CN112767315A

CN112767315A - Determination method and display method for delineation quality of target area and electronic equipment

Info

Publication number: CN112767315A
Application number: CN202011635011.9A
Authority: CN
Inventors: 贾乐成
Original assignee: Shenzhen United Imaging Research Institute of Innovative Medical Equipment
Current assignee: Shenzhen United Imaging Research Institute of Innovative Medical Equipment
Priority date: 2020-12-31
Filing date: 2020-12-31
Publication date: 2021-05-07
Anticipated expiration: 2040-12-31
Also published as: CN112767315B

Abstract

The embodiment of the invention discloses a determination method and a display method of target area delineation quality and electronic equipment. The determination method may include: acquiring a target area delineation and a quality determination object of a delineated target area in a medical image, wherein the quality determination object comprises a prior appearance of a dangerous part and/or the delineated target area; determining delineation information delineated by the target area and object information of a quality determination object, wherein the object information comprises a dangerous image of a dangerous part and/or prior morphology characteristics of the prior morphology; and determining the punishment relation between the target area sketching and the quality determination object according to the sketching information and the object information, and determining the sketching quality of the target area sketching according to the punishment relation. According to the technical scheme of the embodiment of the invention, the target area delineation quality manually determined by the user can be objectively represented, so that the effect of automatically determining and determining the target area delineation quality consistent with the target area delineation quality manually determined by the user is realized, and the method has a good clinical application value.

Description

Determination method and display method for delineation quality of target area and electronic equipment

Technical Field

The embodiment of the invention relates to the field of image processing, in particular to a method for determining the delineation quality of a target area, a display method and electronic equipment.

Background

The target delineation of the target area of cancer radiotherapy (hereinafter referred to as target area) is an important part of the radiotherapy process, and the accuracy thereof will directly influence the execution effect of the subsequent radiotherapy plan, therefore, the determination of the target area delineation quality is indispensable in the radiotherapy process.

At present, the determination of the delineation quality of a target area is mainly completed manually by a doctor and depends on the clinical experience and subjective judgment of the doctor. It should be noted that although the target delineation has a uniform delineation consensus, different doctors have different understandings of the consensus, which may cause that different doctors and even the same doctor may give different quality evaluations to the same target delineation in different times, i.e. the consistency and the repeatability are poor. In addition, the above-mentioned solutions require manual work by the doctor, which consumes a lot of human resources and cannot process large-scale data.

Disclosure of Invention

The embodiment of the invention provides a method for determining the delineation quality of a target area, a display method and electronic equipment, and aims to achieve the effect of automatically determining the delineation quality of the target area.

In a first aspect, an embodiment of the present invention provides a method for determining quality of a target delineation, which may include:

acquiring a target area delineation and a quality determination object of a delineated target area in a medical image, wherein the quality determination object comprises a prior appearance of a dangerous part and/or the delineated target area;

determining delineation information delineated by the target area and object information of a quality determination object, wherein the object information comprises a dangerous image of a dangerous part and/or prior morphology characteristics of the prior morphology;

determining the punishment relation between the target area sketching and the quality determination object according to the sketching information and the object information, and determining the sketching quality of the target area sketching according to the punishment relation.

In a second aspect, an embodiment of the present invention further provides a display method for quality of target delineation, which may include:

receiving a medical image and a target area sketch of a sketched target area in the medical image;

outputting the processed medical image, and displaying quality information related to the target delineation quality of the target delineation on the processed medical image, wherein the quality information comprises at least one of scores of the target delineation quality, sub-delineations to be modified in the target delineation and modification reasons of the sub-delineations to be modified;

the quality information is determined according to the penalty relationship between the target delineation and the quality determination object in the medical image, the penalty relationship is determined according to the delineation information delineated by the target and the object information of the quality determination object, the quality determination object comprises the prior morphology of the endangered part and/or the delineated target, and the object information comprises the prior morphology characteristics of the endangered image of the endangered part and/or the prior morphology.

In a third aspect, an embodiment of the present invention further provides an electronic device, which may include:

one or more processors;

a memory for storing one or more programs;

when the one or more programs are executed by the one or more processors, the one or more processors implement the method for determining the target delineation quality or the method for displaying the target delineation quality provided by any of the embodiments of the present invention.

According to the technical scheme of the embodiment of the invention, the target area sketching and the quality determination object of the target area in the acquired medical image can determine the sketching information of the target area sketching and the object information of the quality determination object, since the quality determination object is an object provided to simulate how a user manually determines the quality of the target delineation, such as an a priori topography of the endangered part and/or the delineated target, therefore, the punishment relationship between the target delineation determined according to the delineation information and the object information and the quality determination object can indicate whether the target delineation invades a dangerous part which should not be invaded and/or whether a part which does not accord with the prior morphology exists, so that the target delineation quality which is consistent with the target delineation quality of the target delineation manually determined by a user can be automatically determined according to the punishment relationship, and the method has a good clinical application value. According to the technical scheme, the quality determination object is set from the angle of how the user manually determines the target area delineation quality, and subsequently, when the target area delineation quality is determined based on the punishment relation between the target area delineation and the quality determination object, objective representation can be carried out on the target area delineation quality manually determined by the user, so that the effect of automatically determining and determining the target area delineation quality consistent with the target area delineation quality manually determined by the user is realized.

Drawings

Fig. 1 is a flowchart of a method for determining a target delineation quality according to a first embodiment of the present invention;

fig. 2 is a flowchart of a method for determining target delineation quality according to a second embodiment of the present invention;

fig. 3 is a flowchart of a method for determining the target delineation quality according to a third embodiment of the present invention;

fig. 4 is a flowchart of a method for determining the quality of a target delineation in a fourth embodiment of the present invention;

fig. 5 is a flowchart of an alternative example of a method for determining target delineation quality according to a fourth embodiment of the present invention;

fig. 6 is a flowchart of a method for displaying target delineation quality in a fifth embodiment of the present invention;

fig. 7 is a schematic structural diagram of an electronic device in an eighth embodiment of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.

Before the embodiment of the present invention is described, an application scenario of the embodiment of the present invention is exemplarily described: in order to solve the problems of poor consistency and repeatability and high human resource consumption caused by the fact that doctors manually determine the target area delineation quality, researchers provide an automatic determination scheme of the target area delineation quality. Specifically, the target area delineation is compared with a "gold standard" (i.e., a gold standard delineation), the comparison result is quantified based on a preset quality determination index, and the quantified result is used as the target area delineation quality. The quality determination index can be an Intersection Over Unit (IOU), accuracy, precision, recall rate, Dice coefficient, Hausdorff distance and the like, and has the advantages of high calculation speed, high repeatability, strong consistency and less human resource consumption.

However, such automatic determination schemes rely on the accuracy of the "gold standard", are susceptible to tag noise, and are difficult to embody the various factors that a physician considers in manually determining the target delineation quality. Illustratively, the target delineation is a multi-layer delineation result, and if each layer has more or less points than the "gold standard", the automatically determined target delineation quality is still good, but poor for the physician. As another example, a certain pixel point is usually present or absent, once the target area delineates less than 1 layer, even if the other layers are fine, the quality of the target area delineation determined by the method is still very poor, but the method is still good for doctors. In addition, the automatic determination scheme determines the target delineation quality from a pixel level, and cannot intuitively give concrete situations of where the target delineation is good, where the target delineation is bad, and the like, or give a feasible modification scheme, as if a doctor manually determines the target delineation quality.

On the basis, on the basis that the target area can comprise a primary tumor focus, a metastatic focus, a subclinical focus thereof and a part possibly invaded by the tumor, and the greatest characteristic of the embodiments is that the position and the size are not fixed and no clear boundary exists in a medical image, if various factors considered in the process of manually determining the target area delineation quality by a doctor are fused in the automatic determination process, an automatic determination result consistent with the target area delineation quality manually determined by the doctor can be obtained, namely, the manual determination process of the doctor is objectively represented. Accordingly, the following method for determining the target delineation quality is proposed, and the specific implementation process is as follows.

Example one

Fig. 1 is a flowchart of a method for determining a target delineation quality according to an embodiment of the present invention. The embodiment is applicable to the situation of automatically determining the target delineation quality, and is particularly applicable to the situation of automatically determining the target delineation quality from the perspective of how the user manually determines the target delineation quality. The method may be performed by an apparatus for determining the target delineation quality provided by the embodiment of the present invention, and the apparatus may be implemented by software and/or hardware, and the apparatus may be integrated on an electronic device.

Referring to fig. 1, the method of the embodiment of the present invention specifically includes the following steps:

s110, obtaining a target area delineation of a target area in the medical image and obtaining a quality determination object, wherein the quality determination object comprises an endangered part and/or a prior appearance of the delineated target area.

Wherein, the target area of having sketched out can be the target area at certain tumour place in the medical image, and the target area sketching can be this sketching out result of target area of having sketched out, and it can be by the user to the sketching out result that the target area of having sketched out carried out the manual sketching out and obtained, also can be the sketching out result that obtains after inputing to the target area sketching out the model with medical image, and the target area sketching model can be neural network model or mathematical model. On the basis, an optional target delineation determining scheme is provided, wherein segmented images of all parts in a medical image are determined, and target delineation is determined according to the segmented images and prior delineation knowledge of the delineated target, wherein the prior delineation knowledge can comprise target delineation and delineation position relation between all parts. Specifically, the delineation position relationship may indicate a position of a position to which each boundary of the target region needs to reach, for example, an upper boundary of the target region needs to reach a lower side of the clavicle, the target region cannot invade the pectoralis major muscle, and the like, so that it is necessary to determine positions of the clavicle and the pectoralis major muscle in the medical image, and then determine the delineation of the target region according to the specific positions. Therefore, a segmented image of each part in the medical image, where the part may be an organ, a tissue, or the like, may be determined first, the segmented image may be an image obtained by segmenting each part from the medical image, then the position of each part on the medical image is determined according to each segmented image, and then a target delineation of the delineated target area is obtained by combining with a priori delineation knowledge. In addition, the segmented image of each part may be obtained based on different image segmentation models, or may be obtained based on the same image segmentation model, and in this case, different parts may be represented in the image segmentation model based on different labels, which is not specifically limited herein.

The quality determination object may be an object related to determining a quality of the target delineation, such as an at-risk region, a prior topography of the delineated target, and the like. The endangered part can be a part which cannot be sketched by the target area in each part, or a part to be protected under the influence of radiation, such as endangered tissues, endangered organs and the like; the a priori topography may be the appearance of the shape that the target region should have, such as a triangle, butterfly, ellipse, etc.

S120, determining delineation information delineated by the target area and object information of the quality determination object, wherein the object information comprises an image at the position at risk and/or prior topography characteristics of the prior topography.

The delineation information is information related to delineation of the target area, and may be, for example, a delineation position of the target area on the medical image, a delineation feature of the target area, and the like. The object information may be information of the quality determination object, for example, when the quality determination object is a critical part, the object information may be a critical image of the critical part, the critical image may be a segmented image of the critical part on the medical image, that is, the critical image may show positioning information of the critical part on the medical image and/or description information of the critical part; further illustratively, when the quality determination object is a prior topography, the object information may be a prior topography feature of the prior topography, and the prior topography feature may be a feature obtained by feature extraction of the prior topography, and may optionally be represented by a number (e.g. 01). The object information may be obtained by inputting the medical image into a corresponding neural network model, or may be obtained in other manners, which is not specifically limited herein.

S130, determining punishment relation between the target area delineation and the quality determination object according to the delineation information and the object information, and determining the target area delineation quality of the target area delineation according to the punishment relation.

The penalty relationship may be a relative relationship between the target region delineation and the quality determination object, and may have different meanings in different application scenarios. For example, when the quality determination object is a site of risk, the penalty relationship may be whether the target region delineates the site of risk that should not be violated, and specifically may be that the site of risk is not violated, that the site of risk is violated, and what the extent and/or location of the violation is; as another example, when the quality determination object is a prior topography, the penalty relationship may be whether a delineation topography delineated by the target region conforms to the prior topography, specifically may be complete conformance, partial non-conformance and degree of non-conformance, and/or what the non-conformance position is, and the like, i.e., it may indicate whether a target sub-delineation not conforming to the prior topography is included in the target region delineation, and the like. The penalty relationship can be determined in various ways, such as inputting the delineation information and the object information into a pre-trained neural network model or a pre-set mathematical model; for example, comparing the delineation information with the object information, comparing the delineation position with the endangered position of the endangered image in the medical image, comparing the delineation topographic features with the prior topographic features, and the like; of course, the penalty relationship may also be determined in other ways, which are not specifically limited herein.

Furthermore, because punishment relation can show whether have the information that can produce negative effects to the target area delineation quality in the target area delineation, consequently can determine the target area delineation quality that the target area delineated according to punishment relation. Illustratively, continuing with the above example as an example, if it is determined from the positional relationship that the target delineation violates a compromised site that should not be violated, then the target delineation quality is relatively low; if the target area sub-delineation which does not accord with the prior morphology exists in the target area delineation according to the punishment relation, the target area delineation quality is lower; etc., and are not specifically limited herein. Of course, at least two of the cases can be combined together to determine the target delineation quality, which is determined from different angles, thereby improving the accuracy of determining the target delineation quality.

On the basis, optionally, besides determining the target area delineation quality according to the penalty relationship, the target area delineation quality can also be determined, and the sub-delineation to be modified and the modification reason of the sub-delineation to be modified in the target area delineation can be determined, so that the sub-delineation to be modified and the modification reason can be displayed. The sub-sketch to be modified can be a part which does not meet the requirements of the related sketch in the target area sketch, and the modification reason is the reason why the sub-sketch to be modified does not meet the requirements of the related sketch. Illustratively, continuing with the above example as an example, if it is determined that a part of the target delineation violates the at-risk part according to the penalty relationship, the part may be regarded as a sub-delineation to be modified, and the modification reason may be that the part violates the at-risk part; as another example, if it is determined from the penalty relationship that there is a portion of the target delineation that does not conform to the prior topography, the portion may be used as a sub-delineation to be modified, and the modification reason may be that the portion does not conform to the prior topography. Furthermore, the sub-sketching and the modification reason to be modified are contents which can be directly understood by the user, so that the sub-sketching and the modification reason to be modified can be displayed, the user can intuitively know whether a problem exists in the target sketching, what problem exists, which part of the problem appears and the like according to the display result, and can intuitively know a feasible modification scheme, and therefore the actual application value of determining the target sketching quality is improved. The user may be a doctor or the like who is involved in the target delineation.

In practical application, optionally, in order to implement digitization of the target region delineation quality, a preset quality determination index may be obtained first, then an index score of the quality determination index is calculated according to a penalty relationship, and the index score is used as the target region delineation quality of the target region delineation. The quality determination index may be an index used for determining the target delineation quality, such as an IOU, an accuracy rate, a recall rate, a Dice coefficient, a Hausdorff distance, and the like. The specific implementation process of calculating the index score of the quality determination index according to the penalty relationship may be related to the type of the quality determination index, and for example, assuming that the quality determination index is an IOU, the intersection area of the intersection region and the union area of the union region between the target region delineation and the quality determination object may be determined according to the penalty relationship, and then the numerical value of the IOU may be determined according to the intersection area and the union area; further exemplarily, assuming that the quality determination index is a Dice coefficient, the similarity between the character string corresponding to the target region and the character string corresponding to the quality determination object may be determined according to the penalty relationship, and the similarity is used as a numerical value of the Dice coefficient; etc., which are not specifically limited herein. The concrete expression form of the index score may be related to the type of the quality determination index, or may be a preset form, for example, assuming that the calculation result of the quality determination index is a numerical value, but each numerical value is mapped with "excellent", "good", "medium" and "poor" in advance, the index score may be expressed by "excellent", "good", "medium" and "poor".

Example two

Fig. 2 is a flowchart of a method for determining target delineation quality according to a second embodiment of the present invention. The present embodiment is optimized based on the above technical solutions. In this embodiment, optionally, the quality determination object includes an endangered part and the object information includes an endangered image, and the delineation information includes a delineation position of the target region in the medical image; correspondingly, confirm the punishment relation between target area sketch and the quality determination object according to sketch information and object information, confirm the target area sketch quality that the target area sketch was drawn according to punishment relation, include: determining a punishment relation between the target delineation and the endangered part according to the delineation position and the endangered position of the endangered image in the medical image; and judging whether the target area sketch invades the endangered part according to the punishment relation, and if so, determining the target area sketch quality of the target area sketch according to the invasion degree. The same or corresponding terms as those in the above embodiments are not explained in detail herein.

Referring to fig. 2, the method of the present embodiment may specifically include the following steps:

s210, target area delineation of a target area already delineated in the medical image and a position at risk in the medical image are obtained.

S220, determining a delineation position of the target region in the medical image and an endangered image of the endangered part.

And S230, determining a penalty relation between the target region delineation and the endangered position according to the delineation position and the endangered position of the endangered image in the medical image.

Wherein, the delineation position may be a position where the target region is delineated on the medical image, which may indicate which positions on the medical image the target region delineation exists; and the location of the compromise may be the location of the compromise image on the medical image, which may indicate which locations on the medical image the compromised site is present. The position in the medical image is used as an intermediate medium, and various implementation modes for determining the penalty relationship according to the endangered position and the delineating position exist, for example, whether the penalty relationship invades or does not invade the endangered part can be determined according to whether the positions of the endangered position and the delineating position are coincident or not; further, the invasion degree can be determined according to the number of the coincident positions, namely, the invasion to the endangered part and the invasion degree can be used as a punishment relation; and so on.

S240, judging whether the target area sketching invades to a dangerous part according to the punishment relation, and if so, determining the target area sketching quality of the target area sketching according to the invasion degree.

Since the target delineation should not invade the endangered part in principle, when the target delineation is determined to invade the endangered part according to the penalty relationship, the invasion degree of the target delineation aiming at the endangered part can be further determined, and the invasion degree can be represented by various forms, such as area, distance, area ratio, distance ratio and the like, and is not specifically limited herein. Further, the target delineation quality of the target delineation can be determined according to the invasion degree, if the invasion degree is high, the target delineation quality is very poor, and at the moment, information such as warning, modification suggestion and the like can be given, if the target delineation invades to the bladder inner boundary by 5 mm; for example, when the invasion degree is low, the target region delineation quality is good; and so on. Of course, if the target delineation does not impinge on the site of risk, then it is considered from this perspective that no serious problems exist in the target delineation. In practical applications, optionally, the target delineation quality may be represented by a score, where the score may be at least one of a score, a pass-fail, a good-medium-poor, a + a-, and the like, and then after the invasion degree is determined, the score of the target delineation quality may be determined according to the invasion degree and a preset degree score mapping relationship.

According to the technical scheme, when the quality determination object comprises the dangerous part, the object information comprises the dangerous image, and the delineation information comprises the delineation position of the target region on the medical image, whether the target region delineation invades the dangerous part can be judged according to the dangerous position and the delineation position of the dangerous image on the medical image, if so, the delineation quality of the target region delineation of the target region can be determined according to the invasion degree of the target region delineation relative to the dangerous part, and therefore the effect that the delineation quality of the target region consistent with the target region delineation quality determined manually by a user is automatically determined and determined from the angle of the invasion degree of the target region delineation relative to the dangerous part frequently involved in the process of manually determining the target region delineation quality by the user is achieved.

EXAMPLE III

Fig. 3 is a flowchart of a method for determining the target delineation quality provided in the third embodiment of the present invention. The present embodiment is optimized based on the above technical solutions. In this embodiment, optionally, the quality determination object includes a priori topography and the object information includes a priori topography feature, and the delineation information includes delineation topography feature of a delineation topography delineated by the target region; correspondingly, confirm the punishment relation between target area sketch and the quality determination object according to sketch information and object information, confirm the target area sketch quality that the target area sketch was drawn according to punishment relation, include: determining a punishment relation between the sketched morphology and the prior morphology according to the sketched morphology features and the prior morphology features; and if it is determined according to the punishment relationship that target sub-delineations which do not conform to the prior morphology exist in the target delineation, determining the target delineation quality of the target delineation according to the target sub-delineations and the target delineation. The same or corresponding terms as those in the above embodiments are not explained in detail herein.

Referring to fig. 3, the method of this embodiment may specifically include the following steps:

s310, obtaining the target area delineation of the delineated target area and the prior appearance of the delineated target area in the medical image.

It is considered that although there may be differences in the positions and sizes of different tumors, the target regions where the tumors are located are likely to have similar a priori morphologies because they are limited by the distribution of surrounding regions, and they are generally three-dimensionally continuous, wherein the a priori morphologies may be the shape appearances that the target regions should have been delineated. Therefore, when determining the target delineation quality, the prior morphology of the corresponding delineated target can be used as a determining factor, which is also one of the factors frequently referred to in the process of manually determining the target delineation quality by a user.

S320, determining the delineation morphological characteristics of the delineation morphology delineated by the target area and the prior morphological characteristics of the prior morphology.

The target area delineation is the actual topography of the target area delineation, and the prior topography is the topography that the target area delineation should have, in other words, when the delineation topography approaches to the prior topography, the target area delineation quality of the target area delineation is higher, namely when the delineation topography is not in accordance with the prior topography, there is a problem in the target area delineation. On this basis, it should be noted that, in order to ensure the realizability in engineering, the delineation morphology and the prior morphology may be digitally represented, for example, the delineation morphology may be represented by delineation morphology features, and the prior morphology may be represented by the prior morphology features.

On the basis, the sketched feature may be obtained in various ways, for example, after feature extraction is performed on the sketched feature, the target region sketching is input into a pre-trained appearance feature generation model, and the like, which is not specifically limited herein. Correspondingly, the prior morphological characteristics can also be obtained in various ways, for example, in consideration of the fact that the gold standard drawing can be used as the most accurate target area drawing, the characteristics of the gold standard drawing on the shape appearance can be used as the prior morphological characteristics; as another example, since the three-dimensional continuous prior morphology can be well fitted based on the neural network, the medical image can be input into the trained morphology feature generation model, and the prior morphology feature is determined according to the output result of the morphology feature generation model, where the morphology feature generation model can be trained in advance through the following steps: acquiring a historical image and the labeled morphology features of the historical target area in the historical image, inputting the historical image into a generator, and obtaining the historical morphology features of the historical target area according to the output result of the generator; and inputting the historical morphology features and the labeled morphology features into a discriminator, and adjusting network parameters in a generator according to the output result of the discriminator to obtain a morphology feature generation model, namely, the generator after network parameter adjustment can be used as the morphology feature generation model.

S330, determining a punishment relation between the sketched morphology and the prior morphology according to the sketched morphology features and the prior morphology features.

The delineation feature and the prior feature are compared, and the punishment relationship between the delineation feature and the prior feature can be determined according to the comparison result. Illustratively, if the delineation topographic features and the prior topographic features are completely consistent, the corresponding penalty relationship may be that the delineation topographic features and the prior topographic features are completely consistent; otherwise, the punishment relation can be that the two do not completely accord with each other, and further, which part in the delineation morphology does not accord with the prior morphology can be determined.

And S340, if it is determined according to the punishment relationship that target sub-delineations which do not conform to the prior morphology exist in the target delineation, determining the target delineation quality of the target delineation according to the target sub-delineations and the target delineation.

If a certain part in the delineation morphology is not coincident with the prior morphology according to the punishment relationship, the corresponding part of the part in the target delineation can be used as the target sub-delineation, namely the target sub-delineation is the part which is not coincident with the prior morphology in the target delineation. For example, assuming that the target delineation is a sphere and the a priori delineations corresponding to the a priori topographies are triangles, the non-overlapping portions of the two may be referred to as target sub-delineations. Further, the target area delineation quality can be determined according to the target area sub-delineation and the target area delineation, for example, the ratio of the area of the sub-delineation area formed according to the target area sub-delineation to the area of the delineation area formed according to the target area delineation is used as the target area delineation quality; then, the ratio of the number of the pixel points contained in the target region sketch to the number of the pixel points contained in the target region sketch is used as target region sketch quality; and so on.

According to the technical scheme, when the quality determination object comprises the prior appearance, the object information comprises the prior appearance characteristics, and the delineation information comprises the delineation appearance characteristics of the delineation appearance delineated by the target area, whether target area sub-delineation which does not accord with the prior appearance exists in the target area delineation can be determined according to the comparison result between the delineation appearance characteristics and the prior appearance characteristics, and if so, the target area delineation quality delineated by the target area can be determined according to the target area sub-delineation and the target area delineation, so that the effect of automatically determining and determining the target area delineation quality consistent with the target area delineation quality manually determined by a user from the angle of whether the delineation appearance delineated by the target area frequently involved in the target area in the process of manually determining the target area delineation quality by the user accords with the prior appearance is achieved.

Example four

Fig. 4 is a flowchart of a method for determining the target delineation quality provided in the fourth embodiment of the present invention. The present embodiment is optimized based on the above technical solutions. In this embodiment, optionally, the method for determining the target delineation quality may further include: obtaining a gold standard sketch of a sketched target area and a target area image of a target area part in a medical image; determining an overlapping area between the target area sketching and the gold standard sketching, and determining the target area weight of each target area pixel point in the target area image based on a preset weight determination function; determining the overlapping weight of each overlapping pixel point in the overlapping area according to the overlapping position of the overlapping area in the medical image and the target position of the target image in the medical image; determining the basic quality delineated by the target area according to the overlapping area and each overlapping weight; correspondingly, determining the target area delineation quality of the target area delineation according to the penalty relationship may include: and determining punishment quality of target area sketching according to the punishment relation, and determining the target area sketching quality of the target area sketching according to the basic quality and the punishment quality. The same or corresponding terms as those in the above embodiments are not explained in detail herein.

Referring to fig. 4, the method of this embodiment may specifically include the following steps:

s410, obtaining a quality determination object in the medical image, a target area image of the target area, and a target area sketch and a gold standard sketch of the sketched target area, wherein the quality determination object comprises a prior appearance of the endangered part and/or the sketched target area.

The target area in the medical image is a tumor location, the target area image is a segmentation image of the target area in the medical image, and the acquisition mode of the target area image can refer to a endangered image, which is not described herein any more. Of course, the target region image and the endangered image may be acquired in the same or different manners, and are not specifically limited herein. The gold standard delineation of the delineated target area can be the most standard delineation result of the delineated target area, which can be obtained by manually delineating the delineated target area by a user, or can be obtained by inputting medical images into the target area delineation model and evaluating the target area delineation model by the user, and the like, and is not specifically limited here.

S420, determining an overlapping area between the target area delineation and the gold standard delineation, and determining the target area weight of each target area pixel point in the target area image based on a preset weight determination function.

When determining the quality of the target area delineation, whether a part which does not meet the requirements of the related delineation exists in the target area delineation or not is considered, and whether a part which is consistent with the gold standard delineation exists in the target area delineation or not can be considered, so that the quality of the target area delineation can be determined from two angles of the front surface and the back surface, therefore, the overlapping area between the target area delineation and the gold standard delineation can be determined, and particularly the overlapping area between the gold standard areas formed by the gold standard delineation of the target area region formed by the target area delineation is determined.

Further, in a clinical actual scene, there is a high possibility that there is a difference in the effect when performing radiotherapy on different positions of the target region, and the reason for this is that each tumor at least includes a primary focus and a subclinical focus, and the subclinical focus is a portion that is expanded from the primary focus, taking a rectal tumor as an example, the primary focus is in the rectum and the subclinical focus is around the primary focus, and since the subclinical focus cannot detect the tumor on the medical image, but the tumor does exist on the subclinical focus, but there are many and few problems, so the effect of performing radiotherapy on the position where the primary focus is located in the target region is stronger than performing radiotherapy on the position where the subclinical focus is located. Since the primary and sub-clinical foci are located at different positions of the target region respectively, this means that the influence of the target region delineation on different positions of the target region on the target region delineation quality is different. In other words, if the overlap region is located on the primary foci, its impact on the target delineation quality may be greater; accordingly, if the overlap region is located on a subclinical focus, its impact on the quality of the target delineation may be smaller.

In view of the above, in order to show that the importance degrees at different positions of the target region are different, the target region weights of the target region pixel points in the target region image of the target region can be determined based on the preset weight determining function, the weight determining function can be a function set based on linear weights, gaussian weights and the like, illustratively, the rectal tumor is taken as an example continuously, the target region weights of the target region pixel points of the rectum in the target region image are the highest, and the target region weights of the target region pixel points other than the target region pixel points of the rectum can be gradually reduced along with the increase of the distance between the target region pixel points and the rectum. In other words, in principle, the target region outlines the parts except for the rectum which should be wrapped by more than a few of the rectum, so the target region weight of the target region pixel point where the rectum which should be wrapped by more than a few of the rectum is located in the target region image may be higher, and the target region weight where the parts which should be wrapped by less than a few of the rectum is located may be lower. It should be noted that, the weight determination function may be set according to the target weights of the target pixels located at the positions, and then the positions of the respective positions on the medical image may be determined according to the segmented images of the respective positions.

S430, determining the overlapping weight of each overlapping pixel point in the overlapping area according to the overlapping position of the overlapping area in the medical image and the target position of the target image in the medical image; and determining the basic quality of the target delineation according to the overlapping areas and the overlapping weights.

Wherein, to the overlapping position with the coincidence of target area position, can regard the target area weight of the target area pixel point that is located on this target area position as the overlapping weight of the overlapping pixel point on this overlapping position. Therefore, the basic quality of the target delineation can be determined according to the overlapping region and each overlapping weight, for example, the basic quality is determined according to the overlapping area of the overlapping region, the overlapping weight of each overlapping pixel point in the overlapping region and the delineation area of the target delineation, and the basic quality can be the quality which has positive influence on the target delineation quality.

S440, determining the sketching information of the target area and the object information of the quality determination object, determining the punishment relation between the target area sketching and the quality determination object according to the sketching information and the object information, and determining the punishment quality of the target area sketching according to the punishment relation.

The object information may include an image at the location of the target and/or a priori topographic features of the priori topography, and the penalty quality may be a quality that negatively affects the delineation quality of the target.

S450, determining the target area delineation quality delineated by the target area according to the basic quality and the punishment quality.

The punishment quality which has negative influence on the target area delineation quality and the basic quality which has positive influence on the target area delineation quality are comprehensively considered, so that the accuracy of the obtained target area delineation quality is higher. For example, taking the example of representing the quality by the score, when the penalty qualities include a first penalty quality obtained when the at-risk portion is involved and a second penalty quality obtained when the feature is involved, assuming that the specific score of the base quality is 90, the specific score of the first penalty quality is 10, and the specific score of the second penalty quality is 13, the specific score of the final target-region delineation quality is 90-10-13-67.

According to the technical scheme of the embodiment of the invention, because the importance degrees of different positions on the target area part relative to radiotherapy are different, in order to reflect the difference, the target area weight of each target area pixel point in the target area image of the target area part can be determined based on the preset weight determining function, and the overlapping position of the overlapping area between the target area sketch and the gold standard sketch in the medical image and the target area position of the target area image in the medical image are combined, so that the overlapping weight of each overlapping pixel point in the overlapping area can be obtained; therefore, the target area delineation quality can be determined from a plurality of angles in the positive and negative aspects according to the basic quality of the target area delineation determined based on the overlapping region and each overlapping weight and the punishment quality of the target area delineation determined based on the punishment relation, and the determination accuracy of the target area delineation quality is improved.

In order to better understand the specific implementation process of the above steps, the following describes an exemplary method for determining the target delineation quality according to this embodiment with reference to a specific example. Illustratively, considering the complexity of the target area, the present example considers the target area distribution, the organs at risk protection and other constraint information in combination, and automatically and objectively determines the target area delineation quality of the target area delineation. The method can be applied to various application scenarios such as control of target delineation quality, doctor target delineation training, target segmentation model performance evaluation and the like. The present example may include a plurality of mutually independent modules, such as a digital medical image representation module, a target weight determination module, an organ protection determination module, a target morphology prior module, a comprehensive determination module, and a visualization module, which may be flexibly invoked according to different application scenarios.

Specifically, as shown in fig. 5, a medical image is input to a medical image digital representation module, so that the medical image digital representation module inputs the medical image into a pre-trained image segmentation model, obtains segmented images of organs and tissues in the medical image according to an output result of the image segmentation model, and outputs the segmented images, wherein the segmented images belonging to the organs at risk may be referred to as images at risk, and the segmented images belonging to the target organs and the target tissues may be referred to as images at target. The method comprises the steps of obtaining target area sketching and gold standard sketching of sketched target areas in medical images, inputting each segmented image, the target area sketching and the gold standard sketching into a target area weight determining module, enabling the target area weight determining module to determine basic quality of the sketched target areas by combining input data and preset weight determining functions, and outputting the basic quality. The at-risk image and the target volume delineation are input into an organ protection determination module to cause the organ protection determination module to determine a first penalty quality from the input data. Inputting the medical image and the delineation morphology into a target area morphology prior module, so that the target area morphology prior module inputs the medical image into a morphology generation model trained in advance, obtaining prior morphology characteristics according to an output result of the morphology generation model, comparing the prior morphology characteristics with the prior morphology characteristics obtained after extracting the characteristics of the delineation morphology, and obtaining second punishment quality according to a comparison result. And inputting the basic quality, the first punishment quality and the second punishment quality into the comprehensive determination module so that the comprehensive determination module obtains the target area sketching quality sketched in the target area, the sub-sketches to be modified in the target area sketching and the modification reasons of the sub-sketches to be modified according to the input data, and outputs the target area sketching quality, the sub-sketches to be modified and the modification reasons. The target area delineation quality, the sub-delineations to be modified and the modification reasons are input into the visualization module, so that the visualization module displays the input data, for example, the sub-delineations to be modified are highlighted in the medical image, and the modification reasons are marked.

EXAMPLE five

Fig. 6 is a flowchart of a method for displaying the target delineation quality provided in the fifth embodiment of the present invention. The embodiment is applicable to the situation of displaying the delineation quality of the target area on the medical image. The method can be executed by a presentation apparatus for target delineation quality provided by the embodiment of the present invention, and the apparatus can be implemented by software and/or hardware, and the apparatus can be integrated on an electronic device. The same or corresponding terms as those in the above embodiments are not explained in detail herein.

Referring to fig. 6, the method of the embodiment of the present invention specifically includes the following steps:

s510, receiving the medical image and the target area delineation of the target area already delineated in the medical image.

Wherein, the user can input medical image and the target area sketch of the target area in this medical image through electronic equipment, this moment this electronic equipment can receive corresponding medical image and target area sketch to can carry out corresponding processing according to the two, obtain the quality information that is relevant with the target area sketch quality that the target area sketch was described, for example score, the sub-sketch that treats the modification in the target area sketch, the modification reason that treats the sub-sketch that treats the modification in the target area sketch and so on of target area sketch quality, do not specifically limit here.

S520, outputting the processed medical image, and displaying quality information related to the target delineation quality of the target delineation on the processed medical image, wherein the quality information comprises at least one of scores of the target delineation quality, sub-delineations to be modified in the target delineation and modification reasons of the sub-delineations to be modified, the quality information is determined according to the penalty relationship between the target delineation and the quality determination objects in the medical image, the penalty relationship is determined according to the delineation information of the target delineation and the object information of the quality determination objects, the quality determination objects comprise the prior topography of the endangered parts and/or the delineated target areas, and the object information comprises the prior topography characteristics of the endangered parts and/or the prior topography.

After the electronic device processes the obtained corresponding quality information, it may process the quality information onto the medical image, so that the electronic device may output the processed medical image, and display the quality information on the processed medical image, such as displaying the score at a preset score position on the medical image, displaying the sub-sketch to be modified on the medical image based on a preset display mode (such as highlighting display, color-changing display, and the like), displaying the modification reason of the sub-sketch to be modified near the sub-sketch to be modified, and the like.

On this basis, optionally, if the quality information is the sub-sketch to be modified, the sub-sketch to be modified may be displayed on the processed medical image based on the first display mode, and the target modification sub-sketch corresponding to the sub-sketch to be modified may be displayed on the processed medical image based on the second display mode, where the target modification sub-sketch may be a modification target of the sub-sketch to be modified determined according to a modification reason, which may help a user to directly determine which part of the target region sketch needs to be modified and a feasible modification scheme, thereby improving convenience of user operation.

On this basis, further optionally, the first display mode may include a highlight mode based on a first color, and the second display mode may include a highlight mode based on a second color, so that besides displaying the quality information, description information of the highlight mode of the first color and description information of the highlight mode of the second color may be displayed on the processed medical image, which may help a user to intuitively know which highlight part of the color is a part to be modified and which highlight part of the color is a modification target, thereby improving friendliness of user interface interaction.

Above-mentioned technical scheme has realized the effect that target area delineation quality demonstrates on medical image through the interactive mode of user interface, and this makes the user can confirm fast and directly perceivedly whether have the problem in score, the target area delineation with target area delineation quality, what problem exists, the problem appears in which part grade with target area delineation quality relevant quality information, has better practical application and worth.

EXAMPLE six

The device for determining the target delineation quality provided by the sixth embodiment of the present invention is used for executing the method for determining the target delineation quality provided by any of the above embodiments. The device and the method for determining the target delineation quality of each embodiment belong to the same inventive concept, and details which are not described in detail in the embodiment of the device for determining the target delineation quality of each embodiment can refer to the embodiment of the method for determining the target delineation quality of each embodiment. The device may specifically comprise: the device comprises a quality determination object acquisition module, an object information determination module and a target area delineation quality determination module.

The quality determination object acquisition module is used for acquiring a target area delineation of a delineated target area in a medical image and a quality determination object, wherein the quality determination object comprises a dangerous part and/or a prior appearance of the delineated target area;

the object information determining module is used for determining delineation information delineated by the target area and object information of a quality determination object, wherein the object information comprises a dangerous image of a dangerous part and/or prior morphology characteristics of the prior morphology;

and the target area delineation quality determination module is used for determining the punishment relation between the target area delineation and the quality determination object according to the delineation information and the object information, and determining the target area delineation quality of the target area delineation according to the punishment relation.

Optionally, the quality determination object includes an endangered part and the object information includes an endangered image, and the delineation information includes delineation positions of the target region in the medical image;

the target delineation quality determination module specifically may include:

the first punishment relation determining unit is used for determining punishment relation between the target delineation and the endangered part according to the delineation position and the endangered position of the endangered image in the medical image;

and the first target area delineation quality determination unit is used for judging whether the target area delineation invades the endangered part according to the punishment relation, and if so, determining the target area delineation quality of the target area delineation according to the invasion degree.

Optionally, the quality determination object includes a prior topography and the object information includes a prior topography feature, and the delineation information includes a delineation topography feature of a delineation topography delineated by the target region;

the target delineation quality determination module specifically may include:

the second punishment relation determining unit is used for determining punishment relation between the delineation morphology and the prior morphology according to the delineation morphology characteristics and the prior morphology characteristics;

and the second target area delineation quality determining unit is used for determining the target area delineation quality of the target area delineation according to the target area sub-delineation and the target area delineation if the target area sub-delineation which does not conform to the prior appearance exists in the target area delineation according to the punishment relation.

Optionally, the target delineation quality determination module may specifically include:

and the third target area delineation quality determination unit is used for acquiring preset quality determination indexes, calculating index scores of the quality determination indexes according to the penalty relationship, and taking the index scores as target area delineation quality of the target area delineation.

Optionally, the target delineation quality determination apparatus may further include:

the target area image acquisition module is used for acquiring a gold standard sketch of the sketched target area and a target area image of the target area part in the medical image; the target area weight determining module is used for determining an overlapping area between the target area delineation and the gold standard delineation, and determining the target area weight of each target area pixel point in the target area image based on a preset weight determining function; the overlapping weight determining module is used for determining the overlapping weight of each overlapping pixel point in the overlapping area according to the overlapping position of the overlapping area in the medical image and the target position of the target image in the medical image; a basic quality determination module for determining the basic quality delineated by the target area according to the overlapping region and each overlapping weight;

correspondingly, the target delineation quality determination module may specifically include:

and the fourth target area delineation quality determining unit is used for determining punishment quality of the target area delineation according to the punishment relation and determining the target area delineation quality of the target area delineation according to the basic quality and the punishment quality.

the target area delineation determining module is used for determining segmentation images of all parts in the medical image, and determining target area delineation according to the segmentation images and prior delineation knowledge of the delineated target area, wherein the prior delineation knowledge comprises the target area delineation and the delineation position relation between all parts.

According to the device for determining the target area delineation quality provided by the sixth embodiment of the invention, the quality determination object acquisition module, the object information determination module and the target area delineation quality determination module are matched with each other, so that the effect of automatically determining and determining the target area delineation quality consistent with the target area delineation quality manually determined by a user is realized.

The device for determining the target delineation quality provided by the embodiment of the invention can execute the method for determining the target delineation quality provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

It should be noted that, in the embodiment of the device for determining the target delineation quality, each unit and each module included in the device are only divided according to functional logic, but are not limited to the above division, as long as the corresponding function can be implemented; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.

EXAMPLE seven

The display device for the target delineation quality provided by the seventh embodiment of the invention is used for executing the display method for the target delineation quality provided by any embodiment. The device and the method for displaying the target delineation quality of each embodiment belong to the same inventive concept, and details which are not described in detail in the embodiment of the device for displaying the target delineation quality of each embodiment can refer to the embodiment of the method for displaying the target delineation quality of each embodiment. The device may specifically comprise: the target area sketching receiving module and the target area sketching quality display module.

The target area delineation receiving module is used for receiving the medical image and the target area delineation of the delineated target area in the medical image;

the target area delineation quality display module is used for outputting the processed medical image and displaying quality information related to the target area delineation quality delineated by the target area on the processed medical image, wherein the quality information comprises at least one of scores of the target area delineation quality, sub delineations to be modified in the target area delineation and modification reasons of the sub delineations to be modified;

Optionally, the quality information includes a sub-sketch to be modified, and the target area sketch quality display module includes:

and the target area delineation quality display unit is used for displaying the sub-delineation to be modified on the processed medical image based on the first display mode and displaying the target modification sub-delineation corresponding to the sub-delineation to be modified on the processed medical image based on the second display mode, wherein the target modification sub-delineation is a modification target of the sub-delineation to be modified determined according to the modification reason.

On this basis, optionally, the first display mode includes a highlighting display mode based on a first color, and the second display mode includes a highlighting display mode based on a second color, and the display device for delineating the quality of the target area may further include: and the description information display module is used for respectively displaying the description information of the highlight display mode with the first color and the highlight display mode with the second color on the processed medical image.

According to the display device for the target delineation quality provided by the seventh embodiment of the invention, the target delineation receiving module and the target delineation quality display module are mutually matched, and the effect of displaying the target delineation quality on the medical image is realized through a user interface interaction mode, so that a user can quickly and intuitively determine quality information related to the target delineation quality, such as whether problems exist in the target delineation, what problems exist, where the problems appear, and the like, and the display device has a good practical application value.

The display device for the target delineation quality provided by the embodiment of the invention can execute the display method for the target delineation quality provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

It should be noted that, in the embodiment of the display apparatus for delineating quality of the target area, the included units and modules are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.

Example eight

Fig. 7 is a schematic structural diagram of an electronic device according to an eighth embodiment of the present invention, as shown in fig. 7, the electronic device includes a memory 610, a processor 620, an input device 630, and an output device 640. The number of processors 620 in the device may be one or more, and one processor 620 is taken as an example in fig. 7; the memory 610, processor 620, input device 630, and output device 640 in the apparatus may be connected by a bus or other means, such as by bus 650 in fig. 7.

The memory 610 is used as a computer readable storage medium and may be used to store a software program, a computer executable program, and modules, such as program instructions/modules corresponding to the method for determining the target delineation quality in the embodiment of the present invention (for example, a quality determination object obtaining module, an object information determination module, and a target delineation quality determination module in the apparatus for determining the target delineation quality), or program instructions/modules corresponding to the method for displaying the target delineation quality in the embodiment of the present invention (for example, a target delineation receiving module and a target delineation quality displaying module in the apparatus for displaying the target delineation quality). The processor 620 executes software programs, instructions and modules stored in the memory 610 to execute various functional applications and data processing of the device, that is, to implement the above-mentioned method for determining the target delineation quality or the method for displaying the target delineation quality.

The memory 610 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application program required for at least one function; the storage data area may store data created according to use of the device, and the like. Further, the memory 610 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some examples, memory 610 may further include memory located remotely from processor 620, which may be connected to devices through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.

The input device 630 may be used to receive input numeric or character information and generate key signal inputs related to user settings and function controls of the device. The output device 640 may include a display device such as a display screen.

From the above description of the embodiments, it is obvious for those skilled in the art that the present invention can be implemented by software and necessary general hardware, and certainly, can also be implemented by hardware, but the former is a better embodiment in many cases. With this understanding, the technical solutions of the present invention may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a floppy disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a FLASH Memory (FLASH), a hard disk or an optical disk of a computer, and includes instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute the methods according to the embodiments of the present invention.

It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.

Claims

1. A method for determining the quality of a target delineation is characterized by comprising the following steps:

acquiring a target delineation and a quality determination object of a delineated target in a medical image, wherein the quality determination object comprises a prior appearance of an endangered part and/or the delineated target;

determining delineation information delineated by the target region and object information of the quality determination object, wherein the object information comprises an at-risk image of the at-risk region and/or a priori topographic features of the a priori topography;

and determining a punishment relation between the target area sketch and the quality determination object according to the sketch information and the object information, and determining the target area sketch quality of the target area sketch according to the punishment relation.

2. The method of claim 1, wherein the quality determination object comprises the at-risk site and the object information comprises the at-risk image, the delineation information comprising delineation locations of the target volume in the medical image;

correspondingly, the determining a punishment relationship between the target delineation and the quality determination object according to the delineation information and the object information, and determining the target delineation quality of the target delineation according to the punishment relationship include:

determining a punishment relation between the target delineation and the endangered position according to the delineation position and the endangered position of the endangered image in the medical image;

and judging whether the target area delineation invades the endangered part or not according to the punishment relation, and if so, determining the target area delineation quality of the target area delineation according to the invasion degree.

3. The method of claim 1, wherein the quality determination object comprises the prior topography and the object information comprises the prior topography features, the delineation information comprising delineation topography features of a delineation topography delineated by the target region;

determining a punishment relation between the delineation morphology and the prior morphology according to the delineation morphology features and the prior morphology features;

and if it is determined according to the punishment relationship that a target sub-delineation which does not accord with the prior morphology exists in the target delineation, determining the target delineation quality of the target delineation according to the target sub-delineation and the target delineation.

4. The method according to claim 1, wherein said determining a target delineation quality of the target delineation according to the penalty relationship comprises:

and acquiring a preset quality determination index, calculating an index score of the quality determination index according to the punishment relation, and taking the index score as the target area delineation quality of the target area delineation.

5. The method of claim 1, further comprising:

acquiring a gold standard sketch of the sketched target area and a target area image of a target area part in the medical image;

determining an overlapping area between the target area delineation and the gold standard delineation, and determining target area weights of target area pixel points in the target area image based on a preset weight determination function;

determining the overlapping weight of each overlapping pixel point in the overlapping region according to the overlapping position of the overlapping region in the medical image and the target position of the target image in the medical image;

determining a basis mass of the target delineation from the overlap regions and each of the overlap weights;

correspondingly, the determining the target area delineation quality of the target area delineation according to the penalty relationship includes:

and determining punishment quality of the target area sketching according to the punishment relation, and determining the target area sketching quality of the target area sketching according to the basic quality and the punishment quality.

6. The method of claim 1, further comprising:

determining segmented images of all parts in the medical image, and determining the target area delineation according to the segmented images and the prior delineation knowledge of the delineated target area, wherein the prior delineation knowledge comprises the target area delineation and the delineation position relation between all the parts.

7. A display method for target delineation quality is characterized by comprising the following steps:

receiving a medical image and a target region delineation of a target region already delineated in the medical image;

outputting the processed medical image, and displaying quality information related to the target delineation quality of the target delineation on the processed medical image, wherein the quality information comprises at least one of a score of the target delineation quality, a sub-delineation to be modified in the target delineation, and a modification reason of the sub-delineation to be modified;

wherein the quality information is determined from a penalty relationship between the target delineation and a quality determination object in the medical image, the penalty relationship being determined from delineation information delineated by the target and object information of the quality determination object, the quality determination object comprising a prior topography of the at-risk site and/or the delineated target, and the object information comprising a prior topography feature of the at-risk site at-risk image and/or the prior topography.

8. The method according to claim 7, wherein the quality information comprises the sub-delineation to be modified, and wherein the presenting quality information on the target delineation quality of the target delineation on the processed medical image comprises:

displaying the sub-sketch to be modified on the processed medical image based on a first display mode, and displaying a target modification sub-sketch corresponding to the sub-sketch to be modified on the processed medical image based on a second display mode, wherein the target modification sub-sketch is a modification target of the sub-sketch to be modified determined according to the modification reason.

9. The method of claim 8, wherein the first presentation comprises a first color-based highlighting and the second presentation comprises a second color-based highlighting, the method further comprising: and displaying the description information of the highlighting mode of the first color and the highlighting mode of the second color on the processed medical image respectively.

10. An electronic device, comprising:

one or more processors;

a memory for storing one or more programs;

when executed by the one or more processors, cause the one or more processors to implement a method of determining target delineation quality as claimed in any one of claims 1-6, or a method of demonstrating target delineation quality as claimed in claims 7-9.