Method, device and equipment for evaluating quality of X-ray image and storage medium

Abstract

The invention relates to a method for evaluating the quality of an X-ray image, a training method of an image imaging performance scoring model, a device, equipment, a computer readable storage medium and a computer program product. The method for evaluating the quality of the X-ray image comprises the following steps: acquiring a to-be-evaluated image; based on the to-be-evaluated image, a shooting condition score of the to-be-evaluated image is determined, and the shooting condition score is used for evaluating at least one of the following items: the position accuracy of a shot object and whether a foreign matter exists in a shooting area; and based on the to-be-evaluated image, using a trained image imaging performance scoring model to determine an imaging performance score of the to-be-evaluated image, the imaging performance score being used to evaluate performance of images generated at different dose levels and noise levels.

Description

Technical Field

[0001] The present disclosure relates to the field of radiographic imaging technology, and in particular to a method for evaluating X-ray image quality, a training method for an image imaging performance scoring model, a training device for an image imaging performance scoring model, a device for evaluating X-ray image quality, medical imaging equipment, a computer-readable storage medium, and a computer program product. Background Art

[0002] Radiographic imaging technology can be used to assist medical diagnosis. For example, radiographic imaging technology can be used to photograph a subject, thereby obtaining an image that can reflect the subject's physical characteristics, and professionals can then use the image to assist medical diagnosis.

[0003] The methods described in this section are not necessarily methods that have been previously conceived or employed. Unless otherwise indicated, it should not be assumed that any method described in this section is considered to be prior art simply because it is included in this section. Similarly, unless otherwise indicated, the issues mentioned in this section should not be considered to have been recognized in any prior art. Summary of the invention

[0004] In view of this, the first aspect of the present disclosure proposes a method for evaluating the quality of X-ray images. The method comprises: acquiring an image to be evaluated; determining a shooting condition score of the image to be evaluated based on the image to be evaluated, wherein the shooting condition score is used to evaluate at least one of the following: the position accuracy of the object to be evaluated, and whether there is a foreign body in the shooting area; and determining an imaging performance score of the image to be evaluated based on the image to be evaluated using a trained image imaging performance scoring model, wherein the imaging performance score is used to evaluate the performance of images generated at different dose levels and noise levels.

[0005] The second aspect of the present disclosure proposes a training method for an image imaging performance scoring model, which is used in the method for evaluating the quality of X-ray images according to the first aspect of the present disclosure, and the training method includes: acquiring at least one sample image; performing the following processing on the at least one sample image: performing a first processing on the at least one sample image to generate a plurality of first sample images corresponding to images with a plurality of different noise levels to obtain a first data set containing images with different noise levels, and performing a second processing on the at least one sample image to generate a plurality of second sample images corresponding to images with a plurality of different dose levels to obtain a second data set containing images with different dose levels; feeding the first data set and the second data set as the first input and the second input respectively into a first network and a second network sharing weights, and training by calculating the loss function to obtain a function representing the image imaging performance score.

[0006] A third aspect of the present disclosure proposes a device for evaluating the quality of an X-ray image, comprising: an acquisition module, the acquisition module being configured to acquire an image to be evaluated; a shooting condition score determination module, the shooting condition score determination module being configured to determine a shooting condition score of the image to be evaluated based on the image to be evaluated, the shooting condition score being used to evaluate at least one of the following: the position accuracy of the subject, and whether there are foreign objects in the shooting area; and an imaging performance score determination module, the imaging performance score determination module being configured to determine an imaging performance score of the image to be evaluated based on the image to be evaluated using a trained image imaging performance score model, the imaging performance score being used to evaluate the performance of images generated at different dose levels and noise levels.

[0007] In a fourth aspect, the present disclosure proposes a training device for an image imaging performance scoring model, comprising: a sample image acquisition module, the sample image acquisition module being configured to acquire at least one sample image; a sample image processing module, the sample image processing module being configured to perform the following processing on the at least one sample image: performing a first processing on the at least one sample image to generate a plurality of first sample images corresponding to images of a plurality of different noise levels, so as to obtain a first data set containing images of different noise levels, and performing a second processing on the at least one sample image to generate a plurality of second sample images corresponding to images of a plurality of different dose levels, so as to obtain a second data set containing images of different dose levels; a training module, the training module being configured to feed the first data set and the second data set as the first input and the second input respectively into a first network and a second network sharing weights, and to perform training by calculating a loss function to obtain a function characterizing the image imaging performance score.

[0008] The fifth aspect of the present disclosure proposes a medical imaging device, comprising: a memory, a processor, and a computer program stored on the memory, wherein the processor is configured to execute the computer program to implement the steps of the method for evaluating the quality of X-ray images according to the first aspect of the present disclosure or to implement the steps of the training method of the image imaging performance scoring model according to the second aspect of the present disclosure.

[0009] The sixth aspect of the present disclosure proposes a non-transitory computer-readable storage medium storing a computer program, wherein the computer program, when executed by a processor, implements the steps of the method for evaluating the quality of X-ray images according to the first aspect of the present disclosure or implements the steps of the training method of the image imaging performance scoring model according to the second aspect of the present disclosure.

[0010] The seventh aspect of the present disclosure proposes a computer program product, comprising a computer program, wherein, when the computer program is executed by a processor, implements the steps of the method for evaluating the quality of X-ray images according to the first aspect of the present disclosure or implements the steps of the training method of the image imaging performance scoring model according to the second aspect of the present disclosure.

[0011] Through the embodiments of the present disclosure, the efficiency and accuracy of X-ray image quality assessment can be improved. The cost of manually judging the quality of X-ray images can be effectively reduced, and the assessment of X-ray image quality can be made more objective and accurate. In addition, by adjusting the image shooting based on the obtained X-ray image quality assessment results, higher quality images can be obtained.

[0012] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings, so that those skilled in the art can better understand the above and other features and advantages of the present disclosure. In the accompanying drawings:

[0014] Figure 1 A schematic flow chart showing a method for evaluating X-ray image quality according to some embodiments of the present disclosure;

[0015] Figure 2 A schematic flow chart showing a method for evaluating X-ray image quality according to some other embodiments of the present disclosure;

[0016] Figure 3 Other embodiments of the present disclosure are shown Figure 1 or Figure 2 A schematic flow chart of a part of the process of the method for evaluating the quality of an X-ray image;

[0017] Figure 4 A schematic diagram showing the application of the method for evaluating the quality of X-ray images according to an embodiment of the present disclosure to detect foreign matter is shown;

[0018] Figure 5 Other embodiments of the present disclosure are shown Figure 1 or Figure 2 A schematic flow chart of a part of the process of the method for evaluating the quality of an X-ray image;

[0019] Figure 6(a)-6(d) A schematic diagram showing the application of the method for evaluating the quality of an X-ray image according to an embodiment of the present disclosure to perform position recognition of a photographed object;

[0020] Figure 7 A schematic flow chart of a training method 700 of an image imaging performance scoring model according to some embodiments of the present disclosure is shown;

[0021] Figure 8(a)-8(b) A schematic diagram showing image processing using the training method of the image imaging performance scoring model according to an embodiment of the present disclosure;

[0022] Fig. 9 A schematic flow chart showing a method for training an image imaging performance scoring model according to other embodiments of the present disclosure;

[0023] Fig.10 A structural block diagram of an apparatus for evaluating X-ray image quality according to an embodiment of the present disclosure is shown;

[0024] Fig.11 A structural block diagram of a training device for an image imaging performance scoring model according to an embodiment of the present disclosure is shown; and

[0025] Fig.12 An example configuration of a medical imaging device is shown that may be used to implement the methods described herein. DETAILED DESCRIPTION

[0026] In order to have a clearer understanding of the technical features, purposes and effects of the present disclosure, specific embodiments of the present disclosure are now described with reference to the accompanying drawings, in which the same reference numerals represent the same parts.

[0027] In this document, “exemplary” means “serving as an example, instance or illustration”, and any diagram or implementation described in this document as “exemplary” should not be interpreted as a more preferred or more advantageous technical solution.

[0028] In the present disclosure, unless otherwise specified, the use of the terms "first", "second", etc. to describe various elements is not intended to limit the positional relationship, timing relationship, or importance relationship of these elements, and such terms are only used to distinguish one element from another element. In some examples, the first element and the second element may refer to the same instance of the element, and in some cases, based on the description of the context, they may also refer to different instances.

[0029] The terms used in the description of various examples in this disclosure are only for the purpose of describing specific examples and are not intended to be limiting. Unless the context clearly indicates otherwise, if the number of elements is not specifically limited, the element can be one or more. In addition, the term "and / or" used in this disclosure covers any one of the listed items and all possible combinations.

[0030] Medical imaging research uses some kind of medium (such as X-rays, electromagnetic fields, ultrasound, etc.) to interact with the human body, and displays the structure and density of internal tissues and organs in the form of images, so that professionals can make judgments based on the information provided by the images and evaluate the health of the human body.

[0031] Taking X-ray imaging equipment as an example, X-ray imaging equipment includes an X-ray generator and an X-ray detector. The X-ray generator is responsible for generating X-rays, and the X-ray detector is responsible for receiving X-rays that pass through the human body. Since different structures of the human body have different densities and different degrees of blocking (attenuation), the intensity of rays passing through different structures is different. Depending on the amount of penetrating rays, different images of light and dark will be left on the X-ray film, and professionals can diagnose patients with bone injuries or other diseases based on the images.

[0032] As mentioned above, good medical imaging images, such as X-rays, need to fully show the differences in the internal structure of the body, and their image quality is directly related to the accuracy of medical diagnosis. The inventors of the present disclosure have noticed that there is currently a lack of standards in the industry for quality assessment of images obtained through radiographic imaging technology. In some scenarios, the number of patients who need to be re-photographed increases year by year due to quality problems with the captured images (such as incorrect placement of the part to be photographed and non-standard selection of X-ray doses, etc.). In other scenarios, although the hospital has stipulated corresponding reading standards, doctors will still evaluate the quality of medical images based on their own reading preferences. This will put hospitals or medical institutions under tremendous cost pressure, and the subjective judgment of image quality will also lead to reduced judgment efficiency and accuracy, and may also lead to unnecessary re-photographing of patients.

[0033] Based on this, the present disclosure provides a method for evaluating X-ray image quality, a training method for an image imaging performance scoring model, a training device for an image imaging performance scoring model, a device for evaluating X-ray image quality, a medical imaging device, a computer-readable storage medium, and a computer program product. Through the embodiments of the present disclosure, the quality of X-ray images can be automatically evaluated to improve the efficiency and accuracy of X-ray image quality evaluation. The cost of manually judging the quality of X-ray images can be effectively reduced, and the X-ray image quality evaluation can be more objective and accurate, avoiding unnecessary reshooting of patients. In addition, by adjusting the shooting of the image based on the obtained X-ray image quality evaluation results, a higher quality image can be obtained.

[0034] The embodiments of the present disclosure will be described in detail below with reference to the accompanying drawings.

[0035] Figure 1 A schematic flow chart of a method 100 for evaluating X-ray image quality according to an embodiment of the present disclosure is shown.

[0036] Step S110, obtaining an image to be evaluated;

[0037] Step S120: determining a shooting condition score of the image to be evaluated based on the image to be evaluated, wherein the shooting condition score is used to evaluate at least one of the following: position accuracy of the object to be evaluated, and whether there is a foreign object in the shooting area; and

[0038] Step S130: Based on the image to be evaluated, use the trained image imaging performance scoring model to determine the imaging performance score of the image to be evaluated, where the imaging performance score is used to evaluate the performance of images generated at different dose levels and noise levels.

[0039] It should be understood that steps S120 and S130 are not necessarily performed in sequence. Instead, the steps may be processed in reverse order or simultaneously.

[0040] In some embodiments, the quality of the image to be evaluated can be evaluated from two aspects: the shooting conditions of the image to be evaluated and the image imaging performance of the image to be evaluated. The scoring of the shooting conditions of the image to be evaluated may at least include scoring the accuracy of the position of the subject and scoring whether there are foreign objects (e.g., external objects such as rings) in the shooting area. The scoring of the image imaging performance uses a trained image imaging performance scoring model, so that the performance of images generated at different dose levels and noise levels can be objectively and accurately scored.

[0041] In some embodiments, the shooting condition score determined in step S120 and the imaging performance score determined in step S130 can be displayed to the user separately. In other embodiments, method 100 may further include: determining an overall image quality score of the image to be evaluated based on the determined shooting condition score of the image to be evaluated, the determined imaging performance score of the image to be evaluated, and their respective pre-assigned weights (to be described in detail below). Thus, the overall image quality score can be displayed to the user.

[0042] Figure 2 A schematic flow chart of a method 200 for evaluating X-ray image quality according to some other embodiments of the present disclosure is shown.

[0043] In some embodiments, reference Figure 2 , step S110, obtaining an image to be evaluated, comprising:

[0044] Step S211: obtaining an original image to be evaluated, and

[0045] Step S212: pre-process the original image.

[0046] A plurality of images may be captured by a radiographic imaging device (e.g., an X-ray imaging device), and the captured original images are unprocessed original images. The images captured by the radiographic imaging device may be preprocessed, for example, by an imaging system of the radiographic imaging device, such as performing noise reduction processing, cropping processing, etc. on the captured images, to obtain images to be evaluated.

[0047] In some embodiments, reference Figure 3 Step S120, determining the shooting condition score of the image to be evaluated, including:

[0048] Step S310, detecting suspected edges in the image to be evaluated;

[0049] Step S320, calculating the internal and external gradient differences of multiple points on the suspected edge;

[0050] Step S330, determining whether the suspected edge is a foreign body edge based on whether the internal and external gradient difference is greater than or equal to a preset gradient difference threshold, so as to identify whether there is a foreign body in the image to be evaluated; and

[0051] Step S340: Determine a shooting condition score of the image to be evaluated at least according to whether there is a foreign object in the image to be evaluated.

[0052] In step S340, the suspected edge in the image to be evaluated is detected, for example, the suspected edge in the image is extracted by edge detection methods such as the Canny operator. The suspected edge may be the edge of the object itself or the edge of a foreign object that is not the object. Figure 4 A schematic diagram of foreign body identification in a method for evaluating X-ray image quality according to an embodiment of the present disclosure is shown. Figure 4 The suspected edge in the image to be evaluated can be a pixel with a dramatic change in brightness, such as a point with a dramatic change in grayscale value. These pixels with dramatic changes in brightness can usually indicate discontinuity in depth, discontinuity in surface direction, and changes in material properties. If a pixel with a dramatic change in brightness is detected in the image to be evaluated, there may be foreign matter. For example, Figure 4 The portion encircled by the light-colored solid circle in the image to be evaluated has a white portion whose brightness is significantly higher than that of the hand bones adjacent to it in the image to be evaluated. This portion may be caused by the subject wearing a ring on his hand. The edge of the significantly brighter white portion is the edge of the foreign object that we want to identify.

[0053] Continue to refer Figure 3 as well as Figure 4After detecting that there is a suspected edge in the image to be evaluated, in order to determine whether the suspected edge is a foreign body edge or to identify a foreign body edge from multiple suspected edges, the step proceeds to S320, calculating the internal and external gradient differences of multiple points on the suspected edge. Figure 4 The boundary line 430 between the inner portion 410 and the outer portion 420, shown in the middle right portion, defines the foreign body edge. Figure 4 As shown in the left part of the figure, the outer part 420 can be, for example, a white part shown as brighter in the image to be evaluated, that is, a part of the ring worn by the subject in the image to be evaluated. The inner part 410 can be a hand bone part with darker brightness next to the outer part 420 in the image to be evaluated. Whether the suspected edge is a foreign body edge can be determined based on the following formula to identify whether there is a foreign body in the image to be evaluated:

[0054]

[0055] f(θ 1 ,θ 2 ,θ 3 …θ n =τ) (2)

[0056] In formula (1), G 外 Represents the external gradient of a point on the suspected edge, G 内 represents the internal gradient of a point on the suspected edge, T G represents the preset internal and external gradient difference threshold. 1 ,θ 2 ,θ 3 …θ n Represents a set of multiple points on the suspected edge, by determining each point θ in the suspected edge 1 ,θ 2 ,θ 3 …θ n The relationship between the inner and outer gradient difference τ and the preset inner and outer gradient difference threshold, formula (2) can determine whether the suspected edge is a foreign body edge. In step S330, when the absolute value of the inner and outer gradient difference of multiple points on the suspected edge determined in step S320 is greater than or equal to the preset gradient difference threshold, the suspected edge is determined to be a foreign body edge, and the image to be evaluated is identified as containing foreign matter; if it is less than the gradient difference threshold, it is determined that the suspected edge is not a foreign body edge, but a normal edge, and there is no foreign matter in the image to be evaluated.

[0057] In step S340, in determining the shooting condition score of the image to be evaluated, if there is no foreign matter in the image to be evaluated, the shooting condition score is not reduced; on the contrary, if the image to be evaluated is identified as containing foreign matter, the shooting condition score is appropriately reduced. For example, according to the proportion of the condition of whether there is a foreign matter in the shooting area in the shooting condition, the shooting condition score is subtracted by the proportion. Exemplarily, the score of whether there is a foreign matter in the shooting area is 50 points. If there is no foreign matter, the 50 points are retained, and if there is a foreign matter, the 50 points are subtracted.

[0058] In some embodiments, reference Figure 5 Step S120, determining the shooting condition score of the image to be evaluated, including:

[0059] Step S510, determining the position of the photographed object in the image to be evaluated; and

[0060] Step S520: Determine a shooting condition score of the image to be evaluated at least according to the positioning.

[0061] In some embodiments, reference Figure 5 Step S520: determining a shooting condition score of the image to be evaluated at least according to the positioning, including:

[0062] Step S521, determining the center of the X-ray field in the image to be evaluated as the first center;

[0063] Step S522, determining the center of the circumscribed figure of the photographed object in the image to be evaluated as the second center;

[0064] Step S523, calculating the distance between the first center and the second center; and

[0065] Step S540: Determine a shooting condition score of the image to be evaluated at least according to the degree to which the distance exceeds a preset distance threshold.

[0066] Figure 6(a)-6(d) FIG. 1 is a schematic diagram showing the method for evaluating the quality of an X-ray image according to an embodiment of the present disclosure for identifying the position of an object. Figure 5 As shown in FIG6(a), in step S510, the position of the photographed object in the image to be evaluated is determined. The image to be evaluated may be an image that has been preprocessed, for example, in step S212, or an original image that has not been preprocessed. Figure 5As shown in FIG6(b), in step S521, the center of the X-ray field in the image to be evaluated is determined as the first center. The X-ray field refers to the range of the X-ray beam irradiation. If the image to be evaluated is an unprocessed original image taken by a radiographic imaging device and the center of the X-ray field is aligned with the center of the X-ray detector, the center of the X-ray field in the image to be evaluated may be, for example, the center of the image to be evaluated. If the image to be evaluated is a preprocessed image, such as a cropped image, the position of the center of the X-ray field in the image to be evaluated needs to be determined accordingly if necessary.

[0067] Continue to refer Figure 5 as well as Figure 6(c)-6(d) , in step S522, the center of the circumscribed figure of the subject in the image to be evaluated is determined as the second center. The circumscribed figure refers to an image circumscribed to the subject in the image, such as a figure surrounded by a set of straight lines or curves, which can be used to locate the object in the image. The circumscribed figure is, for example, a circumscribed rectangle in FIG6(c). For the determination of the rectangular circumscribed figure, all the contours of the subject in the image can be first found, and then the highest and lowest points of the contour in the vertical direction, as well as the leftmost and rightmost points in the horizontal direction are calculated, and then the coordinates of the four vertices of the rectangular circumscribed figure are calculated based on these four points. After determining the circumscribed figure of the subject in the image to be evaluated, continue to refer to FIG6(d) and determine the center of the circumscribed figure as the second center.

[0068] Return to reference Figure 5 In step S523, the following formula (3) can be used to calculate the distance D between the first center and the second center, and the distance D exceeds the preset distance threshold T 距离 to determine the shooting condition score of the image to be evaluated,

[0069]

[0070] In formula (3), loc 1 is the first center position of the X-ray field center, loc 2 is a second center position representing the center of the circumscribed figure of the object in the image to be evaluated.

[0071] Step S524: at least according to the distance between the first center position and the second center position exceeding the preset distance threshold T 距离 The degree of the shooting condition score of the image to be evaluated is determined, wherein the distance exceeds the preset threshold T 距离 The degree of can represent the position accuracy of the object. If the distance D between the first center and the second center is less than the preset distance threshold T 距离The position of the object in the image to be evaluated is determined to be at the center of the X-ray field, that is, the position of the object is accurate, and the shooting condition score of the image to be evaluated can be determined to be a relatively high score. For example, if the position accuracy of the object accounts for 30 points in the shooting condition score, the 30 points can be obtained. If the distance D between the first center and the second center is greater than or equal to the preset distance threshold T 距离 , then the positioning of the subject in the image to be evaluated is determined to be not at the center position, that is, the positioning of the subject is inaccurate, and the degree to which the distance D beyond the first center and the second center exceeds the preset distance threshold can represent the degree of inaccuracy, and the more it exceeds, the more inaccurate it is. The shooting condition score of the image to be evaluated can be determined as a relatively low score. For example, the score can be reduced by a corresponding proportion according to the degree of exceeding the preset distance threshold. If the score of the subject's position accuracy in the shooting condition score is 50 points, it can be reduced by 10 points, 20 points, 30 points, etc. according to the degree of exceeding the preset distance threshold until the score about the position accuracy is reduced to 0 points.

[0072] Return to reference Figure 2 In some embodiments, step S120, based on the image to be evaluated, determines a shooting condition score of the image to be evaluated, wherein the shooting condition score is used to evaluate at least one of the following: the position accuracy of the object being photographed, and whether there is a foreign object in the shooting area, including:

[0073] Step S221, determining a foreign body detection score indicating whether there is a foreign body in the image to be evaluated;

[0074] Step S222, determining a position score representing the location of the object in the image to be evaluated; and

[0075] Step S223: Determine the shooting condition score of the image to be evaluated based on the foreign object detection score, the position score, and their respective pre-assigned weights.

[0076] In steps S221 to S223, a foreign object detection score and a position score are assigned according to whether a foreign object exists and the position accuracy of the object, and then the shooting condition score of the image to be evaluated is obtained according to the weights of the foreign object detection score and the position score. That is, after each shooting condition is quantified and scored, the overall shooting condition score is obtained.

[0077] After the foreign object detection score and the position score are determined, in step S223 , the shooting condition score of the image to be evaluated is determined based on the foreign object detection score, the position score and their respective pre-assigned weights.

[0078] In some embodiments, the foreign body detection score and the location score may be pre-assigned corresponding weights, for example, 50% respectively, without any limitation herein, and other weight assignments shall also fall within the protection scope of the present invention. The shooting condition score of the image to be evaluated may be calculated based on the following formula:

[0079] M c =w a ×a+(1-w a )×b (4)

[0080] Among them, M c represents the shooting condition score of the image to be evaluated, a and b represent the foreign body detection score and the position score respectively, and w a and (1-w a ) represent the weights assigned to the foreign body detection score and the location score, respectively.

[0081] In some embodiments, for example, when the foreign object detection score and the position score are respectively assigned a weight of 50%, when a foreign object is detected in the image to be evaluated and the foreign object detection score is 0, and the subject is determined to be at the center of the image to be evaluated and the position score is 100, the shooting condition score of the image to be evaluated can be determined to be 50 points out of a full score of 100 points.

[0082] In other embodiments, different from steps S221 to S223, instead of assigning a score to each shooting condition separately, the shooting condition score is obtained according to each shooting condition and the preset correspondence between each shooting condition and the overall shooting condition score. For example, when a shooting condition is not met, the shooting condition score is reduced accordingly, such as reducing the shooting condition score from 100 points by 50 points, or 10 points, 20 points, 30 points, etc.

[0083] Return to reference Figure 2 In step S130, based on the image to be evaluated, the imaging performance score of the image to be evaluated is determined using the trained image imaging performance scoring model, and the imaging performance score is used to evaluate the performance of images generated at different dose levels and noise levels.

[0084] Compared with the shooting condition scoring, it is difficult to accurately judge the scores of the imaging performance of different images. In some embodiments, in order to obtain the imaging performance score of the image to be evaluated, a training method 700 of an image imaging performance scoring model is proposed. Figure 7 A schematic flow chart of a training method 700 for an image imaging performance scoring model according to some embodiments of the present disclosure is shown, the training method comprising:

[0085] Step S710, obtaining at least one sample image;

[0086] Step S720: for the at least one sample image, perform the following steps:

[0087] Step S721: performing a first process on the at least one sample image to generate a plurality of first sample images corresponding to the images with different noise levels, so as to obtain a first data set including images with different noise levels; and

[0088] Step S722, performing a second process on the at least one sample image to generate a plurality of second sample images corresponding to the images at a plurality of different dose levels, so as to obtain a second data set including images at different dose levels;

[0089] Step S730: Feed the first data set and the second data set as the first input and the second input respectively into the first network and the second network sharing the weights, and train by calculating the loss function to obtain a function representing the image imaging performance score. Specifically, the first network and the second network can be twin neural networks.

[0090] As mentioned above, it is difficult to judge the scores of different images for their imaging performance. However, for a group of images, such as a pair of images, which are processed from the same sample image, it is easier to judge which image has better imaging performance and can be assigned a higher score. Therefore, a training method for the above-mentioned image imaging performance scoring model is proposed, and the trained model can be used to obtain an imaging performance score.

[0091] Figure 8(a)-8(b) A schematic diagram showing image processing using a training method for an image imaging performance scoring model according to an embodiment of the present disclosure.

[0092] In some embodiments, in step S710, at least one sample image is obtained. The sample image may be an unprocessed original image taken by a radiographic imaging device, or may be an image pre-processed by an imaging system of the radiographic imaging device, such as an image subjected to noise reduction processing, cropping processing, etc. For example, the left image in FIG8(a) and the left image in FIG8(b) are the sample images, which are original images taken by a radiographic imaging device or are images pre-processed from the original images.

[0093] In steps S721 and S722, the first processing and the second processing are respectively performed on the at least one sample image. The first processing and the second processing can be various distortion operations, such as Gaussian blur processing, Gaussian noise processing, etc.

[0094] In step S721, the at least one sample image is first processed to generate a plurality of first sample images corresponding to images with different noise levels, so as to obtain a first data set containing images with different noise levels. In some embodiments, the sample image (i.e., the image on the left) in FIG. 8(a) is first processed to obtain a first sample image (i.e., the image on the right), and the first processing may be Gaussian blur processing or Gaussian noise processing. After the at least one sample image is first processed, a plurality of first sample images corresponding to different noise levels (i.e., different blurriness) may be obtained, such as the image on the right in FIG. 8(a), which has a higher noise level (e.g., higher blurriness) than the image on the left in FIG. 8(a). In some embodiments, a plurality of first sample images corresponding to different noise levels may be used as the first data set.

[0095] In step S722, the at least one sample image is subjected to a second processing to generate a plurality of second sample images corresponding to the images at a plurality of different dose levels, so as to obtain a second data set containing images at different dose levels. In some embodiments, the sample image (i.e., the image on the left) in FIG8(b) is subjected to a second processing to obtain a second sample image (i.e., the image on the right), and the second processing may be a Gaussian blur processing, a Gaussian noise processing, etc. After the at least one sample image is subjected to the second processing, a plurality of second sample images corresponding to different dose levels (i.e., different exposure degrees) may be obtained, such as the image on the right in FIG8(b), which has a lower dose level (e.g., a lower exposure degree) than the image on the left in FIG8(b). In some embodiments, a plurality of second sample images corresponding to different dose levels may be used as a second data set.

[0096] By performing the first processing and the second processing on the original image taken by the radiation imaging device or the image after pre-processing the original image, and constructing the first data set and the second data set based on the multiple first and second sample images obtained thereby, on the one hand, it is possible to avoid shooting the subject multiple times with different doses, and on the other hand, for a group of images, especially paired images, it is easier to judge which image has better quality, for example, an undistorted image is better than a distorted image, and a severely distorted image is better than a less distorted image.

[0097] In other embodiments, the first data set and the second data set may also respectively include a plurality of images captured by the radiation imaging device at different noise levels and different dose levels.

[0098] In some embodiments, for example, a first process and / or a second process may be performed on a sample image. Take the first process of an image as an example: an image with a certain noise level after the first process (e.g., Gaussian blur process, Gaussian noise process) is a first sample image. A plurality of first sample images with different noise levels constitute a first data set. When making an image quality judgment about the noise level of the image, a group of images, especially paired images, which are processed from the same sample image can be used. Since there is only a difference in noise level, the judgment can be made more accurate. Referring to the example in FIG8(a), the left and right images with different noise levels in FIG8(a) can be used as a pair of images to determine the image quality. As shown in FIG8(a), the image on the right has a higher noise level than the image on the left. Therefore, the image on the left has significantly better quality than the image on the right.

[0099] It should be understood that although a pair of images, i.e., two images, is used as an example in Figure 8 of this article, the present disclosure is not limited to this; using another number of groups (e.g., three or more) of images to judge the image imaging performance score should also be included in the scope of protection of the present disclosure.

[0100] Continue to refer Figure 7 In step S730, the first data set and the second data set are fed as the first input and the second input to the first network and the second network sharing the weights, respectively, and the training is performed by calculating the loss function to obtain a function representing the image imaging performance score. The first network and the second network can be the same network, for example, a twin neural network.

[0101] Fig. 9 A schematic flow chart of a training method for an image imaging performance scoring model according to other embodiments of the present disclosure is shown. Figure 7 and Fig. 9 In some embodiments, the first network 911 and the second network 912 are two networks that share weights, for example, they can be two completely identical networks. The first data set 901 and the second data set 902 are fed as the first input and the second input to the first network 911 and the second network 922, respectively, wherein the first network 901 and the second network 902 respectively map the input to a new space to form a representation of the input in the new space, for example, the first network and the second network can respectively convert the first input and the second input into vectors, and in the new space, the similarity can be determined by judging the distance between different inputs. In some embodiments, by calculating the loss function Loss indicated by box 930, the distance between two inputs with high similarity can be made as small as possible, while the distance between two inputs with low similarity can be made as large as possible. Through this training, a scoring model F(x) for evaluating image imaging performance indicated by box 940 can be obtained.

[0102] In some embodiments, in step S720, the following processing is further performed for the at least one sample image:

[0103] Comparing imaging performances of a group of first sample images among the plurality of first sample images and assigning corresponding labels thereto; and

[0104] The imaging performances of a group of second sample images among the plurality of second sample images are compared and corresponding labels are assigned thereto.

[0105] Continue to combine Fig. 9 In some embodiments, the imaging performance of a group of first sample images among the multiple first sample images can be compared and corresponding labels can be assigned to them. The group of first sample images can be, for example, a pair of first sample images with different noise levels obtained by the same sample image after the first processing. Specifically, for example, for the first pair of first sample images corresponding to different noise levels, they can be a sample image A with a lower noise level (for example, the left side of Figure 8 (a)) and a corresponding sample image B with a higher noise level (for example, the right side of Figure 8 (a)). When evaluating the pair of first sample images, the sample image A with a lower noise level can be assigned a label of 1 to represent its lower blurriness, and the sample image B with a higher noise level can be assigned a label of 0 to represent its higher blurriness.

[0106] Similarly, in some embodiments, the imaging performance of a group of second sample images among the plurality of second sample images may be compared and a corresponding label may be assigned to them. The group of second sample images may, for example, be a pair of second sample images having different noise levels obtained by the second processing of the same sample image. Specifically, for example, for a first pair of second sample images corresponding to different dose levels, they may be a sample image C corresponding to a higher dose level (e.g., the left side of FIG. 8( b)) and a corresponding sample image D corresponding to a lower dose level (e.g., the right side of FIG. 8( b)). When evaluating the pair of second sample images, the sample image C having a higher dose level may be assigned a label 1 to characterize its overexposure, while the sample image D having a lower dose level may be assigned a label 0 to characterize its underexposure.

[0107] The first sample image pair and the second sample image pair obtained above and their respective assigned labels are sent to the first network and the second network, and after training, an image imaging performance scoring model is obtained.

[0108] Return to reference Figure 1 and Figure 2In step S130, based on the image to be evaluated, the imaging performance score of the image to be evaluated is determined using the trained image imaging performance scoring model, and the imaging performance score is used to evaluate the performance of images generated at different dose levels and noise levels.

[0109] In some embodiments, the imaging performance score of the image to be evaluated can be obtained by inputting the image to be evaluated into the trained image imaging performance scoring model, wherein the imaging performance score can characterize the performance of the image generated at different dose levels and noise levels.

[0110] Continue to refer Figure 2 According to some embodiments of the present disclosure, the method 200 for evaluating the quality of an X-ray image further includes:

[0111] Step S240: Determine the overall image quality score of the image to be evaluated based on the determined shooting condition score of the image to be evaluated, the determined imaging performance score of the image to be evaluated, and their respective pre-assigned weights.

[0112] In some embodiments, the shooting condition score of the image to be evaluated and the imaging performance score of the image to be evaluated may be pre-assigned corresponding weights, for example, 50% respectively, without any limitation here, and other weight assignments shall also fall within the protection scope of the present invention. The shooting condition score of the image to be evaluated may be calculated based on the following formula:

[0113] M f =w c ×c+(1-w c )×d (4)

[0114] Among them, M f represents the overall image quality score of the image to be evaluated, c and d represent the shooting condition score of the image to be evaluated and the imaging performance score of the image to be evaluated, respectively, and w c and (1-w c ) respectively represent the weights assigned to the shooting condition score of the image to be evaluated and the imaging performance score of the image to be evaluated.

[0115] In some embodiments, for example, when the shooting condition score of the image to be evaluated and the imaging performance score of the image to be evaluated are respectively assigned a weight of 50%, when the shooting condition score of the image to be evaluated is 70 points (out of 100) and the imaging performance score of the image to be evaluated is 90 points (out of 100), the overall image quality score of the image to be evaluated can be determined as 80 points out of the full 100 points.

[0116] In some embodiments, the overall image quality score can be displayed to the user through a display interface. The display interface can display the image to be evaluated, the overall image quality score of the image to be evaluated, and the corresponding processing suggestions, which can be "foreign matter detected in the exposure area", "dose meets requirements", etc. Based on this, professionals can adjust the image shooting more specifically, such as reminding patients to remove foreign matter from their hands before the next shooting.

[0117] According to another aspect of the present disclosure, an apparatus for evaluating X-ray image quality is provided. Fig.10 FIG. 1 shows a structural block diagram of an apparatus 1000 for evaluating X-ray image quality according to an embodiment of the present disclosure. Fig.11 As shown, the apparatus 1000 for evaluating the quality of an X-ray image comprises:

[0118] An acquisition module 1010, wherein the acquisition module is configured to acquire an image to be evaluated;

[0119] A shooting condition score determination module 1020, the shooting condition score determination module is configured to determine a shooting condition score of the image to be evaluated based on the image to be evaluated, the shooting condition score being used to evaluate at least one of the following: position accuracy of the subject, and whether there is a foreign object in the shooting area; and

[0120] An imaging performance score determination module 1030 is configured to determine an imaging performance score of the image to be evaluated based on the image to be evaluated using a trained image imaging performance score model, wherein the imaging performance score is used to evaluate the performance of images generated at different dose levels and noise levels.

[0121] It should be understood that Fig.10 The various modules of the device 1000 shown in FIG. Figure 1 The method described in 100, reference Figure 2 Therefore, the operations, features and advantages described above for methods 100 and 200 are also applicable to the apparatus 1000 and the units included therein. For the sake of brevity, some operations, features and advantages are not described in detail here.

[0122] According to another aspect of the present disclosure, a training device for an image imaging performance scoring model is also provided. Fig.11 FIG. 1 shows a structural block diagram of a training device 1100 for an image imaging performance scoring model according to an embodiment of the present disclosure. Fig.11 As shown, the training device 1100 of the image imaging performance scoring model includes:

[0123] A sample image acquisition module 1110, wherein the sample image acquisition module is configured to acquire at least one sample image;

[0124] The sample image processing module 1120 is configured to perform the following processing on the at least one sample image:

[0125] performing a first process on the at least one sample image to generate a plurality of first sample images corresponding to images with different noise levels, so as to obtain a first data set containing images with different noise levels; and

[0126] Performing a second process on the at least one sample image to generate a plurality of second sample images corresponding to the images at a plurality of different dose levels, so as to obtain a second data set including images at different dose levels;

[0127] A training module 1130 is configured to feed the first data set and the second data set as the first input and the second input respectively into the first network and the second network sharing weights, and perform training by calculating the loss function to obtain a function characterizing the image imaging performance score.

[0128] It should be understood that Fig.11 The various modules of the apparatus 1100 shown in FIG. 1 can be used in conjunction with the reference Figure 7 Therefore, the operations, features and advantages described above for method 700 are also applicable to device 1100 and the units included therein. For the sake of brevity, some operations, features and advantages are not described in detail here.

[0129] According to another aspect of the present disclosure, a medical imaging device is provided, comprising: a memory, a processor, and a computer program stored on the memory, wherein the processor is configured to execute the computer program to implement the steps of the above-mentioned method for evaluating the quality of X-ray images or to implement the steps of the above-mentioned method for training the image imaging performance scoring model.

[0130] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing a computer program is provided, wherein the computer program, when executed by a processor, implements the steps of the above-mentioned method for evaluating X-ray image quality or implements the steps of the above-mentioned method for training the image imaging performance scoring model.

[0131] According to another aspect of the present disclosure, a computer program product is provided, comprising a computer program, wherein the computer program, when executed by a processor, implements the steps of the above-mentioned method for evaluating X-ray image quality or the steps of the above-mentioned method for training the image imaging performance scoring model.

[0132] In the following, combined Fig.12 Illustrative examples of such a computer device, non-transitory computer-readable storage medium, and computer program product are described.

[0133] Fig.12 An example configuration of a medical imaging device 1200 that can be used to implement the methods described herein is shown. The above-mentioned apparatus for evaluating X-ray image quality can also be fully or at least partially implemented by the medical imaging device 1200 or similar devices or apparatuses.

[0134] The medical imaging device 1200 can be a variety of different types of devices, such as a server of a service provider, a device associated with a client (e.g., a client device), a system on a chip, and / or any other suitable computer device or computing system. Examples of the medical imaging device 1200 include, but are not limited to, a desktop computer, a server computer, a laptop or netbook computer, a mobile device (e.g., a tablet computer, a cellular or other wireless phone (e.g., a smart phone), a notepad computer, a mobile station), and the like.

[0135] The medical imaging device 1200 may include at least one processor 1202, memory 1204, communication interface(s) 1206, a display device 1208, other input / output (I / O) devices 1210, and one or more mass storage devices 1212, which can communicate with each other, such as via a system bus 1214 or other appropriate connection.

[0136] The processor 1202 may be a single processing unit or multiple processing units, all of which may include a single or multiple computing units or multiple cores. The processor 1202 may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuits, and / or any device that manipulates signals based on operating instructions. Among other capabilities, the processor 1202 may be configured to obtain and execute computer-readable instructions stored in the memory 1204, mass storage device 1212, or other computer-readable media, such as program codes for operating devices 1216, program codes for application programs 1218, program codes for other programs 1220, and the like.

[0137] The memory 1204 and the mass storage device 1212 are examples of computer-readable storage media for storing instructions that are executed by the processor 1202 to implement the various functions described above. For example, the memory 1204 may generally include both volatile memory and non-volatile memory (e.g., RAM, ROM, etc.). In addition, the mass storage device 1212 may generally include a hard drive, a solid-state drive, a removable medium, including external and removable drives, a memory card, a flash memory, a floppy disk, an optical disk (e.g., a CD, a DVD), a storage array, a network attached storage, a storage area network, etc. The memory 1204 and the mass storage device 1212 may all be collectively referred to herein as memory or computer-readable storage media, and may be a non-transitory medium capable of storing computer-readable, processor-executable program instructions as computer program code, which may be executed by the processor 1202 as a specific machine configured to implement the operations and functions described in the examples herein.

[0138] A plurality of program modules may be stored on the mass storage device 1212. These programs include an operating system 1216, one or more application programs 1218, other programs 1220, and program data 1222, and they may be loaded into the memory 1204 for execution. Examples of such application programs or program modules may include, for example, computer program logic (e.g., computer program code or instructions) for implementing the following components / functions: the apparatus 1000 (including the acquisition module 1010, the shooting condition score determination module 1020, and the imaging performance score determination module 1030), the apparatus 1100 (including the sample image acquisition module 1110, the sample image processing module 1120, and the training module 1130), the methods 100, 200, 700 (including any suitable steps of the methods 100, 200, 700), and / or other embodiments described herein.

[0139] Although in Fig.12 1204 of the computer device 1200, but modules 1216, 1218, 1220, and 1222, or portions thereof, may be implemented using any form of computer-readable media accessible by the computer device 1200. As used herein, "computer-readable media" includes at least two types of computer-readable media, namely, computer storage media and communication media.

[0140] Computer storage media includes volatile and nonvolatile, removable and non-removable media implemented by any method or technology for storage of information, such as computer readable instructions, data structures, program modules or other data. Computer storage media includes but is not limited to RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disk (DVD), or other optical storage devices, magnetic cassettes, magnetic tapes, magnetic disk storage devices or other magnetic storage devices, or any other non-transmission media that can be used to store information for access by a computer device.

[0141] In contrast, communication media may embody computer readable instructions, data structures, program modules or other data in a modulated data signal such as a carrier wave or other transport mechanism. Computer storage media as defined herein does not include communication media.

[0142] The medical imaging device 1200 may also include one or more communication interfaces 1206 for exchanging data with other devices, such as through a network, direct connection, etc., as previously discussed. Such communication interfaces may be one or more of: any type of network interface (e.g., a network interface card (NIC)), a wired or wireless (such as an IEEE 802.11 wireless LAN (WLAN)) wireless interface, a Worldwide Interoperability for Microwave Access (Wi-MAX) interface, an Ethernet interface, a Universal Serial Bus (USB) interface, a cellular network interface, a Bluetooth™ interface, a Near Field Communication (NFC) interface, etc. The communication interface 1206 may facilitate communication within a variety of network and protocol types, including wired networks (e.g., LAN, cable, etc.) and wireless networks (e.g., WLAN, cellular, satellite, etc.), the Internet, etc. The communication interface 1206 may also provide communication with an external storage device (not shown) such as a storage array, a network attached storage, a storage area network, etc.

[0143] In some examples, a display device 1208 such as a monitor may be included for displaying information and images to the user, such as displaying an X-ray image quality assessment score and image processing suggestions. Other I / O devices 1210 may be devices that receive various inputs from the user and provide various outputs to the user, and may include a touch input device, a gesture input device, a camera, a keyboard, a remote control, a mouse, a printer, an audio input / output device, and the like.

[0144] Although the embodiments or examples of the present disclosure have been described with reference to the accompanying drawings, it should be understood that the above-mentioned methods, systems and devices are merely exemplary embodiments or examples, and the scope of the present invention is not limited by these embodiments or examples, but only by the claims after authorization and their equivalent scope. Various elements in the embodiments or examples can be omitted or replaced by their equivalent elements. In addition, each step can be performed in an order different from that described in the present disclosure. Further, the various elements in the embodiments or examples can be combined in various ways. It is important that with the evolution of technology, many elements described herein can be replaced by equivalent elements that appear after the present disclosure.

Claims

1. A method for evaluating the quality of an X-ray image, comprising: Obtaining an image to be evaluated; Based on the image to be evaluated, determining a shooting condition score of the image to be evaluated, the shooting condition score being used to evaluate at least one of the following: position accuracy of the object being photographed, and whether there is a foreign object in the shooting area; and Based on the image to be evaluated, an imaging performance score of the image to be evaluated is determined using a trained image imaging performance scoring model, wherein the imaging performance score is used to evaluate the performance of images generated at different dose levels and noise levels.

2. The method according to claim 1, wherein: Determining the shooting condition score of the image to be evaluated includes: Detecting suspected edges in the image to be evaluated; Calculate the internal and external gradient differences of multiple points on the suspected edge; Determining whether the suspected edge is a foreign object edge based on whether the internal and external gradient difference is greater than or equal to a preset gradient difference threshold, so as to identify whether there is a foreign object in the image to be evaluated; and The shooting condition score of the image to be evaluated is determined at least according to whether there is a foreign object in the image to be evaluated.

3. The method according to claim 1 or 2, wherein: The step of determining the shooting condition score of the image to be evaluated further includes: Determining the location of the object in the image to be evaluated; and A shooting condition score of the image to be evaluated is determined at least according to the positioning.

4. The method according to claim 3, wherein: Determining a shooting condition score of the image to be evaluated at least according to the positioning, including: Determining a center of the X-ray field in the image to be evaluated as a first center; Determine the center of the circumscribed figure of the photographed object in the image to be evaluated as a second center; calculating the distance between the first center and the second center; and The shooting condition score of the image to be evaluated is determined at least according to the extent to which the distance exceeds a preset distance threshold.

5. The method according to claim 1 or 2, wherein: The shooting condition score of the image to be evaluated includes a foreign object detection score indicating whether there is a foreign object in the image to be evaluated and a position score indicating the location of the subject in the image to be evaluated. The determining of the shooting condition score of the image to be evaluated includes: determining the shooting condition score of the image to be evaluated based on the foreign object detection score, the position score, and their respective pre-assigned weights.

6. The method according to claim 1 or 2, wherein: The method further comprises: Based on the determined shooting condition score of the image to be evaluated, the determined imaging performance score of the image to be evaluated, and their respective pre-assigned weights, an overall image quality score of the image to be evaluated is determined.

7. A method for training an image imaging performance scoring model, the image imaging performance scoring model being used in the method according to any one of claims 1 to 6, the training method comprising: acquiring at least one sample image; For the at least one sample image, the following processing is performed: performing a first process on the at least one sample image to generate a plurality of first sample images corresponding to images with different noise levels, so as to obtain a first data set containing images with different noise levels; and Performing a second process on the at least one sample image to generate a plurality of second sample images corresponding to the images at a plurality of different dose levels, so as to obtain a second data set including images at different dose levels; The first data set and the second data set are fed as the first input and the second input respectively into the first network and the second network which share weights, and the training is performed by calculating the loss function to obtain the function which characterizes the image imaging performance score.

8. The training method according to claim 7, wherein: The first processing or the second processing includes at least one of the following: Gaussian blur processing, Gaussian noise processing.

9. The training method according to claim 7, wherein: For the at least one sample image, the following processing is further performed: Comparing imaging performances of a group of first sample images among the plurality of first sample images and assigning corresponding labels thereto; and The imaging performances of a group of second sample images among the plurality of second sample images are compared and corresponding labels are assigned thereto.

10. A device for evaluating the quality of an X-ray image, comprising: An acquisition module, wherein the acquisition module is configured to acquire an image to be evaluated; a shooting condition score determination module, the shooting condition score determination module being configured to determine a shooting condition score of the image to be evaluated based on the image to be evaluated, the shooting condition score being used to evaluate at least one of the following: position accuracy of the object being photographed, and whether there is a foreign object in the shooting area; and An imaging performance score determination module is configured to determine an imaging performance score of the image to be evaluated based on the image to be evaluated using a trained image imaging performance score model, wherein the imaging performance score is used to evaluate the performance of images generated at different dose levels and noise levels.

11. A training device for an image imaging performance scoring model, comprising: A sample image acquisition module, wherein the sample image acquisition module is configured to acquire at least one sample image; A sample image processing module, wherein the sample image processing module is configured to perform the following processing on the at least one sample image: performing a first process on the at least one sample image to generate a plurality of first sample images corresponding to images with different noise levels, so as to obtain a first data set containing images with different noise levels; and Performing a second process on the at least one sample image to generate a plurality of second sample images corresponding to the images at a plurality of different dose levels, so as to obtain a second data set including images at different dose levels; A training module is configured to feed the first data set and the second data set as the first input and the second input respectively into the first network and the second network sharing weights, and to perform training by calculating the loss function to obtain a function characterizing the image imaging performance score.

12. A medical imaging device comprising: a memory, a processor and a computer program stored on the memory, The processor is configured to execute the computer program to implement the steps of the method according to any one of claims 1 to 6 or to implement the steps of the training method according to claim 7 or 8.

13. A non-transitory computer-readable storage medium storing a computer program, wherein: When the computer program is executed by a processor, the computer program implements the method according to any one of claims 1 to 6 or implements the steps of the training method according to claim 7 or 8.

14. A computer program product comprising a computer program, wherein: When the computer program is executed by a processor, the computer program implements the method according to any one of claims 1 to 6 or implements the steps of the training method according to claim 7 or 8.

Images