CN118096845A - Medical image processing method and device, electronic equipment and storage medium - Google Patents
Medical image processing method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN118096845A CN118096845A CN202410519698.1A CN202410519698A CN118096845A CN 118096845 A CN118096845 A CN 118096845A CN 202410519698 A CN202410519698 A CN 202410519698A CN 118096845 A CN118096845 A CN 118096845A
- Authority
- CN
- China
- Prior art keywords
- target
- image
- ternary diagram
- voxel
- medical image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 18
- 238000010586 diagram Methods 0.000 claims abstract description 201
- 230000003902 lesion Effects 0.000 claims abstract description 144
- 238000012545 processing Methods 0.000 claims abstract description 78
- 238000001959 radiotherapy Methods 0.000 claims abstract description 77
- 239000011159 matrix material Substances 0.000 claims abstract description 39
- 238000000034 method Methods 0.000 claims description 37
- 238000004590 computer program Methods 0.000 claims description 6
- 230000008602 contraction Effects 0.000 claims description 5
- 238000005516 engineering process Methods 0.000 abstract description 3
- 230000036285 pathological change Effects 0.000 abstract 2
- 231100000915 pathological change Toxicity 0.000 abstract 2
- 238000002600 positron emission tomography Methods 0.000 description 34
- 238000004422 calculation algorithm Methods 0.000 description 11
- 230000011218 segmentation Effects 0.000 description 11
- 230000008569 process Effects 0.000 description 6
- 230000005855 radiation Effects 0.000 description 6
- 238000012549 training Methods 0.000 description 6
- 206010028980 Neoplasm Diseases 0.000 description 4
- 238000013135 deep learning Methods 0.000 description 4
- 230000003044 adaptive effect Effects 0.000 description 3
- 238000003384 imaging method Methods 0.000 description 3
- 210000000920 organ at risk Anatomy 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000000295 complement effect Effects 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 206010060862 Prostate cancer Diseases 0.000 description 1
- 208000000236 Prostatic Neoplasms Diseases 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000013475 authorization Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000009545 invasion Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 210000002307 prostate Anatomy 0.000 description 1
- 210000001519 tissue Anatomy 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Landscapes
- Image Analysis (AREA)
Abstract
The application discloses a medical image processing method, a medical image processing device, electronic equipment and a storage medium, and relates to the field of medical science and technology. The medical image processing method comprises the following steps: determining a ternary diagram aligned with the medical image according to the medical image shot by the target object before radiotherapy, wherein the ternary diagram is used for representing the probability that voxels in the medical image belong to a lesion area of the target object; acquiring a target image shot by a target object during radiotherapy; obtaining a target ternary diagram aligned with the target image by registering the medical image and the target image; and determining the probability that the voxels in the target ternary diagram belong to the lesion region of the target object by calculating the spatial correlation matrix of different voxels in the target ternary diagram. The application solves the technical problem of poor accuracy in the prior art when determining the pathological change area and the non-pathological change area in the image shot during the radiotherapy of the patient.
Description
Technical Field
The present application relates to the field of medical science and technology, and in particular, to a medical image processing method, device, electronic apparatus, and storage medium.
Background
The self-adaptive radiotherapy can observe the change of the shape, the size and the position of the tumor according to the image of the current day in the treatment process, and adjust the radiotherapy plan in real time so as to improve the accuracy and the effect of the radiotherapy.
At present, one technical difficulty in the adaptive radiotherapy process is to delineate the target region on the image of the current day. The mainstream approach is MR (Magnetic Resonance ) guided adaptive radiotherapy.
However, for some target areas and organs at risk, such as the intralesional lesions of the prostate, which are the most common sites for local recurrence after treatment of prostate cancer, special delineation is required for high dose radiation treatment. For such target areas with particularly unclear boundaries, precise delineation on MR images remains a difficulty, usually requiring additional PET images to be taken for auxiliary delineation, however, PET images are very expensive to take and have radiation, and for cost and patient health considerations, usually can only be taken before treatment, and the current day PET (Positron Emission Tomography ) images for each treatment in a session cannot be acquired. Therefore, in this case, the existing adaptive radiotherapy technique based on MR guidance has a technical problem of poor accuracy in determining a lesion region and a non-lesion region in an image photographed during radiotherapy of a patient.
In view of the above problems, no effective solution has been proposed at present.
Disclosure of Invention
The application provides a medical image processing method, a medical image processing device, electronic equipment and a storage medium, which at least solve the technical problem of poor accuracy in the prior art when determining a lesion area and a non-lesion area in an image shot during radiotherapy of a patient.
According to an aspect of the present application, there is provided a medical image processing method, including: determining a ternary diagram aligned with the medical image according to the medical image shot by the target object before radiotherapy, wherein the ternary diagram is used for representing the probability that voxels in the medical image belong to a lesion area of the target object; acquiring a target image shot by a target object during radiotherapy; obtaining a target ternary diagram aligned with the target image by registering the medical image and the target image; and determining the probability that the voxels in the target ternary diagram belong to the lesion region of the target object by calculating the spatial correlation matrix of different voxels in the target ternary diagram.
Optionally, the medical image processing method further includes: acquiring a medical image of a target object photographed before radiotherapy, wherein the medical image comprises a first image and a second image, and the resolution of the first image is higher than that of the second image; the first image is taken as a reference, and a lesion area of the target object is sketched in the second image; and obtaining a ternary diagram aligned with the medical image by carrying out an inward contraction operation and an outward expansion operation on the delineated lesion area of the target object in the second image.
Optionally, the ternary diagram includes a background region, an unknown region, and a foreground region, wherein a probability that a voxel in the background region belongs to a lesion region of the target object is 0, a probability that a voxel in the foreground region belongs to a lesion region of the target object is 100%, and a probability that a voxel in the unknown region belongs to a lesion region of the target object is greater than or equal to 0 and less than or equal to 100%.
Optionally, the medical image processing method further includes: determining a correlation characteristic value of each voxel in an unknown region of the target ternary diagram and voxels except the voxel in the target ternary diagram based on the spatial correlation matrix; and determining the probability that each voxel in the unknown region of the target ternary diagram belongs to the lesion region according to the correlation characteristic value of each voxel in the unknown region of the target ternary diagram and the voxels except the voxel in the target ternary diagram.
Optionally, the medical image processing method further includes: taking a correlation characteristic value between an i-th voxel in an unknown region of the target ternary diagram and a voxel in a foreground region of the target ternary diagram as a first characteristic value, wherein the i-th voxel is any voxel in the unknown region of the target ternary diagram; taking a correlation eigenvalue between an ith voxel in an unknown region of the target ternary diagram and a voxel in a background region of the target ternary diagram as a second eigenvalue; taking a correlation characteristic value between an ith voxel in an unknown region of the target ternary diagram and other voxels in the unknown region of the target ternary diagram as a third characteristic value, wherein the other voxels are voxels different from the ith voxel; and calculating the probability that the ith voxel belongs to the lesion area according to the weight corresponding to the first characteristic value, the weight corresponding to the second characteristic value and the weight corresponding to the third characteristic value.
Optionally, the medical image processing method further includes: under the condition that the target image is a multi-mode image, calculating to obtain a spatial correlation matrix of different voxels in the target ternary diagram according to image information among the multi-mode images as reference information, and determining the probability that the voxels in the target ternary diagram belong to a lesion area of the target object.
Optionally, the medical image processing method further includes: detecting whether the registration precision of the medical image and the target image is larger than a preset threshold value; under the condition that the registration precision of the medical image and the target image is larger than a preset threshold value, calculating according to the medical image and the target image to obtain a spatial correlation matrix of different voxels in the target ternary diagram, and determining the probability that the voxels in the target ternary diagram belong to a lesion area of the target object; under the condition that the registration precision of the medical image and the target image is smaller than or equal to a preset threshold value, calculating according to the target image to obtain a spatial correlation matrix of different voxels in the target ternary diagram, and determining the probability that the voxels in the target ternary diagram belong to a lesion area of the target object.
Optionally, the medical image processing method further includes: after determining the probability that voxels in the target ternary diagram belong to a lesion area of a target object by calculating the spatial correlation matrix of different voxels in the target ternary diagram, representing the voxels in the target ternary diagram, the probability of which belongs to the lesion area is greater than a preset probability, by using a first pixel value; and representing voxels with probability less than or equal to a preset probability of belonging to the lesion area in the target ternary diagram by using a second pixel value.
According to another aspect of the present application, there is also provided a medical image processing apparatus, wherein the medical image processing apparatus includes: a ternary diagram determining unit for determining a ternary diagram aligned with the medical image according to the medical image shot by the target object before radiotherapy, wherein the ternary diagram is used for representing the probability that voxels in the medical image belong to a lesion area of the target object; a target image acquisition unit for acquiring a target image of a target subject photographed during radiotherapy; the first processing unit is used for obtaining a target ternary diagram aligned with the target image in a mode of registering the medical image and the target image; and the second processing unit is used for determining the probability that the voxels in the target ternary diagram belong to the lesion area of the target object by calculating the spatial correlation matrix of different voxels in the target ternary diagram.
Optionally, the ternary diagram determining unit includes: the system comprises a first acquisition subunit, a sketching subunit and a processing subunit. The first acquisition subunit is used for acquiring a medical image of the target object shot before radiotherapy, wherein the medical image comprises a first image and a second image, and the resolution of the first image is higher than that of the second image; a sketching subunit, configured to sketch a lesion area of the target object in the second image with the first image as a reference; and the processing subunit is used for obtaining a ternary diagram aligned with the medical image by carrying out an inward contraction operation and an outward expansion operation on the delineated lesion area of the target object in the second image.
Optionally, the ternary diagram includes a background region, an unknown region, and a foreground region, wherein a probability that a voxel in the background region belongs to a lesion region of the target object is 0, a probability that a voxel in the foreground region belongs to a lesion region of the target object is 100%, and a probability that a voxel in the unknown region belongs to a lesion region of the target object is greater than or equal to 0 and less than or equal to 100%.
Optionally, the second processing unit includes: a first determination subunit and a second determination subunit. The first determining subunit is used for determining a correlation characteristic value of each voxel in the unknown area of the target ternary diagram and the voxels except the voxel in the target ternary diagram based on the spatial correlation matrix; and the second determining subunit is used for determining the probability that each voxel in the unknown region of the target ternary diagram belongs to the lesion region according to the correlation characteristic value of each voxel in the unknown region of the target ternary diagram and the voxels except the voxel in the target ternary diagram.
Optionally, the second determining subunit includes: the system comprises a first processing subunit, a second processing subunit, a third processing subunit and a computing subunit. The first processing subunit is configured to take a correlation characteristic value between an i-th voxel in an unknown region of the target ternary diagram and a voxel in a foreground region of the target ternary diagram as a first characteristic value, where the i-th voxel is any voxel in the unknown region of the target ternary diagram; a second processing subunit, configured to take, as a second feature value, a correlation feature value between an i-th voxel in an unknown region of the target ternary diagram and a voxel in a background region of the target ternary diagram; a third processing subunit, configured to take, as a third feature value, a correlation feature value between an i-th voxel in an unknown region of the target ternary diagram and other voxels in the unknown region of the target ternary diagram, where the other voxels are voxels different from the i-th voxel; the calculating subunit is configured to calculate, according to the weight corresponding to the first feature value, the weight corresponding to the second feature value, and the weight corresponding to the third feature value, a probability that the i-th voxel belongs to the lesion area.
Optionally, the second processing unit includes: and the third determination subunit is used for calculating to obtain the spatial correlation matrix of different voxels in the target ternary diagram according to the image information among the multi-mode images as the reference information under the condition that the target image is the multi-mode image, and determining the probability that the voxels in the target ternary diagram belong to the lesion area of the target object.
Optionally, the second processing unit includes: the detecting subunit, the fourth determining subunit and the fifth determining subunit. The detection subunit is used for detecting whether the registration accuracy of the medical image and the target image is greater than a preset threshold value; a fourth determining subunit, configured to calculate, according to the medical image and the target image, a spatial correlation matrix of different voxels in the target ternary diagram and determine a probability that a voxel in the target ternary diagram belongs to a lesion region of the target object, where the registration accuracy of the medical image and the target image is greater than a preset threshold; and the fifth determining subunit is used for calculating according to the target image to obtain a spatial correlation matrix of different voxels in the target ternary diagram and determining the probability that the voxels in the target ternary diagram belong to the lesion area of the target object under the condition that the registration precision of the medical image and the target image is smaller than or equal to a preset threshold value.
Optionally, the medical image processing device further includes: a fourth processing subunit and a fifth processing subunit. The fourth processing subunit is configured to represent voxels in the target ternary diagram, where the probability of the voxels belonging to the lesion region is greater than a preset probability, by using a first pixel value; and a fifth processing subunit, configured to represent, by using the second pixel value, voxels in the target ternary diagram that have a probability of belonging to the lesion area that is less than or equal to a preset probability.
According to another aspect of the present application, there is also provided a computer readable storage medium, wherein the computer readable storage medium has stored therein a computer program, wherein the computer program is arranged to execute the method of processing a medical image of any one of the items when run.
According to another aspect of the present application, there is also provided an electronic device, wherein the electronic device includes one or more processors; and a storage means for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to implement a method for running the program, wherein the program is arranged to perform the method for processing medical images of any of the above.
In the application, a ternary image registration alignment mode is adopted, and firstly, a ternary image aligned with a medical image is determined according to the medical image shot by a target object before radiotherapy, wherein the ternary image is used for representing the probability that voxels in the medical image belong to a lesion area of the target object. Then, after the target image shot by the target object during radiotherapy is acquired, a target ternary diagram aligned with the target image is obtained by registering the medical image and the target image, and then the probability that the voxels in the target ternary diagram belong to the lesion region of the target object is determined by calculating the spatial correlation matrix of different voxels in the target ternary diagram.
From the foregoing, it is apparent that the present application determines a ternary image aligned with a medical image based on a medical image before radiotherapy (e.g., a PET image before radiotherapy and an MR image), and then, after acquiring a target image of a target subject photographed during radiotherapy, by registering the medical image and the target image, a target ternary image aligned with the target image can be obtained, and since the medical image before radiotherapy can be acquired only once, the cost at the time of acquiring the medical image before radiotherapy can be effectively controlled, and the patient does not receive excessive radiation.
In addition, in the application, the medical image before radiotherapy can be an MR image and a PET image with higher resolution than the MR image, so that an accurate ternary image can be obtained through registration and sketching of the PET image and the MR image (because the resolution of the PET image is very high and no fuzzy area exists), on the basis, after the target ternary image aligned with the target image is obtained through registration of the medical image and the target image, the accurate positioning of a lesion area on the target image is equivalent (because the reliability of the target ternary image is sufficiently ensured as a registration result of an image area), and therefore, the application can determine the probability that each voxel belongs to the lesion area according to the target ternary image (namely, the probability that each voxel belongs to a non-lesion area) without registration or manual sketching on the target image, thereby solving the technical problems that in the prior art, when the lesion area and the non-lesion area in the image shot during radiotherapy of a patient are determined, the problem of poor accuracy exists and reducing the cost of determining the lesion area is solved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute a limitation on the application. In the drawings:
FIG. 1 is a flow chart of an alternative medical image processing method according to an embodiment of the application;
FIG. 2 is a schematic illustration of an alternative ternary diagram according to an embodiment of the application;
FIG. 3 is a flow chart of an alternative method of determining the probability that a voxel belongs to a lesion region in accordance with an embodiment of the application;
fig. 4 is a schematic view of an alternative medical image processing apparatus according to an embodiment of the present application.
Detailed Description
In order that those skilled in the art will better understand the present application, a technical solution in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in which it is apparent that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present application without making any inventive effort, shall fall within the scope of the present application.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present application and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the application described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
It should be further noted that, the information (including but not limited to user equipment information, user personal information, etc.) and the data (including but not limited to data for analysis, stored data, displayed data, etc.) collected by the present application are information and data authorized by the user or fully authorized by each party, and the related data are collected, stored, used, processed, transmitted, provided, disclosed, applied, etc. processed, all in compliance with the related laws and regulations and standards of the related region, necessary security measures are taken, no prejudice to the public order and custom are provided with corresponding operation entries for the user to select authorization or rejection. For example, an interface is provided between the system and the relevant user or institution, before acquiring the relevant information, the system needs to send an acquisition request to the user or institution through the interface, and acquire the relevant information after receiving the consent information fed back by the user or institution.
In addition, in the application, the client information is collected, the client information is analyzed, a corresponding operation entrance is provided for the user, and the user can choose to agree or reject the automatic decision result; if the user selects refusal, the expert decision flow is entered.
According to an embodiment of the present application, there is provided an embodiment of a method of processing medical images, it being noted that the steps shown in the flowcharts of the figures may be performed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowcharts, in some cases the steps shown or described may be performed in an order different from that shown or described herein.
Fig. 1 is a flowchart of a medical image processing method according to an embodiment of the present application, as shown in fig. 1, the method includes the following steps:
Step S101, determining a ternary diagram aligned with the medical image according to the medical image of the target object photographed before radiotherapy.
In step S101, the ternary diagram is used to characterize the probability that a voxel in the medical image belongs to a lesion region of the target object.
In an alternative embodiment, a medical image processing system (hereinafter referred to as a processing system for short) may be used as an execution subject of the medical image processing method in the embodiment of the present application. The processing system may be a software system or an embedded system combining software and hardware.
Alternatively, the medical image of the target object taken before the radiotherapy may be a plurality of medical images of different types, for example, an MR image and a PET image of the target object taken before the radiotherapy, wherein the PET image is an image having a higher resolution than the MR image.
Optionally, taking an MR image and a PET image of the target object taken before radiotherapy as an example, the processing system may perform a registration operation and a delineating operation on the MR image and the PET image of the target object taken before radiotherapy, so as to obtain a lesion area of the target object, where, due to the high resolution of the PET image, the delineating accuracy may be ensured, that is, a clear and accurate lesion area may be obtained. The processing system may then adaptively contract and expand the ternary image to obtain a ternary image aligned with the medical image, wherein the pre-radiotherapy MR image and the PET image have been substantially aligned due to the registration performed with respect to the pre-radiotherapy MR image and the PET image, and the ternary image obtained on the basis is also aligned with respect to each other.
It should be noted that, the ternary diagram is used to characterize the probability that a voxel in the medical image belongs to a lesion region of a target object, for example, the ternary diagram includes: the method comprises the steps of a background area, an unknown area and a foreground area, wherein the probability that a voxel in the background area belongs to a lesion area of a target object is 0, the probability that the voxel in the foreground area belongs to the lesion area of the target object is 100%, and the probability that the voxel in the unknown area belongs to the lesion area of the target object is greater than or equal to 0 and less than or equal to 100%.
Step S102, acquiring a target image of a target object photographed during radiotherapy.
Alternatively, the target image is typically a resolution-limited image, for example, the target image is a MR image of the day, wherein the target image is typically an MR image rather than a PET image, because the PET image is very expensive to take and has a strong radiation, and because during radiotherapy, the target subject typically needs to take multiple target images to facilitate the doctor to follow up with the tumor change and treatment effect, the PET image is typically taken only before the radiotherapy, and not during the radiotherapy, for cost and patient health considerations.
However, since the resolution of the MR image on the same day is insufficient and there is no PET image on the same day, the definition of the boundary of the lesion region such as the intraprostate lesion region on the MR image is insufficient, and the accuracy of the automatic delineation on the MR image on the same day is not high. Even if the doctor of the radiotherapy department manually checks the sketching result, it is difficult to objectively distinguish the sketching accuracy of the target area with unclear boundaries and the organs at risk, and in addition, the labor cost is high. In addition, the scheme of transferring the sketched image before treatment to the MR image on the same day by means of registration inevitably has registration errors, so that the problem of insufficient accuracy of the sketched information on the MR image on the same day is caused.
In order to solve the above-mentioned problems, the present application proposes a method of registration transfer by ternary diagram, and the detailed description will be given below in step S103 and step S104.
In addition, it should be noted that the above medical image is an MR image and a PET image, and in practical application, the medical image may be another type of image, and an optional general implementation scenario is: the target object has rich images before radiotherapy, so that accurate sketching (for example, sketching PET images) can be performed, and the images on the same day during treatment (for example, MR images during treatment) cannot be performed on tumors or organs at risk with high sketching difficulty due to the limitation of image resolution. Therefore, the method is not limited to the specific medical image type and the target image type, and can be adopted to solve the technical problem of poor accuracy in determining the lesion area and the non-lesion area in the image shot during the radiotherapy of the patient as long as the method corresponds to the general implementation scene.
Step S103, obtaining a target ternary diagram aligned with the target image by registering the medical image and the target image.
Optionally, in order to facilitate explanation of the solution of the embodiment of the present application, the following description will take, as an example, an MR image and a PET image taken before radiotherapy of a medical image as a target object, and a current-day MR image taken during radiotherapy of the target image as the target object.
Optionally, during radiotherapy of the target object, only the MR image of the current day is provided, and no PET image of the current day is provided, at this time, the processing system applies a registration algorithm to register the MR image captured before radiotherapy and the MR image of the current day, and since the ternary image is aligned with the MR image captured before radiotherapy, after registration, a target ternary image corresponding to the MR image of the current day is obtained, and after the target ternary image is obtained, accurate positioning of the lesion area of the target object on the MR image of the current day is achieved.
It should be noted that, in the prior art, a technical solution is mentioned that the sketching information before treatment is transferred to the MR image on the same day by means of registration, and the essential content of the solution is that the sketching information is directly registered to the MR image on the same day for use, which requires very accurate registration, and cannot allow more registration errors. The scheme of the application only registers the ternary diagram on the target image as the location of the lesion area, and very accurate registration is not needed, so the technical scheme of the application greatly reduces the registration difficulty and improves the fault tolerance of scheme implementation.
Step S104, determining the probability that the voxels in the target ternary diagram belong to the lesion region of the target object by calculating the spatial correlation matrix of different voxels in the target ternary diagram.
Optionally, the present application further provides a multi-modal semantic soft segmentation algorithm, wherein the processing system may calculate a three-dimensional spatial correlation matrix between voxels in the feature space by the algorithm. For a voxel in the target ternary diagram, the probability that the voxel belongs to the lesion area of the target object may be determined by calculating the correlation between the voxel and voxels in the target ternary diagram other than the voxel.
It should be noted that, although a deep learning neural network model may be introduced to solve the problem that MR images on the same day are unclear for lesion areas and non-lesion areas, deep learning requires a large amount of training models of labeling data, and because of limitation of image resolution on the same day during treatment, it is difficult to obtain high-precision manual sketching for objects with unclear boundaries, so that it is difficult to obtain a large amount of training models of labeling data, and a method based on semantic soft segmentation does not need to train models, has mathematical analytic solutions, and therefore does not have the problem of preparing high-quality model samples.
In addition, when the deep learning algorithm is deployed in a model, the accuracy of the algorithm is reduced due to the difference between training data and actual data, and the radiation therapy has high requirements on the accuracy. The core algorithm of the scheme is semantic soft segmentation, and a training model is not needed, so that the problem is avoided. Although a registration algorithm is used in the process, the step of registration does not need to have particularly high precision, even a traditional rigid registration algorithm can be used, and a training model is not needed. In general, the technical scheme of the application has more advantages compared with a method for training and deploying a deep learning neural network model.
Based on the above-mentioned contents of step S101 to step S104, in the present application, a ternary image registration alignment manner is adopted, and first, a ternary image aligned with a medical image is determined according to a medical image captured by a target object before radiotherapy, where the ternary image is used to characterize a probability that a voxel in the medical image belongs to a lesion region of the target object. Then, after the target image shot by the target object during radiotherapy is acquired, a target ternary diagram aligned with the target image is obtained by registering the medical image and the target image, and then the probability that the voxels in the target ternary diagram belong to the lesion region of the target object is determined by calculating the spatial correlation matrix of different voxels in the target ternary diagram.
From the foregoing, it is apparent that the present application determines a ternary image aligned with a medical image based on a medical image before radiotherapy (e.g., a PET image before radiotherapy and an MR image), and then, after acquiring a target image of a target subject photographed during radiotherapy, by registering the medical image and the target image, a target ternary image aligned with the target image can be obtained, and since the medical image before radiotherapy can be acquired only once, the cost at the time of acquiring the medical image before radiotherapy can be effectively controlled, and the patient does not receive excessive radiation.
In addition, in the application, the medical image before radiotherapy can be an MR image and a PET image with higher resolution than the MR image, so that an accurate ternary image can be obtained through registration and sketching of the PET image and the MR image (because the resolution of the PET image is very high and no fuzzy area exists), on the basis, after the target ternary image aligned with the target image is obtained through registration of the medical image and the target image, the accurate positioning of a lesion area on the target image is equivalent (because the reliability of the target ternary image is sufficiently ensured as a registration result of an image area), and therefore, the application can determine the probability that each voxel belongs to the lesion area according to the target ternary image (namely, the probability that each voxel belongs to a non-lesion area) without registration or manual sketching on the target image, thereby solving the technical problems that in the prior art, when the lesion area and the non-lesion area in the image shot during radiotherapy of a patient are determined, the problem of poor accuracy exists and reducing the cost of determining the lesion area is solved.
In an alternative embodiment, the processing system may acquire a medical image of the target object taken prior to the radiotherapy, wherein the medical image comprises a first image and a second image, wherein the resolution of the first image is higher than the resolution of the second image. And then, the processing system uses the first image as a reference, outlines a lesion area of the target object in the second image, and performs an inward contraction operation and an outward expansion operation on the outlined lesion area of the target object in the second image to obtain a ternary diagram corresponding to the medical image.
Alternatively, the first image may be a PET image of the target object taken before radiotherapy, and the second image may be an MR image of the target object taken before radiotherapy, wherein the lesion region may be delineated with reference to the PET image before radiotherapy and the MR image before radiotherapy. The sketching operation can use a manual sketching mode or an automatic sketching mode, and the sketching precision can be ensured because of the PET image.
Alternatively, the processing system may perform a telescoping operation and a telescoping operation on the delineated lesion region to obtain a ternary diagram aligned with the medical image.
Optionally, the ternary diagram includes: the method comprises the steps of a background area, an unknown area and a foreground area, wherein the probability that a voxel in the background area belongs to a lesion area of a target object is 0, the probability that the voxel in the foreground area belongs to the lesion area of the target object is 100%, and the probability that the voxel in the unknown area belongs to the lesion area of the target object is greater than or equal to 0 and less than or equal to 100%.
Wherein, fig. 2 is a schematic diagram of an alternative ternary diagram according to an embodiment of the present application, as shown in fig. 2, the ternary diagram includes a background area, an unknown area, and a foreground area, wherein, voxels located in the background area do not necessarily belong to a lesion area; part of voxels in the position area belong to the target area, part of voxels do not belong to the target area, and an algorithm is needed to determine the attribution of each voxel; voxels located in the foreground region must belong to the lesion region.
It should be noted that, the scaling ratio may be set by user-defined operation, such as determining the scaling ratio based on the size of the lesion region delineated from the medical image, setting a smaller scaling ratio if the size of the lesion region is smaller, and setting a larger scaling ratio if the size of the lesion region is larger.
It should be noted that, by acquiring a ternary image aligned with the medical image, it is equivalent to determining the boundaries of the lesion region and the non-lesion region by using the ternary image, in other words, it is equivalent to precisely locating the lesion region on the basis of the medical image.
In an alternative embodiment, the processing system may further determine a correlation characteristic value of each voxel in the unknown region of the target ternary diagram with a voxel other than the voxel in the target ternary diagram based on the spatial correlation matrix, and determine a probability that each voxel in the unknown region of the target ternary diagram belongs to the lesion region based on the correlation characteristic value of each voxel in the unknown region of the target ternary diagram with the voxel other than the voxel in the target ternary diagram.
Optionally, the target ternary diagram is substantially a result of the ternary diagram registering onto the target image, and therefore the target ternary diagram also includes an unknown region, a background region, and a foreground region, and the unknown region, the background region, and the foreground region of the target ternary diagram are in one-to-one correspondence with the unknown region, the background region, and the foreground region of the ternary diagram, respectively.
Optionally, for each voxel in the unknown region of the target ternary diagram, the processing system calculates the confidence that it belongs to the lesion region based on the correlation of that voxel with other voxels of the unknown region, voxels of the foreground region, voxels of the background region in the target ternary diagram. Wherein the confidence level characterizes the correlation characteristic value.
Optionally, the confidence level is in a range of [0,1], where 0 indicates that 100% does not belong to the lesion area, and 1 indicates that 100% belongs to the lesion area. By calculating the confidence of each voxel in the unknown region of the target ternary diagram, the probability that each voxel belongs to the lesion region can be objectively presented, and the presentation mode can well represent the invasion degree of the tumor to the body tissue.
In an alternative embodiment, fig. 3 is a flowchart of an alternative method for determining the probability that a voxel belongs to a lesion region according to an embodiment of the application, as shown in fig. 3, comprising the steps of:
step S301 takes, as a first feature value, a correlation feature value between an i-th voxel in an unknown region of the target ternary diagram and a voxel in a foreground region of the target ternary diagram.
In step S301, the i-th voxel is any one voxel in the unknown region of the target ternary diagram.
Step S302, taking a correlation eigenvalue between the i-th voxel in the unknown region of the target ternary diagram and the voxel in the background region of the target ternary diagram as a second eigenvalue.
Step S303, taking a correlation eigenvalue between the i-th voxel in the unknown region of the target ternary diagram and other voxels in the unknown region of the target ternary diagram as a third eigenvalue.
In step S303, the other voxels are voxels different from the i-th voxel.
Step S304, calculating to obtain the probability that the ith voxel belongs to the lesion area according to the weight corresponding to the first characteristic value, the weight corresponding to the second characteristic value and the weight corresponding to the third characteristic value.
Alternatively, setting the corresponding weights for the above-described first feature value, second feature value, and third feature value, respectively, may improve the accuracy of probability calculation, for example, the correlation feature value between the i-th voxel in the unknown region and the voxel in the foreground region of the target ternary diagram is taken as the first feature value, the correlation feature value between the i-th voxel and the voxel in the background region of the target ternary diagram is taken as the second feature value, the weights corresponding to the two feature values may be equal, and the weights may be set to be larger than the weights corresponding to the third feature value, because the foreground region and the background region are already clear regions, and the reliability of the calculation result obtained when they are taken as the reference information is higher, so the weights may be set to be larger. Of course, the weight values of the three feature values may also be set to be the same.
In an alternative embodiment, in the case that the target image is a multi-modal image, the spatial correlation matrix of different voxels in the target ternary diagram is calculated according to image information between the multi-modal images as reference information, and the probability that the voxels in the target ternary diagram belong to the lesion region of the target object is determined.
Alternatively, in an embodiment of the present application, the target image may be multi-modal, for example, the MR image of the day may include: MR T1 weighted imaging, MR T1 enhanced imaging, MR T2 weighted imaging and the like, and image information on each MR mode image is complementary, so that the images can be used as input information of a multi-mode semantic soft segmentation algorithm together after registration and alignment of multi-mode target images, a processing system can calculate and obtain spatial correlation matrixes of different voxels in a target ternary diagram after image information fusion of the multi-mode target images, and the probability that the voxels in the target ternary diagram belong to a lesion region of a target object is determined.
It should be noted that, the image information between the multi-mode target images may be complementary, so that the accuracy of calculating the spatial correlation matrix of the target ternary diagram may be improved.
In an alternative embodiment, the processing system may detect whether the registration accuracy of the medical image and the target image is greater than a preset threshold, and calculate, according to the medical image and the target image, a spatial correlation matrix of different voxels in the target ternary diagram and determine a probability that a voxel in the target ternary diagram belongs to a lesion region of the target object if the registration accuracy of the medical image and the target image is greater than the preset threshold; under the condition that the registration precision of the medical image and the target image is smaller than or equal to a preset threshold value, calculating according to the target image to obtain a spatial correlation matrix of different voxels in the target ternary diagram, and determining the probability that the voxels in the target ternary diagram belong to a lesion area of the target object.
Optionally, if the registration accuracy of the medical image before treatment and the target image during treatment is higher, the processing system may also add the medical image before treatment as input information of the multi-mode semantic soft segmentation algorithm, so as to obtain more supplementary information to obtain a soft segmentation result with higher accuracy. Therefore, under the condition that the registration precision of the medical image and the target image is larger than a preset threshold value, calculating according to the medical image and the target image to obtain a spatial correlation matrix of different voxels in the target ternary diagram, and determining the probability that the voxels in the target ternary diagram belong to a lesion area of a target object; under the condition that the registration accuracy of the medical image and the target image is smaller than or equal to a preset threshold value, the spatial correlation matrix of different voxels in the target ternary diagram is obtained only according to the target image, and the probability that the voxels in the target ternary diagram belong to the lesion region of the target object is determined.
In an alternative embodiment, after determining the probability that a voxel in the target ternary diagram belongs to the lesion area of the target object by calculating the spatial correlation matrix of different voxels in the target ternary diagram, the processing system may represent the voxel in the target ternary diagram having a probability of belonging to the lesion area greater than a preset probability with a first pixel value and then represent the voxel in the target ternary diagram having a probability of belonging to the lesion area less than or equal to the preset probability with a second pixel value.
Alternatively, the results of the semantic soft segmentation may be used to qualitatively analyze the accuracy of the segmentation for the physician to check the accuracy of the segmentation. In the embodiment of the present application, the processing system may perform binarization processing on the result of semantic soft segmentation, and convert the pixel value of each voxel into 0 or 1 by adopting a certain threshold (corresponding to the above preset probability, for example, 0.5), to obtain a conventional segmentation format (binarization) for the subsequent radiotherapy process. For example, voxels in the target ternary diagram having a probability of belonging to the lesion region greater than a preset probability (e.g., 0.5) are represented using a first pixel value (e.g., 0), and then voxels in the target ternary diagram having a probability of belonging to the lesion region less than or equal to the preset probability (e.g., 0.5) are represented using a second pixel value (e.g., 1).
From the foregoing, it is apparent that the present application determines a ternary image aligned with a medical image based on a medical image before radiotherapy (e.g., a PET image before radiotherapy and an MR image), and then, after acquiring a target image of a target subject photographed during radiotherapy, by registering the medical image and the target image, a target ternary image aligned with the target image can be obtained, and since the medical image before radiotherapy can be acquired only once, the cost at the time of acquiring the medical image before radiotherapy can be effectively controlled, and the patient does not receive excessive radiation.
In addition, in the application, the medical image before radiotherapy can be an MR image and a PET image with higher resolution than the MR image, so that an accurate ternary image can be obtained through registration and sketching of the PET image and the MR image (because the resolution of the PET image is very high and no fuzzy area exists), on the basis, after the target ternary image aligned with the target image is obtained through registration of the medical image and the target image, the accurate positioning of a lesion area on the target image is equivalent (because the reliability of the target ternary image is sufficiently ensured as a registration result of an image area), and therefore, the application can determine the probability that each voxel belongs to the lesion area according to the target ternary image (namely, the probability that each voxel belongs to a non-lesion area) without registration or manual sketching on the target image, thereby solving the technical problems that in the prior art, when the lesion area and the non-lesion area in the image shot during radiotherapy of a patient are determined, the problem of poor accuracy exists and reducing the cost of determining the lesion area is solved.
According to another aspect of the present application, there is also provided a medical image processing apparatus, wherein, as shown in fig. 4, the medical image processing apparatus includes: a ternary diagram determining unit 401, configured to determine a ternary diagram aligned with the medical image according to a medical image captured by the target object before radiotherapy, where the ternary diagram is used to characterize a probability that a voxel in the medical image belongs to a lesion area of the target object; a target image acquisition unit 402 for acquiring a target image of a target subject photographed during radiotherapy; a first processing unit 403, configured to obtain a target ternary diagram aligned with the target image by registering the medical image and the target image; the second processing unit 404 is configured to determine a probability that the voxel in the target ternary diagram belongs to the lesion area of the target object by calculating a spatial correlation matrix of different voxels in the target ternary diagram.
Alternatively, the ternary diagram determining unit 401 includes: the system comprises a first acquisition subunit, a sketching subunit and a processing subunit. The first acquisition subunit is used for acquiring a medical image of the target object shot before radiotherapy, wherein the medical image comprises a first image and a second image, and the resolution of the first image is higher than that of the second image; a sketching subunit, configured to sketch a lesion area of the target object in the second image with the first image as a reference; and the processing subunit is used for obtaining a ternary diagram aligned with the medical image by carrying out an inward contraction operation and an outward expansion operation on the delineated lesion area of the target object in the second image.
Optionally, the ternary diagram includes a background region, an unknown region, and a foreground region, wherein a probability that a voxel in the background region belongs to a lesion region of the target object is 0, a probability that a voxel in the foreground region belongs to a lesion region of the target object is 100%, and a probability that a voxel in the unknown region belongs to a lesion region of the target object is greater than or equal to 0 and less than or equal to 100%.
Optionally, the second processing unit 404 includes: a first determination subunit and a second determination subunit. The first determining subunit is used for determining a correlation characteristic value of each voxel in the unknown area of the target ternary diagram and the voxels except the voxel in the target ternary diagram based on the spatial correlation matrix; and the second determining subunit is used for determining the probability that each voxel in the unknown region of the target ternary diagram belongs to the lesion region according to the correlation characteristic value of each voxel in the unknown region of the target ternary diagram and the voxels except the voxel in the target ternary diagram.
Optionally, the second determining subunit includes: the system comprises a first processing subunit, a second processing subunit, a third processing subunit and a computing subunit. The first processing subunit is configured to take a correlation characteristic value between an i-th voxel in an unknown region of the target ternary diagram and a voxel in a foreground region of the target ternary diagram as a first characteristic value, where the i-th voxel is any voxel in the unknown region of the target ternary diagram; a second processing subunit, configured to take, as a second feature value, a correlation feature value between an i-th voxel in an unknown region of the target ternary diagram and a voxel in a background region of the target ternary diagram; a third processing subunit, configured to take, as a third feature value, a correlation feature value between an i-th voxel in an unknown region of the target ternary diagram and other voxels in the unknown region of the target ternary diagram, where the other voxels are voxels different from the i-th voxel; the calculating subunit is configured to calculate, according to the weight corresponding to the first feature value, the weight corresponding to the second feature value, and the weight corresponding to the third feature value, a probability that the i-th voxel belongs to the lesion area.
Optionally, the second processing unit 404 includes: and the third determination subunit is used for calculating to obtain the spatial correlation matrix of different voxels in the target ternary diagram according to the image information among the multi-mode images as the reference information under the condition that the target image is the multi-mode image, and determining the probability that the voxels in the target ternary diagram belong to the lesion area of the target object.
Optionally, the second processing unit 404 includes: the detecting subunit, the fourth determining subunit and the fifth determining subunit. The detection subunit is used for detecting whether the registration accuracy of the medical image and the target image is greater than a preset threshold value; a fourth determining subunit, configured to calculate, according to the medical image and the target image, a spatial correlation matrix of different voxels in the target ternary diagram and determine a probability that a voxel in the target ternary diagram belongs to a lesion region of the target object, where the registration accuracy of the medical image and the target image is greater than a preset threshold; and the fifth determining subunit is used for calculating according to the target image to obtain a spatial correlation matrix of different voxels in the target ternary diagram and determining the probability that the voxels in the target ternary diagram belong to the lesion area of the target object under the condition that the registration precision of the medical image and the target image is smaller than or equal to a preset threshold value.
Optionally, the medical image processing device further includes: a fourth processing subunit and a fifth processing subunit. The fourth processing subunit is configured to represent voxels in the target ternary diagram, where the probability of the voxels belonging to the lesion region is greater than a preset probability, by using a first pixel value; and a fifth processing subunit, configured to represent, by using the second pixel value, voxels in the target ternary diagram that have a probability of belonging to the lesion area that is less than or equal to a preset probability.
According to another aspect of the embodiments of the present application, there is also provided a computer readable storage medium, including a stored computer program, where the computer program when executed controls a device in which the computer readable storage medium is located to perform a method for processing a medical image according to any one of the above.
According to another aspect of the embodiment of the present application, there is also provided an electronic device, including: a processor; and a memory for storing executable instructions of the processor; wherein the processor is configured to execute the medical image processing method of any one of the above via execution of the executable instructions.
The foregoing embodiment numbers of the present application are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
In the foregoing embodiments of the present application, the descriptions of the embodiments are emphasized, and for a portion of this disclosure that is not described in detail in this embodiment, reference is made to the related descriptions of other embodiments.
In the several embodiments provided in the present application, it should be understood that the disclosed technology may be implemented in other manners. The above-described embodiments of the apparatus are merely exemplary, and the division of the units, for example, may be a logic function division, and may be implemented in another manner, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interfaces, units or modules, or may be in electrical or other forms.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present application may be embodied essentially or in part or all of the technical solution or in part in the form of a software product stored in a storage medium, including instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present application. And the aforementioned storage medium includes: a usb disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a removable hard disk, a magnetic disk, or an optical disk, or other various media capable of storing program codes.
The foregoing is merely a preferred embodiment of the present application and it should be noted that modifications and adaptations to those skilled in the art may be made without departing from the principles of the present application, which are intended to be comprehended within the scope of the present application.
Claims (11)
1. A method for processing medical images, comprising:
Determining a ternary diagram aligned with a medical image of a target object before radiotherapy according to the medical image, wherein the ternary diagram is used for representing the probability that voxels in the medical image belong to a lesion area of the target object;
acquiring a target image shot by a target object during radiotherapy;
Obtaining a target ternary diagram aligned with the target image by registering the medical image and the target image;
And determining the probability that the voxels in the target ternary diagram belong to the lesion area of the target object by calculating the spatial correlation matrix of different voxels in the target ternary diagram.
2. The method of processing medical images according to claim 1, wherein determining a ternary image aligned with the medical image from a medical image taken of a target object prior to radiotherapy, comprises:
Acquiring a medical image of the target object photographed before radiotherapy, wherein the medical image comprises a first image and a second image, and the resolution of the first image is higher than that of the second image;
Drawing a lesion area of the target object in the second image by taking the first image as a reference;
And performing inward contraction operation and outward expansion operation on the delineated lesion area of the target object in the second image to obtain a ternary diagram aligned with the medical image.
3. The method of processing a medical image according to claim 1, wherein the ternary diagram includes a background region, an unknown region, and a foreground region, wherein a probability that a voxel in the background region belongs to a lesion region of the target object is 0, a probability that a voxel in the foreground region belongs to a lesion region of the target object is 100%, and a probability that a voxel in the unknown region belongs to a lesion region of the target object is greater than or equal to 0 and less than or equal to 100%.
4. A method of processing a medical image according to claim 3, wherein determining the probability that a voxel in the target ternary diagram belongs to a lesion region of the target object by calculating a spatial correlation matrix of different voxels in the target ternary diagram comprises:
determining a correlation characteristic value of each voxel in an unknown region of the target ternary diagram and voxels except the voxel in the target ternary diagram based on a spatial correlation matrix;
And determining the probability that each voxel in the unknown region of the target ternary diagram belongs to the lesion region according to the correlation characteristic value of each voxel in the unknown region of the target ternary diagram and the voxels except the voxel in the target ternary diagram.
5. The method according to claim 4, wherein in determining a probability that each voxel in an unknown region of the target ternary diagram belongs to the lesion region based on a correlation between each voxel in the unknown region of the target ternary diagram and voxels other than the voxel in the target ternary diagram, the method further comprises:
Taking a correlation characteristic value between an ith voxel in an unknown region of the target ternary diagram and a voxel in a foreground region of the target ternary diagram as a first characteristic value, wherein the ith voxel is any voxel in the unknown region of the target ternary diagram;
Taking a correlation characteristic value between an ith voxel in an unknown region of the target ternary diagram and a voxel in a background region of the target ternary diagram as a second characteristic value;
taking a correlation characteristic value between an ith voxel in an unknown region of the target ternary diagram and other voxels in the unknown region of the target ternary diagram as a third characteristic value, wherein the other voxels are voxels different from the ith voxel;
and calculating the probability that the ith voxel belongs to the lesion area according to the weight corresponding to the first characteristic value, the weight corresponding to the second characteristic value and the weight corresponding to the third characteristic value.
6. The method of processing a medical image according to claim 1, wherein determining the probability that a voxel in the target ternary diagram belongs to a lesion region of the target object by calculating a spatial correlation matrix of different voxels in the target ternary diagram comprises:
and under the condition that the target image is a multi-mode image, calculating to obtain a spatial correlation matrix of different voxels in the target ternary diagram according to image information among the multi-mode images as reference information, and determining the probability that the voxels in the target ternary diagram belong to a lesion area of the target object.
7. The method of processing a medical image according to claim 1, wherein determining the probability that a voxel in the target ternary diagram belongs to a lesion region of the target object by calculating a spatial correlation matrix of different voxels in the target ternary diagram comprises:
detecting whether the registration precision of the medical image and the target image is larger than a preset threshold value;
Under the condition that the registration precision of the medical image and the target image is larger than the preset threshold, calculating to obtain a spatial correlation matrix of different voxels in the target ternary diagram according to the medical image and the target image, and determining the probability that the voxels in the target ternary diagram belong to a lesion area of the target object;
And under the condition that the registration precision of the medical image and the target image is smaller than or equal to the preset threshold value, calculating according to the target image to obtain a spatial correlation matrix of different voxels in the target ternary diagram, and determining the probability that the voxels in the target ternary diagram belong to the lesion area of the target object.
8. The method of processing a medical image according to claim 1, wherein after determining the probability that a voxel in the target ternary diagram belongs to a lesion region of the target object by calculating a spatial correlation matrix of different voxels in the target ternary diagram, the method of processing a medical image further comprises:
Representing voxels in the target ternary diagram, which belong to the lesion area and have a probability larger than a preset probability, by using a first pixel value;
and representing voxels with probability less than or equal to the preset probability of belonging to the lesion area in the target ternary diagram by using a second pixel value.
9. A medical image processing apparatus, comprising:
A ternary diagram determining unit, configured to determine a ternary diagram aligned to a medical image of a target object captured before radiotherapy, where the ternary diagram is used to characterize a probability that voxels in the medical image belong to a lesion region of the target object;
a target image acquisition unit for acquiring a target image of a target subject photographed during radiotherapy;
A first processing unit, configured to obtain a target ternary diagram aligned with the target image by registering the medical image and the target image;
And the second processing unit is used for determining the probability that the voxels in the target ternary diagram belong to the lesion area of the target object by calculating the spatial correlation matrix of different voxels in the target ternary diagram.
10. A computer readable storage medium, characterized in that the computer readable storage medium has stored therein a computer program, wherein the computer program is arranged to execute the medical image processing method according to any of the claims 1 to 8 at run-time.
11. An electronic device, the electronic device comprising one or more processors; storage means for storing one or more programs which when executed by the one or more processors cause the one or more processors to implement a method for running a program, wherein the program is arranged to perform the method for processing a medical image according to any one of claims 1 to 8 when run.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410519698.1A CN118096845A (en) | 2024-04-28 | 2024-04-28 | Medical image processing method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410519698.1A CN118096845A (en) | 2024-04-28 | 2024-04-28 | Medical image processing method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118096845A true CN118096845A (en) | 2024-05-28 |
Family
ID=91148034
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410519698.1A Pending CN118096845A (en) | 2024-04-28 | 2024-04-28 | Medical image processing method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118096845A (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115136189A (en) * | 2020-03-13 | 2022-09-30 | 基因泰克公司 | Automated detection of tumors based on image processing |
| US20230078011A1 (en) * | 2021-09-13 | 2023-03-16 | Northwestern University | Method and system for labeling medical images |
| CN116258725A (en) * | 2023-05-16 | 2023-06-13 | 福建自贸试验区厦门片区Manteia数据科技有限公司 | Medical image processing method and device based on feature images and storage medium |
-
2024
- 2024-04-28 CN CN202410519698.1A patent/CN118096845A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115136189A (en) * | 2020-03-13 | 2022-09-30 | 基因泰克公司 | Automated detection of tumors based on image processing |
| US20230005140A1 (en) * | 2020-03-13 | 2023-01-05 | Genentech, Inc. | Automated detection of tumors based on image processing |
| US20230078011A1 (en) * | 2021-09-13 | 2023-03-16 | Northwestern University | Method and system for labeling medical images |
| CN116258725A (en) * | 2023-05-16 | 2023-06-13 | 福建自贸试验区厦门片区Manteia数据科技有限公司 | Medical image processing method and device based on feature images and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 殷恺铭;闫士举;宋成利;: "基于改进局部三元模式的乳腺癌预测模型", 中国医学影像技术, no. 04, 20 April 2018 (2018-04-20) * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10769791B2 (en) | Systems and methods for cross-modality image segmentation | |
| US20230267611A1 (en) | Optimization of a deep learning model for performing a medical imaging analysis task | |
| EP3123443B1 (en) | Method and device for generating one or more computer tomography images based on magnetic resonance images with the help of tissue class separation | |
| Pasquier et al. | Automatic segmentation of pelvic structures from magnetic resonance images for prostate cancer radiotherapy | |
| RU2436161C2 (en) | Recording images at deformation for image-based control beam therapy | |
| JP2023507109A (en) | Automated tumor identification and segmentation from medical images | |
| US20160016007A1 (en) | Method for improved surface tracking-based motion management and dynamic planning in adaptive external beam radiation therapy | |
| US20100067768A1 (en) | Method and System for Physiological Image Registration and Fusion | |
| US20040071325A1 (en) | Registration of multi-modality data in imaging | |
| Chen et al. | Combining registration and active shape models for the automatic segmentation of the lymph node regions in head and neck CT images | |
| Kalinosky et al. | White matter structural connectivity is associated with sensorimotor function in stroke survivors | |
| US20110148861A1 (en) | Pet data processing system, an arrangement, a method and a computer program product for determining a distribution of a tracer uptake | |
| EP2689344B1 (en) | Knowledge-based automatic image segmentation | |
| CN114881848A (en) | Method for converting multi-sequence MR into CT | |
| Sanaat et al. | A cycle-consistent adversarial network for brain PET partial volume correction without prior anatomical information | |
| CN118096845A (en) | Medical image processing method and device, electronic equipment and storage medium | |
| Fortunati et al. | Hyperthermia critical tissues automatic segmentation of head and neck CT images using atlas registration and graph cuts | |
| CN118302791A (en) | Machine learning based segmentation of biological objects in medical images | |
| Kim et al. | Co-registration of MRI and autoradiography of rat brain in three-dimensions following automatic reconstruction of 2D data set | |
| CN117177704A (en) | System and method for template-based automatic detection of anatomical structures | |
| Liu et al. | Employing symmetry features for automatic misalignment correction in neuroimages | |
| CN112465739A (en) | Method and provision system for generating a surrogate marker based on medical image data | |
| Costa et al. | A maximum-likelihood approach to PET emission/attenuation image registration | |
| US11823399B2 (en) | Multi-scan image processing | |
| Thieu et al. | Segmentation by a local and global fuzzy Gaussian distribution energy minimization of an active contour model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |