CN117765044A - Registration method, system and device for medical image - Google Patents

Abstract

The embodiment of the specification provides a registration method, a registration system and a registration device of medical images, which comprise the following steps: acquiring a medical image sequence, wherein the medical image sequence comprises a plurality of frames of medical images; selecting mask images from the medical image sequence; acquiring at least one key feature point which is positioned in a target area of a multi-frame medical image; matching key feature points of the multi-frame medical image and the mask image, and determining a first offset of at least one key feature point; smoothing the first offset to obtain a second offset of the key feature points of the medical image of the target frame; and determining a target structure image corresponding to the target frame medical image based on the second offset.

Description

Registration method, system and device for medical image

Technical Field

The present disclosure relates to the field of medical scanning, and in particular, to a method, system, and apparatus for registering medical images.

Background

Digital subtraction angiography (Digital Subtraction Angiography, DSA for short) is widely applied to clinical diagnosis of vascular diseases due to the advantages of high contrast resolution, short examination time, small contrast agent consumption and the like. The DSA technique removes bone and soft tissue structures by subtracting mask images from the contrast-bearing mask images, and visualizes contrast-filled vessels in a subtraction map. However, the use of contrast agents may cause adverse reactions, which may result in the patient not being able to remain completely stationary. In addition, local movements of autonomous tissues, which are difficult to avoid, such as breathing and swallowing, of the patient may also result in the patient not remaining stationary. Further resulting in motion artifacts in the DSA image. In order to solve the artifacts generated by the motion, the feature matching is generally performed on the tissue structure information of the vascular phase and the mask phase, and then subtraction is performed. However, the existing technical means only pursues the similarity of the mask and the interference pattern, ignores the difference of image registration between adjacent frames, and causes the inconsistent phenomenon of the whole sequence after registration, and has strong flickering and shaking feeling in vision, thereby influencing the user experience.

Therefore, the present disclosure provides a method, a system and a device for registering medical images, which can remove motion artifacts and simultaneously consider image stability.

Disclosure of Invention

One of the embodiments of the present specification provides a registration method of medical images, the method comprising: acquiring a medical image sequence, wherein the medical image sequence comprises a plurality of frames of medical images; selecting mask images from the medical image sequence; acquiring at least one key feature point, wherein the at least one key feature point is positioned in a target area of the multi-frame medical image; performing key feature point matching on the multi-frame medical image and the mask image, and determining a first offset of the at least one key feature point; smoothing the first offset to obtain a second offset of the key feature points of the medical image of the target frame; and determining a target structure image corresponding to the target frame medical image based on the second offset.

In some embodiments, the multi-frame medical image comprises at least one historical frame medical image and/or at least one subsequent frame medical image, the at least one historical frame medical image being an image preceding the target frame medical image, the at least one subsequent frame medical image being an image following the target frame medical image, the first offset comprising an offset of at least one key feature point of the target region in the at least one historical frame medical image and/or an offset of at least one key feature point of the target region in the at least one subsequent frame medical image.

In some embodiments, the medical image comprises a DSA image.

In some embodiments, the determining, based on the second offset, a target structure image corresponding to a target frame medical image includes: obtaining a second mask image of the target frame medical image based on the second offset; and subtracting the second mask image from the target frame medical image to determine the target structure image corresponding to the target frame medical image.

In some embodiments, the smoothing of the first offset is achieved by an image smoothing technique including at least one of gaussian kernel filtering, kalman filtering, parabolic filtering.

In some embodiments, the method further comprises: and performing preprocessing operation on the medical image sequence, wherein the preprocessing operation comprises at least one of log transformation, vst transformation, regularization and noise reduction.

In some embodiments, the matching the multi-frame medical image with the mask image for key feature points, determining the first offset of the at least one key feature point includes: and matching the at least one key feature point of the target area in the multi-frame medical image with the at least one key feature point of the target area in the mask image, and determining a first offset of the at least one key feature point.

In some embodiments, the at least one key feature point is selected uniformly in each of the plurality of medical images or based on a shape of the target structure.

One of the embodiments of the present specification provides a registration system for medical images, the system comprising: a first acquisition module for acquiring a medical image sequence comprising a plurality of frames of medical images; the selecting module is used for selecting mask images from the medical image sequence; the second acquisition module is used for acquiring at least one key characteristic point, and the at least one key characteristic point is positioned in a target area of the multi-frame medical image; the matching module is used for matching key feature points of the multi-frame medical image and the mask image and determining a first offset of the at least one key feature point; the smoothing module is used for smoothing the first offset to obtain a second offset of the key feature points of the medical image of the target frame; and the determining module is used for determining a target structure image corresponding to the target frame medical image based on the second offset.

One of the embodiments of the present specification provides a registration apparatus of medical images, the apparatus comprising: at least one storage medium storing computer instructions; at least one processor executes the computer instructions to implement a method of registration of the medical images.

By the registration method of medical images described in some embodiments of the present disclosure, at least the following beneficial effects may be achieved: 1) The artifact correction can be realized on the premise of ensuring the stability of the image, the offset condition of the multi-frame image is reserved in the artifact correction process, and bad impressions such as flickering, jumping, shaking and the like of the image are avoided; 2) The offset determining process adopts two technologies of real-time processing and post-processing, and provides various alternatives for actual operation; 3) And the new mask is subjected to image subtraction with the medical image of the target frame to obtain an artifact removal image, the new mask has the integral deviation condition of the medical images of multiple frames, and the artifact removal image has high association degree with the medical images of the adjacent frames and other frames.

Drawings

The present specification will be further elucidated by way of example embodiments, which will be described in detail by means of the accompanying drawings. The embodiments are not limiting, in which like numerals represent like structures, wherein:

FIG. 1 is a schematic illustration of an application scenario of a registration system for medical images shown in accordance with some embodiments of the present description;

FIG. 2 is an exemplary block diagram of a registration system for medical images shown in accordance with some embodiments of the present description;

FIG. 3 is an exemplary flowchart of a method of registration of medical images shown in accordance with some embodiments of the present description;

FIG. 4 is an exemplary flow chart for determining a target structure image according to some embodiments of the present disclosure;

FIG. 5 is a schematic illustration of registration of medical images shown in accordance with some embodiments of the present description;

fig. 6 is a schematic diagram of a smoothing filtering technique shown in accordance with some embodiments of the present description.

Detailed Description

In order to more clearly illustrate the technical solutions of the embodiments of the present specification, the drawings that are required to be used in the description of the embodiments will be briefly described below. It is apparent that the drawings in the following description are only some examples or embodiments of the present specification, and it is possible for those of ordinary skill in the art to apply the present specification to other similar situations according to the drawings without inventive effort. Unless otherwise apparent from the context of the language or otherwise specified, like reference numerals in the figures refer to like structures or operations.

It will be appreciated that "system," "apparatus," "unit" and/or "module" as used herein is one method for distinguishing between different components, elements, parts, portions or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

As used in this specification and the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.

A flowchart is used in this specification to describe the operations performed by the system according to embodiments of the present specification. It should be appreciated that the preceding or following operations are not necessarily performed in order precisely. Rather, the steps may be processed in reverse order or simultaneously. Also, other operations may be added to or removed from these processes.

Digital subtraction angiography (Digital Subtraction Angiography, DSA for short) is widely applied to clinical diagnosis of vascular diseases due to the advantages of high contrast resolution, short examination time, small contrast agent consumption and the like. The DSA technique removes bone and soft tissue structures by subtracting mask images from the contrast-bearing mask images, and visualizes contrast-filled vessels in a subtraction map. However, the use of contrast agents may cause adverse reactions, which may result in the patient not being able to remain completely stationary. In addition, local movements of autonomous tissues, which are difficult to avoid, such as breathing and swallowing, of the patient may also result in the patient not remaining stationary. Further, DSA images may be caused to produce motion artifacts that may obscure critical vessel information, leading to vessel edge distortion and blurring. In order to solve the artifacts generated by the motion, the feature matching is generally performed on the tissue structure information of the vascular phase and the mask phase, and then subtraction is performed. However, the existing technical means only pursues the similarity of the mask and the interference pattern, ignores the difference of image registration between adjacent frames, and causes the inconsistent phenomenon of the whole sequence after registration, and has strong flickering and shaking feeling in vision, thereby influencing the user experience.

Therefore, the present disclosure provides a method, a system and a device for registering medical images, which can remove motion artifacts and simultaneously consider image stability.

Fig. 1 is a schematic illustration of an application scenario of a registration system for medical images according to some embodiments of the present description.

As shown in fig. 1, an application scenario 100 of a registration system of medical images may include an imaging device 110, a processing device 120, a terminal 130, a network 140, and a storage device 150.

The imaging device 110 may be used to acquire a sequence of medical images of a region of interest of a target object. The imaging device 110 may be a medical imaging device. For example, DSA (Digital Subtraction Angiography ) devices. In some embodiments, imaging device 110 may include components (not shown in part) such as an X-ray tube, a high voltage generator, an image intensifier, an optical system, a controller, a gantry, and the like. In some embodiments, the target object may be an individual being scanned, such as a person (patient), an animal, or the like. In some embodiments, the region of interest may be a focal region of the target object, such as the head, neck, chest, abdomen, etc.

The processing device 120 may process data and/or information obtained from the imaging device 110, the terminal 130, and/or the storage device 150. For example, the processing device 120 may acquire a sequence of medical images comprising a plurality of frames of medical images through the imaging device 110; selecting mask images from the medical image sequence; acquiring at least one key feature point, wherein the at least one key feature point is positioned in a target area of the multi-frame medical image; performing key feature point matching on the multi-frame medical image and the mask image, and determining a first offset of the at least one key feature point; smoothing the first offset to obtain a second offset of the key feature points of the medical image of the target frame; and determining a target structure image corresponding to the target frame medical image based on the second offset. The target area may be a blood vessel, heart or other tissue.

In some embodiments, the processing device 120 may be a single server or a group of servers. The server group may be centralized or distributed. In some embodiments, the processing device 120 may be local or remote. For example, processing device 120 may access information and/or data from imaging device 110, terminal 130, and/or storage device 150 via network 140. As another example, processing device 120 may be directly connected to imaging device 110, terminal 130, and/or storage device 150 to access information and/or data. In some embodiments, the processing device 120 may be implemented on a cloud platform. For example, the cloud platform may include one or a combination of several of private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, cross-cloud, multi-cloud, and the like.

The terminal 130 may include a mobile device 131, a tablet 132, a notebook 133, or the like, or any combination thereof. In some embodiments, terminal 130 may interact with other components through network 140. For example, the terminal 130 may send one or more control instructions to the imaging device 110 to control the imaging device 110 to scan the target object in accordance with the instructions. For another example, the terminal 130 may also receive and present the target structure image corresponding to the target frame medical image sent by the processing device 120 to the user, and perform other operations in response to user feedback. In some embodiments, terminal 130 may be part of processing device 120. The user may be a user of the present system. Such as doctors, researchers, etc.

Network 140 may include any suitable network capable of facilitating the exchange of information and/or data. In some embodiments, one or more components (e.g., imaging device 110, processing device 120, terminal 130, storage device 150, etc.) may exchange information and/or data with one or more components over network 140. For example, the processing device 120 may obtain a sequence of medical images from the imaging device 110 over the network 140. The network 140 may include one or a combination of several of public networks (e.g., the internet), private networks (e.g., local Area Network (LAN), wide Area Network (WAN)), etc.), wired networks (e.g., ethernet), wireless networks, cellular networks, frame relay networks, virtual private networks, satellite networks, telephone networks, routers, hubs, server computers, etc.

Storage device 150 may store data, instructions, and/or any other information. In some embodiments, the storage device 150 may store data obtained from the imaging device 110, the terminal 130, and/or the processing device 120, e.g., the storage device 150 may store a sequence of medical images obtained from the imaging device 110. In some embodiments, the storage device 150 may store data and/or instructions for execution or use by the processing device 120 to perform the exemplary methods described herein. In some embodiments, the storage device 150 may include one or a combination of several of mass storage, removable storage, volatile read-write memory, read-only memory (ROM), and the like. In some embodiments, storage device 150 may be implemented by a cloud platform as described herein. For example, the cloud platform may include one or a combination of several of private cloud, public cloud, hybrid cloud, community cloud, distributed cloud, cross-cloud, multi-cloud, and the like.

In some embodiments, the storage device 150 may be connected to the network 140 to enable communication with one or more components (e.g., the processing device 120, the terminal 130, etc.). One or more components may read data or instructions in storage device 150 over network 140. In some embodiments, the storage device 150 may be part of the processing device 120 or may be separate and directly or indirectly connected to the processing device. It should be noted that the application scenario 100 of the registration system of medical images is provided for illustration purposes only and is not intended to limit the scope of the present application. Many modifications and variations will be apparent to those of ordinary skill in the art in light of the present description. For example, the application scenario 100 of the registration system of medical images may also comprise a database. As another example, the application scenario 100 of the registration system of medical images may implement similar or different functions on other devices. However, such changes and modifications do not depart from the scope of the present application.

Fig. 2 is an exemplary block diagram of a registration system for medical images shown in accordance with some embodiments of the present description.

In some embodiments, the registration system 200 of medical images may include a first acquisition module 210, a selection module 220, a second acquisition module 230, a matching module 240, a smoothing module 250, and a determination module 260.

The first acquisition module 210 is configured to acquire a medical image sequence, the medical image sequence comprising a plurality of frames of medical images.

The selecting module 220 is configured to select a mask image from the medical image sequence.

The second obtaining module 230 is configured to obtain at least one key feature point, where the at least one key feature point is located in a target area of the multi-frame medical image.

The matching module 240 is configured to match the multi-frame medical image with the mask image to determine a first offset of the at least one key feature point.

The smoothing module 250 is configured to smooth the first offset to obtain a second offset of the key feature point of the medical image of the target frame.

The determining module 260 is configured to determine a target structure image corresponding to the target frame medical image based on the second offset.

Fig. 3 is an exemplary flowchart of a method of registration of medical images shown in accordance with some embodiments of the present description. As shown in fig. 3, the process 300 includes the following steps. In some embodiments, the process 300 may be performed by a processing device (e.g., the processing device 120) and its components.

Step 310, a sequence of medical images is acquired. In some embodiments, step 310 may be performed by the first acquisition module 210.

The sequence of medical images is a video image obtained during injection of the contrast agent into the target object. The medical image may comprise a DSA image. I.e. X-ray images before and after injection of the contrast agent. In some embodiments, the sequence of medical images may be acquired by the imaging device 110. In DSA techniques, the target object is injected with contrast agent to better visualize vascular conditions through medical imaging means. In some embodiments, the medical image sequence comprises a plurality of frames of medical images. For example, multiple frames of medical images may be arranged linearly. Further, multiple frames of medical images may be arranged sequentially based on the temporal order of acquisition, constituting a medical image sequence (or video). For example only, the frame rate of the medical image sequence may be 10 frames, 15 frames, etc. In some embodiments, the multi-frame medical image includes an image not injected with contrast agent (i.e., a mask image) and an image injected with contrast agent (i.e., a contrast image).

In some embodiments, the processing device 120 performs a preprocessing operation on the sequence of medical images, the preprocessing operation including at least one of logarithmic transformation, variance stabilization (Variance Stabilizing Transform, VST) transformation, regularization, noise reduction. Wherein the logarithmic transformation is used to enhance the contrast and brightness of the image. The VST transform is used to adjust the hue and saturation of an image. Regularization is used to ensure that the data is within a certain range and to reduce the effects of noise. Noise reduction is used to reduce noise points and artifacts in images. It should be noted that, the preprocessing operation may also set other operations according to actual needs, which is not limited in this specification.

Step 320, selecting mask images from the medical image sequence. In some embodiments, step 320 may be performed by the pick module 220.

Mask images are a reference standard for DSA scanning. In some embodiments, the mask image is a reference mask. For example, the mask image is an X-ray image without contrast injected. In some embodiments, the mask image may be selected manually, or the system may be selected based on a particular rule. For example, processing device 120 may designate an image frame not injected with contrast agent as a mask image.

In some embodiments, the difference between the mask image and other medical images of the sequence of medical images is less than a preset range. The preset range may be a preset gray value range, a distance range, and the like.

In some embodiments, any one of a gray scale difference between the mask image and other medical images of the medical image sequence, a target structure position difference is less than a preset range. The mask image is selected to take into account as many image features as possible that can cover multiple frames of medical images, so the differences between the mask image and other medical images of the medical image sequence need to be as small as possible. The gray scale difference may be a difference in pixel values, with smaller gray scale differences indicating closer pixel values for each pixel. The target structure position difference may be a difference in position coordinates of a target structure (e.g., a blood vessel region) in the image, the smaller the target structure position difference, the closer the target structure is positioned in the image. In some embodiments, the differences further comprise target structure shape differences. The shape difference of the target structure can be the shape difference of the target structure in the image, and the smaller the shape difference of the target structure, the smaller the difference of the shooting angle of the target structure in the image and the shape change (such as uncontrollable physiological shape change such as breathing) of the target structure. In some embodiments, the differences may be determined by detecting pixel values of individual pixels of the image, detecting edges of the image, and the like.

At step 330, at least one key feature point is obtained. In some embodiments, step 330 may be performed by the second acquisition module 230.

The key feature points are points reflecting the image features of the target region in the image. For example, the key feature points may be corner points of an organ, edge points of an organ, or other points where the gray value changes significantly. In some embodiments, the at least one key feature point is located in a target region of the multi-frame medical image. In some embodiments, the target region of the same frame of medical image may include at least one key feature point. It can be understood that the more the key feature points are selected, the more accurate the effect of the subsequent offset calculation, but the higher the calculation requirement and the memory requirement on the system.

In some embodiments, the at least one key feature point is selected uniformly in each of the plurality of medical images or based on the shape of the target structure. The uniform selection can be uniform sampling or grid sampling, namely key feature points are selected on a plurality of frames of medical images according to certain intervals or rules, and the key feature points are ensured to be uniformly distributed. The shape selection based on the target structure may be to select key feature points related to the shape of the target structure by means of corner detection, edge intersection detection, etc. after performing structural analysis (such as edge detection, contour extraction, etc.) on the target structure. The selection of key feature points generally requires the selection of points which are easily discernable and have significant changes in the gray scale values of the image. For example, the processing device 120 may extract the same edge location point of the abdomen in each of the multiple frames of medical images as the key feature point. In some embodiments, key feature points may also be manually selected.

And step 340, matching key feature points of the multi-frame medical image and the mask image, and determining a first offset of at least one key feature point. In some embodiments, step 340 may be performed by matching module 240.

The first offset reflects an offset of key feature points between the mask image and other medical images. In some embodiments, the first offset may be represented by a pixel value, a distance value (e.g., millimeters). For example, when a pixel is used as a measure of the size of a frame of medical image (e.g., 1000 x 1000 pixels), the first offset may be a pixel value of 20; when using the length-width (or specific coordinate values) as a measure of the size of a frame of medical image (e.g., 500 mm-500 mm), the first offset may be a distance value of 2mm. It will be appreciated that the metrics described above may be converted to one another by a certain mapping (e.g., resolution). In some embodiments, processing device 120 may determine a first offset of at least one key feature point between the mask image and each of the other medical images. For example, assuming that the multi-frame medical image includes N frames, the other medical images than the mask image are N-1 frames, and the determined first offset amount is N-1. For example only, the N-1 first offsets may be ordered based on the corresponding frame medical image order, and the N-1 first offsets may have an index i indicating which first offset the target first offset is. In some embodiments, the processing device 120 matches at least one key feature point of the target region in the multi-frame medical image with at least one key feature point of the target region in the mask image, determining a first offset of the at least one key feature point. For example, processing device 120 may determine, based on a mask image and a film image of a location of a blood vessel, a distance deviation of the location in the two images as a first offset. When a plurality of key feature points are included in one frame of the medical image and the mask image, the processing apparatus 120 may determine a distance deviation of each feature point, respectively, and average all the distance deviations as the first offset of the frame of the image.

In some embodiments, the multi-frame medical image comprises at least one historical frame medical image, the at least one historical frame medical image being an image prior to the target frame medical image, and the first offset comprises an offset of at least one key feature point of the target region in the at least one historical frame medical image. The target frame medical image is a current frame medical image or a medical image needing artifact correction. The history frame medical image is an already acquired medical image. In this embodiment, the multiple frames of medical images are acquired in real time, and the processing device 120 may acquire the multiple frames of medical images in real time and extract the first offset corresponding to each frame of medical image (i.e. the history frame of medical image) in real time. For example, the processing device 120 may extract the i-n first offsets to the current i-th first offset (i > =n > =2, n represents any frame before the i-th frame, and i represents the index of the current first offset), respectively. The real-time offset condition of the medical image can be obtained by extracting the first offset in real time.

In some embodiments, the multi-frame medical image comprises at least one historical frame medical image and at least one subsequent frame medical image, the at least one historical frame medical image is an image prior to the target frame medical image, the at least one subsequent frame medical image is an image subsequent to the target frame medical image, and the first offset comprises an offset of at least one key feature point of the target region in the at least one historical frame medical image and an offset of at least one key feature point of the target region in the at least one subsequent frame medical image. In this embodiment, the first offset is extracted after the multi-frame medical image is acquired, and the processing device 120 may extract the first offset corresponding to each frame of medical image (i.e. the history frame medical image and the subsequent frame medical image) after the multi-frame medical image is acquired. For example, the processing device 120 may obtain the i-n first offsets to the current i+m first offsets (n > =1, m represents any frame after the i-th frame, and i represents the index of the current key point), respectively. By extracting the first offset after acquiring the multi-frame medical image, the whole offset condition of the medical image can be obtained through a post-processing technology by combining the historical medical image and the subsequent medical image. It will be appreciated that the multi-frame medical image may also comprise only at least one subsequent frame medical image, and that the first offset comprises an offset of at least one key feature point of the target region in the at least one subsequent frame medical image, accordingly.

And step 350, smoothing the first offset to obtain a second offset of the key feature points of the medical image of the target frame. In some embodiments, step 350 may be performed by the smoothing module 250.

The second offset is an offset obtained by smoothing the first offset. In some embodiments, the second offset may be represented by a pixel value, a distance value (e.g., millimeters). For example, when a pixel is used as a measure of the size of a frame of medical image (e.g., 1000 x 1000 pixels), the second offset may be 20 pixels; when using the length-width (or specific coordinate values) as the size measure of a frame of medical image (e.g., 500 mm-500 mm), the second offset may be a distance value of 2mm. It will be appreciated that the metrics described above may be converted to one another by a certain mapping (e.g., resolution). In some embodiments, the second offset may be obtained by smoothing at least two first offsets. The second offset reflects the offset condition of the whole multi-frame medical image, and artifact correction can be achieved and the whole characteristics of the multi-frame medical image are reserved by combining the offset condition of the whole multi-frame medical image on the target frame medical image.

In some embodiments, smoothing the first offset is achieved by an image smoothing technique including at least one of gaussian kernel filtering, kalman filtering, parabolic filtering. The image smoothing technology analyzes an offset trend based on the medical image of the current frame and other medical images, and then carries out smoothing processing to offset pixel points needing to be offset. The image smoothing technique requires a plurality of first offsets (e.g., N-1 first offsets) to output a second offset.

Step 360, determining a target structure image corresponding to the target frame medical image based on the second offset. In some embodiments, step 360 may be performed by determination module 260.

The target structure image is a target image after registration of the images. For example, the target structure image may be a vascular structure image after artifact correction. In some embodiments, the processing device 120 may image shift the target frame medical image based on the second offset, or determine the target structure image by way of deep learning, or the like. For a detailed description of the determination of the target structure image, see fig. 4 and its associated description.

According to the medical image registration method disclosed by some embodiments of the specification, artifact correction can be realized on the premise of ensuring the stability of images, the offset condition of multi-frame images is reserved in the artifact correction process, and bad impressions such as flickering, jumping and shaking of the images are avoided; in addition, the offset determination process adopts two technologies of real-time processing and post-processing, and provides various alternatives for actual operation.

Fig. 4 is an exemplary flow chart for determining a target structure image according to some embodiments of the present description. As shown in fig. 4, in some embodiments, the process 400 may be performed by a processing device (e.g., the processing device 120) or a component thereof.

Step 410, obtaining a second mask image of the medical image of the target frame based on the second offset.

The second mask image is an updated DSA scan reference standard. In some embodiments, the second mask image is an updated reference mask. For example, the second mask image is an X-ray image without contrast agent injected. In some embodiments, the second mask image may be obtained by shifting the mask image described above. For example, processing device 120 may shift the reference mask based on the second shift amount to obtain a second mask image. In some embodiments, the processing device 120 may acquire the second mask image through a deep learning model. For example, processing device 120 inputs the reference mask and the second offset to the deep learning model, and outputs a second mask image. The deep learning model may be a convolutional neural network, a deep neural network, or the like.

Step 420, subtracting the second mask image from the target frame medical image to determine a target structure image corresponding to the target frame medical image.

In some embodiments, the subtraction of the target frame medical image and the second mask image may be performed by image processing software. The interference of contrast agent development such as tissue structures, bones and the like can be subtracted through image subtraction, and the vascular structures and the flow of the contrast agent are highlighted.

In the embodiment, the image subtraction is performed between the new mask and the medical image of the target frame to obtain the artifact removal image, the new mask has the integral deviation condition of the medical images of multiple frames, and the association degree of the artifact removal image and the medical images of the adjacent frames and other frames is high, so that bad impressions such as flickering and jumping do not occur in the whole image sequence.

Fig. 5 is a schematic illustration of registration of medical images shown according to some embodiments of the present description. As shown in fig. 5, the process 500 includes the steps of:

at step 510, the processing device 120 performs a preprocessing operation on the acquired DSA image sequence. For example, preprocessing operations such as log transformation, regularization, noise reduction, etc. are performed on the image. Selecting mask images from the preprocessed DSA image sequence;

at step 520, the processing device 120 extracts image feature points of the critical region from the DSA image sequence by using a feature extraction method, wherein the critical region may be a lesion region or other specific organ tissue;

step 530, the processing device 120 performs image feature matching on the feature points to obtain corresponding feature points;

step 540, the processing device 120 determines a first offset for the i- (N-1) th frame, … … i th frame, … … i+m th frame medical image;

step 550, the processing device 120 obtains a second offset using a smoothing technique;

step 560, the processing device 120 updates the matching result based on the second offset;

in step 570, the processing device 120 performs an image transformation on the target frame medical image to obtain a target structure image. Processing of the sequence of medical images may be achieved through the above-described process 500.

Fig. 6 is a schematic diagram of a smoothing filtering technique shown in accordance with some embodiments of the present description. As shown in fig. 6, the abscissa is time (frame) and the ordinate is offset (pixel value). Fig. 6 includes a base (not smoothed) curve 610, a gaussian kernel 3 curve 620, and a gaussian kernel 7 curve 630. Wherein, part of peak value of the base curve 610 is marked by a circle; part of the peak value of the gaussian kernel 3 curve 620 is marked by a rectangle; part of the peak of the gaussian kernel 7 curve 630 is marked by triangles. Gaussian kernel 7 is larger than gaussian kernel 3 kernel, gaussian kernel 7 is the smoothest, and second, gaussian kernel 3, the different frame deviations of base curve 610 without smoothing are apparent (e.g., the difference in the ordinate of the partial peak of base curve 610 is the largest in fig. 6). As can be seen from fig. 6, the difference between the ordinate of the partial peak values of the gaussian kernel 7 curve 630 is the smallest, i.e. the offset variation is the smallest, and the gaussian kernel 7 can obtain the smoothest result.

By the registration method of medical images described in some embodiments of the present disclosure, at least the following beneficial effects may be achieved: 1) According to the medical image registration method disclosed by some embodiments of the specification, artifact correction can be realized on the premise of ensuring the stability of images, the offset condition of multi-frame images is reserved in the artifact correction process, and bad impressions such as flickering, jumping and shaking of the images are avoided; 2) The offset determining process adopts two technologies of real-time processing and post-processing, and provides various alternatives for actual operation; 3) And the new mask is subjected to image subtraction with the medical image of the target frame to obtain an artifact removal image, the new mask has the integral deviation condition of the medical images of multiple frames, and the artifact removal image has high association degree with the medical images of the adjacent frames and other frames.

Some embodiments of the present specification also provide a registration apparatus for medical images, the apparatus comprising at least one processor and at least one memory; at least one memory for storing computer instructions; at least one processor is configured to execute at least some of the computer instructions to implement the above-described medical image registration method.

While the basic concepts have been described above, it will be apparent to those skilled in the art that the foregoing detailed disclosure is by way of example only and is not intended to be limiting. Although not explicitly described herein, various modifications, improvements, and adaptations to the present disclosure may occur to one skilled in the art. Such modifications, improvements, and modifications are intended to be suggested within this specification, and therefore, such modifications, improvements, and modifications are intended to be included within the spirit and scope of the exemplary embodiments of the present invention.

Meanwhile, the specification uses specific words to describe the embodiments of the specification. Reference to "one embodiment," "an embodiment," and/or "some embodiments" means that a particular feature, structure, or characteristic is associated with at least one embodiment of the present description. Thus, it should be emphasized and should be appreciated that two or more references to "an embodiment" or "one embodiment" or "an alternative embodiment" in various positions in this specification are not necessarily referring to the same embodiment. Furthermore, certain features, structures, or characteristics of one or more embodiments of the present description may be combined as suitable.

Furthermore, the order in which the elements and sequences are processed, the use of numerical letters, or other designations in the description are not intended to limit the order in which the processes and methods of the description are performed unless explicitly recited in the claims. While certain presently useful inventive embodiments have been discussed in the foregoing disclosure, by way of various examples, it is to be understood that such details are merely illustrative and that the appended claims are not limited to the disclosed embodiments, but, on the contrary, are intended to cover all modifications and equivalent arrangements included within the spirit and scope of the embodiments of the present disclosure. For example, while the system components described above may be implemented by hardware devices, they may also be implemented solely by software solutions, such as installing the described system on an existing server or mobile device.

Likewise, it should be noted that in order to simplify the presentation disclosed in this specification and thereby aid in understanding one or more inventive embodiments, various features are sometimes grouped together in a single embodiment, figure, or description thereof. This method of disclosure, however, is not intended to imply that more features than are presented in the claims are required for the present description. Indeed, less than all of the features of a single embodiment disclosed above.

In some embodiments, numbers describing the components, number of attributes are used, it being understood that such numbers being used in the description of embodiments are modified in some examples by the modifier "about," approximately, "or" substantially. Unless otherwise indicated, "about," "approximately," or "substantially" indicate that the number allows for a 20% variation. Accordingly, in some embodiments, numerical parameters set forth in the specification and claims are approximations that may vary depending upon the desired properties sought to be obtained by the individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and employ a method for preserving the general number of digits. Although the numerical ranges and parameters set forth herein are approximations that may be employed in some embodiments to confirm the breadth of the range, in particular embodiments, the setting of such numerical values is as precise as possible.

Each patent, patent application publication, and other material, such as articles, books, specifications, publications, documents, etc., referred to in this specification is incorporated herein by reference in its entirety. Except for application history documents that are inconsistent or conflicting with the content of this specification, documents that are currently or later attached to this specification in which the broadest scope of the claims to this specification is limited are also. It is noted that, if the description, definition, and/or use of a term in an attached material in this specification does not conform to or conflict with what is described in this specification, the description, definition, and/or use of the term in this specification controls.

Finally, it should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments of this specification. Other variations are possible within the scope of this description. Thus, by way of example, and not limitation, alternative configurations of embodiments of the present specification may be considered as consistent with the teachings of the present specification. Accordingly, the embodiments of the present specification are not limited to only the embodiments explicitly described and depicted in the present specification.

Claims (10)

1. A method of registration of medical images, the method comprising:

acquiring a medical image sequence, wherein the medical image sequence comprises a plurality of frames of medical images;

selecting mask images from the medical image sequence;

acquiring at least one key feature point, wherein the at least one key feature point is positioned in a target area of the multi-frame medical image;

performing key feature point matching on the multi-frame medical image and the mask image, and determining a first offset of the at least one key feature point;

smoothing the first offset to obtain a second offset of the key feature points of the medical image of the target frame;

and determining a target structure image corresponding to the target frame medical image based on the second offset.

2. The method of claim 1, wherein the multi-frame medical image comprises at least one frame of history frame medical image and/or at least one frame of subsequent frame medical image, the at least one frame of history frame medical image being an image preceding the target frame medical image, the at least one frame of subsequent frame medical image being an image following the target frame medical image, the first offset comprising an offset of at least one key feature point of the target region in the at least one frame of history frame medical image and/or an offset of at least one key feature point of the target region in the at least one frame of subsequent frame medical image.

3. The method of claim 1, wherein the medical image comprises a DSA image.

4. The method of claim 1, wherein the determining a target structure image corresponding to a target frame medical image based on the second offset comprises:

obtaining a second mask image of the target frame medical image based on the second offset;

and subtracting the second mask image from the target frame medical image to determine the target structure image corresponding to the target frame medical image.

5. The method of claim 1, wherein the smoothing of the first offset is achieved by an image smoothing technique including at least one of gaussian kernel filtering, kalman filtering, parabolic filtering.

6. The method of claim 1, wherein the method further comprises:

and performing preprocessing operation on the medical image sequence, wherein the preprocessing operation comprises at least one of log transformation, vst transformation, regularization and noise reduction.

7. The method of claim 1, wherein the key feature point matching the multi-frame medical image with the mask image, determining the first offset of the at least one key feature point comprises:

and matching the at least one key feature point of the target area in the multi-frame medical image with the at least one key feature point of the target area in the mask image, and determining a first offset of the at least one key feature point.

8. The method of claim 1, wherein the at least one key feature point is selected uniformly in each of the plurality of medical images or based on a shape of the target structure.

9. A registration system for medical images, the system comprising:

a first acquisition module for acquiring a medical image sequence comprising a plurality of frames of medical images;

the selecting module is used for selecting mask images from the medical image sequence;

the second acquisition module is used for acquiring at least one key characteristic point, and the at least one key characteristic point is positioned in a target area of the multi-frame medical image;

the matching module is used for matching key feature points of the multi-frame medical image and the mask image and determining a first offset of the at least one key feature point;

the smoothing module is used for smoothing the first offset to obtain a second offset of the key feature points of the medical image of the target frame;

and the determining module is used for determining a target structure image corresponding to the target frame medical image based on the second offset.

10. A registration apparatus for medical images, the apparatus comprising:

at least one storage medium storing computer instructions;

at least one processor executing the computer instructions to implement the method of any one of claims 1-8.