Labeling system and labeling method for medical image

Abstract

The present disclosure provides an annotation system and annotation method for medical images, the annotation system having a display device presenting an annotation unit, the annotation unit simultaneously displaying a plurality of panels, a first panel of the plurality of panels being configured to display image data through a first image indicator; a second panel of the plurality of panels is configured to display image data through second image indicators, in which a maximum number of selectable first image indicators in the first panel is smaller than a maximum number of selectable second image indicators in the second panel; a third panel of the plurality of panels is configured to display tag data through a tag indicator; when a determination command is received, the tagging system is configured to determine a selected medical image and a selected tag based on a selection state of the first image indicator, a selection state of the second image indicator, and a selection state of the tag indicator, and to determine a tagging result of the selected medical image based on the selected tag. According to the invention, the labeling efficiency can be improved.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and particularly to a labeling system and a labeling method for medical images. Background Art

[0002] In recent years, computer technology has been widely applied in the fields of medical diagnosis, artificial intelligence, or other fields. In these fields, computer technology can generally be used to label medical images so as to utilize the obtained labeling results to assist in lesion analysis, train neural networks, or perform other processing. For example, in the field of artificial intelligence, a neural network can learn features from the labeled medical images.

[0003] Generally, after obtaining a dataset to be labeled, a labeler can use a labeling system to carefully read and analyze the content of the medical images in the dataset, and then label the medical images to obtain labeling results. For example, for the labeling of gastroscope images, the lesions and abnormal areas of the stomach can be marked and located.

[0004] However, labeling medical images generally takes a long time, and if a large number of medical images need to be labeled, it will greatly increase the manual burden. For example, in the training of neural networks, a large number of medical images are usually required so that the neural network can learn enough features from them to improve the generalization performance of the neural network. Therefore, if a large number of medical images need to be labeled, how to improve the labeling efficiency remains to be studied. Summary of the Invention

[0005] The present disclosure is proposed in view of the above situation, and its purpose is to provide a labeling system and a labeling method for medical images that can improve the labeling efficiency.

[0006] To this end, a first aspect of the present disclosure provides a labeling system for medical images. The labeling system has a display device configured to present a labeling unit, and the labeling unit is divided into a plurality of panels that are displayed simultaneously. A first panel among the plurality of panels is configured to display image data through at least one first image indicator, and each first image indicator represents a medical image in the image data. The image data is obtained by an acquisition device acquiring medical images in a time series. A second panel among the plurality of panels is configured to display the image data through at least one second image indicator, and each second image indicator represents a medical image in the image data. Wherein, the maximum number of selectable first image indicators in the first panel is less than the maximum number of selectable second image indicators in the second panel. A third panel among the plurality of panels is configured to display label data through a plurality of label indicators, and each label indicator represents a label in the label data. When a determination command is received, the labeling system is configured to determine the selected medical image and the selected label based on the selection status of the first image indicator in the first panel, the selection status of the second image indicator in the second panel, and the selection status of the label indicator in the third panel, and determine the labeling result of the selected medical image based on the selected label.

[0007] In the first aspect of the present disclosure, by comprehensively determining the selected medical image based on the selection status of the first image indicator in the first panel and the selection status of the second image indicator in the second panel, it is convenient for the user to flexibly select the corresponding panel for labeling according to needs, and the labeling efficiency can be improved. In addition, the maximum number of selectable first image indicators in the first panel is less than the maximum number of selectable second image indicators in the second panel, which can make the first panel tend to perform more detailed labeling on a small number of medical images, and the second panel tend to perform batch labeling on a large number of similar or lesion-free medical images. Furthermore, by cooperating with the first panel and the second panel, it is possible to flexibly switch between a small number of labelings and a large number of labelings, and thus improve the labeling efficiency while reducing the misjudgment rate of batch labeling.

[0008] In addition, in the labeling system for medical images according to the first aspect of the present disclosure, optionally, the selected status of the first image indicator in the first panel is mutually exclusive with the selected status of the second image indicator in the second panel. In this case, the conflict between the first panel and the second panel can be reduced, the labeling unit can be simplified compared with setting additional controls to clear the selected image indicator, and the process of determining the selected medical image can also be simplified.

[0009] In addition, in the annotation system for medical images according to the first aspect of the present disclosure, optionally, the second image indicator includes an associated icon configured to affect the first panel to display the image data based on the medical image represented by the second image indicator, and the affecting includes resetting the first image indicator at a preset position in the first panel to represent the medical image represented by the second image indicator, and synchronously adjusting the first image indicators at other positions in the first panel except the preset position to display the image data in the time series. Thereby, it is convenient to perform more detailed analysis and annotation of the medical image that needs to be focused on in the second panel through the first panel.

[0010] In addition, in the annotation system for medical images according to the first aspect of the present disclosure, optionally, if the number of selected medical images is one, the annotation system is configured to, after determining the annotation result, display the annotation result in the third panel and in each corresponding image indicator of the selected medical image; and / or if the number of selected medical images is multiple, the annotation system is configured to, after determining the annotation result, display the annotation result in each corresponding image indicator of the selected medical images. In this case, after determining the annotation result, displaying the annotation result in each corresponding image indicator of the selected medical images facilitates viewing the annotation result. In addition, when the number of selected medical images is one, displaying it in the third panel at the same time can highlight the annotation result of a single annotation and facilitate further confirmation of the annotation result.

[0011] In addition, in the annotation system for medical images according to the first aspect of the present disclosure, optionally, when the data structure of the tag data is a tree data structure, the third panel is configured to display the tag data in a tree structure, where the root node and / or branch nodes of the tree structure represent the annotation task categories, and the leaf nodes of the tree structure represent the tags. In this case, displaying the tag data in a tree structure facilitates adding a hierarchical structure to the tags to expand the annotation tasks and can improve the scalability of the annotation system.

[0012] In addition, in the annotation system for medical images according to the first aspect of the present disclosure, optionally, the first panel is configured to display the image data in a first mode, and the second panel is configured to display the image data in a second mode. The first mode is a mode in which the display quantity is switchable, and the second mode is a mode in which the display quantity is non-switchable. In this case, the second panel can be made to focus more on batch annotation of similar or lesion-free medical images. In addition, in cooperation with the first panel that displays the image data in the first mode, it is convenient to switch back to the first panel as needed to perform single annotation or batch annotation of medical images through the first panel. Correspondingly, when performing single annotation or batch annotation of medical images through the first panel, it is also convenient to switch back to the second panel as needed to perform batch annotation of medical images through the second panel. Thus, through the cooperation of the two modes in the first panel and the second panel, it is convenient to flexibly switch between single annotation, small-batch annotation, and large-batch annotation, thereby improving the annotation efficiency.

[0013] In addition, in the annotation system for medical images according to the first aspect of the present disclosure, optionally, when the switched display quantity meets a preset condition, the first panel is further configured to display a plurality of operation controls. The plurality of operation controls are configured to adjust the parameters of the medical image displayed by the first image indicator and mark an area of interest on the medical image displayed by the first image indicator. In this case, it is convenient to discover the area of interest information in the medical image by adjusting the parameters, and it is also convenient to perform more detailed analysis and annotation on the medical image in the first panel. In addition, the area of interest can also provide more basis for training or validating the network.

[0014] In addition, in the annotation system for medical images according to the first aspect of the present disclosure, optionally, the display device is further configured to present a task list unit and a picture list unit. The task list unit is configured to display a plurality of annotation tasks in a table, and one row in the table represents one annotation task. The picture list unit is configured to display multiple medical images of a single annotation task. When the task list unit receives a details command, the display device is configured to display the picture list unit to replace the task list unit. When the task list unit receives an annotation command, the display device is configured to display the annotation unit to replace the task list unit. When the picture list unit receives the annotation command, the display device is configured to display the annotation unit to replace the picture list unit. In this case, it is possible to enter the annotation unit in multiple ways, thereby further improving the annotation efficiency.

[0015] In addition, in the annotation system for medical images according to the first aspect of the present disclosure, optionally, the first panel and the second panel are arranged in sequence along the vertical direction of the annotation unit, the length of the first panel is less than or equal to the length of the second panel, and the height of the first panel is greater than the height of the second panel; or the first panel and the second panel are arranged in sequence along the horizontal direction of the annotation unit, the length of the first panel is greater than the length of the second panel, and the height of the first panel is less than or equal to the height of the second panel. In this case, it is convenient to display more detailed information of the medical image through the first panel and display more medical images through the second panel. Thus, by cooperating with the first panel and the second panel, it is possible to further improve the annotation efficiency while reducing the misjudgment rate of batch annotation.

[0016] The second aspect of the present disclosure provides a method for annotating medical images, including: displaying image data through at least one first image indicator in a first panel of an annotation unit, each first image indicator representing a medical image in the image data, and the image data being obtained by a collection device collecting medical images in a time series; displaying the image data through at least one second image indicator in a second panel of the annotation unit, each second image indicator representing a medical image in the image data, wherein the maximum number of selectable first image indicators in the first panel is less than the maximum number of selectable second image indicators in the second panel; displaying label data through a plurality of label indicators in a third panel of the annotation unit, each label indicator representing a label in the label data; and when a determination command is received, determining the selected medical image and the selected label based on the selection status of the first image indicator in the first panel, the selection status of the second image indicator in the second panel, and the selection status of the label indicator in the third panel, and determining the annotation result of the selected medical image based on the selected label.

[0017] In the second aspect of the present disclosure, determining the selected medical image by integrating the selection status of the first image indicator in the first panel and the selection status of the second image indicator in the second panel facilitates the user to flexibly select the corresponding panel for annotation according to needs and can improve the annotation efficiency. In addition, the maximum number of selectable first image indicators in the first panel is less than the maximum number of selectable second image indicators in the second panel, which can make the first panel tend to perform more detailed annotation on a small number of medical images, and the second panel tend to perform batch annotation on a large number of similar or lesion-free medical images. Furthermore, by cooperating with the first panel and the second panel, it is possible to flexibly switch between a small number of annotations and a large number of annotations, and thus improve the annotation efficiency while reducing the misjudgment rate of batch annotation.

[0018] In addition, in the annotation system for medical images according to the first aspect of the present disclosure, optionally, the image data is displayed in a first mode on the first panel and in a second mode on the second panel. The first mode is a mode in which the display quantity is switchable, and the second mode is a mode in which the display quantity is not switchable. In this case, the second panel can be made to focus more on batch annotation of similar or lesion-free medical images. In addition, by cooperating with the first panel that displays the image data in the first mode, it is convenient to switch back to the first panel as needed to perform single or batch annotation of medical images through the first panel. Correspondingly, when performing single or batch annotation of medical images through the first panel, it is also convenient to switch back to the second panel as needed to perform batch annotation of medical images through the second panel. Thus, through the cooperation of the two modes in the first panel and the second panel, it is convenient to flexibly switch between single annotation, small batch annotation, and large batch annotation, and thus the annotation efficiency can be improved.

[0019] According to the present disclosure, there is provided an annotation system for medical images that can improve the annotation efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The present disclosure will now be further explained in detail only by way of examples with reference to the accompanying drawings.

[0021] Figure 1 is an exemplary schematic diagram showing the annotation environment of the medical images involved in the examples of the present disclosure.

[0022] Figure 2 is an exemplary block diagram showing the annotation system involved in the examples of the present disclosure.

[0023] Figure 3 is an exemplary schematic diagram showing the display device involved in the examples of the present disclosure.

[0024] Figure 4 is an exemplary block diagram showing the main unit involved in the examples of the present disclosure.

[0025] Figure 5A is an exemplary schematic diagram showing the annotation unit involved in the examples of the present disclosure.

[0026] Figure 5B is an exemplary schematic diagram showing the annotation unit after determining the annotation result involved in the examples of the present disclosure.

[0027] Figure 6 is an exemplary schematic diagram showing the first image indicator involved in the examples of the present disclosure.

[0028] Figure 7 is an exemplary schematic diagram showing the second image indicator involved in the examples of the present disclosure.

[0029] Figure 8 It is an exemplary schematic diagram showing the third panel related to the examples of the present disclosure.

[0030] Figure 9 It is an exemplary schematic diagram showing the task list unit related to the examples of the present disclosure.

[0031] Figure 10 It is an exemplary schematic diagram showing the picture list unit related to the examples of the present disclosure.

[0032] Figure 11 It is an exemplary flowchart showing the annotation method related to the examples of the present disclosure. Detailed Embodiments

[0033] Hereinafter, with reference to the accompanying drawings, the preferred embodiments of the present disclosure will be described in detail. In the following description, the same reference numerals are given to the same components, and repeated descriptions are omitted. In addition, the drawings are only schematic diagrams, and the proportional relationship of the sizes between components or the shapes of components may be different from the actual ones. It should be noted that the terms "include" and "have" in the present disclosure and any variations thereof, for example, a process, method, system, product or device including or having a series of steps or units do not necessarily have to be limited to those steps or units clearly listed, but may include or have other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0034] First, an introduction to the related terms involved in the present disclosure will be made.

[0035] "Annotation result" may refer to the label of an image determined after annotating the image. In the field of artificial intelligence, the annotation result can be used to indicate the content that a neural network needs to predict or learn.

[0036] "Annotation" may refer to the process of manually marking or annotating a data set (such as training data). Annotation can include operations such as classifying, bounding box marking, or semantic segmentation of images, text, or other types of data. For example, for medical images, the labels provided by annotation can be marks for target regions (such as lesion regions) on medical images, or classifications for the entire medical image.

[0037] "Tree - shaped data structure" may refer to a hierarchical data structure composed of nodes, presenting a shape similar to a tree.

[0038] "User interface" may refer to an access point for a user to interact and communicate with a device or system. It can allow the user to communicate and operate with the device or system through forms such as a graphical interface, a command - line interface, voice interaction, a touch screen, or virtual reality.

[0039] "Screen space" may refer to the 2D space on a display device.

[0040] Various commands involved in the examples of the present disclosure, such as determination commands, detail commands, annotation commands, selection commands, cancellation commands, return commands, may, without special specification, be from the user interface of an annotation system. For example, various commands may be triggered in response to receiving an operation from a user.

[0041] A single annotation involved in the examples of the present disclosure may refer to annotating a single image, a small batch or a small number of annotations may refer to simultaneously annotating multiple images, and a large batch or a large number of annotations may refer to simultaneously annotating more images compared to a relatively small batch of annotations.

[0042] The selection control involved in the examples of the present disclosure may be used to receive selection commands and cancellation commands to determine the selection status of corresponding controls (such as the first image indicator, the second image indicator, and the label indicator described later). Additionally, the selection status may include at least one of a selected state, an unselected state, and an unknown state.

[0043] As described above, if a large number of medical images need to be annotated, how to improve the annotation efficiency remains to be studied. Through research, the inventors have provided some solutions, and the corresponding embodiments can at least improve the annotation efficiency, thereby being able to increase the number of images with annotation results. Additionally, in some scenarios, the efficiency of locating key images can also be improved. The examples of the present disclosure will be described in detail below.

[0044] An annotation system for images involved in the examples of the present disclosure may be configured to annotate an image to determine the annotation result of the image. Additionally, the annotation system for images involved in the examples of the present disclosure may also be referred to as an annotation system, an image processing system, a film reading system, etc. In some examples, the image may be a medical image, and correspondingly, the annotation system may also be referred to as an annotation system for medical images. In other examples, the image may also be other images other than medical images.

[0045] Additionally, the annotation system can facilitate batch annotation and is particularly suitable for datasets with similar or identical content. In some examples, for medical images, the annotation system is particularly suitable for annotating image data collected by an in-vivo acquisition device. Specifically, the annotation system may be configured to annotate image data collected by an in-vivo acquisition device in a time series to determine the annotation result of the medical image in the image data. In this case, the image data collected by the in-vivo acquisition device generally has a correlation between medical images and relatively small changes in image content within a certain period of time, being suitable for batch annotation, and the annotation efficiency can be improved through the annotation system.

[0046] In some examples, the image data can be a medical image. For example, the image data can be a gastroscope image.

[0047] For convenience of description hereinafter, some examples are described by taking the image as a medical image, and the medical image is further taken as a gastroscope image for example. It should be noted that this does not represent a limitation to the present disclosure. Unless there is a contradiction, the relevant description also applies to other types of images.

[0048] Examples of the present disclosure will be described in detail below with reference to the accompanying drawings. Figure 1 is an exemplary schematic diagram showing the annotation environment of the medical image involved in the examples of the present disclosure.

[0049] Reference Figure 1 , the annotation environment may include a terminal device 700. The terminal device 700 can be configured to integrate at least a part of the annotation system 1 (described later) to provide a user unit 100 for annotation. In addition, the terminal device 700 can be a desktop terminal or a mobile terminal. In some examples, the mobile terminal may include at least one of a mobile phone, a tablet computer, and a laptop computer.

[0050] In some examples, the annotation environment may further include a server 800. The server 800 can be configured to provide remote data services for the annotation system 1. For example, the terminal device 700 can request data (such as an annotation task, image data, or label data, etc.) from the server 800 and / or store data to the server 800. In other examples, for the image data, it can also be directly stored in the terminal device 700.

[0051] In some examples, the annotation environment may further include a collection device 900. The collection device 900 can be configured to collect medical images to obtain image data. In some examples, the collection device 900 can be configured to collect medical images in a time series. In some examples, the image data obtained by the collection device 900 can be transmitted to the server 800 and / or the terminal device 700.

[0052] In some examples, the collection device 900 may include an in-vivo collection device and / or an in-vitro collection device. In addition, the in-vivo collection device can also be called a medical endoscope. At least a part of the in-vivo collection device can enter the interior of the subject. Such a device usually consists of two parts, an endoscope and a camera, and can enter the body through the natural cavity or incision of the subject to observe the internal situation of the subject, thereby helping the doctor make a more accurate judgment. In some examples, the in-vivo collection device may include at least one of a capsule endoscope, a tubular endoscope, and an ultrasonic endoscope. In addition, the in-vitro collection device can be a device located outside the subject. For example, the in-vitro collection device may include at least one of a fundus camera, an electrocardiograph, and an ultrasonic endoscope.

[0053] Continue to refer to Figure 1 , the terminal device 700, the server 800, and the acquisition device 900 can be connected through a computer network.

[0054] Figure 2 is an exemplary block diagram showing the annotation system 1 involved in the examples of the present disclosure. Figure 3 is an exemplary schematic diagram showing the display device 10 involved in the examples of the present disclosure. Figure 4 is an exemplary block diagram showing the main unit 110 involved in the examples of the present disclosure.

[0055] Refer to Figure 2 , in some examples, the annotation system 1 may include a display device 10. That is, the annotation system 1 may have a display device 10. Additionally, the display device 10 can be any device capable of displaying information. In some examples, the display device 10 can be a monitor. For example, the display device 10 can be the monitor of the terminal device 700.

[0056] In some examples, the display device 10 can be configured to present the user unit 100. Additionally, the number of user units 100 can be one or more. In some examples, the user unit 100 can include the main unit 110. In some examples, the main unit 110 can be configured to display content related to the annotation service.

[0057] In some examples, the user unit 100 may further include a navigation unit 120. In some examples, the navigation unit 120 can be configured to switch and display different main units 110. In some examples, the navigation unit 120 can be hidden. Thus, it is convenient to increase the screen space of the main unit 110. In some examples, the navigation unit 120 can be located to the left or above the main unit 110. Thus, it can reserve a convenient screen space for the user to operate on the main unit 110.

[0058] As an example, Figure 3 shows a schematic diagram in which the display device 10 is configured to simultaneously present the main unit 110 and the navigation unit 120, where the navigation unit 120 can be located to the left of the main unit 110. In the navigation unit 120, the entrances for switching and displaying different main units 110 can be arranged in a tree structure. It should be noted that this does not represent a limitation to the present disclosure, and the display device 10 can be configured to display any combination of user units 100 or a single user unit 100 in any layout.

[0059] As described above, the user unit 100 can include the main unit 110. Refer to Figure 4, in some examples, the main unit 110 may include an annotation unit 170. In some examples, the main unit 110 may further include a task list unit 180 and a picture list unit 190. In some examples, the display can be switched among the units included in the main unit 110. That is, only one unit can be displayed at the same moment.

[0060] In some examples, the task list unit 180 can be triggered to display through the navigation unit 120. For example, it can be triggered to display the task list unit 180 for gastroscopy by clicking on the area corresponding to "gastroscopy annotation" in Figure 3 . The picture list unit 190 can be triggered to display through the task list unit 180. In addition, the annotation unit 170 can be triggered to display in any way. In some examples, the annotation unit 170 can be triggered to display through at least one of the task list unit 180 and the picture list unit 190. In some examples, the annotation unit 170 can also be triggered to display through the navigation unit 120.

[0061] In some examples, the annotation system 1 may further include a storage device. In some examples, the storage device can be a memory. In some examples, the annotation system 1 may further include a processing device. In some examples, the processing device can be a processor.

[0062] In some examples, the display device 10 can be configured to present the annotation unit 170. In some examples, the display device 10 can be configured to only present the annotation unit 170 and switch to other units in response to the annotation unit 170 receiving a return command. In this case, more screen space can be provided for the annotation unit 170, and thus the annotation unit 170 can display more information (such as more medical images), facilitating the determination of annotation results within the annotation unit 170.

[0063] In addition, the annotation unit 170 can be configured to display image data and label data to determine the annotation result of the medical image in the image data. Hereinafter, for convenience of description, the annotation unit 200 is used to represent the annotation unit 170.

[0064] Figure 5A is an exemplary schematic diagram showing the annotation unit 200 involved in the examples of the present disclosure. Figure 5B is an exemplary schematic diagram showing the annotation unit 200 involved in the examples of the present disclosure after determining the annotation result. Figure 6 is an exemplary schematic diagram showing the first image indicator 2111 involved in the examples of the present disclosure. Figure 7 is an exemplary schematic diagram showing the second image indicator 2121 involved in the examples of the present disclosure. Figure 8 is an exemplary schematic diagram showing the third panel 213 involved in the examples of the present disclosure. Among them,Figure 5A The viewing mode in Figure 5B is to display 6 medical images, where Figure 5A and 5B the content of some panels 210 in

[0065] is not fully shown, which does not represent a limitation to the present disclosure. Figure 5A In some examples, referring to

[0066] the annotation unit 200 can be divided into multiple panels 210 that are displayed simultaneously. In some examples, the multiple panels 210 can include a first panel 211, a second panel 212, and a third panel 213. In some examples, the multiple panels 210 can further include a fourth panel 214. These panels 210 mentioned above will be described in detail later. Figure 5B In some examples, the first panel 211 and the second panel 212 can be configured to display medical images in the image data, and the third panel 213 can be configured to display label data. In some examples, when a determination command is received, the annotation system 1 can be configured to determine an annotation result of the medical image based on the first panel 211, the second panel 212, and the third panel 213. Referring to

[0067] Figure 5A As described above, in some examples, the multiple panels 210 can include a first panel 211. Continuing to refer to Figure 6 In some examples, the first panel 211 can be configured to display image data through at least one first image indicator 2111. Additionally, each first image indicator 2111 can represent a medical image in the image data. Referring to

[0068] Figure 6 Figure 5B

[0069] Figure 5B ).

[0069] In some examples, batch selection of the first image indicator 2111 can be achieved through the first selection control 2111b. For example, if it is recognized that multiple medical images have the same label, the first selection controls 2111b of the corresponding multiple first image indicators 2111 can be set to the selected state.

[0070] Continuing to refer to Figure 6 , in some examples, the first image indicator 2111 can include a first associated icon 2111c, and the first associated icon 2111c can be configured to affect the image data displayed on the second panel 212 based on the medical image represented by the first image indicator 2111. Additionally, the first associated icon 2111c of the first image indicator 2111 can also be referred to as the associated icon of the first image indicator 2111. That is, the first image indicator 2111 can include an associated icon.

[0071] In some examples, the influence of the above-mentioned first associated icon 2111c can include resetting the second image indicator 2121 at a preset position in the second panel 212 to represent the medical image represented by the first image indicator 2111, and synchronously adjusting the second image indicators 2121 at other positions in the second panel 212 except the preset position to display the image data in chronological order. In this case, it is convenient to identify more similar medical images of the medical images supporting batch annotation in the first panel 211 through the second panel 212, and thus it is convenient to perform batch annotation on more medical images. For example, when identifying more similar or lesion-free medical images, the first associated icon 2111c can be used to switch to the second panel 212 for annotation.

[0072] In some examples, the preset position can be the starting position, the middle position, or the ending position, etc. of the second panel 212. Specifically, it can be set according to needs. For example, if it is necessary to analyze other images near the corresponding medical image in the first panel 211, the preset position can be set to the middle position of the second panel 212.

[0073] In some examples, the image indicator can include a text control. The text control can be used to display the annotation result. Taking the first image indicator 2111 as an example, returning to the reference Figure 5B , the first image indicator 2111 can include a first text control 2111d. The first text control 2111d can be used to display the annotation result. The same applies to the second image indicator 2121, which will not be elaborated hereinafter.

[0074] Refer to Figure 6, in some examples, the first image indicator 2111 may include a viewing control 2111e. The viewing control 2111e may be configured to switch to the single-image mode of the medical image corresponding to the viewing control 2111e when the number of switched displays is greater than 1 (i.e., in the multi-image mode). That is, by clicking on the viewing control 2111e corresponding to a medical image, it can be switched to display only that medical image in the first panel 211.

[0075] In some examples, the first panel 211 may be configured to display image data in a first mode, and the first mode may be a mode with a switchable display number. Additionally, there may be multiple switchable display numbers. In this case, compared with the traditional single-image browsing method, there may be certain limitations in locating key images (such as abnormal images), especially when the lesion is small or the location is complex, which further increases the difficulty of location. The mode with a switchable display number facilitates combining different numbers of medical images to locate key images and can improve the efficiency of locating key images. In some examples, the switchable display numbers may include 1, 2, 4, 6, and 8. Additionally, the display number may be the number of first image indicators 2111 displayed simultaneously. As an example, Figure 5A A schematic diagram showing 6 first image indicators 2111 displayed simultaneously is shown.

[0076] In some examples, the display number may be set by a switching control 2141 in the fourth panel 214. In this case, the switching control 2141 is set in the fourth panel 214, which is convenient for the user to focus on analyzing and annotating the medical images in the first panel 211. In some other examples, the display number may also be set by any control that implements the function of the switching control 2141, and the position of this control may not be limited to the fourth panel 214. For example, this control may also be set around the first panel 211.

[0077] Reference Figure 5B , in some examples, when the switched display number meets a preset condition, the first panel 211 may also be configured to display multiple operation controls 2112, and the multiple operation controls 2112 may be configured to adjust the parameters of the medical image displayed by the first image indicator 2111 and mark the region of interest on the medical image displayed by the first image indicator 2111. In this case, it is convenient to discover the information of interest in the medical image by adjusting the parameters, and it is also convenient to perform more detailed analysis and annotation on the medical images in the first panel 211. Additionally, the region of interest can also provide more basis for training or validating the network.

[0078] Preferably, the preset condition may be when the switched display number is equal to 1. In this case, it is convenient to analyze and annotate the details of key or more difficult medical images.

[0079] In addition, the information of interest may be information related to the annotation result. In addition, the region of interest may be a region related to the annotation result. In some examples, the information of interest may be lesion information, and the region of interest may be a lesion region.

[0080] Reference Figure 5B , in some examples, the multiple operation controls 2112 may include at least one of a contrast control 2112a, a brightness control 2112b, a zoom control 2112c, and at least one annotation control 2112d. The contrast control 2112a may be configured to adjust the contrast of the medical image. The brightness control 2112b may be configured to adjust the brightness of the medical image. The zoom control 2112c may be configured to adjust the size of the medical image in the first image indicator 2111. The annotation control 2112d may be configured to mark the region of interest in the medical image. Additionally, the figure marked by the annotation control 2112d may be erasable.

[0081] In some examples, the annotation control 2112d may include at least one of a point control, a line control, a rectangle control, an ellipse control, and a center point control. The point control may be configured to draw a point. The line control may be configured to draw a line. The rectangle control may be configured to draw a rectangle. The ellipse control may be configured to draw an ellipse. The center point control may be configured to draw a point at the center of the medical image.

[0082] Continue to refer to Figure 5B , in some examples, the first panel 211 is further configured to display a first progress control 2113, and the first progress control 2113 may be configured to switch the display of the medical image.

[0083] In some examples, the first progress control 2113 may be configured to clear the selected state of the first image indicator 2111 that is switched to non-display when switching the display of the medical image. Thereby, the possibility of the undesired first image indicator 2111 being selected can be reduced.

[0084] In some examples, the first progress control 2113 may be configured to implement at least one of the functions of drag jump, forward and backward jump, pause, play, and adjust play speed.

[0085] In addition, the function of drag jump may be to make the first image indicator 2111 display the jumped-to medical image by dragging the progress bar on the first progress control 2113. Thereby, it is convenient to quickly locate the corresponding medical image.

[0086] In addition, the function of forward and backward jumping can make the first image indicator 2111 jump from displaying the current medical image to displaying the previous or next medical image. In this case, it is possible to quickly locate medical images with similar positions, facilitating the annotation of relevant medical images.

[0087] In addition, the function of playing can make the medical images in the image data be automatically displayed in sequence on the first panel 211. The function of pausing can be to pause the playback. The function of adjusting the playback speed can be to adjust the display speed of the medical images on the first panel 211. In this case, by coordinating functions such as playing, pausing, and adjusting the playback speed, it is convenient to browse the image data as needed to quickly locate the medical images to be annotated, check the annotation results, or perform other processing. For example, compared with a system without the first progress control 2113, users may need to spend more time and effort when browsing and observing consecutive medical images.

[0088] Reference Figure 5B , in some examples, the first progress control 2113 can also be configured to display time information related to the acquisition time of the medical images. The time information can at least include at least one of the start acquisition time, end acquisition time, and current acquisition time of the image data.

[0089] In some examples, the first panel 211 can also include a select all control 2114, and the first image indicator 2111 in the first panel 211 can be all selected through the select all control 2114. Specifically, the select all control 2114 can be configured to set the selection state of the selection controls in the first panel 211 to the selected state.

[0090] As described above, in some examples, the plurality of panels 210 can include a second panel 212. Return to reference Figure 5A , the second panel 212 can be configured to display the image data through at least one second image indicator 2121. In addition, each second image indicator 2121 can represent a medical image in the image data. In some examples, reference Figure 7 , the second image indicator 2121 can at least include a second image control 2121a. The second image control 2121a can be configured to display the medical image.

[0091] In some examples, the second panel 212 can be configured to display image data in a second mode. The second mode can be a mode with a non-switchable display quantity. In this case, the second panel 212 can be made to focus more on batch annotation of similar or lesion-free medical images. Additionally, by cooperating with the first panel 211 that displays image data in the first mode, it is convenient to switch back to the first panel 211 as needed to perform single annotation or batch annotation of medical images through the first panel 211. Correspondingly, when performing single annotation or batch annotation of medical images through the first panel 211, it is also convenient to switch back to the second panel 212 as needed to perform batch annotation of medical images through the second panel 212. Thus, through the cooperation of the two modes in the first panel 211 and the second panel 212, it is convenient to flexibly switch between single annotation, small batch annotation, and large batch annotation, thereby improving the annotation efficiency.

[0092] For example, when a medical image that needs to be focused on is recognized in the second panel 212, it can be switched back to the first panel 211 to facilitate single annotation or small batch annotation. Also, for example, when similar or lesion-free medical images are continuously recognized in the first panel 211, it can be switched back to the second panel 212 to facilitate large batch annotation.

[0093] Continuing to refer to Figure 7 , in some examples, the second image indicator 2121 can include a second selection control 2121b. The second selection control 2121b can be configured to determine the selection state of the second image indicator 2121. Refer to the above description for the selection control. In some examples, batch selection of the second image indicator 2121 can be achieved through the second selection control 2121b for batch annotation.

[0094] In some examples, the second image indicator 2121 can include a second associated icon 2121c. The second associated icon 2121c can be configured to affect the display of image data on the first panel 211 based on the medical image represented by the second image indicator 2121. Additionally, the second associated icon 2121c of the second image indicator 2121 can also be referred to as the associated icon of the second image indicator 2121. That is, the second image indicator 2121 can include an associated icon.

[0095] In some examples, the influence of the above second associated icon 2121c can include resetting the first image indicator 2111 at a preset position in the first panel 211 to represent the medical image represented by the second image indicator 2121, and synchronously adjusting the first image indicators 2111 at other positions in the first panel 211 except the preset position to display image data in a time series. Thus, it is convenient to perform more detailed analysis and annotation of the medical images that need to be focused on in the second panel 212 through the first panel 211.

[0096] In some examples, the preset position may be the starting position, the middle position, or the end position of the first panel 211. The preset position may be set as required. For example, if it is necessary to focus on other images near the corresponding medical image in the second panel 212, the preset position may be set to the middle position of the first panel 211.

[0097] In some examples, the second image indicator 2121 may include a second text control. The text control may be used to display the annotation result. For related descriptions, see the first text control 2111d.

[0098] In some examples, the second image indicator 2121 is also configured to display the acquisition time. In this case, it is convenient for the first panel 211 to locate the corresponding medical image by the acquisition time, thereby facilitating more detailed analysis and annotation in the first panel 211. In some examples, the second image indicator 2121 may include a third text control 2121d. The third text control 2121d may be configured to display the acquisition time (see Figure 7 ).

[0099] In some examples, the second image indicator 2121 may also be configured to display the image number, thereby making it easier to understand the marking progress.

[0100] Return to reference Figure 5A In some examples, the second panel 212 may include a second progress control 2122, and the second progress control 2122 may be configured to switch the display of the medical image. In some examples, the second progress control 2122 may be configured to not clear the selected state of the second image indicator 2121 that is switched to non-display until a confirmation command is received when switching the display of the medical image. Thus, it is convenient to select more second image indicators 2121 to further improve the annotation efficiency. In some examples, the second progress control 2122 may be a scroll bar. For example, the second progress control 2122 may be a scroll bar that comes with the browser. Thus, it is convenient to implement the second progress control 2122.

[0101] As described above, the first panel 211 and the second panel 212 can be configured to display medical images in the image data. The first image indicator 2111 in the first panel 211 and the second image indicator 2121 in the second panel 212 can both adjust the selection state to a selected state in response to receiving a selection command. That is, the first panel 211 and the second panel 212 can both be used to annotate the medical image.

[0102] In some examples, the maximum number of selectable first image indicators 2111 in the first panel 211 can be less than the maximum number of selectable second image indicators 2121 in the second panel 212. In this case, it is possible to make the first panel 211 tend to perform more detailed annotation on a small number of medical images, and the second panel 212 tend to perform batch annotation on a large number of similar or lesion-free medical images. Furthermore, by cooperating with the first panel 211 and the second panel 212, it is possible to flexibly switch between a small number of annotations and a large number of annotations, and thus improve the annotation efficiency while reducing the misjudgment rate of batch annotation. For example, if batch annotation is performed using the first panel 211, a specific number of medical images can be annotated at most at one time, but using the second panel 212, more medical images than the specific number can be selected for batch annotation.

[0103] In some examples, the selected state of the first image indicator 2111 in the first panel 211 and the selected state of the second image indicator 2121 in the second panel 212 can be mutually exclusive. That is, if the first image indicator 2111 in the first panel 211 is selected, the selected second image indicator 2121 in the second panel 212 is cleared; if the second image indicator 2121 in the second panel 212 is selected, the selected first image indicator 2111 in the first panel 211 is cleared. In other words, the selected image indicators between the first panel 211 and the second panel 212 can cover each other. In this case, the conflict between the first panel 211 and the second panel 212 can be reduced, and the annotation unit 200 can be simplified compared with setting additional controls to clear the selected image indicators. In addition, the process of determining the selected medical images can be simplified.

[0104] In some examples, the first panel 211 can be in front of the second panel 212. In this case, the user attention to the first panel 211 can be improved, which is convenient for guiding the user to preferentially perform more detailed analysis and annotation through the first panel 211.

[0105] Continue to refer to Figure 5A In some examples, the first panel 211 and the second panel 212 can be arranged in sequence along the vertical direction of the annotation unit 200. The length of the first panel 211 can be less than or equal to the length of the second panel 212, and the height of the first panel 211 can be greater than the height of the second panel 212. In this case, the higher first panel 211 is convenient for displaying more detailed information of medical images, and the longer second panel 212 is convenient for displaying more medical images. Thus, by cooperating with the first panel 211 and the second panel 212, the annotation efficiency can be further improved while reducing the misjudgment rate of batch annotation.

[0106] In some examples, for the first panel 211 and the second panel 212 arranged in sequence along the vertical direction of the annotation unit 200, the first panel 211 may be located above the second panel 212.

[0107] In some examples, for the first panel 211 and the second panel 212 arranged in sequence along the vertical direction of the annotation unit 200, the first panel 211 may support displaying the first image indicator 2111 in a row-column manner, and the second panel 212 may support displaying the second image indicator 2121 in a single-row manner. Thus, the utilization rate of the screen space can be improved when arranged in sequence in the vertical direction.

[0108] In some other examples, the first panel 211 and the second panel 212 may also be arranged in sequence along the horizontal direction of the annotation unit 200. The length of the first panel 211 may be greater than the length of the second panel 212, and the height of the first panel 211 may be less than or equal to the height of the second panel 212. In this case, the longer first panel 211 is convenient for displaying more detailed information of the medical image, and the higher second panel 212 is convenient for displaying more medical images. Thus, by coordinating the first panel 211 and the second panel 212, the annotation efficiency can be further improved while reducing the misjudgment rate of batch annotation.

[0109] In some examples, for the first panel 211 and the second panel 212 arranged in sequence along the horizontal direction of the annotation unit 200, the first panel 211 may be located to the left of the second panel 212.

[0110] In some examples, for the first panel 211 and the second panel 212 arranged in sequence along the horizontal direction of the annotation unit 200, the first panel 211 may support displaying the first image indicator 2111 in a row-column manner, and the second panel 212 may support displaying the second image indicator 2121 in a single-column manner. Thus, the utilization rate of the screen space can be improved when arranged in sequence in the horizontal direction.

[0111] In some examples, the size of the first image indicator 2111 in the first panel 211 may be greater than the size of the second image indicator 2121 in the second panel 212. In this case, the larger first image indicator 2111 is convenient for displaying more detailed information of the medical image, and the smaller second image indicator 2121 is convenient for displaying more medical images. Thus, by coordinating the first panel 211 and the second panel 212, the annotation efficiency can be further improved while reducing the misjudgment rate of batch annotation. Specifically, the size of the medical image displayed by the first image indicator 2111 may be greater than the size of the medical image displayed by the second image indicator 2121.

[0112] In addition, the present disclosure example also provides an example of using the annotation system 1. Specifically, taking a gastroscope image as an example, if there is no lesion in a certain part of the stomach, annotating one by one in the first panel 211 may be too inefficient. Therefore, it can be switched to an 8-image display mode, so that 8 gastroscope images can be analyzed simultaneously at one time. When a lesion is found in a certain part, it can be switched to a 1-image, 2-image, 3-image, or 4-image mode. In addition, if the switching speed is considered too slow during the switching process, the image data can be played through the first progress control 2113 and the playback speed can be controlled by combining the function of adjusting the playback speed. If more similar or lesion-free medical images are further recognized, the second panel 212 can be used for annotation. Similarly, if a lesion is found in a certain part during the annotation in the second panel 212, it can be switched back to the first panel 211 again.

[0113] As described above, in some examples, the multiple panels 210 may include a third panel 213. Refer to Figure 8 , in some examples, the third panel 213 may be configured to display label data through a plurality of label indicators 2131. In addition, the label data may at least include labels. In some examples, the label data may further include annotation task categories. Each label indicator 2131 may represent one label in the label data.

[0114] In some examples, when the data structure of the label data is a tree data structure, the third panel 213 may be configured to display the label data in a tree structure. In this case, displaying the label data in a tree structure facilitates adding a hierarchical structure of labels to expand the annotation task and can improve the scalability of the annotation system 1. For example, it is possible to easily add new annotation tasks. In some examples, the root node and / or branch nodes of the tree structure may represent annotation task categories, and the leaf nodes of the tree structure may represent labels. That is to say, the leaf nodes of the tree structure may be the label indicators 2131. For example, refer to Figure 8 , "image quality" and "clarity" may respectively represent different hierarchical annotation task categories, and "0 points" may represent a label.

[0115] In other examples, the data structure of the label data may also be a linear data structure. When the data structure of the label data is a linear data structure, the third panel 213 may be configured to display the label data through a plurality of juxtaposed label indicators 2131.

[0116] In some examples, the annotation task categories may include at least one of tasks such as image quality, lesion characteristics, lesion categories, and tissue parts.

[0117] In some examples, the label data can be configurable. In this case, it is convenient to add or modify the label data, thus facilitating the expansion of the annotation task. In some examples, mutual exclusion relationships and association relationships can be supported between the labels in the label data. In this case, the operations for the user to manually control the mutual exclusion relationships and association relationships can be reduced, and the annotation efficiency can be further improved.

[0118] In some examples, the labels in the label data can also support configuring the displayed color, whether it is a required item, whether it is a single selection, and whether it is a multiple selection, etc. Thus, more annotation scenarios can be adapted.

[0119] Continuing to refer to Figure 8 , in some examples, the label indicator 2131 can include a third selection control 2131a. The third selection control 2131a can be configured to determine the selection state of the label indicator 2131. Refer to the above description for the selection control.

[0120] As described above, in some examples, when a determination command is received, the annotation system 1 can be configured to determine the annotation result of the medical image based on the first panel 211, the second panel 212, and the third panel 213.

[0121] In some examples, the annotation system 1 can be configured to determine the selected medical image and the selected label based on the selection state of the first image indicator 2111 in the first panel 211, the selection state of the second image indicator 2121 in the second panel 212, and the selection state of the label indicator 2131 in the third panel 213, and determine the annotation result of the selected medical image based on the selected label. In this case, determining the selected medical image by integrating the selection state of the first image indicator 2111 in the first panel 211 and the selection state of the second image indicator 2121 in the second panel 212 facilitates the user to flexibly select the corresponding panel 210 for annotation according to needs, and can improve the annotation efficiency.

[0122] In some examples, the selected medical image can be determined based on the selection state of the first image indicator 2111 in the first panel 211 and the selection state of the second image indicator 2121 in the second panel 212. In some examples, the medical image corresponding to the image indicator with the selected state among the first image indicator 2111 and the second image indicator 2121 can be used as the selected medical image.

[0123] In some examples, for the case where the image indicators between the first panel 211 and the second panel 212 are mutually exclusive, any one of the image indicators in the first panel 211 and the second panel 212 can be traversed first. If the selection state of the image indicator in one panel 210 is recognized as the selected state, then continue to traverse the image indicators in that panel 210 to determine the selected medical image, and skip traversing the image indicators in the other panel 210. Thereby, the number of times of traversing the image indicators in the first panel 211 and the second panel 212 can be reduced.

[0124] In some examples, the first image indicator 2111 with the selection state of the selected state in the first panel 211 can be obtained as the first selection list, and the second image indicator 2121 with the selection state of the selected state in the second panel 212 can be obtained as the second selection list. The medical image in the non-empty list among the first selection list and the second selection list is selected as the selected medical image.

[0125] In some examples, for the case where the first selection control 2111b is not displayed or not included in the first panel 211 as described above, the selection state of the first image indicator 2111 displayed in the first panel 211 can be default set to the selected state to determine the selected medical image.

[0126] In some examples, the selected label can be determined based on the selection state of the label indicator 2131 in the third panel 213. In some examples, the label corresponding to the label indicator 2131 with the selection state of the selected state can be used as the selected label.

[0127] In some examples, the selected label can be used as the annotation result of the selected medical image to determine the annotation result.

[0128] As described above, in some examples, the determined annotation result can be displayed in the corresponding panel 210 in the annotation unit 200. In some examples, the annotation system 1 can be configured to display the annotation result in each corresponding image indicator of the selected medical image after determining the annotation result. Thereby, it is convenient to view the annotation result.

[0129] In some examples, if the number of selected medical images is one, the annotation system 1 can be configured to display the annotation result in the third panel 213 and in each corresponding image indicator of the selected medical image after determining the annotation result. Thereby, the annotation result of a single annotation can be highlighted, facilitating further confirmation of the annotation result.

[0130] In some examples, if the number of selected medical images is multiple, the annotation system 1 can be configured to, after determining the annotation result, display the annotation result in each corresponding image indicator of the selected medical images.

[0131] In some examples, the annotation result can be displayed on the medical image in the image indicator. In some examples, the annotation result can partially cover the medical image. Thus, the screen space corresponding to the image indicator can be fully utilized. As an example, referring to Figure 5B , the annotation result can be displayed in the corresponding image indicator through a text control.

[0132] In some examples, the annotation result displayed in the image indicator can be hidden, and the hidden annotation result can be displayed. In this case, the annotation result and the medical image can be viewed as needed.

[0133] As described above, in some examples, the annotation system 1 can include a processing device. The processing device can be configured to display corresponding data in the first panel 211, the second panel 212, and the third panel 213, determine the annotation result of the medical image according to the first panel 211, the second panel 212, and the third panel 213, and display the determined annotation result in the corresponding panel 210 of the annotation unit 200.

[0134] As described above, in some examples, the multiple panels 210 can further include a fourth panel 214.

[0135] Returning to the reference Figure 5A , in some examples, the fourth panel 214 can be configured to display information related to at least one of the multiple panels 210.

[0136] In some examples, the fourth panel 214 can be located at the periphery of the annotation unit 200. Thus, it can facilitate reducing interference with the annotation. For example, the fourth panel 214 can be located at the leftmost or uppermost side of the annotation unit 200 (refer to Figure 5A ).

[0137] Continuing to refer to Figure 5A , in some examples, the fourth panel 214 can include a switching control 2141, and the switching control 2141 can be configured to set the display quantity of the first panel 211. In response to the triggering of the switching control 2141, the first panel 211 can be configured to simultaneously display the corresponding quantity of medical images.

[0138] In some examples, the fourth panel 214 may include a return control 2142. The return control 2142 may be configured to leave the annotation unit 200 and return to the previous unit. In some examples, the fourth panel 214 may include a task control 2143, and the task control 2143 may be configured to display information about the annotation task. In some examples, the fourth panel 214 may include a submission control 2144, and the submission control 2144 may be configured to receive a determination command to determine the annotation result. In some examples, the fourth panel 214 may include a modification control 2145, and the modification control 2145 may be configured to receive a modification command to modify the determined annotation result.

[0139] Figure 9 is an exemplary schematic diagram showing the task list unit 180 involved in the examples of the present disclosure.

[0140] As described above, in some examples, the main unit 110 may include a task list unit 180, and the task list unit 180 may be configured to display a plurality of annotation tasks. In some examples, the display device 10 may be configured to present the task list unit 180. In some examples, when the task list unit 180 receives a details command, the display device 10 may be configured to display a picture list unit 190 instead of the task list unit 180. In some examples, when the task list unit 180 receives an annotation command, the display device 10 may be configured to display an annotation unit 200 instead of the task list unit 180.

[0141] Reference Figure 9 , in some examples, the task list unit 180 may be configured to display a plurality of annotation tasks in a table 181, and a row 1811 in the table 181 may represent an annotation task. In some examples, the row 1811 in the table 181 may further include a details button 1811a, and the details button 1811a may be configured to receive a details command. In some examples, the row 1811 in the table 181 may include an annotation button 1811b, and the annotation button 1811b may be configured to receive an annotation command.

[0142] Figure 10 is an exemplary schematic diagram showing the picture list unit 190 involved in the examples of the present disclosure.

[0143] As described above, in some examples, the main unit 110 may include a picture list unit 190. The picture list unit 190 may be configured to display multiple medical images of a single annotation task. In some examples, the display device 10 may be configured to present the picture list unit 190. In some examples, when the picture list unit 190 receives an annotation command, the display device 10 may be configured to display an annotation unit 200 instead of the picture list unit 190.

[0144] In some examples, the picture list unit 190 may include a third image indicator 191, and the third image indicator 191 may represent a medical image in a single annotation task. In some examples, the third image indicator 191 may be configured to receive an annotation command. As described above, when the annotation command is received, the display device 10 is configured to display the annotation unit 200 in place of the picture list unit 190. For example, when the third image indicator 191 is clicked, the annotation unit 200 can be directly entered. In some examples, when the annotation unit 200 is displayed, the first first image indicator 2111 of the annotation unit 200 may represent the medical image represented by the third image indicator 191 that triggers the display of the annotation unit 200.

[0145] As described above, when the task list unit 180 receives an annotation command, the display device 10 is configured to display the annotation unit 200 in place of the task list unit 180, and when the picture list unit 190 receives an annotation command, the display device 10 is configured to display the annotation unit 200 in place of the picture list unit 190. In this case, the annotation unit 200 can be entered in multiple ways, and thus the annotation efficiency can be further improved.

[0146] Figure 11 It is an exemplary flowchart showing the annotation method related to the examples of the present disclosure.

[0147] In addition, the examples of the present disclosure also provide a method for annotating medical images. By executing the annotation method, the image data can be annotated to determine the annotation result of the medical image in the image data. As described above, in some examples, the image data may be obtained by the acquisition device 900 acquiring medical images in a time series. In addition, the annotation method may be implemented by at least some components in the above-mentioned annotation system 1. For example, the annotation method may be implemented by the processing device of the annotation system 1. It should be noted that unless there is a contradiction, the above-related description of the annotation system 1 also applies to the annotation method. In addition, the annotation method may also be referred to as an image processing method or a film reading method, etc.

[0148] Reference Figure 11, the annotation method may include: displaying image data through at least one first image indicator 2111 in the first panel 211 of the annotation unit 200 (step S110), displaying image data through at least one second image indicator 2121 in the second panel 212 of the annotation unit 200 (step S120), displaying label data through a plurality of label indicators 2131 in the third panel 213 of the annotation unit 200 (step S130), and when a determination command is received, determining an annotation result of the image data based on the selection state of the first image indicator 2111 in the first panel 211, the selection state of the second image indicator 2121 in the second panel 212, and the selection state of the label indicator 2131 in the third panel 213 (step S140).

[0149] Continue to refer to Figure 11 , in some examples, in step S110, each first image indicator 2111 may represent a medical image in the image data.

[0150] In some examples, the image data may be displayed in the first panel 211 in a first mode, and the first mode may be a mode with a switchable display quantity. In some examples, when the switched display quantity meets a preset condition, a plurality of operation controls 2112 may be displayed in the first panel 211.

[0151] In some examples, the selection state of the first image indicator 2111 may be determined by the first selection control 2111b. In some examples, the second panel 212 may be affected to display image data based on the medical image represented by the first image indicator 2111 through the first association icon 2111c. In some examples, when the switched display quantity is greater than 1, the viewing control 2111e may be used to switch to the single-image mode of the medical image corresponding to the viewing control 2111e. In some examples, the medical images may be switched for display through the first progress control 2113. In some examples, all the first image indicators 2111 in the first panel 211 may be selected through the all-selection control 2114.

[0152] Continue to refer to Figure 11 , in some examples, in step S120, each second image indicator 2121 may represent a medical image in the image data.

[0153] In some examples, image data can be displayed in a second mode on the second panel 212. The second mode can be a mode with a non-switchable display quantity. In some examples, the selection state of the second image indicator 2121 can be determined by the second selection control 2121b. In some examples, based on the medical image represented by the second image indicator 2121, the first panel 211 can be affected to display image data through the second associated icon 2121c. In some examples, the maximum quantity of selectable first image indicators 2111 in the first panel 211 can be less than the maximum quantity of selectable second image indicators 2121 in the second panel 212. In some examples, the display of medical images can be switched through the second progress control 2122.

[0154] Continuing to refer to Figure 11 , in some examples, in step S130, each label indicator 2131 can represent one label in the label data.

[0155] In some examples, the selection state of the label indicator 2131 can be determined by the third selection control 2131a. In some examples, the third panel 213 can be displayed simultaneously with the first panel 211 and the second panel 212.

[0156] Continuing to refer to Figure 11 , in some examples, in step S140, the determined annotation result can be displayed in the corresponding panel 210 of the annotation unit 200.

[0157] In some examples, the annotation method can further include displaying information related to at least one of the plurality of panels 210 in the fourth panel 214 of the annotation unit 200.

[0158] In some examples, the annotation method can further include when the task list unit 180 receives an annotation command, displaying the annotation unit 200 to replace the task list unit 180, and when the picture list unit 190 receives an annotation command, displaying the annotation unit 200 to replace the picture list unit 190.

[0159] Examples of the present disclosure also disclose an electronic device, including a processor and a memory. The processor executes a program stored in the memory to implement one or more steps of the above-mentioned annotation method.

[0160] Examples of the present disclosure also disclose a computer-readable storage medium, which can store at least one instruction, and when the at least one instruction is executed by a processor, one or more steps of the above-mentioned annotation method are implemented.

[0161] Although the present disclosure has been specifically described above in connection with the accompanying drawings and examples, it is to be understood that the above description does not limit the present disclosure in any way. Those skilled in the art can make modifications and variations to the present disclosure as needed without departing from the spirit and scope of the present disclosure, and these modifications and variations all fall within the scope of the present disclosure.

Claims

1. An annotation system for medical images, characterized in that, The annotation system has a display device configured to present annotation units, and the annotation units are divided into a plurality of panels that are displayed simultaneously; the first panel among the plurality of panels is configured to display image data through at least one first image indicator, and each first image indicator represents a medical image in the image data, and the image data is obtained by a collection device collecting medical images in a time series; The second panel among the plurality of panels is configured to display the image data through at least one second image indicator, and each second image indicator represents a medical image in the image data, wherein the maximum number of selectable first image indicators in the first panel is less than the maximum number of selectable second image indicators in the second panel; the third panel among the plurality of panels is configured to display label data through a plurality of label indicators, and each label indicator represents a label in the label data; when a determination command is received, the annotation system is configured to determine the selected medical image and the selected label based on the selection state of the first image indicator in the first panel, the selection state of the second image indicator in the second panel, and the selection state of the label indicator in the third panel, and determine the annotation result of the selected medical image based on the selected label.

2. The annotation system for medical images according to claim 1, wherein The selected state of the first image indicator in the first panel is mutually exclusive with the selected state of the second image indicator in the second panel.

3. The annotation system for medical images according to claim 1, characterized in that, The second image indicator includes an associated icon, and the associated icon is configured to affect the first panel to display the image data based on the medical image represented by the second image indicator, and the influence includes resetting the first image indicator at a preset position in the first panel to represent the medical image represented by the second image indicator, and synchronously adjusting the first image indicators at other positions in the first panel except the preset position to display the image data in the time series.

4. The annotation system for medical images according to claim 1, wherein If the number of selected medical images is one, the annotation system is configured to, after determining the annotation result, display the annotation result in the third panel and in each corresponding image indicator of the selected medical image respectively; and / or if the number of selected medical images is multiple, the annotation system is configured to, after determining the annotation result, display the annotation result in each corresponding image indicator of the selected medical images.

5. The annotation system for medical images according to claim 1, characterized in that, When the data structure of the label data is a tree data structure, the third panel is configured to display the label data in a tree structure, wherein the root node and / or branch node of the tree structure represents an annotation task category, and the leaf node of the tree structure represents the label.

6. The annotation system for medical images according to any one of claims 1 to 5, characterized in that, The first panel is configured to display the image data in a first mode, the second panel is configured to display the image data in a second mode, the first mode is a mode with a switchable display quantity, and the second mode is a mode with a non-switchable display quantity.

7. The annotation system for medical images according to claim 6, wherein, When the number of displayed images to be switched meets a preset condition, the first panel is further configured to display a plurality of operation controls, and the plurality of operation controls are configured to adjust parameters of the medical image displayed by the first image indicator and mark regions of interest on the medical image displayed by the first image indicator.

8. The annotation system for medical images according to any one of claims 1 to 5 and 7, characterized in that, The display device is further configured to present a task list unit and a picture list unit; the task list unit is configured to display a plurality of annotation tasks in a table, and one row in the table represents one annotation task; the picture list unit is configured to display multiple medical images of a single annotation task; when the task list unit receives a detail command, the display device is configured to display the picture list unit to replace the task list unit, when the task list unit receives an annotation command, the display device is configured to display the annotation unit to replace the task list unit, and when the picture list unit receives the annotation command, the display device is configured to display the annotation unit to replace the picture list unit.

9. The annotation system for medical images according to claim 1, characterized in that, The first panel and the second panel are arranged in sequence along the vertical direction of the annotation unit, the length of the first panel is less than or equal to the length of the second panel, and the height of the first panel is greater than the height of the second panel; or the first panel and the second panel are arranged in sequence along the horizontal direction of the annotation unit, the length of the first panel is greater than the length of the second panel, and the height of the first panel is less than or equal to the height of the second panel.

10. A method for annotating medical images, characterized in that, Comprising: Image data is displayed in the first panel of the annotation unit through at least one first image indicator, and each first image indicator represents a medical image in the image data. The image data is obtained by an acquisition device acquiring medical images in a time series. The image data is displayed in the second panel of the annotation unit through at least one second image indicator, and each second image indicator represents a medical image in the image data. Among them, the maximum number of selectable first image indicators in the first panel is less than the maximum number of selectable second image indicators in the second panel; label data is displayed in the third panel of the annotation unit through a plurality of label indicators, and each label indicator represents a label in the label data; and when a determination command is received, the selected medical image and the selected label are determined based on the selection states of the first image indicators in the first panel, the second image indicators in the second panel, and the label indicators in the third panel, and the annotation result of the selected medical image is determined based on the selected label.

11. The annotation method for medical images according to claim 10, wherein, The image data is displayed in the first panel in a first mode and in the second panel in a second mode. The first mode is a mode with a switchable display number, and the second mode is a mode with a non-switchable display number.

Images