Medical image processing method and computer readable storage medium

Abstract

The application provides a medical image processing method and a computer readable storage medium, comprising: acquiring a plurality of medical images to be detected; performing preliminary screening on the plurality of medical images to determine candidate medical images; and performing image quality evaluation on the candidate medical images to obtain a quality evaluation result. The medical image processing method provided by the application can determine candidate medical images through preliminary screening before quality evaluation on medical images, can quickly screen the image quality, saves computing power, and improves the detection efficiency of the image quality.

Description

Technical Field

[0001] This invention relates to the field of medical image processing technology, and in particular to a medical image processing method and a computer-readable storage medium. Background Technology

[0002] In medical imaging systems, image quality depends on many factors, such as spatial resolution, tissue contrast, signal-to-noise ratio, contrast-to-noise ratio, and image defects. To achieve the best image quality, hardware and scanning parameters are optimized for different organs or pathologies. However, during scanning, physiological factors of the subject (respiration, heartbeat, body structure) and movement can degrade image quality, thus failing to meet clinical diagnostic requirements.

[0003] To ensure scan quality, users need to manually retrieve and observe the scanned images during the scan, assessing the integrity of the information contained within the images to ensure they have acceptable quality before data analysis and determining whether a rescan is necessary to meet requirements. This is especially true for whole-body imaging, where hardware limitations necessitate separate acquisition at several beds. Furthermore, each bed's routine MRI scan includes images with different weights, such as T1, T2, and DWI, as well as acquisitions of the same weighted images in different orientations, such as T2 transverse and T2 coronal views, and acquisitions with different parameters. Additionally, for different diseases, additional targeted MRI sequences may be scanned at certain sites for differential diagnosis. Therefore, a routine MRI examination ultimately involves at least twenty sequences, requiring considerable time and effort to review and assess image quality. This process undoubtedly increases the burden on physicians.

[0004] Artifacts are various forms of images that appear in a scanned image but do not actually exist. Artifacts are a significant factor contributing to the deterioration of medical image quality and can even greatly affect doctors' analysis and diagnosis of lesions. Therefore, medical image artifact recognition is crucial as a basis for medical diagnosis. Traditionally, learning methods (machine / deep learning) can be used to predict artifacts; higher accuracy requires a larger model and greater computing power. In practical applications, less than 5% of images have quality issues. Performing image quality analysis on every scanned medical image would be computationally intensive and would affect the operation of other algorithms. Therefore, it is necessary to provide a rapid image quality screening method to save computing power and improve the efficiency of image quality detection. Summary of the Invention

[0005] The purpose of this invention is to provide a medical image processing method and a computer-readable storage medium to achieve rapid screening of image quality, save computing power, and improve the efficiency of image quality detection.

[0006] To achieve the above objectives, the present invention provides a medical image processing method, comprising:

[0007] Acquire multiple medical images to be inspected;

[0008] A preliminary screening of multiple medical images is performed to identify candidate medical images;

[0009] The candidate medical images are subjected to image quality assessment to obtain the quality assessment results.

[0010] Optionally, a preliminary screening of multiple medical images may be performed to determine candidate medical images, including:

[0011] Segmenting tissue regions in at least one of the aforementioned medical images;

[0012] Based on the segmented medical image, a first threshold and a second threshold are set respectively, and an evaluation parameter is calculated. The evaluation parameter is the ratio of the number of pixels in the medical image whose pixel value is between the first threshold and the second threshold to the number of pixels in the tissue region.

[0013] Candidate medical images are obtained by preliminary screening from multiple medical images based on the relationship between the evaluation parameters and a set threshold.

[0014] Optionally, the maximum pixel value in the background of the medical image is extracted as a first threshold, and the minimum pixel value in the tissue region is extracted as a second threshold.

[0015] Optionally, the identification of body parts in the medical image may be included before segmenting the tissue regions in the medical image.

[0016] Optionally, before identifying body parts in the medical image, the medical image may be preprocessed, including downsampling, filtering, or normalization.

[0017] Optionally, when the evaluation parameter is greater than the set threshold, the medical image is determined to be the candidate medical image.

[0018] Optionally, the quality of the candidate medical images is evaluated using a trained neural network model, wherein the quality is characterized by at least one image attribute parameter among image integrity, image contrast, image signal-to-noise ratio, and image resolution.

[0019] Optionally, the quality assessment results may include normal, average, moderately abnormal, or severely abnormal.

[0020] Optionally, the medical image is obtained by scanning the object with a scanner, and the medical image processing method further includes: determining the status parameters of the scanner based on the quality assessment results.

[0021] The present invention also provides a computer-readable storage medium storing at least one instruction executable by a processor, wherein when the at least one instruction is executed by the processor, the medical image processing method described in any one of the preceding claims is implemented.

[0022] In summary, this invention provides a medical image processing method and a computer-readable storage medium, comprising: acquiring multiple medical images to be detected; performing preliminary screening on the multiple medical images to determine candidate medical images; and performing image quality assessment on the candidate medical images to obtain quality assessment results. The medical image processing method provided by this invention, by first determining candidate medical images through preliminary screening before performing quality assessment on the medical images, can quickly screen image quality, save computing power, and improve the efficiency of image quality detection. Attached Figure Description

[0023] Figure 1 A flowchart of a medical image processing method provided in an embodiment of the present invention;

[0024] Figure 2 A flowchart illustrating the process of acquiring candidate medical images in a medical image processing method provided in an embodiment of the present invention;

[0025] Figure 3 This is a schematic diagram illustrating the segmentation of non-tissue regions and tissue regions in a medical image using a medical image processing method provided in an embodiment of the present invention.

[0026] Figure 4 The structural block diagram of a two-layer CNN convolutional neural network used for artifact quality assessment in a medical image processing method provided in an embodiment of the present invention;

[0027] Figure 5 This is a flowchart illustrating the establishment of an artifact screening network model in a medical image processing method according to an embodiment of the present invention;

[0028] Figure 6 A flowchart of a medical image processing method provided in another embodiment of the present invention;

[0029] Figure 7 This is a flowchart illustrating the process of obtaining quality assessment results in a medical image processing method according to an embodiment of the present invention. Detailed Implementation

[0030] The medical image processing method and computer-readable storage medium of the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The advantages and features of the present invention will become clearer from the following description and drawings; however, it should be noted that the concept of the technical solution of the present invention can be implemented in many different forms and is not limited to the specific embodiments described herein. The accompanying drawings are all in a very simplified form and use non-precise proportions, and are only used to facilitate and clarify the illustration of the embodiments of the present invention.

[0031] The terms "first," "second," etc., used in this specification are used to distinguish between similar elements and are not necessarily used to describe a specific order or chronological sequence. It should be understood that these terms, used so in this way, may be replaced where appropriate, for example, to allow embodiments of the invention described herein to operate in a different order than that described or shown herein. Similarly, if the methods described herein comprise a series of steps, and the order of these steps presented herein is not necessarily the only possible order in which these steps can be performed, and some described steps may be omitted and / or some other steps not described herein may be added to the method. If a component in one figure is identical to a component in another figure, although these components are readily identifiable in all figures, this specification will not label all identical components in every figure for the sake of clarity.

[0032] This embodiment provides a medical image processing method. Figure 1 A flowchart of the medical image processing method provided in this embodiment is shown below. Figure 1 As shown, the medical image processing method provided in this embodiment includes:

[0033] Step S11: Acquire multiple medical images to be detected;

[0034] Step S12: Perform preliminary screening on multiple medical images to identify candidate medical images;

[0035] Step S13: Perform image quality assessment on the candidate medical images to obtain the quality assessment results.

[0036] Optionally, multiple medical images can be reconstructed in real time using a scanner from a medical device, or obtained through a medical image data management system such as a Picture Archiving and Communication System (PACS). Optionally, medical image processing methods can be performed using post-processing software. Optionally, multiple medical images can be stored in a file conforming to the DICOM (Digital Imaging and Communication in Medicine) standard. The DICOM standard standard standardizes the format and structure of descriptive parameters used for radiographic images, as well as the commands for exchanging these images, and also standardizes other data objects such as image sequences, examination sequences, and examination reports.

[0037] Optionally, the process of determining candidate medical images can be equivalent to the process of preliminary identification of multiple medical images. Optionally, preliminary screening of multiple medical images can be achieved by calculating the resolution of each medical image, by calculating the signal-to-noise ratio of each medical image, by calculating the contrast of each medical image, or by determining whether each medical image is complete by calculating whether there is a region of interest (ROI) in each medical image.

[0038] Optionally, preliminary screening of multiple medical images to determine candidate medical images includes: segmenting the tissue region in at least one medical image; setting a first threshold and a second threshold based on the segmented medical image, and calculating an evaluation parameter, which is the ratio of the number of pixels in the medical image whose pixel values ​​are between the first threshold and the second threshold to the number of pixels in the tissue region; and obtaining candidate medical images from multiple medical images based on the relationship between the evaluation parameter and a set threshold. The first threshold and the second threshold are different. For example, the first threshold can be determined based on the pixel values ​​of pixels contained in the background (e.g., non-tissue region) of the medical image, while the second threshold can be determined based on the tissue region in the medical image, specifically based on the pixel values ​​of pixels contained in the tissue region of the medical image. The evaluation parameter in this embodiment can be a medical image quality parameter. The evaluation parameter can be an artifact parameter, resolution parameter, contrast parameter, or integrity parameter, etc. Of course, the first threshold and the second threshold can also be set by empirical values, and this embodiment does not impose any limitations.

[0039] Optionally, the quality assessment results of the candidate medical images may include, for example, the degree / level of artifacts appearing in the medical images, the contrast level corresponding to the medical images, and whether the medical images are suitable / can be used for current clinical diagnosis. Optionally, artifacts in medical images may include: motion artifacts caused by voluntary motion, folding artifacts caused by phase entanglement; gradient spark artifacts; metal artifacts caused by the object being detected carrying metal; shading artifacts caused by scattering; artifacts caused by contrast agents / tracers; truncation artifacts; respiratory motion artifacts, etc.

[0040] In one embodiment, candidate medical images are obtained by initially screening medical images for motion artifacts that cause blurring of the image background. Figure 2 As shown, the acquisition of candidate medical images includes:

[0041] Step S21: Acquire the medical image to be detected;

[0042] Step S22: Segment the tissue region and non-tissue region in at least one of the medical images;

[0043] Step S23: Extract any pixel value from a designated region of the background of the medical image as a first threshold M1, and any pixel value from the tissue region as a second threshold M2, and calculate the artifact parameter P, whereby the artifact parameter P is the ratio of the number M of pixels in the medical image whose pixel values ​​are between the first threshold M1 and the second threshold M2 to the number N of pixels in the tissue region; and,

[0044] Step S24: Determine that the medical image is free of artifacts based on the relationship between the artifact parameter P and a set threshold.

[0045] Specifically, firstly, the medical image to be detected is acquired. This medical image can be a multi-source image, meaning it comes from different types of medical images, such as CT scans and MRIs; or from medical images from different hospital equipment. For example, if the medical image is an MRI, the scanning field strength may include 1.5T (Tesla), 3.0T, etc., and the scanning locations may include transverse, coronal, and sagittal views of the head, transverse, coronal, and sagittal views of the neck, etc. The scanning directions may include coronal, sagittal, and transverse views, and the scanning sequence may include gradient echo pulse (GRE) or fast spin echo pulse (FSE). The image filtering method provided in this embodiment can support artifact recognition in different locations, such as the head, spine, and lower limbs; and supports the recognition of different artifact types, such as motion artifacts and metal artifacts.

[0046] Next, the medical image is preprocessed. This preprocessing includes downsampling, filtering, or normalization. For example, the medical image may be downsampled to a size of 320*320.

[0047] Next, body parts in the medical image are identified. For example, segmentation can be achieved based on the morphology of the human body region. Since human organs or tissues typically share common morphological characteristics—for example, the kidney is bean-shaped, and the breast is hill-shaped—the grayscale distribution characteristics of organs or tissues in the human body region image usually fall within a certain preset range. Based on these grayscale distribution characteristics, non-tissue regions can be separated from the human body region image, facilitating rapid candidate image selection.

[0048] Next, as Figure 3 As shown, the medical image is segmented into tissue and non-tissue regions. For example, a thresholding method is used to segment tissue and non-tissue regions. The threshold selected during the segmentation process is determined based on the background noise of the medical image. Whether the medical image contains background noise or whether the background noise is flat determines the stability of the image processing-based artifact screening method. In a clean background noise environment, even a simple thresholding segmentation detection method can achieve good artifact detection results. However, in general, the medical image signals we obtain will contain background noise. Therefore, the medical image processing method we establish must have good robustness to noise. That is, the threshold for thresholding segmentation is determined based on the background noise of the medical image during the segmentation of tissue and non-tissue regions. In addition, the background noise extraction method will also differ for image data acquired by different acquisition methods.

[0049] Next, any pixel value from a defined region of the background of the medical image is extracted as a first threshold M1, and any pixel value from the tissue region is extracted as a second threshold M2. An artifact parameter P is then calculated, where P is the ratio of the number M of pixels in the medical image whose pixel values ​​fall between the first threshold M1 and the second threshold M2 to the number N of pixels in the tissue region, i.e., P = M / N. Optionally, the defined region of the background can be a region containing corner information. This corner information can be, for example, display mode information, image attribute information, etc., contained in the upper left, lower left, upper right, and upper right corners of the medical image.

[0050] Next, refer to Figure 2 As shown, the medical image is determined to be free of artifacts based on the relationship between the artifact parameter P and a set threshold. Specifically, the relationship between the artifact parameter P and the set threshold is determined. When the artifact parameter P is greater than the set threshold, the medical image is determined to have artifacts (output 1). When the artifact parameter P is less than or equal to the set threshold, the medical image is determined to be free of artifacts (output 0).

[0051] Image quality assessment of candidate medical images, to obtain quality assessment results, can be achieved using machine learning network models. These models can be DNNs (Deep Neural Networks), CNNs (Convolutional Neural Networks), RNNs (Recurrent Neural Networks), etc. When the machine learning network model is a CNN, it can be a V-Net model, a U-Net model, a Generative Adversarial Network (GAN) model, etc. Quality can be characterized by at least one image attribute parameter among image integrity, image contrast, image signal-to-noise ratio, and image resolution.

[0052] In one embodiment, quality is characterized by the image signal-to-noise ratio. When artifacts are determined to exist in the medical image, the quality of the artifacts also needs to be evaluated. Neural networks can be used to evaluate the quality of the artifacts. ResNet 18 or ResNet 50 are typically used for multi-class artifact quality evaluation; for example, a simple two-layer CNN can be used for artifact quality evaluation. Figure 4 As shown. The results of artifact quality assessment are expressed as the degree of artifact, which includes normal, average, moderately abnormal, or severely abnormal.

[0053] Alternatively, in one embodiment, to facilitate determining the degree of impact of artifacts in candidate medical images on the quality of the medical image to be processed, the quality assessment result output by the machine learning network model is artifact severity indication information. Optionally, the computer device can classify the artifact severity indication information into four levels: Level 1, Level 2, Level 3, and Level 4. Level 1 indicates that the candidate medical image is normal and unaffected by artifacts; Level 2 indicates that artifacts have a slight impact on the candidate medical image and it can still be used; Level 3 indicates that artifacts have a moderate impact on the candidate medical image and it cannot be used clinically; and Level 4 indicates that artifacts have a severe impact on the candidate medical image and it cannot be used clinically.

[0054] The quality assessment of artifacts can also be performed using artifact screening network models, such as... Figure 5 As shown, the process of establishing the artifact screening network model includes:

[0055] Step S51: Collect image data of medical images, construct datasets with different scenes and different levels of artifacts, and preprocess them to obtain training datasets;

[0056] Step S52: Construct a body part recognition network and obtain the scene number corresponding to the medical image based on the image data;

[0057] Step S53: Construct an artifact screening network model based on a deep convolutional neural network;

[0058] Step S54: Train the artifact screening network model according to the training dataset and the scene number, and calculate the artifact degree of the medical image through the trained artifact screening network model;

[0059] Step S55: Adjust the dataset appropriately based on the degree of artifacts calculated by the artifact screening network model.

[0060] In step S51, the scenario includes the scanning site and the scanning sequence. The scanning site includes transverse head view, coronal head view, sagittal head view, transverse neck view, coronal neck view, or sagittal neck view. The scanning sequence includes gradient echo pulse sequence (GRE) or fast spin echo pulse sequence (FSE). The degree of artifacts includes normal, moderate, moderately abnormal, or severely abnormal.

[0061] Based on the quality assessment results, a prompt message can be generated. In one embodiment, if the impact of artifacts in the medical image to be processed on the image quality of the candidate medical image is greater than or equal to a preset artifact impact threshold, the computer device outputs a prompt message.

[0062] The prompt information can be a prompt icon to prompt the user to confirm whether to accept the artifacts in the candidate medical image and whether to rescan the corresponding scanned area of ​​the candidate medical image; alternatively, the prompt information can be a warning icon to indicate the presence of artifacts affecting image quality in the medical image to be processed; the prompt information can be a specific sequence corresponding to the medical images affected by artifacts, and this sequence is a time sequence within the entire medical imaging scan. The computer device can also output the prompt information by emitting a prompt sound, emitting a prompt red light, or displaying a rescan prompt text on the screen. This application embodiment does not specifically limit the method by which the computer device outputs the prompt information.

[0063] In one embodiment, the medical image is obtained by scanning and reconstructing the object in real time using a scanner, such as... Figure 6 As shown, medical image processing methods include:

[0064] Step S61: Acquire multiple medical images to be detected;

[0065] Step S62: Perform preliminary screening on multiple medical images to identify candidate medical images;

[0066] Step S63: Perform image quality assessment on the candidate medical images to obtain the quality assessment results;

[0067] Step S64: Determine the scanner's status parameters based on the quality assessment results.

[0068] The scanner's status parameters can be the device's own operational status parameters, such as gradient parameters, radio frequency parameters, and main magnetic field parameters; the scanner's status parameters can also be the status of the scanned object, such as whether it carries metal or exhibits autonomous movement. Optionally, determining the scanner's status parameters can include: if the quality assessment results indicate that the resolution of the candidate medical image has not reached the set value, the scanner's scanning time is set too short, requiring an extension of the scanner's scanning time or an increase in the sampling rate; or, if the quality assessment results indicate that the candidate medical image contains spark artifacts, requiring a change in the gradient waveform; or, if the quality assessment results indicate that the candidate medical image contains metal artifacts, then the scanned object carries metal, requiring the addition of an artifact suppression algorithm.

[0069] In one embodiment, taking the presence of motion artifacts in a candidate medical image as an example, such as... Figure 7 As shown, the process of obtaining quality assessment results may include:

[0070] Step S71: Input the candidate medical image into the target artifact recognition model to obtain the target artifact attribute information output by the target artifact recognition model.

[0071] After processing candidate medical images, the target artifact recognition model can output target artifact attribute information, which indicates the attribute characteristics of artifacts in the candidate medical images. The target artifact attribute information can include at least one of the following: artifact size, artifact location, artifact quantity, and artifact type. Artifact types can include zipper artifacts, spark artifacts, involuntary motion artifacts, breathing artifacts, and vascular pulsation artifacts, etc. Artifacts can also be categorized by their source into equipment artifacts and human artifacts. Equipment artifacts include, for example, measurement error artifacts of the imaging system, X-ray beam hardening artifacts, high voltage fluctuation artifacts of the imaging system, temperature drift artifacts of electronic circuits, and detector drift artifacts; human artifacts include, for example, artifacts caused by the movement of the detected object, artifacts caused by the peristalsis of internal organs, and artifacts caused by internal metal implants.

[0072] Step S72: Input the candidate medical image and target artifact attribute information into the target artifact degree recognition model to obtain the artifact degree indication information output by the target artifact degree recognition model.

[0073] Among them, the artifact level indicator information is used to indicate the degree of impact of artifacts in the candidate medical image on the image quality of the candidate medical image.

[0074] Specifically, after inputting the candidate medical image into the target artifact recognition model and obtaining the target artifact attribute information output by the model, the computer device can input the candidate medical image and the target artifact attribute information into the target artifact severity recognition model. The target artifact severity recognition model can determine the degree of influence of artifacts in the candidate medical image on the image quality of the candidate medical image based on the target artifact attribute information.

[0075] Optionally, the target artifact recognition model can identify candidate medical images and divide them into regions of interest (ROI) and non-ROI regions. The ROI can be a scanned area within the medical image. For example, when the scanned area is the brain, the image includes both brain-related image information and a small portion of the neck area. The target artifact recognition model classifies the neck area as a non-ROI region and the brain area as a ROI region.

[0076] After identifying the region of interest (ROI) in a candidate medical image, the target artifact identification model can determine the degree of impact of artifacts on image quality based on the location information of the ROI, the attribute information of the ROI, and the attribute information of the target artifacts.

[0077] For example, if the candidate medical image corresponds to the brain, the target artifact severity identification model identifies brain tissues such as white matter and gray matter as regions of interest (ROIs) and the neck as a non-ROI. Based on the target artifact attribute information, the model determines that the target artifact in the candidate medical image is a neck motion artifact. Since neck motion artifacts have a minimal impact on brain tissue, the target artifact severity identification model determines that the artifacts in the candidate medical image have a relatively small impact on image quality.

[0078] Optionally, the target artifact severity identification model can determine the degree of influence of artifacts in the medical image to be processed on the image quality of the medical image to be processed based on the difference between the signal-to-noise ratio, contrast, etc. of the medical image to be processed containing target artifacts and the set image quality thresholds (e.g., signal-to-noise ratio threshold, contrast threshold). If the difference between the signal-to-noise ratio, contrast, etc. of the medical image to be processed containing target artifacts and the set image quality thresholds is larger, the target artifact severity identification model determines that the artifacts in the medical image to be processed have a greater influence on the image quality of the medical image to be processed; conversely, if the difference between the signal-to-noise ratio, contrast, etc. of the medical image to be processed containing target artifacts and the set image quality thresholds is smaller, the target artifact severity identification model determines that the artifacts in the medical image to be processed have a smaller influence on the image quality of the medical image to be processed.

[0079] Optionally, the target artifact severity identification model can determine the degree of influence of artifacts in the medical image being processed on the image quality of the medical image based on the positional relationship between the target artifact and the scanned area. If the distance between the target artifact and the scanned area is less than a preset distance threshold, the target artifact severity identification model determines that the artifacts in the medical image being processed have a significant impact on the image quality of the medical image being processed; if the distance between the target artifact and the scanned area is greater than or equal to the preset distance threshold, the target artifact severity identification model determines that the artifacts in the medical image being processed have a relatively small impact on the image quality of the medical image being processed.

[0080] Optionally, the target artifact severity identification model can determine the degree of influence of artifacts in the medical image being processed on the image quality based on the area size of the target artifacts. If the area of ​​the target artifact exceeds a preset area threshold, the target artifact severity identification model determines that the artifacts in the medical image being processed have a significant impact on the image quality; if the area of ​​the target artifact is less than the preset area threshold, the target artifact severity identification model determines that the artifacts in the medical image being processed have a relatively small impact on the image quality.

[0081] Optionally, the target artifact severity identification model can determine the degree of impact of artifacts in the medical image being processed on the image quality based on the number of target artifacts. If the number of target artifacts exceeds a preset threshold, the target artifact severity identification model determines that the artifacts in the medical image being processed have a significant impact on the image quality; if the number of target artifacts is less than the preset threshold, the target artifact severity identification model determines that the artifacts in the medical image being processed have a relatively small impact on the image quality.

[0082] Optionally, the target artifact identification model can determine the degree of influence of artifacts in candidate medical images on image quality based on the type of target artifact. If the type of target artifact is one that is unavoidable during the scanning process of the scanned area, the target artifact identification model determines that the artifacts in the medical image to be processed have a relatively small impact on the image quality of the medical image to be processed; if the type of target artifact is one that is avoidable during the scanning process of the scanned area, the target artifact identification model determines that the artifacts in the medical image to be processed have a relatively large impact on the image quality of the candidate medical image.

[0083] Accordingly, the present invention also provides a computer-readable storage medium storing at least one instruction executable by a processor, wherein when the at least one instruction is executed by the processor, the medical image processing method described above is implemented.

[0084] In summary, this invention provides a medical image processing method and a computer-readable storage medium, comprising: acquiring multiple medical images to be detected; performing preliminary screening on the multiple medical images to determine candidate medical images; and performing image quality assessment on the candidate medical images to obtain quality assessment results. The medical image processing method provided by this invention, by first determining candidate medical images through preliminary screening before performing quality assessment on the medical images, can quickly screen image quality, save computing power, and improve the efficiency of image quality detection.

[0085] The above description is merely a description of preferred embodiments of the present invention and is not intended to limit the scope of the present invention in any way. Any changes or modifications made by those skilled in the art based on the above disclosure shall fall within the protection scope of the claims.

Claims

1. A medical image processing method, characterized in that, include: Acquire multiple medical images to be inspected; A preliminary screening of multiple medical images is performed to identify candidate medical images; The candidate medical images are subjected to image quality assessment to obtain quality assessment results; A preliminary screening of multiple medical images was performed to identify candidate medical images, including: Segmenting tissue regions in at least one of the aforementioned medical images; Based on the segmented medical image, a first threshold and a second threshold are set respectively, and an evaluation parameter is calculated. The evaluation parameter is the ratio of the number of pixels in the medical image whose pixel value is between the first threshold and the second threshold to the number of pixels in the tissue region. Candidate medical images are obtained by preliminary screening from multiple medical images based on the relationship between the evaluation parameters and a set threshold. The first threshold is determined based on the pixel values ​​of the pixels contained in the background of the medical image, and the second threshold is determined by the tissue region in the medical image.

2. The medical image processing method according to claim 1, characterized in that, The maximum pixel value in the background of the medical image is extracted as a first threshold, and the minimum pixel value in the tissue region is extracted as a second threshold.

3. The medical image processing method according to claim 1, characterized in that, The process includes identifying body parts within the medical image before segmenting tissue regions.

4. The medical image processing method according to claim 3, characterized in that, Before identifying body parts in the medical image, the medical image is preprocessed, including downsampling, filtering, or normalization.

5. The medical image processing method according to claim 1, characterized in that, When the evaluation parameter is greater than the set threshold, the medical image is determined to be the candidate medical image.

6. The medical image processing method according to claim 1, characterized in that, The quality of the candidate medical images is evaluated using a machine learning model, and the quality is characterized by at least one image attribute parameter among image integrity, image contrast, image signal-to-noise ratio, and image resolution.

7. The medical image processing method according to claim 6, characterized in that, The quality assessment results include normal, average, moderately abnormal, or severely abnormal.

8. The medical image processing method according to claim 1, characterized in that, The medical image is obtained by scanning the object using a scanner. The medical image processing method further includes: determining the status parameters of the scanner based on the quality assessment results.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one instruction executable by a processor, which, when executed by the processor, implements the medical image processing method as described in any one of claims 1 to 8.

Images