plurality" refers to at least two.

[0106] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present application includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present application belong.

[0107] It should be understood that various parts of the present application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used to implement: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0108] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0109] In addition, the functional units in the various embodiments of the present application may be integrated into a processing module, or each unit may exist physically separately, or two or more units may be integrated into a module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of a software functional module. If the integrated module is implemented in the form of a software functional module and sold or used as an independent product, it may also be stored in a computer-readable storage medium.

[0110] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0111] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0112] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A ureteral stent positioning method, characterized in that: The following steps are involved: Acquiring CT sequence image information of the ureteral stent, wherein the CT sequence image information includes each CT slice image and its sequence; Inputting each CT slice image into a two-dimensional neural network to obtain a detection frame of the ureteral stent; Obtaining a main branch sequence of the ureteral stent according to the order of the CT slice images and the detection frame; Supplementing the main branch sequence according to the CT sequence images to obtain a complete detection frame sequence, including: obtaining, in the order of the CT slice images, first detection frames that are not in the main branch sequence and appear continuously on the CT slice images, wherein the number of CT slice images in which the first detection frame appears continuously is greater than a third preset number; supplementing, in the order of the CT slice images, first detection frames that intersect with the main branch sequence to the main branch sequence; obtaining, in the order opposite to the order of the CT slice images, second detection frames that are not in the main branch sequence and appear continuously on the CT slice images, wherein the number of CT slice images in which the second detection frame appears continuously is greater than a fourth preset number; supplementing, in the order opposite to the order of the CT slice images, the second detection frames that intersect with the supplemented main branch sequence to the main branch sequence to obtain a complete detection frame sequence; The position of the ureteral stent is determined according to the complete detection frame sequence.

2. The method according to claim 1, wherein: Obtaining the main branch sequence of the ureteral stent according to the order of the CT slice images and the detection frame includes: Acquiring a first number of CT slice images in which the same detection frame appears continuously according to the order of the CT slice images; When the first number is greater than a first preset number, the continuous identical detection frames are used as the main branch sequence of the ureteral stent.

3. The method according to claim 2, wherein: The step of obtaining the first number of CT slice images in which the same detection frame appears continuously further includes: acquiring, according to the order of the CT slice images, a second number of target CT slice images between two CT slice images in which the same detection frame appears; When the second number is not greater than a second preset number, the same detection frame is added to the target CT slice image that does not have the same detection frame.

4. The method according to claim 1, wherein: When there are at least two main branch sequences, and any two of them intersect, it includes: Get the target detection frame of the intersection part; The target detection frame is taken as a detection frame belonging only to any one of the two intersecting main branch sequences; and the target detection frame is deleted from the detection frame belonging to the other main branch sequence.

5. The method according to claim 1, wherein: Inputting each CT slice image into a two-dimensional neural network to obtain a detection frame of the ureteral stent includes: Inputting the CT slice image into a target detection network to extract features of the CT slice image; The extracted features are used to generate regions of interest through RPN; Mapping the coordinates of the region of interest back to the extracted features through the ROIAlign operation; Then, the detection frame of the ureteral stent is obtained through the FC layer.

6. A ureteral stent positioning device, characterized in that: include: An image information acquisition module, configured to acquire CT sequence image information of the ureteral stent, wherein the CT sequence image information includes each CT slice image and its sequence; A detection frame acquisition module is used to input each CT slice image into a two-dimensional neural network to obtain a detection frame of the ureteral stent; a main branch sequence acquisition module, configured to obtain a main branch sequence of the ureteral stent according to the order of the CT slice images and the detection frame; a complete detection frame acquisition module, configured to supplement the main branch sequence according to the CT sequence image to obtain a complete detection frame sequence; specifically, to acquire, in accordance with the order of the CT slice images, first detection frames of non-main branch sequences that appear consecutively on the CT slice images, wherein the number of CT slice images in which the first detection frame appears consecutively is greater than a third preset number; In accordance with the order of the CT slice images, first detection frames intersecting with the main branch sequence are added to the main branch sequence; and second detection frames of the main branch sequence that appear continuously on the CT slice images but are not added are acquired in an order reverse to the order of the CT slice images, wherein the number of CT slice images in which the second detection frames appear continuously is greater than a fourth preset number. In an order opposite to the order of the CT slice images, the second detection frame that intersects with the supplemented main branch sequence is supplemented to the main branch sequence to obtain a complete detection frame sequence; A position determination area is used to determine the position of the ureteral stent according to the complete detection frame sequence.

7. A computer device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the method according to any one of claims 1 to 5.

8. A robotic surgery system, characterized in that: include: a positioning module, configured to execute the method according to any one of claims 1 to 5 to obtain the position of the ureteral stent; A surgery module is used to perform surgery according to the position obtained by the positioning module.

Images