Method, device and storage medium for segmenting brain metastases

Abstract

The present application discloses a method, device, and storage medium for segmenting brain metastases. The method includes: acquiring multi-modal nuclear magnetic resonance images of the brain and registering the multi-modal nuclear magnetic resonance images; based on the registered multi-modal nuclear magnetic resonance images, using a whole-brain segmentation model to perform whole-brain segmentation to obtain whole-brain images after registration of each modality; and using a brain metastasis segmentation model to perform brain metastasis segmentation on the whole-brain images after registration of each modality to obtain a brain metastasis segmentation result. By using the solution of the present application, the accuracy and efficiency of brain metastasis segmentation can be improved.

Description

Technical Field

[0001] The present application generally relates to the field of artificial intelligence technology. More specifically, the present application relates to a method, device, and computer-readable storage medium for segmenting brain metastases. Background Art

[0002] Currently, in the treatment and postoperative follow-up of brain metastases, doctors usually need to refer to multi-modal nuclear magnetic resonance data (such as T1 modality, T2 modality) to identify brain metastases, manually search for the lesion area, and outline the lesion contour. Then, a treatment plan for gamma knife treatment is formulated based on the contour. In manually outlining the lesion contour, it often depends on the rich experience and subjective judgment accumulated by doctors over the years, which may lead to recognition errors, resulting in poor treatment planning effects and affecting the treatment effect. In addition, relying on manual operation is time-consuming and laborious.

[0003] Currently, there have been attempts to achieve automated identification and segmentation of brain metastases by using neural networks. However, the current methods only target single-modal nuclear magnetic resonance images and do not consider regions outside the brain, resulting in false positives outside the brain, making the recognition accuracy of brain metastases poor and the segmentation inaccurate.

[0004] In view of this, there is an urgent need to provide a solution for segmenting brain metastases in order to improve the accuracy and efficiency of brain metastasis segmentation. Summary of the Invention

[0005] In order to solve at least one or more of the above-mentioned technical problems, the present application proposes a solution for segmenting brain metastases in multiple aspects.

[0006] In a first aspect, the present application provides a method for segmenting brain metastases, including: collecting multi-modal nuclear magnetic resonance images of the brain and registering the multi-modal nuclear magnetic resonance images; based on the registered multi-modal nuclear magnetic resonance images, using a whole-brain segmentation model to perform whole-brain segmentation to obtain a whole-brain image after registration of each modality; and using a brain metastasis segmentation model to perform brain metastasis segmentation on the whole-brain image after registration of each modality to obtain a brain metastasis segmentation result.

[0007] In some embodiments, the multi-modal nuclear magnetic resonance images at least include nuclear magnetic resonance images of the T1 modality and nuclear magnetic resonance images of the T2 modality.

[0008] In some other embodiments, based on the registered multi-modal nuclear magnetic resonance images, using a whole-brain segmentation model to perform whole-brain segmentation to obtain a whole-brain image after registration of each modality includes: based on the registered multi-modal nuclear magnetic resonance images, using a whole-brain segmentation model to perform whole-brain segmentation to obtain a whole-brain segmentation label; and obtaining a whole-brain image after registration of each modality according to the whole-brain segmentation label.

[0009] In still other embodiments, based on the registered multi-modal nuclear magnetic resonance (NMR) images, whole-brain segmentation is performed using a whole-brain segmentation model to obtain whole-brain segmentation labels, including: selecting a target modal NMR image with the largest imaging area from the registered multi-modal NMR images; and performing whole-brain segmentation on the target modal NMR image using the whole-brain segmentation model to obtain whole-brain segmentation labels.

[0010] In still other embodiments, obtaining the whole-brain images after registration for each modality according to the whole-brain segmentation labels includes: matching the whole-brain segmentation labels with the NMR images of each modality to extract the whole-brain images after registration for each modality.

[0011] In still other embodiments, performing brain metastasis tumor segmentation on the whole-brain images after registration for each modality using a brain metastasis tumor segmentation model to obtain a brain metastasis tumor segmentation result includes: performing brain metastasis tumor segmentation on the whole-brain images after registration for each modality using the brain metastasis tumor segmentation model to obtain brain metastasis tumor segmentation labels; and obtaining the brain metastasis tumor segmentation result according to the brain metastasis tumor segmentation labels.

[0012] In still other embodiments, obtaining the brain metastasis tumor segmentation result according to the brain metastasis tumor segmentation labels includes: performing reverse matching of the metastasis tumor segmentation labels with the multi-modal NMR images to obtain the brain metastasis tumor segmentation result.

[0013] In still other embodiments, the whole-brain segmentation model and the brain metastasis tumor segmentation model include a Unet model, an FPN model, or a PSPNet model.

[0014] In a second aspect, the present application provides a device for segmenting brain metastasis tumors, including: a processor; and a memory storing computer instructions for segmenting brain metastasis tumors, which, when executed by the processor, implement multiple embodiments in the aforementioned first aspect.

[0015] In a third aspect, the present application provides a computer-readable storage medium storing computer program instructions for segmenting brain metastasis tumors, which, when executed by one or more processors, implement multiple embodiments in the aforementioned first aspect.

[0016] Through the scheme for segmenting brain metastases provided above, in the embodiments of the present application, whole-brain segmentation is performed after registering multiple-modal nuclear magnetic resonance images to obtain whole-brain images after registration of each modality, and then brain metastases are segmented in the whole-brain images after registration of each modality to obtain the segmentation results of brain metastases. Based on this, the embodiments of the present application can combine multiple-modal nuclear magnetic resonance images to provide more comprehensive feature information, and only use the registered whole-brain images as the input for brain metastases segmentation, avoiding false positives outside the whole brain, thereby improving the accuracy and efficiency of brain metastases segmentation. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present application will become readily understood. In the drawings, several embodiments of the present application are shown in an exemplary rather than restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0018] Figure 1 is an exemplary flowchart showing a method for segmenting brain metastases according to an embodiment of the present application;

[0019] Figure 2 is an exemplary flowchart showing the overall process for segmenting brain metastases according to an embodiment of the present application;

[0020] Figure 3 is an exemplary schematic diagram showing a whole-brain image segmented in the T1 modality according to an embodiment of the present application;

[0021] Figure 4 is an exemplary schematic diagram showing the segmentation results of brain metastases segmented in the T1 modality according to an embodiment of the present application;

[0022] Figure 5 is an exemplary structural block diagram showing a device for segmenting brain metastases according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0023] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0024] It should be understood that the terms "including" and "comprising" used in the specification and claims of the present application indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0025] It should also be understood that the terms used in the specification of this application are merely for the purpose of describing specific embodiments and are not intended to limit this application. As used in the specification and claims of this application, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include the plural forms. It should be further understood that the term "and / or" used in the specification and claims of this application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0026] As used in this specification and the claims, the term "if" can be interpreted as "when" or "once" or "in response to determining" or "in response to detecting" depending on the context. Similarly, the phrase "if determined" or "if [the described condition or event] is detected" can be interpreted as meaning "once determined" or "in response to determining" or "once [the described condition or event] is detected" or "in response to detecting [the described condition or event]" depending on the context.

[0027] The following will describe in detail the specific embodiments of this application with reference to the accompanying drawings.

[0028] Figure 1 is an exemplary flowchart showing a method 100 for segmenting brain metastases according to an embodiment of this application. As Figure 1 shown, at step S101, multiple-modal nuclear magnetic resonance (NMR) images of the brain are acquired, and the multiple-modal NMR images are registered. In some embodiments, the multiple-modal NMR images of the brain can be acquired by, for example, a magnetic resonance device. The aforementioned multiple-modal NMR images can include, but are not limited to, T1-modal NMR images and T2-modal NMR images. In particular, the aforementioned T1-modal NMR image and T2-modal NMR image can be enhanced T1-modal NMR image and enhanced T2-modal NMR image. Among them, the enhanced T1-modal NMR image can provide an active brain metastasis region, and the enhanced T2-modal NMR image can provide a strong image of the edema by NMR. Thus, more comprehensive image information can be obtained by combining multiple-modal NMR images. In addition, in some embodiments, it may also include, for example, T1 contrast modality (i.e., T1C) and fluid-attenuated inversion recovery modality (i.e., FLAIR), and this application makes no limitation in this regard.

[0029] Based on the collected multi-modal nuclear magnetic resonance (NMR) images of the brain, first, the multi-modal NMR images are registered. Specifically, by calculating the transformation relationship between the multi-modal images, the pixel points on each modality are made to maintain spatial consistency. In some implementation scenarios, the aforementioned transformation relationship includes, for example, translation, rotation, scaling, or non-linear transformation, etc. In some embodiments, registration can be achieved based on regions (such as the whole brain region) or features (such as the edges or corners of the whole brain, etc.) on the NMR images of each modality.

[0030] Next, at step S102, based on the registered multi-modal NMR images, a whole brain segmentation model is used for whole brain segmentation to obtain the whole brain images after registration for each modality. In some embodiments, based on the registered multi-modal NMR images, a whole brain segmentation model can be used for whole brain segmentation to obtain whole brain segmentation labels, and then the whole brain images after registration for each modality can be obtained according to the whole brain segmentation labels. Specifically, in some embodiments, the target modality NMR image with the largest imaging area is selected from the registered multi-modal NMR images to use the whole brain segmentation model for whole brain segmentation of the target modality NMR image to obtain whole brain segmentation labels. Further, the whole brain segmentation labels are matched with the NMR images of each modality to extract the whole brain images after registration for each modality. In some implementation scenarios, the aforementioned whole brain segmentation model can include, but is not limited to, Unet model, FPN model, or PSPNet model, and the present application makes no limitation in this regard.

[0031] That is, in the embodiments of the present application, by taking the modality image with the largest imaging area as the target modality image, inputting it into the whole brain segmentation model, performing whole brain segmentation through the whole brain segmentation model to obtain whole brain labels, and then matching the whole brain labels to other modality images. As an example, assume that the imaging area in the T1 modality NMR image is the largest. Input the T1 modality NMR image into the whole brain segmentation model for whole brain segmentation to obtain whole brain labels (i.e., the whole brain foreground region or the whole brain region). Then, match the whole brain labels to the T2 modality NMR image to obtain the whole brain images after registration for the T1 and T2 modalities. Similarly, if the imaging area in the T2 modality NMR image is the largest, match the whole brain labels after segmentation of the T2 modality NMR image to the T1 modality NMR image to obtain the whole brain images after registration for the T1 and T2 modalities. Based on this, regions outside the imaging area can be deleted and only the whole brain images are retained, avoiding false positives existing outside the whole brain images, thereby improving the accuracy of subsequent brain metastasis tumor segmentation. It can be understood that the matching in the embodiments of the present application is also image registration, and thus the matching process can also be achieved by methods such as regions or features, etc.

[0032] After obtaining the whole-brain images after the above-mentioned multi-modal registrations, at step S103, a brain metastasis segmentation model is used to segment the whole-brain images after multi-modal registrations to obtain the brain metastasis segmentation results. In some embodiments, the brain metastasis segmentation model can be used to segment the whole-brain images after multi-modal registrations to obtain brain metastasis segmentation labels, and the brain metastasis segmentation results can be obtained according to the brain metastasis segmentation labels. Specifically, in some embodiments, the metastasis segmentation labels are reversely matched with the nuclear magnetic resonance images of multiple modalities to obtain the brain metastasis segmentation results. In some implementation scenarios, the aforementioned brain metastasis segmentation model can include, but is not limited to, Unet model, FPN model or PSPNet model, and the present application does not make any restrictions in this regard.

[0033] That is, in the embodiments of the present application, first, the brain metastasis segmentation labels (i.e., the brain metastasis regions) are obtained by segmenting the whole-brain images for brain metastases, and then the brain metastasis segmentation labels are matched to the original nuclear magnetic resonance images of multiple modalities to obtain the final brain metastasis segmentation results. Similarly, this reverse matching is also image registration, and the reverse matching process can be achieved by methods such as regions or features.

[0034] Combined with the above description, it can be seen that in the embodiments of the present application, after registering the nuclear magnetic resonance images of multiple modalities and performing whole-brain segmentation, the whole-brain images after multi-modal registrations are obtained, so as to provide more comprehensive feature information by combining the nuclear magnetic resonance images of multiple modalities, and the regions outside the whole-brain images are deleted to avoid false positives existing outside the whole brain. Then, brain metastasis segmentation is performed on the whole-brain images after multi-modal registrations, so as to obtain accurate brain metastasis segmentation results and improve the accuracy and efficiency of brain metastasis segmentation.

[0035] Figure 2 is an exemplary flowchart showing the overall process for segmenting brain metastases according to an embodiment of the present application. It should be understood that Figure 2 is a specific implementation of the above Figure 1 in Method 100, so the above description about Figure 1 also applies to Figure 2 .

[0036] As Figure 2As shown in the figure, at step S201, nuclear magnetic resonance (NMR) images of the T1 modality and the T2 modality of the brain are acquired, and the NMR images of the T1 modality and the T2 modality are registered. In some embodiments, registration can be achieved based on regions or features on the NMR images of the T1 modality and the T2 modality. Then, at step S202, the target modality image with the largest imaging area is selected from the registered NMR images of the T1 modality and the T2 modality. As an example, assuming that the imaging area on the NMR image of the T1 modality is the largest, at step S203, the NMR image of the T1 modality is input into a whole-brain segmentation model for whole-brain segmentation to obtain whole-brain segmentation labels. In some implementation scenarios, the aforementioned whole-brain segmentation model can be, for example, a Unet model, an FPN model, or a PSPNet model, etc.

[0037] Based on the obtained whole-brain segmentation labels, at step S204, the whole-brain segmentation labels are matched to NMR images of other modalities (such as the T2 modality) to obtain whole-brain images after registration of each modality. Further, at step S205, the whole-brain images after registration of each modality are input into a brain metastasis segmentation model for brain metastasis segmentation to obtain brain metastasis segmentation labels. In some implementation scenarios, the aforementioned brain metastasis segmentation model can include, but is not limited to, a Unet model, an FPN model, or a PSPNet model. At step S206, the brain metastasis segmentation labels are inversely matched to NMR images of multiple modalities to obtain brain metastasis segmentation results. According to the obtained brain metastasis segmentation results, an effective and accurate treatment plan can be provided subsequently to better treat brain metastases.

[0038] Figure 3 is an exemplary schematic diagram showing a whole-brain image segmented in the T1 modality according to an embodiment of the present application. As Figure 3 shown in the figure, the segmented whole-brain images corresponding to the axial, sagittal, and coronal planes of the T1 modality are shown in sequence from left to right. According to the foregoing, the whole-brain labels can be obtained by inputting the NMR image of the T1 modality into a whole-brain segmentation model for whole-brain segmentation, and then matching the whole-brain labels to the NMR image of the T1 modality, the segmented whole-brain image can be obtained. Based on this, regions outside the whole-brain area can be deleted, and only the whole-brain image is retained, avoiding false positives and improving the segmentation accuracy.

[0039] Figure 4 is an exemplary schematic diagram showing the brain metastasis segmentation result segmented in the T1 modality according to an embodiment of the present application. As Figure 4As shown, the segmented brain metastasis segmentation results corresponding to the axial, sagittal, and coronal planes of the T1 modality are shown in sequence from left to right. According to the foregoing, the whole brain image segmented from the T1 modality can be input into the brain metastasis segmentation model for brain metastasis segmentation to obtain the brain metastasis segmentation label. Then, by reverse matching the brain metastasis segmentation label to the nuclear magnetic resonance image of the T1 modality, the segmented brain metastasis segmentation result can be obtained. Through the solution of the embodiment of the present application, an accurate brain metastasis segmentation result can be obtained to provide an effective and accurate treatment plan for the subsequent treatment, thereby better treating brain metastases.

[0040] Figure 5 FIG. is an exemplary structural block diagram showing a device 500 for segmenting brain metastases according to an embodiment of the present application. It can be understood that the device 500 may include the device of the embodiment of the present application, and the device implementing the solution of the present application may be a single device (such as a computing device) or a multifunctional device including various peripheral devices.

[0041] As Figure 5 shown, the device of the present application may further include a central processing unit or central processing unit ("CPU") 511, which may be a general-purpose CPU, a dedicated CPU, or other information processing and program execution units. Further, the device 500 may further include a mass storage 512 and a read-only memory ("ROM") 513, where the mass storage 512 may be configured to store various types of data, including various nuclear magnetic resonance images of multiple modalities, the whole brain images after registration of each modality, brain metastasis segmentation results, algorithm data, intermediate results, and various programs required to operate the device 500. The ROM 513 may be configured to store data and instructions for power-on self-test of the device 500, initialization of each functional module in the system, basic input / output driver programs of the system, and data and instructions required to boot the operating system.

[0042] Optionally, the device 500 may further include other hardware platforms or components, such as the shown tensor processing unit ("TPU") 514, graphics processing unit ("GPU") 515, field programmable gate array ("FPGA") 516, and machine learning unit ("MLU") 517. It can be understood that although various hardware platforms or components are shown in the device 500, this is only exemplary and not restrictive, and those skilled in the art can add or remove the corresponding hardware according to actual needs. For example, the device 500 may only include a CPU, related storage devices, and interface devices to implement the method for segmenting brain metastases of the present application.

[0043] In some embodiments, to facilitate the transfer and interaction of data with an external network, the device 500 of the present application further includes a communication interface 518, so that it can be connected to a local area network / wireless local area network ("LAN / WLAN") 505 through this communication interface 518, and then can be connected to a local server 506 or connected to the Internet ("Internet") 507 through this LAN / WLAN. Alternatively or additionally, the device 500 of the present application can also be directly connected to the Internet or a cellular network based on wireless communication technology through the communication interface 518, such as wireless communication technology based on the third generation ("3G"), the fourth generation ("4G") or the fifth generation ("5G"). In some application scenarios, the device 500 of the present application can also access the server 508 and database 509 of the external network as needed to obtain various known algorithms, data and modules, and can remotely store various data, such as various types of data or instructions for presenting, for example, nuclear magnetic resonance images of each modality, whole brain images after registration of each modality, segmentation results of brain metastases, etc.

[0044] The peripheral devices of the device 500 may include a display device 502, an input device 503, and a data transmission interface 504. In one embodiment, the display device 502 may include, for example, one or more speakers and / or one or more visual displays, which are configured to provide voice prompts and / or image video displays for the brain metastasis segmentation of the present application. The input device 503 may include, for example, a keyboard, a mouse, a microphone, a gesture capture camera, and other input buttons or controls, which are configured to receive the input of audio data and / or user instructions. The data transmission interface 504 may include, for example, a serial interface, a parallel interface, or a universal serial bus interface ("USB"), a small computer system interface ("SCSI"), serial ATA, FireWire, PCI Express, and a high-definition multimedia interface ("HDMI"), etc., which are configured for data transmission and interaction with other devices or systems. According to the solution of the present application, this data transmission interface 504 can receive nuclear magnetic resonance images of each modality collected by an MRI device and transmit various types of data or results including nuclear magnetic resonance images of each modality to the device 500.

[0045] The above-mentioned CPU 511, mass storage 512, ROM 513, TPU 514, GPU 515, FPGA 516, MLU 517, and communication interface 518 of the device 500 of the present application can be interconnected through a bus 519, and data interaction with the peripheral devices is achieved through this bus. In one embodiment, through this bus 519, the CPU 511 can control other hardware components and their peripheral devices in the device 500.

[0046] The above in combination with Figure 5Describes a device that can be used to perform the method for segmenting brain metastases in this application. It should be understood that the device structure or architecture here is only exemplary, and the implementation manner and implementation entity of this application are not limited by it, but can be changed without departing from the spirit of this application.

[0047] According to the above description in conjunction with the drawings, those skilled in the art can also understand that the embodiments of this application can also be implemented through software programs. Therefore, this application also provides a computer-readable storage medium, on which computer-readable instructions for segmenting brain metastases are stored. When the computer-readable instructions are executed by one or more processors, they can be used to implement the method for segmenting brain metastases described in this application in conjunction with the attached Figure 1 drawings.

[0048] It should be noted that although the operations of the method of this application are described in a specific order in the drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. On the contrary, the steps depicted in the flowchart can be changed in the order of execution. Additionally or alternatively, certain steps can be omitted, multiple steps can be combined into one step for execution, and / or one step can be decomposed into multiple steps for execution.

[0049] It should be understood that when terms such as "first", "second", "third", and "fourth" are used in the claims, the specification, and the drawings of this application, they are only used to distinguish different objects, rather than to describe a specific order. The terms "comprising" and "including" used in the specification and claims of this application indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0050] It should also be understood that the terms used in the specification of this application are only for the purpose of describing specific embodiments and are not intended to limit this application. As used in the specification and claims of this application, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms. It should also be further understood that the term "and / or" used in the specification and claims of this application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0051] Although the embodiments of the present application are as described above, the above content is only an example for the convenience of understanding the present application, and is not intended to limit the scope and application scenarios of the present application. Any person skilled in the art within the technical field of the present application may make any modifications and changes in the form of implementation and details without departing from the spirit and scope disclosed by the present application. However, the scope of patent protection of the present application shall still be subject to the scope defined by the appended claims.

[0052] In addition, the collection and acquisition of various data in the present application comply with relevant laws and regulations and are authorized by the data providers. Any organization or individual that needs to obtain external data shall obtain authorization in accordance with the law and ensure data security, and shall not illegally collect, use, process, or transmit unauthorized or unprotected data, nor illegally buy, provide, or disclose unauthorized or unprotected data.

Claims

1. A method for segmenting a brain metastasis, comprising: Acquiring multiple modal nuclear magnetic resonance images of the brain, and registering the multiple modal nuclear magnetic resonance images; Based on the registered MRI images of multiple modalities, the whole brain segmentation model is used to perform whole brain segmentation to obtain the registered whole brain images of each modality; as well as Using the brain metastasis segmentation model to perform brain metastasis segmentation on the whole brain image after the modality registration, so as to obtain a brain metastasis segmentation result, Based on the registered MRI images of multiple modalities, the whole brain segmentation model is used to perform whole brain segmentation to obtain the whole brain images after registration of each modality, including: Selecting the target modality nuclear magnetic resonance image with the largest imaging area from the registered multiple modality nuclear magnetic resonance images; Inputting the target modality MRI image into the whole-brain segmentation model to perform whole-brain segmentation to obtain a whole-brain segmentation label; and The whole-brain segmentation label is registered to the MRI images of modalities other than the target modality MRI image to extract the registered whole-brain images of each modality. 2 . The method according to claim 1 , wherein the multiple modalities of nuclear magnetic images include at least T1 modality nuclear magnetic images and T2 modality nuclear magnetic images.

3. The method according to claim 1, wherein using the brain metastasis segmentation model to perform brain metastasis segmentation on the whole brain image after the modality registration to obtain the brain metastasis segmentation result comprises: Using the brain metastasis segmentation model to perform brain metastasis segmentation on the whole brain image after the modality registration to obtain a brain metastasis segmentation label; as well as A brain metastasis segmentation result is obtained according to the brain metastasis segmentation label.

4. The method according to claim 3, wherein obtaining a brain metastasis segmentation result according to the brain metastasis segmentation label comprises: The metastasis tumor segmentation label is reversely matched with the magnetic resonance imaging images of the multiple modalities to obtain the brain metastasis tumor segmentation result.

5. The method according to claim 1, wherein the whole brain segmentation model and the brain metastasis segmentation model comprise a Unet model, a FPN model or a PSPNet model.

6. An apparatus for segmenting a brain metastasis, comprising: processor; as well as A memory having computer instructions for segmenting brain metastases stored thereon, which, when executed by a processor, implement the method according to any one of claims 1 to 5.

7. A computer-readable storage medium having stored thereon computer program instructions for segmenting brain metastases, wherein when the computer program instructions are executed by one or more processors, the method according to any one of claims 1 to 5 is implemented.

Images