Interventional puncture ultrasonic-assisted positioning method and system

Abstract

The invention relates to the technical field of puncture needle movement evaluation, in particular to an interventional puncture ultrasonic assisted positioning method and system.The method comprises the steps that in the interventional puncture process, ultrasonic images are collected in real time through an ultrasonic probe, and an optical flow vector corresponding to each pixel point in each frame of ultrasonic image is determined; carrying out clustering on pixel points in each frame of ultrasonic image; determining the matching degree between each target cluster and the target cluster in the next frame of ultrasonic image of the ultrasonic image to which the target cluster belongs; screening out mutually matched target clusters from all the ultrasonic images; determining a puncture needle possible index corresponding to each target cluster sequence; screening out a puncture needle region sequence, and screening out a needle point position from each target cluster in the puncture needle region sequence; and evaluating the needle tip movement direction and the needle tip displacement of the puncture needle at the next moment. According to the invention, interventional puncture ultrasonic positioning tracking is realized, and the accuracy of evaluating and positioning the next movement condition of the puncture needle is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of puncture needle movement assessment, and in particular to an interventional puncture ultrasound-assisted positioning method and system. Background Art

[0002] Interventional puncture is a medical procedure that involves puncturing the skin, mucous membranes, or other tissues to insert an instrument for diagnosis or treatment. During interventional puncture, the doctor often observes ultrasound images captured at the current moment to determine the current position of the puncture needle, thereby assessing its movement in the next moment and achieving needle positioning and tracking.

[0003] However, when doctors observe ultrasound images collected at the current moment and evaluate the movement of the puncture needle at the next moment, the following technical problems often arise: During the interventional puncture process, due to various reasons such as poor contact of the ultrasound probe, it is often difficult to ensure that all collected ultrasound images are clear. Therefore, if the ultrasound image collected at the current moment is unclear, it is often difficult for the doctor to accurately evaluate the situation of the puncture needle at the next moment based on the ultrasound image collected at the current moment, resulting in poor accuracy in evaluating the movement of the puncture needle at the next moment, and thus poor accuracy in interventional puncture ultrasound positioning tracking. Summary of the Invention

[0004] In order to solve the technical problem of poor accuracy of interventional puncture ultrasound positioning and tracking due to poor accuracy in evaluating the next moment movement of the puncture needle, the present invention proposes an interventional puncture ultrasound-assisted positioning method and system.

[0005] In a first aspect, the present invention provides an interventional puncture ultrasound-assisted positioning method, the method comprising: During the interventional puncture process, ultrasound images are collected in real time through the ultrasound probe, and the optical flow vector corresponding to each pixel in each frame of ultrasound image is determined based on the movement between each frame of ultrasound image and the next frame of ultrasound image; According to the similarity between the optical flow vectors corresponding to the pixels in each frame of ultrasound image and the distance between the pixels, the pixels in each frame of ultrasound image are clustered to obtain the target cluster; Determine the degree of matching between each target cluster and the target cluster in the next frame of the ultrasound image to which it belongs based on the distribution of non-overlapping pixels between each target cluster and the target cluster in the next frame of the ultrasound image to which it belongs, as well as the center of each target cluster and the optical flow vectors corresponding to the pixels in each target cluster; Based on the matching degree between target clusters, target clusters that match each other are screened out from all ultrasound images to form a target cluster sequence; According to the matching degree between adjacent target clusters in each target cluster sequence, the similarity of optical flow vectors between adjacent target clusters, and the shape characteristics of the target clusters, the possible puncture needle indicators corresponding to each target cluster sequence are determined; Based on all possible indicators of the puncture needle, the puncture needle region sequence is screened out from all target cluster sequences, and the needle tip position is screened out from each target cluster in the puncture needle region sequence; Based on the distance between the needle tip positions in adjacent puncture needle regions in the puncture needle region sequence and the optical flow vector corresponding to the pixel points in the puncture needle region, the needle tip movement direction and needle tip displacement of the next moment are evaluated.

[0006] In combination with the first aspect above, in one possible implementation, clustering the pixels in each frame of ultrasound image based on the similarity between the optical flow vectors corresponding to the pixels in each frame of ultrasound image and the distance between the pixels to obtain the target cluster includes: The normalized value of the cosine similarity between the optical flow vectors corresponding to each two pixel points in each frame of ultrasound image is determined as the vector similarity factor between each two pixel points in each frame of ultrasound image; The absolute value of the difference between the moduli of the optical flow vectors corresponding to every two pixel points in each frame of ultrasound image is determined as the motion distance factor between every two pixel points in each frame of ultrasound image; The consistency coefficient between each two pixels in each frame of ultrasound image is determined based on the vector similarity factor, motion distance factor and Euclidean distance between each two pixels in each frame of ultrasound image. The vector similarity factor is positively correlated with the consistency coefficient, while the motion distance factor and Euclidean distance are negatively correlated with the consistency coefficient. According to the consistency coefficient between all pixels in each frame of ultrasound image, the pixels in each frame of ultrasound image are clustered, and each obtained cluster is recorded as a target cluster.

[0007] In conjunction with the first aspect above, in one possible implementation, determining the degree of match between each target cluster and the target cluster in the next frame of ultrasound image to which it belongs based on the distribution of non-overlapping pixels between each target cluster and the target cluster in the next frame of ultrasound image to which it belongs, and the center of each target cluster and the optical flow vectors corresponding to the pixels within each target cluster, includes: Determine any target cluster as a first marker cluster, and determine any target cluster in a next frame of ultrasound image to which the first marker cluster belongs as a second marker cluster; Determine an area in the next frame of ultrasound image to which the first marker cluster belongs, which is located at the same position as the first marker cluster, as a reference area, and determine the center of the reference area as a reference center; Determine the intersection of the second marker cluster and the reference area as a reference intersection area, and determine the center of the area formed by all pixels in the second marker cluster except the reference intersection area as a candidate center, wherein the pixels in the second marker cluster except the reference intersection area represent non-overlapping pixels between the first marker cluster and the second marker cluster; The direction in which the reference center points to the candidate center is determined as a reference movement direction; Determining the mean of the optical flow vectors corresponding to all pixels in the first marker cluster as a candidate motion vector, and determining the direction of the candidate motion vector as a candidate motion direction; The degree of matching between the first marker cluster and the second marker cluster is determined based on the angle between the reference motion direction and the candidate motion direction, the Euclidean distance between the reference center and the center of the second marker cluster, and the number of pixels in the reference intersection area.

[0008] In conjunction with the first aspect above, in a possible implementation, the formula corresponding to the matching degree between the first marker cluster and the second marker cluster is: ; Wherein, D is the degree of match between the first marker cluster and the second marker cluster; is the normalization function; n is the number of pixels in the reference intersection area; is a natural exponential function; is the angle between the reference motion direction and the candidate motion direction; d is the Euclidean distance between the reference center and the center of the second marker cluster.

[0009] In combination with the first aspect above, in one possible implementation, determining the possible puncture needle indicators corresponding to each target cluster sequence based on the matching degree between adjacent target clusters in each target cluster sequence, the similarity of optical flow vectors between adjacent target clusters, and the shape characteristics of the target clusters includes: According to the matching degree between adjacent target clusters in each target cluster sequence and the difference between the directions of the optical flow vectors between adjacent target clusters, the direction consistency corresponding to each target cluster sequence is determined; The Euclidean distance between each pixel point in each target cluster and the center of the target cluster to which it belongs is determined as the deviation distance corresponding to each pixel point in each target cluster; The variance of the deviation distances corresponding to all pixels in each target cluster is determined as the shape feature index corresponding to each target cluster; According to the directional consistency corresponding to each target cluster sequence and the shape characteristic indices corresponding to all target clusters within each target cluster sequence, the possible puncture needle index corresponding to each target cluster sequence is determined. Among them, the directional consistency and shape characteristic indices are both positively correlated with the possible puncture needle index.

[0010] In combination with the first aspect above, in a possible implementation, the formula corresponding to the directional consistency of the target cluster sequence is: ;in, is the directional consistency corresponding to the rth target cluster sequence; r is the sequence number of the target cluster sequence; is the normalization function; is the number of target clusters in the r-th target cluster sequence; t is the sequence number of the target cluster in the r-th target cluster sequence; is the degree of matching between the tth target cluster and the t+1th target cluster in the rth target cluster sequence; It is the cosine similarity between the motion representative vector corresponding to the t-th target cluster and the motion representative vector corresponding to the t+1-th target cluster in the r-th target cluster sequence; the motion representative vector corresponding to the target cluster is the mean of the optical flow vectors corresponding to all pixels in the target cluster.

[0011] In combination with the first aspect above, in one possible implementation, determining the possible puncture needle index corresponding to each target cluster sequence based on the directional consistency corresponding to each target cluster sequence and the shape feature indexes corresponding to all target clusters within each target cluster sequence includes: The cumulative value of the shape feature indices corresponding to all target clusters in each target cluster sequence is determined as the possible shape factor of the puncture needle corresponding to each target cluster sequence; The product of the directional consistency and the possible factor of the puncture needle shape corresponding to each target cluster sequence is normalized to obtain the possible index of the puncture needle corresponding to each target cluster sequence.

[0012] In combination with the first aspect above, in a possible implementation, screening out the needle tip position from each target cluster in the puncture needle region sequence includes: Any target cluster in the puncture needle region sequence is determined as a marking region, and a marking ray is constructed with the center of the marking region as the starting point and the direction of the mean value of the optical flow vectors corresponding to all pixels in the marking region as the extension direction; The pixel point farthest from the center of the marking area is selected from the intersection of the marking ray and the marking area as the needle tip pixel point, and the position of the needle tip pixel point is determined as the needle tip position.

[0013] In combination with the first aspect above, in a possible implementation, the formulas corresponding to the needle tip movement direction and needle tip displacement of the puncture needle at the next moment are: ; ; Wherein, the direction of R is the estimated direction of movement of the puncture needle tip at the next moment; L is the estimated displacement of the puncture needle tip at the next moment; M is the number of puncture needle areas in the puncture needle area sequence; m is the sequence number of the puncture needle area in the puncture needle area sequence; Q is the cumulative value of the sequence numbers of all puncture needle areas in the puncture needle area sequence; is the mean of the optical flow vectors corresponding to all pixels in the mth puncture needle area in the puncture needle area sequence; is the distance between the needle tip position in the mth puncture needle region and the needle tip position in the m+1th puncture needle region in the puncture needle region sequence.

[0014] In a second aspect, the present invention provides an interventional puncture ultrasound-assisted positioning system, the system comprising: An image acquisition vector determination module is used to acquire ultrasound images in real time using an ultrasound probe during interventional puncture, and determine the optical flow vector corresponding to each pixel in each ultrasound image frame based on the motion between each ultrasound image frame and the next ultrasound image frame; The pixel clustering module is used to cluster the pixels in each frame of ultrasound image according to the similarity between the optical flow vectors corresponding to the pixels in each frame of ultrasound image and the distance between the pixels to obtain the target cluster; a matching degree determination module, configured to determine a matching degree between each target cluster and a target cluster in a next frame of ultrasound image to which it belongs, based on a distribution of non-overlapping pixels between each target cluster and a target cluster in a next frame of ultrasound image to which it belongs, and an optical flow vector corresponding to a center of each target cluster and pixels within each target cluster; A target cluster screening module is used to screen out mutually matching target clusters from all ultrasound images based on the matching degree between target clusters to form a target cluster sequence; The puncture needle possible index determination module is used to determine the puncture needle possible index corresponding to each target cluster sequence based on the matching degree between adjacent target clusters in each target cluster sequence, the similarity of optical flow vectors between adjacent target clusters, and the shape characteristics of the target clusters; A needle tip position screening module is used to screen out a puncture needle region sequence from all target cluster sequences based on all possible puncture needle indicators, and to screen out a needle tip position from each target cluster in the puncture needle region sequence; The needle tip motion direction displacement evaluation module is used to evaluate the needle tip motion direction and needle tip displacement of the next moment based on the distance between the needle tip positions in adjacent puncture needle areas in the puncture needle area sequence and the optical flow vector corresponding to the pixel points in the puncture needle area.

[0015] In a third aspect, a server is provided, comprising a memory and a processor. The memory is configured to store executable program code, and the processor is configured to call and execute the executable program code from the memory, so that the device executes the method of the first aspect or any possible implementation of the first aspect.

[0016] In a fourth aspect, a computer program product is provided, comprising: a computer program code, which, when executed on a computer, enables the computer to execute the method in the first aspect or any possible implementation of the first aspect.

[0017] In a fifth aspect, a computer-readable storage medium is provided, which stores a computer program code. When the computer program code runs on a computer, the computer executes the method in the above-mentioned first aspect or any possible implementation of the first aspect.

[0018] The present invention has the following beneficial effects: The present invention provides an interventional puncture ultrasound-assisted positioning method, which realizes interventional puncture ultrasound positioning and tracking, solves the technical problem of poor accuracy in evaluating the movement of the puncture needle at the next moment, and improves the accuracy of evaluating the movement of the puncture needle at the next moment. Compared with the doctor observing the ultrasound image collected at the current moment to evaluate the movement of the puncture needle at the next moment, the present invention comprehensively considers multiple frames of ultrasound images when evaluating the movement of the puncture needle at the next moment, to a certain extent reducing the adverse effects of the difficulty in accurately positioning the puncture needle due to the unclear ultrasound image of a certain frame, and quantifying the matching degree and possible indicators of the puncture needle, thereby realizing the positioning of the needle tip position, and then relatively objectively realizing the evaluation of the needle tip movement direction and needle tip displacement of the puncture needle at the next moment, thereby improving the accuracy of evaluating the movement of the puncture needle at the next moment, and then improving the accuracy of interventional puncture ultrasound positioning and tracking. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0020] Figure 1 This is a flow chart of an interventional puncture ultrasound-assisted positioning method of the present invention; Figure 2 This is a schematic diagram of the structure of an interventional puncture ultrasound-assisted positioning system of the present invention; Figure 3 The figure is a structural diagram of a computer device of the present invention. DETAILED DESCRIPTION

[0021] To further illustrate the technical means and effects employed by the present invention to achieve its intended objectives, the following, in conjunction with the accompanying drawings and preferred embodiments, describes in detail the specific implementations, structures, features, and effects of the technical solutions proposed by the present invention. In the following description, references to "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.

[0022] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0023] refer to Figure 1 , shows the process of some embodiments of an interventional puncture ultrasound-assisted positioning method of the present invention. The interventional puncture ultrasound-assisted positioning method includes the following steps: Step S1: During the interventional puncture process, ultrasound images are collected in real time through an ultrasound probe, and the optical flow vector corresponding to each pixel point in each frame of ultrasound image is determined based on the movement between each frame of ultrasound image and the next frame of ultrasound image.

[0024] An ultrasound probe is a device that transmits and receives ultrasound waves during ultrasonic testing and can be used to capture ultrasound images. An optical flow vector is the temporal motion vector of a pixel between two consecutive image frames, typically expressed as a vector.

[0025] It should be noted that interventional puncture is a medical procedure that involves puncturing the human body through the skin, mucous membranes, or other tissues to insert an instrument for diagnosis or treatment. It is usually used to locate, sample, treat lesions, or guide other medical devices into the target area. It has the advantages of less trauma and rapid recovery. This technology is widely used in the fields of tumors, visceral diseases, vascular lesions, etc. Ultrasound imaging guidance is a commonly used external guidance method for interventional puncture. Before the interventional puncture, ultrasound images are used to locate and mark the target area to plan the puncture route. The ultrasound image is then used to guide the puncture needle into the target area for sampling or injection. The real-time imaging function of ultrasound can help medical staff accurately locate the target area during the operation to avoid damage to surrounding important structures.

[0026] For example, a doctor can use their experience to place an ultrasound probe at an appropriate location on the patient's skin surface. During the interventional puncture, the ultrasound probe can be used to capture ultrasound images in real time. This means that multiple frames of ultrasound images may have been captured before the current moment. The LK (Lucas-Kanade) optical flow algorithm can then be used to compare the motion of each pixel in each ultrasound image frame with the corresponding pixel in the next ultrasound image frame, generating the optical flow vector corresponding to each pixel in each ultrasound image frame.

[0027] It should be noted that the ultrasound images used in the subsequent calculations can be all ultrasound images acquired in real time during the interventional puncture process. To improve computational efficiency, the ultrasound images used in the subsequent calculations can also be a preset number of ultrasound images acquired recently. The preset number can be a pre-set number, such as 100.

[0028] Step S2: clustering the pixels in each frame of ultrasound image according to the similarity between the optical flow vectors corresponding to the pixels in each frame of ultrasound image and the distance between the pixels to obtain a target cluster.

[0029] It should be noted that within the same image frame, the distribution of the motion directions of pixels belonging to the same structure often exhibits a certain regularity. For structures within the human body that move, such as tissues that undergo peristalsis, contraction, and expansion, the motion directions of pixels within the same structure often do not differ significantly during these movements; that is, the motion directions of pixels within the same structure are often similar. For a puncture needle, a straight, rigid object, the motion direction and distance of all pixels on it often remain consistent during insertion. Therefore, clustering the motion directions and distances between pixels within the same image frame can identify pixels belonging to different tissue structures. The direction of the optical flow vector corresponding to a pixel can, to a certain extent, characterize the pixel's motion direction. The modulus of the optical flow vector corresponding to a pixel can, to a certain extent, characterize the pixel's motion distance. A target cluster can represent a tissue structure or puncture needle.

[0030] As an example, this step may include the following steps: In the first step, the normalized value of the cosine similarity between the optical flow vectors corresponding to every two pixel points in each frame of ultrasound image is determined as the vector similarity factor between every two pixel points in each frame of ultrasound image.

[0031] In the second step, the absolute value of the difference between the moduli of the optical flow vectors corresponding to every two pixel points in each frame of the ultrasound image is determined as the motion distance factor between every two pixel points in each frame of the ultrasound image.

[0032] The third step is to determine the consistency coefficient between every two pixel points in each frame of ultrasound image based on the vector similarity factor, motion distance factor and Euclidean distance between every two pixel points in each frame of ultrasound image.

[0033] Among them, the vector similarity factor can be positively correlated with the consistency coefficient, and the motion distance factor and the Euclidean distance can both be negatively correlated with the consistency coefficient.

[0034] For example, the formula for determining the consistency coefficient between two pixels in the same frame of ultrasound image can be: ; ;in, is the consistency coefficient between the i-th pixel and the j-th pixel in the f-th frame of the ultrasound image. f is the frame number of the ultrasound image. i and j are the serial numbers of different pixels in the f-th frame of the ultrasound image. is the normalization function. is the vector similarity factor between the i-th pixel and the j-th pixel in the f-th frame ultrasound image. It is the cosine similarity between the optical flow vector corresponding to the i-th pixel and the optical flow vector corresponding to the j-th pixel in the f-th frame ultrasound image. is a natural exponential function. is the Euclidean distance between the i-th pixel and the j-th pixel in the f-th frame of ultrasound image. It is the absolute value function. It is the modulus of the optical flow vector corresponding to the j-th pixel in the f-th frame of ultrasound image. It is the modulus of the optical flow vector corresponding to the i-th pixel in the f-th frame of ultrasound image. It is the motion distance factor between the i-th pixel and the j-th pixel in the f-th frame ultrasound image.

[0035] It should be noted that, in actual situations, the smaller the distance between pixels in the same frame of ultrasound image, the more likely these pixels are to belong to the same tissue structure or puncture needle area. The smaller the value, the smaller the Euclidean distance between the i-th pixel and the j-th pixel, which often means that the i-th pixel and the j-th pixel are more likely to belong to the same tissue structure or puncture needle area. In actual practice, during the puncture needle, the movement direction and movement distance of all pixels on it are often consistent. When is larger, it often means that the optical flow vector corresponding to the i-th pixel point is more similar to the optical flow vector corresponding to the j-th pixel point, and often means that the movement direction and movement distance between the i-th pixel point and the j-th pixel point are closer, and often means that the i-th pixel point and the j-th pixel point are more likely to belong to the same tissue structure or puncture needle area. The smaller it is, the closer the movement distance between the i-th pixel and the j-th pixel is, which often means that the i-th pixel and the j-th pixel are more likely to belong to the puncture needle area. The larger it is, the more likely it is that the i-th pixel and the j-th pixel belong to the same tissue structure or puncture needle area.

[0036] The fourth step is to cluster the pixels in each frame of ultrasound image according to the consistency coefficient between all pixels in each frame of ultrasound image, and record each cluster obtained as a target cluster.

[0037] For example, pixels in the same frame of ultrasound image whose consistency coefficient is greater than a preset consistency threshold may be grouped into the same cluster and recorded as a target cluster, wherein the preset consistency threshold may be a pre-set threshold, which may be 0.7.

[0038] Step S3, determining the degree of matching between each target cluster and the target cluster in the next frame of the ultrasound image to which it belongs based on the distribution of non-overlapping pixel points between each target cluster and the target cluster in the next frame of the ultrasound image to which it belongs, and the center of each target cluster and the optical flow vector corresponding to the pixel points in each target cluster.

[0039] As an example, this step may include the following steps: In the first step, any target cluster is determined as a first marker cluster, and any target cluster in the next frame of ultrasound image to which the first marker cluster belongs is determined as a second marker cluster.

[0040] In the second step, the region in the next frame of ultrasound image to which the first marker cluster belongs, which is located at the same position as the first marker cluster, is determined as a reference region, and the center of the reference region is determined as a reference center.

[0041] In the third step, the intersection of the second marker cluster and the reference area is determined as the reference intersection area, and the center of the area formed by all pixels in the second marker cluster except the reference intersection area is determined as the candidate center.

[0042] The pixels in the second marker cluster except for the reference intersection area may represent non-overlapping pixels between the first marker cluster and the second marker cluster.

[0043] The fourth step is to point the reference center in the direction of the candidate center and determine it as the reference movement direction.

[0044] It should be noted that the reference motion direction can represent the motion direction between the first marker cluster and the second marker cluster to a certain extent.

[0045] In the fifth step, the mean value of the optical flow vectors corresponding to all the pixels in the first marked cluster is determined as the candidate motion vector, and the direction of the candidate motion vector is determined as the candidate motion direction.

[0046] It should be noted that the candidate motion direction may represent the motion direction of the first marker cluster.

[0047] In the sixth step, the degree of matching between the first marker cluster and the second marker cluster is determined based on the angle between the reference motion direction and the candidate motion direction, the Euclidean distance between the reference center and the center of the second marker cluster, and the number of pixels in the reference intersection area.

[0048] For example, the formula for determining the degree of matching between the first marker cluster and the second marker cluster may be: ; Where D is the degree of match between the first marker cluster and the second marker cluster. is a normalization function. n is the number of pixels in the reference intersection region. is a natural exponential function. is the angle between the reference motion direction and the candidate motion direction. d is the Euclidean distance between the reference center and the center of the second marker cluster.

[0049] It should be noted that for the same tissue structure in the human body, its position is often relatively fixed. Therefore, the same tissue structure is often located at the same position in different frames of ultrasound images. During the puncture process, the movement speed of the puncture needle is relatively slow, and the time length between adjacent frames of ultrasound images is often short. Therefore, the position of the puncture needle in adjacent frames of ultrasound images is often relatively close. When n is larger, it often means that there are more pixels in the same position in the first marker cluster and the second marker cluster, which often means that the positions of the first marker cluster and the second marker cluster in the corresponding ultrasound image are closer, which often means that the first marker cluster and the second marker cluster are more likely to represent the same tissue structure or puncture needle. When d is smaller, it often means that the positions of the first marker cluster and the second marker cluster in the corresponding ultrasound image are closer, which often means that the first marker cluster and the second marker cluster are more likely to represent the same tissue structure or puncture needle. When The smaller D is, the more likely the movement direction between the first and second marker clusters is to be the same as the first marker cluster's. This also indicates that the second marker cluster is more likely to be the first marker cluster after movement, and the more likely the first and second marker clusters are to represent the same tissue structure or puncture needle. Therefore, the larger D is, the more likely the first and second marker clusters are to represent the same tissue structure or puncture needle.

[0050] Step S4: Based on the matching degree between target clusters, mutually matching target clusters are screened out from all ultrasound images to form a target cluster sequence.

[0051] The target cluster sequence can include regions corresponding to the same tissue structure or puncture needle in different ultrasound images. The number of target cluster sequences can be equal to the number of target clusters in each ultrasound image frame. The earlier a target cluster is in the target cluster sequence, the earlier the ultrasound image to which it corresponds was acquired.

[0052] As an example, if there are 3 frames of ultrasound images, any target cluster in the first frame of ultrasound image can be determined as the first temporary cluster; the target cluster with the greatest degree of match between the first temporary cluster and the second frame of ultrasound image is screened out and determined as the second temporary cluster; the target cluster with the greatest degree of match between the second temporary cluster and the third frame of ultrasound image is screened out and determined as the third temporary cluster; the first temporary cluster, the second temporary cluster and the third temporary cluster can be formed into a target cluster sequence.

[0053] Step S5, determining the possible puncture needle index corresponding to each target cluster sequence according to the matching degree between adjacent target clusters in each target cluster sequence, the similarity of the optical flow vectors between adjacent target clusters, and the shape characteristics of the target clusters.

[0054] As an example, this step may include the following steps: In the first step, the directional consistency corresponding to each target cluster sequence is determined based on the matching degree between adjacent target clusters in each target cluster sequence and the similarity of optical flow vectors between adjacent target clusters.

[0055] For example, the formula for determining the directional consistency corresponding to the target cluster sequence can be: ;in, is the directional consistency corresponding to the rth target cluster sequence. r is the sequence number of the target cluster sequence. is the normalization function. is the number of target clusters in the rth target cluster sequence. t is the sequence number of the target cluster in the rth target cluster sequence. is the degree of matching between the tth target cluster and the t+1th target cluster in the rth target cluster sequence. is the cosine similarity between the motion representative vector corresponding to the tth target cluster and the motion representative vector corresponding to the t+1th target cluster in the rth target cluster sequence. The motion representative vector corresponding to the target cluster can be the mean of the optical flow vectors corresponding to all pixels in the target cluster.

[0056] It should be noted that Can be used as The weight of The larger the value, the better the calculated In actual situations, the insertion position and angle of the puncture needle are usually determined in the initial planning of the interventional puncture surgery. In the middle of the interventional puncture surgery, the direction may be slightly adjusted to avoid dangerous areas such as blood vessels and nerves. The puncture needle itself is a straight object with a certain hardness. In addition, according to the surgical principles and the limitations of actual operations, the movement direction of the puncture needle will not change much during the entire interventional surgery. When is larger, it often means that the motion representative vector corresponding to the t-th target cluster is closer to the motion representative vector corresponding to the t+1-th target cluster, and often means that the motion direction between the t-th target cluster and the t+1-th target cluster is closer, and often means that the motion between the t-th target cluster and the t+1-th target cluster is more consistent with the motion of the puncture needle. The larger it is, the more likely the rth target cluster sequence is to represent the puncture needle.

[0057] In the second step, the Euclidean distance between each pixel point in each target cluster and the center of the target cluster to which it belongs is determined as the deviation distance corresponding to each pixel point in each target cluster.

[0058] In the third step, the variance of the deviation distances corresponding to all pixels in each target cluster is determined as the shape feature index corresponding to each target cluster.

[0059] It should be noted that because a puncture needle is a linear object with a certain degree of hardness, the distance distribution between all pixels on the puncture needle and its center point is often relatively discrete. Therefore, when the shape feature index corresponding to the target cluster is larger, it often indicates that the distance distribution between all pixels in the target cluster and its center point is relatively discrete, which often indicates that the target cluster is more likely to represent the puncture needle region.

[0060] The fourth step is to determine the possible puncture needle indicators corresponding to each target cluster sequence based on the directional consistency corresponding to each target cluster sequence and the shape feature indicators corresponding to all target clusters in each target cluster sequence.

[0061] Among them, the directional consistency and shape characteristic indices can be positively correlated with the possible indices of the puncture needle.

[0062] For example, determining the possible puncture needle indicators corresponding to each target cluster sequence may include the following sub-steps: In the first sub-step, the cumulative value of the shape feature indices corresponding to all target clusters in each target cluster sequence is determined as the possible puncture needle shape factor corresponding to each target cluster sequence.

[0063] In the second sub-step, the product of the directional consistency and the puncture needle shape possibility factor corresponding to each target cluster sequence is normalized to obtain the puncture needle possibility index corresponding to each target cluster sequence.

[0064] For example, the formula for determining the possible indicators of the puncture needle corresponding to the target cluster sequence can be: ;in, is the possible index of the puncture needle corresponding to the rth target cluster sequence. r is the sequence number of the target cluster sequence. is the normalization function. is the directional consistency corresponding to the r-th target cluster sequence. is the possible factor of the puncture needle shape corresponding to the r-th target cluster sequence.

[0065] It should be noted that when The larger the value, the more likely the rth target cluster sequence is to represent the puncture needle. When the value is larger, it often indicates that the distance distribution between all pixels in the rth target cluster sequence and its center point is relatively more discrete, which often indicates that the rth target cluster sequence is more likely to represent the puncture needle area. The larger it is, the more likely the rth target cluster sequence is to represent the puncture needle area.

[0066] Step S6: Based on all possible puncture needle indicators, a puncture needle region sequence is screened out from all target cluster sequences, and a needle tip position is screened out from each target cluster in the puncture needle region sequence.

[0067] As an example, this step may include the following steps: In the first step, the target cluster sequence with the largest possible puncture needle index is selected from all target cluster sequences as the puncture needle region sequence.

[0068] In the second step, any target cluster in the puncture needle region sequence is determined as the marking region. The marking ray is constructed with the center of the marking region as the starting point and the direction of the mean of the optical flow vectors corresponding to all pixels in the marking region as the extension direction.

[0069] In the third step, the pixel point farthest from the center of the marked area is selected from the intersection of the marked ray and the marked area as the needle tip pixel point, and the position of the needle tip pixel point is determined as the needle tip position.

[0070] It should be noted that, in actual situations, unclear ultrasound images collected due to poor contact of the ultrasound probe and other reasons are often relatively few, and when evaluating the needle tip movement direction and needle tip displacement of the puncture needle at the next moment, the embodiment of the present invention not only considers the ultrasound image collected at the current moment, but also considers the ultrasound images collected at multiple frames of historical moments. That is, when more ultrasound images are considered at the same time, the adverse effects caused by the unclear single-frame ultrasound image can be reduced to a certain extent. Therefore, the subsequent evaluation of the needle tip movement direction and needle tip displacement of the puncture needle at the next moment is based on the distance between the needle tip positions of adjacent puncture needle areas in the puncture needle area sequence and the optical flow vector corresponding to the pixel points in the puncture needle area, which can improve the accuracy of the evaluation of the puncture needle's movement at the next moment to a certain extent.

[0071] Step S7: Evaluate the needle tip movement direction and needle tip displacement of the next moment based on the distances between the needle tip positions in adjacent puncture needle regions in the puncture needle region sequence and the optical flow vectors corresponding to the pixels in the puncture needle regions.

[0072] It should be noted that the needle tip movement direction and needle tip displacement of the puncture needle at the next moment are evaluated to realize the positioning of the puncture needle at the next moment, thereby realizing the positioning tracking of the puncture needle, which can assist the doctor in judging the position of the puncture needle at the next moment, and thus make it easier for the doctor to judge whether the puncture needle movement needs to be fine-tuned.

[0073] As an example, the formulas for evaluating the needle tip movement direction and needle tip displacement at the next moment can be: ; Where R is the estimated direction of the needle tip's movement at the next moment. L is the estimated displacement of the needle tip at the next moment. M is the number of puncture needle regions in the puncture needle region sequence. m is the sequence number of the puncture needle region in the puncture needle region sequence. Q is the cumulative value of the sequence numbers of all puncture needle regions in the puncture needle region sequence. It is the mean of the optical flow vectors corresponding to all pixels in the mth puncture needle area in the puncture needle area sequence. is the distance between the needle tip position in the mth puncture needle region and the needle tip position in the m+1th puncture needle region in the puncture needle region sequence. The distance between the needle tip position in the mth puncture needle region and the needle tip position in the m+1th puncture needle region may be obtained by screening out a position identical to the needle tip position in the mth puncture needle region from the ultrasound image corresponding to the m+1th puncture needle region, recording the position as the reference position, and determining the Euclidean distance between the needle tip position in the m+1th puncture needle region and the reference position as the distance between the needle tip position in the mth puncture needle region and the needle tip position in the m+1th puncture needle region.

[0074] It should be noted that, in practice, the puncture needle often moves continuously during the puncture process, and its previous movement often affects its subsequent movement. Therefore, the needle's movement at the next moment can be estimated based on its movement in the acquired ultrasound image. This movement, to a certain extent, reflects the needle's expected position at the next moment. A smaller m indicates that the ultrasound image corresponding to the puncture needle region was acquired earlier, which generally indicates a smaller impact on the next moment's movement. The value range of is (0, 1), which can be used as and The weight of . It can represent the movement of the mth puncture needle area. The larger its value is, the greater the movement of the puncture needle in the ultrasound image of the mth puncture needle area is. It often means that it is more likely to affect the movement of the puncture needle in the next moment. Therefore, the direction of R can represent the movement direction of the puncture needle tip to a certain extent. A larger value indicates a greater distance between the needle tip positions in the mth puncture needle region and the m+1th puncture needle region, which in turn indicates a greater needle displacement between two adjacent ultrasound image frames. This indicates a greater likelihood that the needle tip will experience a larger displacement in the next moment. Therefore, L can, to a certain extent, characterize the needle tip displacement.

[0075] refer to Figure 2 Based on the same inventive concept as the above-mentioned method embodiment, the present invention provides an interventional puncture ultrasound-assisted positioning system, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the steps of an interventional puncture ultrasound-assisted positioning method may specifically include: The image acquisition vector determination module 201 is used to acquire ultrasound images in real time using an ultrasound probe during an interventional puncture procedure, and determine the optical flow vector corresponding to each pixel in each ultrasound image frame based on the motion between each ultrasound image frame and the next ultrasound image frame; The pixel clustering module 202 is configured to cluster the pixels in each frame of ultrasound image according to the similarity between the optical flow vectors corresponding to the pixels in each frame of ultrasound image and the distance between the pixels to obtain a target cluster; a matching degree determination module 203 for determining a matching degree between each target cluster and a target cluster in the next frame of ultrasound image to which it belongs, based on the distribution of non-overlapping pixels between each target cluster and the target cluster in the next frame of ultrasound image to which it belongs, and the center of each target cluster and the optical flow vectors corresponding to the pixels in each target cluster; The target cluster screening module 204 is configured to screen out mutually matching target clusters from all ultrasound images based on the matching degree between target clusters to form a target cluster sequence; The puncture needle possible index determination module 205 is used to determine the puncture needle possible index corresponding to each target cluster sequence based on the matching degree between adjacent target clusters in each target cluster sequence, the similarity of optical flow vectors between adjacent target clusters, and the shape characteristics of the target clusters; The needle tip position screening module 206 is configured to screen out a puncture needle region sequence from all target cluster sequences based on all possible puncture needle indicators, and to screen out a needle tip position from each target cluster in the puncture needle region sequence; The needle tip movement direction displacement evaluation module 207 is used to evaluate the needle tip movement direction and needle tip displacement of the puncture needle at the next moment based on the distance between the needle tip positions in adjacent puncture needle areas in the puncture needle area sequence and the optical flow vectors corresponding to the pixels in the puncture needle areas.

[0076] Figure 3 FIG. 1 is a schematic diagram of the structure of a computer device provided by an embodiment of the present invention. For example, Figure 3 As shown, the computer device 300 includes: a memory 301, a processor 302, and a computer program 303 stored in the memory 301 and running on the processor 302, wherein when the processor 302 executes the computer program 303, the computer device can execute any one of the interventional puncture ultrasound-assisted positioning methods introduced above.

[0077] Based on the same inventive concept as the above-described method embodiment, the present invention provides a server comprising a memory and a processor. The memory is configured to store executable program code, and the processor is configured to retrieve and execute the executable program code from the memory, thereby enabling the device to perform any of the above-described interventional puncture ultrasound-assisted positioning methods.

[0078] Based on the same inventive concept as the above-mentioned method embodiment, the present invention provides a computer program product, which includes: computer program code, which, when running on a computer, enables the computer to execute any one of the above-mentioned interventional puncture ultrasound-assisted positioning methods.

[0079] Based on the same inventive concept as the above-mentioned method embodiment, the present invention provides a computer-readable storage medium, which stores computer program code. When the computer program code runs on a computer, the computer executes any one of the above-mentioned interventional puncture ultrasound-assisted positioning methods.

[0080] In summary, compared with the doctor observing the ultrasound image collected at the current moment to evaluate the movement of the puncture needle at the next moment, the present invention quantifies the matching degree and possible indicators of the puncture needle when evaluating the movement of the puncture needle at the next moment, thereby achieving the positioning of the needle tip position, and then relatively objectively achieving the evaluation of the needle tip movement direction and needle tip displacement of the puncture needle at the next moment, thereby improving the accuracy of the evaluation of the movement of the puncture needle at the next moment, and thus improving the accuracy of interventional puncture ultrasound positioning and tracking.

[0081] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present invention, and should all be included in the scope of protection of the present invention.

Claims

1. An ultrasound-assisted positioning method for interventional puncture, characterized in that: The following steps are involved: During the interventional puncture process, ultrasound images are collected in real time through the ultrasound probe, and the optical flow vector corresponding to each pixel in each frame of ultrasound image is determined based on the movement between each frame of ultrasound image and the next frame of ultrasound image; According to the similarity between the optical flow vectors corresponding to the pixels in each frame of ultrasound image and the distance between the pixels, the pixels in each frame of ultrasound image are clustered to obtain the target cluster; Determine the degree of matching between each target cluster and the target cluster in the next frame of the ultrasound image to which it belongs based on the distribution of non-overlapping pixels between each target cluster and the target cluster in the next frame of the ultrasound image to which it belongs, as well as the center of each target cluster and the optical flow vectors corresponding to the pixels in each target cluster; Based on the matching degree between target clusters, target clusters that match each other are screened out from all ultrasound images to form a target cluster sequence; According to the matching degree between adjacent target clusters in each target cluster sequence, the similarity of optical flow vectors between adjacent target clusters, and the shape characteristics of the target clusters, the possible puncture needle indicators corresponding to each target cluster sequence are determined; Based on all possible indicators of the puncture needle, the puncture needle region sequence is screened out from all target cluster sequences, and the needle tip position is screened out from each target cluster in the puncture needle region sequence; Based on the distance between the needle tip positions in adjacent puncture needle regions in the puncture needle region sequence and the optical flow vector corresponding to the pixel points in the puncture needle region, the needle tip movement direction and needle tip displacement of the next moment are evaluated.

2. The ultrasound-assisted positioning method for interventional puncture according to claim 1, characterized in that: The pixel points in each frame of ultrasound image are clustered according to the similarity between the optical flow vectors corresponding to the pixel points in each frame of ultrasound image and the distance between the pixel points to obtain the target cluster, including: The normalized value of the cosine similarity between the optical flow vectors corresponding to each two pixel points in each frame of ultrasound image is determined as the vector similarity factor between each two pixel points in each frame of ultrasound image; The absolute value of the difference between the moduli of the optical flow vectors corresponding to every two pixel points in each frame of ultrasound image is determined as the motion distance factor between every two pixel points in each frame of ultrasound image; The consistency coefficient between each two pixels in each frame of ultrasound image is determined based on the vector similarity factor, motion distance factor and Euclidean distance between each two pixels in each frame of ultrasound image. The vector similarity factor is positively correlated with the consistency coefficient, while the motion distance factor and Euclidean distance are negatively correlated with the consistency coefficient. According to the consistency coefficient between all pixels in each frame of ultrasound image, the pixels in each frame of ultrasound image are clustered, and each obtained cluster is recorded as a target cluster.

3. The ultrasound-assisted positioning method for interventional puncture according to claim 1, characterized in that: The determining of a matching degree between each target cluster and a target cluster in a next frame of ultrasound image to which the target cluster belongs based on a distribution of non-overlapping pixels between the target cluster and the target cluster in a next frame of ultrasound image to which the target cluster belongs, and an optical flow vector corresponding to a center of each target cluster and pixels within each target cluster, includes: Determine any target cluster as a first marker cluster, and determine any target cluster in a next frame of ultrasound image to which the first marker cluster belongs as a second marker cluster; Determine an area in the next frame of ultrasound image to which the first marker cluster belongs, which is located at the same position as the first marker cluster, as a reference area, and determine the center of the reference area as a reference center; Determine the intersection of the second marker cluster and the reference area as a reference intersection area, and determine the center of the area formed by all pixels in the second marker cluster except the reference intersection area as a candidate center, wherein the pixels in the second marker cluster except the reference intersection area represent non-overlapping pixels between the first marker cluster and the second marker cluster; The direction in which the reference center points to the candidate center is determined as a reference movement direction; Determining the mean of the optical flow vectors corresponding to all pixels in the first marker cluster as a candidate motion vector, and determining the direction of the candidate motion vector as a candidate motion direction; The degree of matching between the first marker cluster and the second marker cluster is determined based on the angle between the reference motion direction and the candidate motion direction, the Euclidean distance between the reference center and the center of the second marker cluster, and the number of pixels in the reference intersection area.

4. The ultrasound-assisted positioning method for interventional puncture according to claim 3, characterized in that: The formula corresponding to the matching degree between the first marker cluster and the second marker cluster is: ; Wherein, D is the degree of match between the first marker cluster and the second marker cluster; is the normalization function; n is the number of pixels in the reference intersection area; is a natural exponential function; is the angle between the reference motion direction and the candidate motion direction; d is the Euclidean distance between the reference center and the center of the second marker cluster.

5. The ultrasound-assisted positioning method for interventional puncture according to claim 1, characterized in that: The method of determining the possible puncture needle indicators corresponding to each target cluster sequence based on the matching degree between adjacent target clusters in each target cluster sequence, the similarity of optical flow vectors between adjacent target clusters, and the shape characteristics of the target clusters includes: According to the matching degree between adjacent target clusters in each target cluster sequence and the difference between the directions of the optical flow vectors between adjacent target clusters, the direction consistency corresponding to each target cluster sequence is determined; The Euclidean distance between each pixel point in each target cluster and the center of the target cluster to which it belongs is determined as the deviation distance corresponding to each pixel point in each target cluster; The variance of the deviation distances corresponding to all pixels in each target cluster is determined as the shape feature index corresponding to each target cluster; According to the directional consistency corresponding to each target cluster sequence and the shape characteristic indices corresponding to all target clusters within each target cluster sequence, the possible puncture needle index corresponding to each target cluster sequence is determined. Among them, the directional consistency and shape characteristic indices are both positively correlated with the possible puncture needle index.

6. The ultrasound-assisted positioning method for interventional puncture according to claim 5, characterized in that: The formula corresponding to the directional consistency of the target cluster sequence is: ;in, is the directional consistency corresponding to the rth target cluster sequence; r is the sequence number of the target cluster sequence; is the normalization function; is the number of target clusters in the r-th target cluster sequence; t is the sequence number of the target cluster in the r-th target cluster sequence; is the degree of matching between the tth target cluster and the t+1th target cluster in the rth target cluster sequence; It is the cosine similarity between the motion representative vector corresponding to the t-th target cluster and the motion representative vector corresponding to the t+1-th target cluster in the r-th target cluster sequence; the motion representative vector corresponding to the target cluster is the mean of the optical flow vectors corresponding to all pixels in the target cluster.

7. The ultrasound-assisted positioning method for interventional puncture according to claim 5, characterized in that: The method of determining the possible puncture needle index corresponding to each target cluster sequence based on the directional consistency corresponding to each target cluster sequence and the shape feature indexes corresponding to all target clusters within each target cluster sequence includes: The cumulative value of the shape feature indices corresponding to all target clusters in each target cluster sequence is determined as the possible shape factor of the puncture needle corresponding to each target cluster sequence; The product of the directional consistency and the possible factor of the puncture needle shape corresponding to each target cluster sequence is normalized to obtain the possible index of the puncture needle corresponding to each target cluster sequence.

8. The ultrasound-assisted positioning method for interventional puncture according to claim 1, characterized in that: The step of selecting the needle tip position from each target cluster in the puncture needle region sequence includes: Any target cluster in the puncture needle region sequence is determined as a marking region, and a marking ray is constructed with the center of the marking region as the starting point and the direction of the mean value of the optical flow vectors corresponding to all pixels in the marking region as the extension direction; The pixel point farthest from the center of the marking area is selected from the intersection of the marking ray and the marking area as the needle tip pixel point, and the position of the needle tip pixel point is determined as the needle tip position.

9. The ultrasound-assisted positioning method for interventional puncture according to claim 1, characterized in that: The formulas corresponding to the needle tip movement direction and needle tip displacement at the next moment are: ； ; Wherein, the direction of R is the estimated direction of movement of the puncture needle tip at the next moment; L is the estimated displacement of the puncture needle tip at the next moment; M is the number of puncture needle areas in the puncture needle area sequence; m is the sequence number of the puncture needle area in the puncture needle area sequence; Q is the cumulative value of the sequence numbers of all puncture needle areas in the puncture needle area sequence; is the mean of the optical flow vectors corresponding to all pixels in the mth puncture needle area in the puncture needle area sequence; is the distance between the needle tip position in the mth puncture needle region and the needle tip position in the m+1th puncture needle region in the puncture needle region sequence.

10. An interventional puncture ultrasound-assisted positioning system, characterized in that: The method comprises a processor and a memory, wherein the processor is used to process instructions stored in the memory to implement an interventional puncture ultrasound-assisted positioning method according to any one of claims 1 to 9.

Images