Method, device and equipment for quality assessment of spine image and storage medium

Abstract

The application provides a quality evaluation method, device and equipment of a spine image and a storage medium. The method comprises the following steps: acquiring a spine image to be evaluated; determining the position coordinates of key points of the spine in the spine image; and determining the shooting posture of the spine image according to the position coordinates of the key points of the spine, so as to evaluate the shooting quality of the spine image. The embodiment of the application can realize automatic evaluation of whether the shooting quality of the spine image meets the standard, correct the spine image that does not meet the standard in time, enhance the usability of the spine image for the diagnosis of the health of the spine of a patient, and improve the diagnosis accuracy of the health problem of the spine of the patient.

Description

Technical Field

[0001] This application relates to the field of image processing technology, specifically to a method, apparatus, device, and storage medium for quality assessment of spinal images. Background Technology

[0002] With increasing pressure in modern life, more and more people are experiencing various spinal health problems, such as scoliosis and spinal skeletal dysplasia. Therefore, to better assist doctors in diagnosing and treating patients' spinal health, digital radiography (DR) technology is usually used to capture images of the patient's spine, thereby analyzing whether the patient has problems such as scoliosis or skeletal dysplasia.

[0003] However, improper DR (Digital Radiography) procedures can lead to problems such as foreign bodies appearing in spinal images or incomplete imaging of critical areas, thus failing to provide accurate structural support for clinicians' diagnoses. Therefore, there is an urgent need to design a method that can accurately assess the standardity and usability of spinal images to assist clinicians in making accurate diagnoses of spinal health issues. Summary of the Invention

[0004] This application provides a method, apparatus, device, and storage medium for quality assessment of spinal images, enabling automated assessment of whether the imaging quality of spinal images meets standards, enhancing the usability of spinal images for diagnosing patients' spinal health, and improving the accuracy of diagnosing patients' spinal health problems.

[0005] In a first aspect, embodiments of this application provide a method for quality assessment of spinal images, the method comprising:

[0006] Obtain images of the spine to be evaluated;

[0007] Determine the position coordinates of key points in the spine in the spinal image;

[0008] Based on the location coordinates of the key points of the spine, the shooting position of the spinal image is determined in order to evaluate the shooting quality of the spinal image.

[0009] Secondly, embodiments of this application provide a spinal imaging quality assessment device, the device comprising:

[0010] The image acquisition module is used to acquire images of the spine to be evaluated.

[0011] The key point determination module is used to determine the position coordinates of key points of the spine in the spinal image;

[0012] The quality assessment module is used to determine the shooting position of the spinal image based on the position coordinates of the key points of the spine, so as to assess the shooting quality of the spinal image.

[0013] Thirdly, embodiments of this application provide an electronic device, which includes:

[0014] A processor and a memory, the memory being used to store a computer program, and the processor being used to call and run the computer program stored in the memory to perform the spinal image quality assessment method provided in the first aspect of this application.

[0015] Fourthly, embodiments of this application provide a computer-readable storage medium for storing a computer program that causes a computer to perform a spinal image quality assessment method as provided in the first aspect of this application.

[0016] Fifthly, embodiments of this application provide a computer program product, including a computer program / instructions, characterized in that, when the computer program / instructions are executed by a processor, they implement the spinal image quality assessment method provided in the first aspect of this application.

[0017] This application provides a method, apparatus, device, and storage medium for quality assessment of spinal images. After acquiring the spinal image to be assessed, the location coordinates of each key point in the spinal spine are first determined. Then, based on the location coordinates of each key point, the imaging position of the spinal image is determined to assess whether the imaging quality of the spinal image meets the standards. This achieves automated assessment of the quality of spinal image imaging, allowing for timely correction of substandard spinal images, enhancing the usability of spinal images for diagnosing patients' spinal health, and improving the accuracy of diagnosing patients' spinal health problems. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating a method for quality assessment of spinal images, as shown in an embodiment of this application.

[0020] Figure 2 This is a schematic diagram of key points of the spine in a spinal image shown in an embodiment of this application.

[0021] Figure 3A flowchart illustrating another method for quality assessment of spinal images as shown in an embodiment of this application;

[0022] Figure 4 This is a schematic diagram illustrating the extraction process of key points of the spine in an embodiment of this application;

[0023] Figure 5 This is a schematic block diagram illustrating a quality assessment device for spinal images according to an embodiment of this application;

[0024] Figure 6 This is a schematic block diagram of an electronic device shown in an embodiment of this application. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or server that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0027] This application applies to various spinal images obtained by using DR equipment to perform X-ray imaging on patients in the medical field, and the spinal images can be DR full spine images.

[0028] Considering that improper operation by staff may result in foreign objects or incomplete imaging of critical areas in spinal images, failing to provide accurate spinal structural support for clinicians' diagnoses, this application provides a scheme to accurately assess the quality of spinal image capture. This scheme determines the imaging position of the spinal image based on the coordinates of key spinal points, thereby assessing the image quality and automating the evaluation of its adequacy. This allows for timely correction of substandard spinal images, enhancing the usability of spinal imaging for diagnosing patients' spinal health and improving the accuracy and versatility of spinal image assessment.

[0029] Figure 1 This is a flowchart illustrating a method for quality assessment of spinal images, as shown in an embodiment of this application. (Refer to...) Figure 1 The method may specifically include the following steps:

[0030] S110, acquire images of the spine to be evaluated.

[0031] When diagnosing whether a patient has spinal health problems, clinicians usually need the patient to have a spinal image taken in advance so that the clinicians can use the image to help analyze the patient's spinal health issues.

[0032] Furthermore, in order to facilitate clinicians' rapid and accurate diagnosis of spinal health problems using spinal imaging, spinal imaging is usually required to meet certain standards. For example, the spinal region in the spinal imaging should be completely captured and presented in the center from the front, so that clinicians can clearly observe the patient's spinal condition from all angles and thus accurately diagnose the patient's spinal health problems.

[0033] Therefore, in order to ensure that the spinal images of each patient meet the standards, this application can assess whether the spinal images meet the standards after each patient has taken spinal images with the DR equipment and before leaving the DR equipment. This allows for timely correction of any substandard spinal images before the patient leaves the DR equipment, avoiding the waste of diagnostic time caused by discovering that the spinal images do not meet the standards only when the patient arrives at the clinic and having to return to the DR equipment to retake the spinal images.

[0034] This application uses a DR device to capture spinal images of each patient, and then uses each captured spinal image as the spinal image to be evaluated in this application, so that the quality of each spinal image can be accurately evaluated in a timely manner at the DR device, and the substandard spinal images can be corrected in a timely manner.

[0035] S120, determine the location coordinates of key points in the spine in the spinal image.

[0036] To accurately determine whether spinal imaging meets standards, key spinal points are typically defined within the images. These key points, when combined, completely surround the entire spinal region, allowing for a comprehensive assessment of its morphology. In a standard spinal imaging profile, these key points will be in relatively standard positions. Therefore, this application analyzes whether spinal imaging meets standards by comparing the differences between the positions of the key spinal points in the images and their standard positions.

[0037] Among them, such as Figure 2 As shown, the key points of the spine in this application may include, but are not limited to, the key points represented by the nose, left ear, right ear, left acromioclavicular edge, right acromioclavicular edge, left iliac wing, right iliac wing, left femoral tuberosity, right femoral tuberosity, left ischial tuberosity, and right ischial tuberosity. These 11 key points of the spine can almost completely surround the entire spine, thus representing the complete shape of the spine.

[0038] In this application, for each acquired spinal image, feature analysis is first performed on the spinal image to determine each spinal key point within the spinal image, thereby obtaining the position coordinates of each spinal key point within the spinal image.

[0039] S130 determines the shooting position for spinal images based on the location coordinates of key points on the spine, in order to assess the quality of spinal image capture.

[0040] After determining the coordinates of each key point in the spinal image, the relative positional relationships between these key points can be analyzed. Then, based on these relative positions and the correlations between spinal structures, the spinal morphology within the image can be determined. This allows analysis of the patient's posture during the image capture, thus obtaining the imaging position for the spinal images.

[0041] Furthermore, by determining the patient's posture during spinal imaging, it's possible to assess whether the entire spine is fully visible within the image and whether it's presented frontally, allowing clinicians to observe the entire spine from all angles. This analysis helps determine the quality of the spinal image. After each spinal imaging session, the DR (Digital Radiography) equipment can promptly assess the image quality. If an image fails to meet standards, it can be corrected immediately, avoiding the need for clinicians to determine the quality and for substandard images to be returned to the DR equipment for recalibration. This improves the accuracy of spinal imaging and reduces the time spent on initial imaging procedures.

[0042] The technical solution provided in this application, after acquiring the spinal image to be evaluated, first determines the positional coordinates of each key point in the spinal image. Then, based on the positional coordinates of each key point, the imaging position of the spinal image is determined to assess whether the imaging quality of the spinal image meets the standards. This achieves automated assessment of whether the imaging quality of the spinal image meets the standards, allowing for timely correction of substandard spinal images, enhancing the usability of spinal images for diagnosing patients' spinal health, and improving the accuracy of diagnosing patients' spinal health problems.

[0043] As an optional implementation scheme in the embodiments of this application, in order to ensure the accuracy of the key points of the spine in the spinal images and the accuracy of the spinal image evaluation, this application can provide a detailed explanation of the specific process of determining the key points of the spine in the spinal images and the specific process of evaluating the spinal images based on the shooting position of the spinal images.

[0044] Figure 3 This is a flowchart illustrating another method for quality assessment of spinal images, as shown in an embodiment of this application.

[0045] like Figure 3 As shown, the method may specifically include the following steps:

[0046] S310, acquire images of the spine to be evaluated.

[0047] S320 scales the spinal image from its original size to the target size, resulting in a scaled spinal image.

[0048] When taking spinal images of any patient using a DR (Digital Radiography) device, the size of the spinal image may vary depending on the patient's distance from the DR device, their height, and the field of view chosen by the technician. Furthermore, when extracting key spinal points from individual spinal images through feature analysis, accurate feature analysis of spinal images at a pre-defined target size is typically supported.

[0049] Therefore, in order to ensure the accuracy of feature analysis of key points of the spine in the spinal images, this application first performs z-score normalization preprocessing on each acquired spinal image, and then performs corresponding scaling operations on the spinal image so that the spinal image can be scaled from the original size to the target size, thereby obtaining the scaled spinal image.

[0050] For example, this application can scale the acquired spinal images to the target size represented by [512,1024] each time and save the original size of the spinal images so as to accurately analyze the position coordinates of each key point of the spine in subsequent analysis.

[0051] S330 inputs the scaled spinal image into the trained key point extraction model to extract the corresponding spinal key points and obtain the predicted coordinates of the spinal key points.

[0052] To ensure accurate extraction of key points in spinal images, this application can pre-train a key point extraction model to analyze the pixel features in the spinal images to determine the corresponding key points in the spine.

[0053] The keypoint extraction model in this application can employ UNet, Fully Convolutional Networks (FCN), or variations of UNet. Furthermore, it can utilize convolutional layers with fewer parameters and separable convolutional representations, combined with batch normalization (BN) layers and activation function layers to construct the keypoint extraction model. The activation function layer can use Leaky ReLU, or other activation functions; this application does not impose any limitations on this.

[0054] Furthermore, to ensure accurate extraction of each spinal key point in the spinal image, this application can construct the same number of thermal channels in the key point extraction model according to a pre-set number of spinal key points. That is, the key point extraction model in this application can include thermal channels constructed for each spinal key point, so that a heat map corresponding to the spinal key point can be generated in the thermal channel under each spinal key point, and the pixel with the highest thermal response value in the heat map corresponding to each spinal key point is taken as the spinal key point.

[0055] In other words, in the key point extraction model of this application, a Gaussian kernel function with a radius r of 3σ and a diameter d of 2r+1 can be constructed in the thermal channel of each spinal key point. This kernel function has a radius r of 3σ and a diameter d of d*d. Then, the Gaussian kernel function constructed in the thermal channel of each spinal key point can generate a heatmap corresponding to that key point. Pixels near the coordinates of the key point in this heatmap have higher thermal response values, while the thermal response values ​​of pixels far from the key point decrease smoothly and rapidly. Furthermore, the Gaussian kernel function constructed in the thermal channel of each spinal key point ensures that the thermal response values ​​of each pixel in the output heatmap of each key point are within the range [0, 1], thus normalizing the mask representing the thermal map of the entire spinal image.

[0056] Therefore, after obtaining the scaled spinal image, it can be input into the trained keypoint extraction model. Then, as... Figure 4As shown, the spinal image is processed through heatmap channels constructed for each spinal keypoint in the keypoint extraction model. Within each heatmap channel, a Gaussian kernel function is used to transform the spinal image according to a Gaussian distribution, generating a corresponding heatmap. Within the heatmap generated from the heatmap channels for each spinal keypoint, the thermal response value of each pixel can be analyzed to identify the target pixel with the highest thermal response value, which is then used as the spinal keypoint corresponding to that heatmap channel. Following this method, each spinal keypoint in the spinal image can be determined from the heatmaps generated from the heatmap channels for each spinal keypoint, and the predicted coordinates of each spinal keypoint in the scaled spinal image can be obtained.

[0057] In some feasible implementations, in order to eliminate the influence of outliers in the heatmap, this application can also use threshold segmentation, median filtering, maximum analysis and other methods to analyze the thermal response value of each pixel in the heatmap generated by the thermal channel under each spinal key point, so as to determine the corresponding spinal key point.

[0058] S340 transforms the predicted coordinates of the spinal key points according to the scaling ratio between the original size and the target size to obtain the position coordinates of the spinal key points.

[0059] After determining the predicted coordinates of each spinal key point in the scaled spinal image, in order to assess the quality of the original spinal image, this application calculates the corresponding scaling ratio based on the original size of the spinal image before scaling and the target size after scaling. Then, according to the scaling ratio, the predicted coordinates of each spinal key point are transformed into coordinates in the original spinal image, thereby obtaining the position coordinates of each spinal key point in the original spinal image.

[0060] S350 determines the visual field defects of spinal images based on the location coordinates of key points in the spine and pre-defined non-key point areas.

[0061] To facilitate accurate diagnosis of spinal health problems by clinicians using spinal imaging, it is generally required that the spinal images include every key point of the spine to ensure a complete field of view during imaging. Furthermore, when spinal imaging meets these standards, each key point of the spine will be in a relatively standard position within the image.

[0062] Therefore, in order to accurately determine whether there is a visual gap in the spinal imaging for each key point of the spine, this application will pre-define a non-key point area where the key points of the spine will not appear, based on the standard shape of the entire spinal region when the spinal imaging meets the standards.

[0063] For example, considering that keypoint extraction models typically start from the top left corner of the heatmap converted from a spinal image, using the pixel in the top left corner as the initial spinal keypoint, and then analyzing the thermal response value of each pixel relative to the spinal keypoint row by row and column by column to continuously update the spinal keypoints, if a spinal keypoint is missing, it means that the thermal response values ​​of all pixels are relatively low, resulting in a pixel in the blank area of ​​the top left corner having the highest thermal response value, thus becoming a false spinal keypoint. Therefore, this application can use the top left corner region in the spinal image, where no spinal part exists under standard morphology, as the corresponding non-keypoint region.

[0064] Then, after determining the location coordinates of each key point in the spine, it can be determined whether the key point is real or false by judging whether its location coordinates are located within a pre-defined non-key point area. If the location coordinates of a key point are located within a pre-defined non-key point area, it means that the key point has not been identified in the spinal image, that is, the key point is missing in the spinal image, and it can be determined that there is a visual field defect in the spinal image.

[0065] However, if the coordinates of each spinal key point are not located within the predefined non-key point area, it means that each spinal key point has been identified in the spinal image, that is, no spinal key points are missing in the spinal image, and it can be determined that there is no visual field loss in the spinal image.

[0066] Furthermore, considering that when spinal imaging meets the standards, each key point of the spine will also be in a relatively standard position within the spinal image, it indicates that there is a relatively standard positional coordinate relationship between each key point of the spine when spinal imaging meets the standards.

[0067] Taking the 11 key spinal points in this application—nose, left ear, right ear, left acromioclavicular edge, right acromioclavicular edge, left iliac wing, right iliac wing, left femoral trochanter, right femoral trochanter, left ischial tuberosity, and right ischial tuberosity—as an example, the relative standard positional coordinate relationship between each key spinal point can be set as follows: the longitudinal positional coordinates of the nose, left and right ears, left and right acromioclavicular edges, left and right iliac wings, left and right femoral trochanters, and left and right ischial tuberosities continuously increase. Furthermore, for any pair of symmetrical key spinal points among the left and right ears, left and right acromioclavicular edges, left and right iliac wings, left and right femoral trochanters, and left and right ischial tuberosities, the lateral positional coordinate of the left key spinal point will be smaller than that of the right key spinal point.

[0068] Therefore, in order to accurately determine whether there is a visual field gap for each key spinal point in the spinal image, this application will further determine whether the positional coordinates of each key spinal point meet the aforementioned relative standard positional coordinate relationship. If the positional coordinates of a certain key spinal point do not meet the aforementioned relative standard positional coordinate relationship, it can be determined that the key spinal point is missing in the spinal image, thus determining that there is a visual field gap in the spinal image. Conversely, if the positional coordinates of each key spinal point meet the aforementioned relative standard positional coordinate relationship, it can be determined that no key spinal point is missing in the spinal image, thus determining that there is no visual field gap in the spinal image.

[0069] S360: If there is no visual field loss in the spinal image, the spinal centering of the spinal image is determined based on the position coordinates of the target key point pair in the spinal key points.

[0070] If there are visual gaps in the spinal imaging, it means that the imaging of the spinal spine did not meet the standards and needs to be corrected to meet the standards in order to help clinicians make accurate diagnoses of the patient's spinal health problems.

[0071] If there are no visual gaps in the spinal images, it is necessary to further determine whether the spinal region is presented centered in the frontal view, allowing clinicians to observe the complete morphology of the entire spine from all angles. Furthermore, considering that some key spinal points are symmetrical when the spinal region is presented frontally in spinal images, this application identifies every two symmetrically existing key spinal points from the various key spinal points in the spinal images as target key point pairs. Then, based on the positional coordinates of the two key spinal points included in each target key point pair, it can be determined whether the target key point pair is symmetrically positioned on either side of the spinal image's center line, thereby determining whether the spinal region is presented centered in the spinal image.

[0072] As an optional implementation of this application, determining the spinal centering of a spinal image can specifically be achieved by: determining the corresponding target key point pair from the spinal key points; determining the center offset of the target key point pair within the spinal image based on the position coordinates of each spinal key point in the target key point pair; and determining the spinal centering of the spinal image based on the center offset.

[0073] In other words, the target key point pairs in this application may include key point pairs consisting of the left iliac wing and the right iliac wing, key point pairs consisting of the left femoral trochanter and the right femoral trochanter, key point pairs consisting of the left ischial tuberosity and the right ischial tuberosity, etc.

[0074] Then, for each pair of target keypoints, based on the position coordinates of the two spinal keypoints that make up the pair, the coordinates of the center point between the two spinal keypoints in the pair can be calculated. Then, by determining the distance between the coordinates of this center point and the vertical centerline of the spinal image, the center offset of the target keypoint pair within the spinal image can be determined. Therefore, a center offset can be determined for each pair of target keypoints.

[0075] Furthermore, by analyzing whether the center offset of each target key point pair exceeds the preset center offset error range, it can be determined whether the spine in the spinal image is centered.

[0076] S370 assesses the image quality of spinal images based on visual field defects and spinal centering.

[0077] After determining the visual field defects and spinal centering of the spinal image, the spinal image is deemed substandard if there are visual field defects or if there are no visual field defects but the spine is not centered.

[0078] In some feasible implementations, if there are gaps in the field of view in the spinal image, or if the spine in the spinal image is not centered, a message indicating that the spinal image quality is substandard will be generated. That is, when the spinal image is substandard, a corresponding message can be presented to the operator to remind them to re-capture the patient's spinal image, thus allowing for timely correction of the substandard spinal image while the patient is still at the DR equipment.

[0079] The technical solution provided in this application, after acquiring the spinal image to be evaluated, first determines the positional coordinates of each key point in the spinal image. Then, based on the positional coordinates of each key point, the imaging position of the spinal image is determined to assess whether the imaging quality of the spinal image meets the standards. This achieves automated assessment of whether the imaging quality of the spinal image meets the standards, allowing for timely correction of substandard spinal images, enhancing the usability of spinal images for diagnosing patients' spinal health, and improving the accuracy of diagnosing patients' spinal health problems.

[0080] Figure 5 This is a schematic block diagram illustrating the principle of a spinal imaging quality assessment device according to an embodiment of this application.

[0081] like Figure 5 As shown, the device 500 may include:

[0082] Image acquisition module 510 is used to acquire spinal images to be evaluated;

[0083] The key point determination module 520 is used to determine the position coordinates of key points of the spine in the spinal image.

[0084] The quality assessment module 530 is used to determine the shooting position of the spinal image based on the position coordinates of the key points of the spine, so as to assess the shooting quality of the spinal image.

[0085] In some possible implementations, the key point determination module 520 can be specifically used for:

[0086] The spinal image is scaled from its original size to the target size to obtain a scaled spinal image;

[0087] The scaled spinal image is input into the trained key point extraction model to extract the corresponding spinal key points and obtain the predicted coordinates of the spinal key points.

[0088] Based on the scaling ratio between the original size and the target size, the predicted coordinates of the spinal key points are transformed to obtain the position coordinates of the spinal key points.

[0089] In some implementations, the key point extraction model includes a thermal channel constructed under each spinal key point to generate a heat map corresponding to that spinal key point, so that the pixel with the highest thermal response value in the heat map corresponding to each spinal key point is taken as that spinal key point.

[0090] In some possible implementations, the quality assessment module 530 may include:

[0091] The visual field defect determination unit is used to determine the visual field defect of the spinal image based on the position coordinates of the key points of the spine and the pre-defined non-key point area.

[0092] The spine centering determination unit is used to determine the spine centering of the spine image based on the position coordinates of the target key point pair in the spine key points if there is no visual field loss in the spine image.

[0093] A quality assessment unit is used to assess the imaging quality of the spinal images based on the visual field defects and spinal centering of the spinal images.

[0094] In some implementations, the spine centering determination unit can be specifically used for:

[0095] Identify the corresponding target key point pairs from the aforementioned spinal key points;

[0096] Based on the position coordinates of each spinal key point in the target key point pair, determine the center offset of the target key point pair within the spinal image;

[0097] The spinal centering of the spinal image is determined based on the center offset.

[0098] In some feasible implementations, the quality assessment unit can be specifically used for:

[0099] If the spinal image has a missing field of view, or if the spine in the spinal image is not centered, a message indicating that the image quality of the spinal image is substandard will be generated.

[0100] In this embodiment, after acquiring the spinal image to be evaluated, the positional coordinates of each key spinal point in the image are first determined. Then, based on the positional coordinates of each key spinal point, the imaging position of the spinal image is determined to assess whether the imaging quality of the spinal image meets the standards. This achieves automated assessment of the imaging quality of the spinal image, allowing for timely correction of substandard spinal images, enhancing the usability of spinal images for diagnosing patients' spinal health, and improving the accuracy of diagnosing patients' spinal health problems.

[0101] It should be understood that the device embodiments and method embodiments can correspond to each other, and similar descriptions can be found in the method embodiments. To avoid repetition, further details are omitted here. Specifically, Figure 5 The apparatus 500 shown can execute any of the method embodiments provided in this application, and the foregoing and other operations and / or functions of each module in the apparatus 500 are respectively for implementing the corresponding processes in the various methods of the embodiments of this application. For the sake of brevity, they will not be described in detail here.

[0102] The apparatus 500 of this application embodiment has been described above from the perspective of functional modules in conjunction with the accompanying drawings. It should be understood that this functional module can be implemented in hardware, in software instructions, or in a combination of hardware and software modules. Specifically, the steps of the method embodiments in this application can be completed by integrated logic circuits in the processor's hardware and / or by software instructions. The steps of the method disclosed in this application embodiment can be directly embodied as being executed by a hardware decoding processor, or by a combination of hardware and software modules in the decoding processor. Optionally, the software module can reside in a mature storage medium in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, etc. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps in the above method embodiments.

[0103] Figure 6 This is a schematic block diagram of an electronic device 600 shown in an embodiment of this application.

[0104] like Figure 6 As shown, the electronic device 600 may include:

[0105] The system includes a memory 610 and a processor 620. The memory 610 stores computer programs and transfers the program code to the processor 620. In other words, the processor 620 can retrieve and run the computer programs from the memory 610 to implement the methods described in the embodiments of this application.

[0106] For example, the processor 620 can be used to execute the above-described method embodiments according to instructions in the computer program.

[0107] In some embodiments of this application, the processor 620 may include, but is not limited to:

[0108] General-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0109] In some embodiments of this application, the memory 610 includes, but is not limited to:

[0110] Volatile memory and / or non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), and Direct Rambus RAM (DR RAM).

[0111] In some embodiments of this application, the computer program may be divided into one or more modules, which are stored in the memory 610 and executed by the processor 620 to perform the method provided in this application. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the electronic device.

[0112] like Figure 6 As shown, the electronic device may further include:

[0113] Transceiver 630, which can be connected to processor 620 or memory 610.

[0114] The processor 620 can control the transceiver 630 to communicate with other devices; specifically, it can send information or data to other devices or receive information or data sent by other devices. The transceiver 630 may include a transmitter and a receiver. The transceiver 630 may further include antennas, and the number of antennas may be one or more.

[0115] It should be understood that the various components in the electronic device are connected through a bus system, which includes a data bus, a power bus, a control bus, and a status signal bus.

[0116] This application also provides a computer storage medium storing a computer program thereon, which, when executed by a computer, enables the computer to perform the methods of the above-described method embodiments. Alternatively, this application also provides a computer program product containing instructions that, when executed by a computer, cause the computer to perform the methods of the above-described method embodiments.

[0117] When implemented using software, it can be implemented entirely or partially as a computer program product. This computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., digital video disc (DVD)), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0118] Those skilled in the art will recognize that the various embodiments described herein are related to the present invention.

[0119] The example modules and algorithm steps can be implemented using electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific technical solution.

[0120] Define application and design constraints. Those skilled in the art may use different methods to implement the described functionality for each specific application, but such implementation should not be considered beyond the scope of this application.

[0121] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices, and

[0122] The method can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of the modules is only a logical functional division, and in actual implementation, it can be...

[0123] There are other ways of classifying them, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed can be indirect coupling or communication connection through some interface, device, or module, and can be electrical, mechanical, or other forms.

[0124] 5. Modules described as separate components may or may not be physically separate.

[0125] The components displayed in a block may or may not be physical modules; that is, they may be located in one place, or...

[0126] It can also be distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment, depending on actual needs. For example, the various functions in the embodiments of this application...

[0127] Modules can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0128] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for quality assessment of spinal images, characterized in that, include: Obtain images of the spine to be evaluated; Determine the position coordinates of key points in the spine in the spinal image; Eleven key points of the spine surround the entire spinal region to represent the complete shape of the spinal region. The 11 key points of the spine include: nose, left ear, right ear, left acromioclavicular edge, right acromioclavicular edge, left iliac wing, right iliac wing, left femoral tuberosity, right femoral tuberosity, left ischial tuberosity, and right ischial tuberosity. Based on the positional coordinates of the key points of the spine and the pre-defined non-key point region, the visual field defects of the spinal image are determined. The non-key point region is the upper left corner of the spinal image where the spine is absent in the standard morphology. If the positional coordinates of a certain key point of the spine are located within the pre-defined non-key point region, it indicates that the key point of the spine is missing in the spinal image, and the visual field defects of the spinal image are determined. Further, it is determined whether the positional coordinates of each key point of the spine meet the set relative standard positional coordinate relationship. If the positional coordinates of a certain key point of the spine do not meet the set relative standard positional coordinate relationship, it indicates that the key point of the spine is missing in the spinal image, and the visual field defects of the spinal image are determined. The set relative standard positional coordinate relationship includes: the longitudinal positional coordinates of the nose, left and right ears, left and right acromioclavicular edges, left and right iliac wings, left and right femoral tuberosities, and left and right ischial tuberosities continuously increase; and for any pair of key points of the spine that are symmetrically located on the left and right sides, the lateral positional coordinates of the left key point of the spine will be smaller than the lateral positional coordinates of the right key point of the spine. If the spinal image does not have any missing visual field, the spinal centering of the spinal image is determined based on the position coordinates of the target key point pairs in the spinal key points. If the spinal image has a missing field of view, or if the spine in the spinal image is not centered, a message indicating that the image quality of the spinal image is substandard will be generated.

2. The method according to claim 1, characterized in that, Determining the position coordinates of key points in the spine in the spinal image includes: The spinal image is scaled from its original size to the target size to obtain a scaled spinal image; The scaled spinal image is input into the trained key point extraction model to extract the corresponding spinal key points and obtain the predicted coordinates of the spinal key points. Based on the scaling ratio between the original size and the target size, the predicted coordinates of the spinal key points are transformed to obtain the position coordinates of the spinal key points.

3. The method according to claim 2, characterized in that, The key point extraction model includes a thermal channel constructed under each spinal key point to generate a heat map corresponding to that spinal key point, so that the pixel with the highest thermal response value in the heat map corresponding to each spinal key point is taken as that spinal key point.

4. The method according to claim 1, characterized in that, Determining the spine centering of the spine image based on the position coordinates of the target key point pairs among the spinal key points includes: Identify the corresponding target key point pairs from the aforementioned spinal key points; Based on the position coordinates of each spinal key point in the target key point pair, determine the center offset of the target key point pair within the spinal image; The spinal centering of the spinal image is determined based on the center offset.

5. A device for quality assessment of spinal images, characterized in that, include: The image acquisition module is used to acquire images of the spine to be evaluated. The key point determination module is used to determine the position coordinates of key points of the spine in the spinal image; Eleven key points of the spine surround the entire spinal region to represent the complete shape of the spinal region. The 11 key points of the spine include: nose, left ear, right ear, left acromioclavicular edge, right acromioclavicular edge, left iliac wing, right iliac wing, left femoral tuberosity, right femoral tuberosity, left ischial tuberosity, and right ischial tuberosity. The quality assessment module is used to determine the visual field defects of the spinal image based on the position coordinates of the key points of the spine and the pre-defined non-key point area. If the position coordinates of a certain key point of the spine are located within the pre-defined non-key point area, it indicates that the key point of the spine is missing in the spinal image, and the visual field defects of the spinal image are determined. The module further determines whether the position coordinates of each key point of the spine meet the set relative standard position coordinate relationship. If the position coordinates of a certain key point of the spine do not meet the set relative standard position coordinate relationship, it indicates that the key point of the spine is missing in the spinal image, and the visual field defects of the spinal image are determined. If the spinal image does not have a visual field defect, the spinal centering of the spinal image is determined based on the position coordinates of the target key point pairs among the key points of the spine. If the spinal image has a visual field defect, or the spine of the spinal image is not centered, a prompt indicating that the image capture quality is substandard is generated. The non-critical area refers to the upper left corner of the spinal image where there is no spinal part under standard morphology. The set relative standard position coordinate relationship includes: the longitudinal position coordinates of the nose, left and right ears, left and right acromioclavicular edges, left and right iliac wings, left and right femoral tuberosities and left and right ischial tuberosities continuously increase, and any pair of spinal critical points among the left and right ears, left and right acromioclavicular edges, left and right iliac wings, left and right femoral tuberosities and left and right ischial tuberosities that exist symmetrically on the left and right sides, the lateral position coordinate of the left spinal critical point will be smaller than the lateral position coordinate of the right spinal critical point.

6. An electronic device, characterized in that, include: A processor and a memory, the memory for storing a computer program, the processor for calling and running the computer program stored in the memory to perform the spinal image quality assessment method according to any one of claims 1-4.

7. A computer-readable storage medium, characterized in that, Used to store computer programs that cause a computer to perform a method for assessing the quality of spinal images as described in any one of claims 1-4.

8. A computer program product comprising a computer program / instructions, characterized in that, When executed by a processor, the computer program / instruction implements the method for quality assessment of spinal images as described in any one of claims 1-4.

Images