Medical image processing method, medical image processing device and medical image processing equipment

Abstract

The invention provides a medical image processing method, a medical image processing device and medical image processing equipment, which are used for creating a set of novel image processing architecture from an image processing level, so that high-quality images can be output and displayed under the condition that motion artifacts are accompanied, multiple image acquisition operations can be effectively avoided, and the image processing efficiency is improved. The detection success rate is improved, the subsequent diagnosis work can be better carried out, and the medical service quality is effectively guaranteed through high-quality image inspection.

Description

Technical Field

[0001] The present application relates to the field of medical imaging, and in particular to a medical image processing method, apparatus, and processing device. Background Art

[0002] In magnetic resonance imaging (MRI), the presence of motion artifacts has always been an important factor restricting the success rate of examinations and has also affected the diagnostic physician's diagnosis of lesions.

[0003] Normally, if the patient has a certain degree of cooperation, the imaging technician will repeatedly collect data during the operation, and the diagnostic physician will make a comparative diagnosis based on the data collected multiple times.

[0004] However, in the face of physiological movements or involuntary shaking of some patients (such as Parkinson's disease and infants and other special patients), although the acquisition of multiple images may prevent motion artifacts from existing at the same level and position, motion artifacts still exist, and low image quality will cause examination failure and diagnostic difficulties. Summary of the Invention

[0005] The present application provides a medical image processing method, apparatus, and processing equipment, which are used to create a novel image processing architecture starting from the image processing level, so that high-quality image output and display can be performed even in the presence of motion artifacts. This can effectively avoid multiple image acquisition operations, improve the detection success rate, and enable subsequent diagnostic work to be carried out better, effectively ensuring the quality of medical services through high-quality image examinations.

[0006] In a first aspect, the present application provides a medical image processing method, the method comprising:

[0007] In the medical image acquisition step, a first medical image is acquired, wherein the first medical image is specifically a magnetic resonance image;

[0008] In the medical image motion artifact processing step, motion artifact processing is performed on the first medical image using a preconfigured motion artifact processing model, wherein the motion artifact processing model is a machine learning model, and the motion artifact processing model is used to determine whether motion artifacts exist in the medical image input to the model, and if it is determined that motion artifacts exist, continue to perform motion artifact correction processing on the medical image input to the model;

[0009] In the medical image repeated processing phase, a motion artifact processing model is used to determine whether a second medical image after the correction of the first medical image still has motion artifacts. If motion artifacts still exist, the medical image acquisition phase and the medical image motion artifact processing phase are triggered to be re-executed until the preset conditions configured for the number of acquisitions are met, and a medical image set consisting of the corrected medical images is formed;

[0010] Display medical image collections.

[0011] In a second aspect, the present application provides a medical image processing device, comprising:

[0012] An acquisition unit, configured to acquire a first medical image in a medical image acquisition step, wherein the first medical image is specifically a magnetic resonance image;

[0013] a primary processing unit, configured to perform motion artifact processing on a first medical image using a preconfigured motion artifact processing model in a medical image motion artifact processing step, wherein the motion artifact processing model is a machine learning model, and the motion artifact processing model is configured to determine whether motion artifacts exist in the medical image input to the model, and to continue to perform motion artifact correction processing on the medical image input to the model if motion artifacts are determined to exist;

[0014] a secondary processing unit configured to determine, during a repeated medical image processing step, whether a second medical image after correction of the first medical image still has motion artifacts by using a motion artifact processing model; if motion artifacts still exist, triggering re-execution of the medical image acquisition step and the medical image motion artifact processing step until a preset condition configured for the number of acquisitions is met, thereby forming a medical image set consisting of the corrected medical images;

[0015] The display unit is used to display the medical image collection.

[0016] In a third aspect, the present application provides a processing device comprising a processor and a memory, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the method provided in the first aspect of the present application or any possible implementation of the first aspect of the present application is executed.

[0017] In a fourth aspect, the present application provides a computer-readable storage medium, which stores multiple instructions, and the instructions are suitable for a processor to load to execute the method provided in the first aspect of the present application or any possible implementation of the first aspect of the present application.

[0018] From the above content, it can be concluded that this application has the following beneficial effects:

[0019] In order to effectively overcome the adverse effects of motion artifacts, this application starts from the image processing level to create a novel image processing architecture, so that high-quality images can be output and displayed even in the presence of motion artifacts. This can effectively avoid multiple image acquisition operations, thereby improving the detection success rate, and subsequent diagnostic work can also be carried out better, thereby effectively ensuring the quality of medical services through high-quality imaging examinations. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0021] Figure 1 A flowchart of the medical image processing method of this application;

[0022] Figure 2 A schematic diagram of the structure of the medical image processing device of the present application;

[0023] Figure 3 This is a structural diagram of the processing equipment for this application. DETAILED DESCRIPTION

[0024] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0025] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or modules is not necessarily limited to those steps or modules clearly listed, but may include other steps or modules that are not clearly listed or that are inherent to these processes, methods, products or devices. The naming or numbering of steps in this application does not mean that the steps in the method flow must be executed in the time / logical sequence indicated by the naming or numbering. The process steps that have been named or numbered can be changed in the execution order according to the technical purpose to be achieved, as long as the same or similar technical effects can be achieved.

[0026] The division of modules in this application is a logical division. In actual application, there may be other division methods. For example, multiple modules can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection between modules can be electrical or other similar forms, which are not limited in this application. Moreover, the modules or submodules described as separate components may or may not be physically separated, may or may not be physical modules, or may be distributed into multiple circuit modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this application.

[0027] Before introducing the medical image processing method provided by this application, the background content involved in this application is first introduced.

[0028] The medical image processing method, device and computer-readable storage medium provided in this application can be applied to processing equipment to create a novel image processing architecture starting from the image processing level, so that high-quality image output can be displayed even in the presence of motion artifacts. Multiple image acquisition operations can be effectively avoided, thereby improving the detection success rate, and subsequent diagnostic work can also be carried out better, thereby effectively ensuring the quality of medical services through high-quality image examinations.

[0029] The medical image processing method mentioned in this application can be executed by a medical image processing device, or a server, physical host, or user equipment (UE) incorporating the medical image processing device, and other processing devices. The medical image processing device can be implemented in hardware or software, and the UE can be a terminal device such as a smartphone, tablet computer, laptop computer, desktop computer, or personal digital assistant (PDA). The processing device can be configured as a device cluster.

[0030] It can be understood that the focus of the present application is on data processing after the medical images, more specifically magnetic resonance images, are collected. Therefore, for the processing equipment that executes the medical image processing method of the present application or is equipped with the corresponding application service of the medical image processing method of the present application, it is usually only necessary to have the required data processing capabilities, and the specific equipment type and equipment deployment form are relatively flexible.

[0031] That is, the processing device can be connected to an external magnetic resonance imaging system / device to achieve the purpose of acquiring the image to be processed through the external magnetic resonance imaging system.

[0032] If the processing equipment also needs to have direct image acquisition capabilities, it needs to integrate the existing magnetic resonance imaging system itself, which involves software and hardware modifications. Alternatively, the existing magnetic resonance imaging system can be incorporated into the equipment cluster of the processing equipment. The processing equipment specifically consists of two major parts: the magnetic resonance imaging acquisition system and the data processing center.

[0033] In addition, this application also involves the processing link of image display. In this regard, the processing device itself can have a display screen (including a touch screen), so that its own hardware and software conditions have the ability to display images. Alternatively, the processing device can also achieve the purpose of displaying images through an external display screen device or other device with a display screen. This is more flexible to meet the flexible and changeable application needs in actual situations.

[0034] Next, we will introduce the medical image processing method provided by this application.

[0035] First, see Figure 1 , Figure 1 A schematic flow chart of the medical image processing method of the present application is shown. The medical image processing method provided by the present application may specifically include the following steps S101 to S104:

[0036] Step S101, in the medical image acquisition phase, acquiring a first medical image, wherein the first medical image is specifically a magnetic resonance image;

[0037] It is understandable that in response to the demand for high-quality medical image acquisition / display in hospital work, specifically magnetic resonance imaging, the present application solution can start from acquiring the medical images in their initial state.

[0038] Among them, the medical image acquisition link here is usually the direct acquisition and processing of medical images. In addition, in some cases, the possibility of manual input or retrieval of medical images from other devices / systems is not ruled out.

[0039] For the convenience of explanation, medical images at different stages are distinguished by first and second.

[0040] As for the medical images themselves, specifically magnetic resonance images, considering that how to collect them is not the focus of this application and falls within the scope of existing technology, this application will not elaborate on this aspect.

[0041] Step S102: In the medical image motion artifact processing step, motion artifact processing is performed on the first medical image using a pre-configured motion artifact processing model, wherein the motion artifact processing model is a machine learning model, and the motion artifact processing model is used to determine whether motion artifacts exist in the medical image input to the model, and if motion artifacts are determined to exist, continue to perform motion artifact correction processing on the medical image input to the model;

[0042] It can be understood that the goal of the present application is to effectively overcome the adverse effects of motion artifacts. The motion artifacts that may exist in medical images are specifically strip or arc-shaped artifacts distributed along the phase encoding direction caused by autonomous and involuntary movements of the human body or vascular pulsation during the acquisition of magnetic resonance signals. The strength of the artifacts is related to the magnetic field strength, movement amplitude and movement direction. It is an existing concept in itself. Generally speaking, normal physiological movements or involuntary shaking of some patients (such as Parkinson's disease and infants and other special patients) may cause motion artifacts in the image. The presence of motion artifacts will obviously affect the image quality. More motion artifacts will easily lead to examination failure. In addition, it will also affect the quality of the subsequent diagnostic work carried out by relevant doctors / medical detection models based on the examination images.

[0043] In this case, the present application involves the identification and correction of motion artifacts that may exist in medical images. For this processing, artificial intelligence (AI) technology can be introduced, and the processing can be completed by a motion artifact processing model with powerful processing performance configured by the corresponding machine learning method.

[0044] It can be understood that the main ability of the motion artifact processing model is to correct motion artifacts in the image so as to correct the image content of the corresponding position / area to normal image content as much as possible. In addition, it can also be used to identify whether there are motion artifacts in the image. For images without motion artifacts or the degree of motion artifacts (quantified by indicators such as the proportion of artifact image areas in the overall image, the number of motion artifact areas, and the severity of motion artifacts) is lower than the preset degree, no processing is required (corresponding to no output), or a prompt indicating that no processing is required is output, or the original image is directly output. For images with motion artifacts, the next correction processing is triggered.

[0045] Correspondingly, the present application solution may also involve pre-model training processing, and correspondingly, the present application method may also include:

[0046] Acquiring sample medical images, wherein the sample medical images include normal medical images and abnormal medical images with motion artifacts;

[0047] Configure corresponding annotations for sample medical images;

[0048] Train a motion artifact processing model based on labeled sample medical images.

[0049] It is understandable that the sample medical images can be real historical images collected, images obtained by further adjustment based on historical images, or directly generated images, and the data source can be adjusted according to actual conditions.

[0050] The annotations are specifically the motion artifact recognition results and motion artifact correction results, which can be understood as effective model processing results (true values), so that the model training effect can be effectively guided during the model training process.

[0051] As for the specific model training processing steps, there are:

[0052] In each round of model training, a sample medical image is input into the model, allowing the model to identify whether there are motion artifacts, and continue to perform motion artifact correction processing when it is determined that there are motion artifacts to achieve forward propagation. Then, based on the model processing results and combined with the annotations, the loss function is calculated, and the model parameters are optimized according to the loss function calculation results to achieve backward propagation. In this way, when the preset model training requirements such as the number of training times, training time or processing accuracy are met, the model training can be completed, and a motion artifact processing model that can be put into practical use can be obtained.

[0053] It is understandable that for the model architecture adopted by the motion artifact processing model, the model training scheme adopted, and the loss function used in the training process, you can use the existing scheme, or you can make further optimization and improvement schemes based on the existing schemes, or even adopt self-developed novel schemes, and make specific adjustments according to actual needs.

[0054] As an example, the specific model architecture adopted by the motion artifact processing model can be a model such as a deep neural network (DNN), a recurrent neural network (RNN), a long short-term memory network (LSTM), etc.

[0055] In addition, it is worth noting that for the first medical image, it can be either a group of images (multiple, continuous images) or a single image. For the motion artifact processing model, the identification and correction of motion artifacts can be completed under pre-configuration.

[0056] For the case where the first medical image is specifically a group of images, the motion artifact processing model can in some cases be configured so that the model input is a group of images rather than a single image. In this way, motion artifacts can be identified and corrected based on the temporal characteristics of a group of images in an image sequence (with temporal continuity / relationship), which also requires corresponding configuration in the pre-training processing of the model.

[0057] Step S103: In the medical image reprocessing phase, the motion artifact processing model is used to determine whether motion artifacts still exist in the second medical image after the correction of the first medical image. If motion artifacts still exist, the medical image acquisition phase and the medical image motion artifact processing phase are triggered to be re-executed until the preset conditions configured for the number of acquisitions are met, and a medical image set consisting of the corrected medical images is formed;

[0058] It can be understood that in actual applications, there are two situations in the processing results of the motion artifact correction processing applied by the motion artifact processing model. One is that the processed image can meet the requirements, and the other is that the processed image fails to meet the requirements, that is, there is still the influence of motion artifacts or there is obvious influence of motion artifacts. For the latter, the inventors of this application believe that the influence of motion artifacts can be effectively solved or significantly alleviated through two or more motion artifact correction processes.

[0059] To this end, it is necessary to repeat the previous steps S101 and S102, that is, to acquire a new third medical image, identify and correct the motion artifacts through the motion artifact processing model, and then determine whether it is necessary to continue repeating the previous steps S101 and S102. If necessary, acquire a new fourth medical image... The number of reprocessing times needs to be adjusted according to the preset conditions configured for the number of acquisitions. If the preset conditions configured for the number of acquisitions are met, the process can be stopped.

[0060] The preset condition for the number of acquisitions can be directly the number of acquisitions, such as 1 (corresponding to the situation where only the third medical image is acquired), or it can be a condition such as judging that the newly acquired medical image does not have motion artifacts, the corrected medical image does not have motion artifacts, or the image quality of the corrected medical image meets the requirements. In addition, it can also be constrained by combining the upper limit of the number of acquisitions to avoid excessive re-acquisition operations.

[0061] Similarly, when performing motion artifact correction processing (including first and non-first processing) through the motion artifact processing model, it is also possible to introduce a method to determine whether the degree of motion artifact existence (quantified by indicators such as the proportion of artifact image areas in the overall image, the number of motion artifact areas, and the severity of motion artifacts) is higher than a preset degree. If it is higher than the preset degree, it means that the motion artifact image is large and difficult to correct or the correction value is very low. At this time, no correction is required. On the contrary, if it is lower than the preset degree, motion artifact correction processing can be triggered.

[0062] In this way, after re-performing the motion artifact correction process once or multiple times, a large number of images with improved image quality can be generated. These images can be recorded as a medical image set consisting of the corrected medical images.

[0063] Step S104: display the medical image collection.

[0064] It can be understood that after obtaining the set of medical images processed for motion artifacts, they can be displayed to provide medical staff (imaging technicians or diagnostic physicians, etc.) with an effective reference for the existence of motion artifacts in the medical images of the examinee (or patient). Medical staff can better understand / know the physical condition of the examinee corresponding to the position of the motion artifact based on the corrected images, thereby helping to better carry out subsequent disease diagnosis work.

[0065] from Figure 1It can be seen from the illustrated embodiments that in order to effectively overcome the adverse effects of motion artifacts, the present application starts from the image processing level to create a novel image processing architecture, so that high-quality images can be output and displayed even in the presence of motion artifacts. This can effectively avoid multiple image acquisition operations, thereby improving the detection success rate, and subsequent diagnostic work can also be better carried out, thereby effectively ensuring the quality of medical services through high-quality image examinations.

[0066] Continue to the above Figure 1 Each step of the illustrated embodiment and its possible implementation in practical applications are described in detail.

[0067] As an exemplary embodiment, in order to further improve the processing effect of the motion artifact processing model on the motion artifact correction processing of the image, the present application can also introduce filtering processing or noise reduction processing.

[0068] Correspondingly, the motion artifact processing model is also used to perform noise reduction processing on the medical images input into the model.

[0069] It can be understood that by introducing noise reduction processing in the motion artifact recognition link and / or in the motion artifact correction link, the signal-to-noise ratio (SINR) of the image can be effectively improved, which can lay the foundation for more delicate data processing in terms of details, thereby further improving the recognition and correction effects of motion artifacts.

[0070] Among them, for the specific algorithm content involved in noise reduction processing, considering that filtering processing itself is a relatively mature concept, the existing related algorithms can be directly adopted. Of course, in actual situations, further improved and optimized or even self-developed algorithms can also be used here.

[0071] In addition, this application also has further optimized configuration for the motion artifact processing model.

[0072] It can be understood that the motion artifact processing model is a model configured through a machine learning algorithm, and for the identification and correction of motion artifacts, generally speaking, the same algorithm is used by default, that is, within the model, a general algorithm is used to identify and correct slightly different motion artifacts in different subjects. The inventors of this application believe that in the two links of motion artifact identification and motion artifact correction, processing algorithms with subtle differences can be configured for different processing situations, so as to achieve a more adaptive, more delicate and more precise processing effect.

[0073] To this end, it is necessary to pre-plan different categories of motion artifact recognition and motion artifact correction situations, and configure corresponding different motion artifact recognition algorithms and motion artifact correction algorithms.

[0074] As an example, the specific category settings here can be configured from two contents that are easy for staff to intuitively access, namely the image location and the specific body parts involved in the image, or it can be configured from deeper contents such as the image working mode and the characteristics of the examinee himself (including examination posture, disease type, physical deformity, side effects of medication, psychological state, etc.).

[0075] Thus, as an exemplary embodiment, the currently adapted category identifier (which may be the category identifier of the most adapted category or a set of multiple adapted categories) may be selected from the pre-configured type identifiers and assigned to the first medical image, and then the first medical image may be input into the model for processing, or the model may be input for processing along with the first medical image. In this way, the model may select the corresponding target motion artifact recognition algorithm and target motion artifact correction algorithm from the algorithm set according to the category identifier (usually different algorithms have specific model branch structures, and of course there are also cases where the same or overlapping model branch structures are used) to perform more discriminative motion artifact recognition processing and motion artifact correction processing, and rely on the high learning ability of the machine learning algorithm to achieve in-depth and intelligent processing effects that technical personnel cannot achieve in actual situations.

[0076] In addition, the inventors of this application also believe that in addition to selecting the optimal motion artifact recognition algorithm and motion artifact correction algorithm to process the images input by the current model, the present application scheme can also select multiple adaptations or even all algorithms for parallel processing, and the processing results of different algorithms can be fused to obtain the final processing result.

[0077] Taking motion artifact recognition processing as an example, after different motion artifact recognition results are obtained through different algorithms, fusion processing can be performed to obtain the target motion artifact recognition result, which is then used as the input for the next stage of motion artifact correction processing.

[0078] In addition, different motion artifact recognition results can also be directly used as input for the next stage of motion artifact correction processing. When the corresponding different motion artifact correction results are obtained, the final fusion processing is performed to obtain the target motion artifact correction result.

[0079] Furthermore, in terms of details, in addition to direct fusion, a weight mechanism can also be introduced. For the currently adapted category (determined by the type identifier), the processing results of the corresponding algorithm can obtain a higher weight and greater contribution / influence in the fusion process, thereby taking into account the overall and local model processing and achieving delicate processing effects.

[0080] For the above settings, corresponding configurations need to be made during the model's pre-training process.

[0081] At the same time, it can be learned from the previous content that the focus of this application is to further provide medical staff with more effective image references by improving image quality. In this regard, this application can also make further detailed optimizations in the result display link, so as to achieve a better user experience of image reference effects through a more delicate image display mechanism in actual applications.

[0082] Specifically, as an exemplary embodiment, the aforementioned display of a medical image collection may include:

[0083] Based on the image quality scores of different medical images in the medical image collection, a target medical image collection is screened out from the medical image collection;

[0084] Display the target medical image collection.

[0085] It can be understood that for the corrected medical images, the image quality can be quantified by the corresponding image quality score under the pre-configured scoring mechanism, such as the degree of motion artifacts introduced earlier (quantified by indicators such as the proportion of artifact image areas in the overall image, the number of motion artifact areas, and the severity of motion artifacts), and other scoring indicators, which can be flexibly adjusted in actual applications.

[0086] In this way, the image quality scores of different medical images are used as a benchmark, combined with screening conditions such as whether the score is greater than the score threshold and whether the ranking is within the screening number range, to screen out medical images that meet the quality requirements and form a target medical image set.

[0087] At this point, the target medical image set that has been screened, selected, effectively simplified, or has a good correction effect can be displayed through a related visualization interface.

[0088] Thus, in this embodiment, by screening the preliminary medical image set, high-quality images are obtained for display, and image reconstruction is effectively achieved, which can not only achieve a more effective image display effect for medical staff to better view and diagnose, but also further avoid the problem of data redundancy, and has better application value.

[0089] In addition, the medical image set may specifically be an image sequence, that is, the medical image set is composed of multiple images of the nature of an image sequence. In this case, the aforementioned display of the medical image set may specifically include:

[0090] Based on the image quality scores of different medical images in the medical image set, and under a preset medical image sequence adjustment strategy, the sequence order of different medical images in the medical image set is adjusted;

[0091] Display the medical image collection after sequence adjustment.

[0092] It can be understood that for a group of images in an image sequence, the images contained therein usually have a corresponding order when they are displayed. Therefore, under the setting of this embodiment, by further adjusting the sequence order of the images, the difference in image quality scores between the images can be reflected through the order of display, and image reconstruction can be effectively achieved. This allows medical staff who view the images to focus more on the high-quality images on the one hand, and on the other hand, to better capture the physical conditions behind the image areas with motion artifacts based on the changes in image quality, thereby providing more delicate data support.

[0093] Among them, priority is given to sorting in a manner that the highest image quality score is ranked first and the lowest image quality score is ranked last (i.e., the image quality score is ranked from large to small). In addition, sorting can also be carried out in a manner that the lowest image quality score is ranked last and the highest image quality score is ranked first (i.e., the image quality score is ranked from small to large). In some cases or for some medical staff, there may be corresponding application needs.

[0094] It can also be seen from here how the present application solution displays a collection of medical images. The various methods involved in the present application solution (involving the previous, here and later embodiments) can be flexibly adjusted according to the real-time application needs of medical staff.

[0095] In addition, based on the more significant reduction and compression of data volume to more significantly avoid the problem of data redundancy, the present application can also introduce an image fusion mechanism for medical image sets involving multiple images.

[0096] In this regard, as an exemplary embodiment, the aforementioned display of the medical image collection may specifically include:

[0097] With the goal of preserving the image content with no motion artifacts or lower motion artifacts as much as possible in the same image area, the corrected medical images are fused to achieve the goal of updating the representation and content of the medical image collection;

[0098] Display the updated medical image collection.

[0099] Among them, it can be understood that the motion artifacts existing in the images collected in different rounds will have certain differences in levels and positions, and the motion artifacts existing in the images also have this characteristic. In the embodiment here, image fusion usually defaults to fusion into one image, thereby retaining the levels and positions where the motion artifacts in multiple acquisition processes are less affected, thereby achieving the purpose of maximizing the possible compression of the data volume involved in the image and integrating all high-quality image content. In addition, when the number of images involved is large and the amount of data is large, the different images involved in the medical image set can be fused into multiple images (the images before fusion may overlap or be repeated), which is also possible in practical applications.

[0100] At the same time, in response to clinical imaging application requirements including display, the processed medical image set can be processed in the form of a report.

[0101] In this regard, a corresponding motion artifact report can be generated based on the medical image collection for display purposes.

[0102] The report usually also includes the corresponding serial number (unique identifier), brief information about the examinee, image acquisition time, image acquisition unit, etc.

[0103] In addition, although the motion artifact report is similar to a medical image collection and is linked to motion artifacts, it can also be stored, displayed, etc. together with normal images and normal image reports, and can be distinguished by identification and other means during processing.

[0104] The motion artifact report configured directly for motion artifacts helps medical staff to directly understand the specific situation of motion artifacts that exist during the examination process and promote subsequent diagnosis work.

[0105] At the same time, in order to further enhance the reporting performance of the motion artifact report or expand the report content, as an exemplary embodiment, the method of the present application may further include:

[0106] Extracting the recorded position of the lesion tissue corresponding to the subject from the system;

[0107] Based on the medical image collection, the content of motion artifact location, diseased tissue location and normal tissue location is combined to generate a corresponding motion artifact report;

[0108] The motion artifact report is displayed.

[0109] It can be understood that the present application also introduces a prompt mechanism for the location of the diseased tissue of the examinee in other examination items or manually entered by medical staff. In this way, the descriptive information of the diseased tissue location and the normal tissue location is added to the report, or the diseased tissue location and the normal tissue location are integrated into the image in the form of labels for more intuitive identification and distinction. This can form a more vivid and vivid report content presentation effect, thereby playing a more intuitive data support role. Medical staff who assist in viewing can more conveniently understand the physical condition of the examinee behind the motion artifact location, and make corresponding grasps and decisions faster and more accurately.

[0110] Furthermore, to continue to enhance the user experience, the report can also provide more intelligent identification suggestions based on the static positions of motion artifacts, diseased tissues and normal tissues, to assist medical staff who view the report. They can use the identification suggestions as a reference to more quickly and accurately grasp and make decisions about the content presented in the report, especially the physical condition of the examinee corresponding to the motion artifact position.

[0111] In this regard, as an exemplary embodiment, based on the medical image set, the motion artifact location, the diseased tissue location, and the normal tissue location are combined to generate a corresponding motion artifact report, which may specifically include:

[0112] Generate identification suggestions for the positional relationships among motion artifacts, diseased tissue, and normal tissue based on a collection of medical images;

[0113] Based on the medical image collection, the corresponding motion artifact report is generated by combining the content of motion artifact location, diseased tissue location and normal tissue location, as well as the identification suggestions based on the positional relationship between the three.

[0114] The identification suggestions mentioned here are intelligent suggestions based on medical staff's habit of observing motion artifacts in image content / reports. For example, they can point out that a certain location has been significantly improved or even has no motion artifacts after one or more corrections. Another example is that they can point out that a certain location has undergone multiple corrections. Another example is that they can point out the fuzzy area between motion artifacts, diseased tissue, and normal tissue at a certain location. Another example is that they can point out the probability that a certain location is a motion artifact, diseased tissue, or normal tissue. Another example is that they can point out areas that require extra attention or are easily overlooked based on the current medical staff's viewing habits. Another example is that they can point out areas that require extra attention or are easily overlooked in the current patient.

[0115] Obviously, in addition to assisting in more convenient and accurate judgment of the position of motion artifacts and the positional relationship between other tissues, it can also further involve the individual situation of the current examinee or the current medical staff (additional collection / extraction of descriptive data of the corresponding individual situation is required) to configure more detailed identification suggestions. Therefore, through the setting of more intelligent identification suggestions, the motion artifact report can have a better and more delicate user experience, and it can also help medical staff to capture special situations that are difficult or even impossible to notice when manually reviewing the report, thereby promoting deeper and more accurate decision-making, which is completely impossible with conventional imaging report settings.

[0116] In addition, the above embodiments are explained from the perspective of image display, but in some cases, these processes can also be separated from the image display processing, that is, before executing the step of displaying the corresponding image, the step of performing corresponding processing on the medical image set is introduced. It can be processed by the motion artifact processing model or by the configured corresponding processing algorithm. Considering that this is relatively easy to understand, no further elaboration will be given.

[0117] The above is an introduction to the medical image processing method provided by this application. To facilitate better implementation of the medical image processing method provided by this application, this application also provides a medical image processing device from the perspective of functional modules.

[0118] See Figure 2 , Figure 2 This is a schematic diagram of the structure of the medical image processing device of the present application. In the present application, the medical image processing device 200 may specifically include the following structure:

[0119] The acquisition unit 201 is configured to acquire a first medical image in a medical image acquisition step, wherein the first medical image is specifically a magnetic resonance image;

[0120] a primary processing unit 202 configured to perform motion artifact processing on a first medical image using a preconfigured motion artifact processing model in a medical image motion artifact processing step, wherein the motion artifact processing model is a machine learning model configured to determine whether motion artifacts exist in the medical image input to the model, and to continue to perform motion artifact correction processing on the medical image input to the model if motion artifacts are determined to exist;

[0121] The secondary processing unit 203 is configured to determine, during the repeated medical image processing phase, whether motion artifacts still exist in the second medical image after the correction of the first medical image using the motion artifact processing model. If motion artifacts still exist, the secondary processing unit 203 triggers the re-execution of the medical image acquisition phase and the medical image motion artifact processing phase until a preset condition configured for the number of acquisitions is met, thereby forming a medical image set consisting of the corrected medical images involved.

[0122] The display unit 204 is used to display the medical image set.

[0123] In an exemplary embodiment, the motion artifact processing model is further used to perform noise reduction processing on a specified noise signal of a medical image input to the model.

[0124] In another exemplary embodiment, the display unit 204 is specifically configured to:

[0125] Based on the image quality scores of different medical images in the medical image collection, a target medical image collection is screened out from the medical image collection;

[0126] Display the target medical image collection.

[0127] In another exemplary embodiment, the display unit 204 is specifically configured to:

[0128] Based on the image quality scores of different medical images in the medical image set, and under a preset medical image sequence adjustment strategy, the sequence order of different medical images in the medical image set is adjusted;

[0129] Display the medical image collection after sequence adjustment.

[0130] In another exemplary embodiment, the display unit 204 is specifically configured to:

[0131] With the goal of preserving the image content with no motion artifacts or lower motion artifacts as much as possible in the same image area, the corrected medical images are fused to achieve the goal of updating the representation and content of the medical image collection;

[0132] Display the updated medical image collection.

[0133] In another exemplary embodiment, the display unit 204 is further configured to:

[0134] Extracting the recorded position of the lesion tissue corresponding to the subject from the system;

[0135] Based on the medical image collection, the content of motion artifact location, diseased tissue location and normal tissue location is combined to generate a corresponding motion artifact report;

[0136] The motion artifact report is displayed.

[0137] In another exemplary embodiment, the display unit 204 is further configured to:

[0138] Generate identification suggestions for the positional relationships among motion artifacts, diseased tissue, and normal tissue based on a collection of medical images;

[0139] Based on the medical image collection, the corresponding motion artifact report is generated by combining the content of motion artifact location, diseased tissue location and normal tissue location, as well as the identification suggestions based on the positional relationship between the three.

[0140] This application also provides a processing device from the perspective of hardware structure, which can be either a single device or a device cluster. For the convenience of explanation, the devices that may be involved in different situations are described as a single processing device. Figure 3 , Figure 3 The schematic diagram of the structure of the processing device of the present application is shown. Specifically, the processing device of the present application may include a processor 301, a memory 302 and an input / output device 303. The processor 301 is used to execute the computer program stored in the memory 302 to implement the following Figure 1 Each step of the medical image processing method in the corresponding embodiment; or, when the processor 301 is used to execute the computer program stored in the memory 302, the following is implemented Figure 2 The memory 302 is used to store the functions of each unit in the embodiment corresponding to the processor 301. Figure 1 The computer program required by the medical image processing method in the corresponding embodiment.

[0141] For example, the computer program may be divided into one or more modules / units, one or more of which are stored in the memory 302 and executed by the processor 301 to complete the present application. One or more modules / units may be a series of computer program instruction segments capable of performing specific functions, and the instruction segments are used to describe the execution process of the computer program in a computer device.

[0142] The processing device may include, but is not limited to, a processor 301, a memory 302, and an input / output device 303. Those skilled in the art will appreciate that the illustrations are merely examples of processing devices and do not limit the processing device. The processing device may include more or fewer components than shown, or a combination of certain components, or different components. For example, the processing device may also include a network access device, a bus, etc., and the processor 301, the memory 302, the input / output device 303, etc. are connected via a bus.

[0143] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor. The processor is the control center of the processing device and connects various parts of the entire device using various interfaces and lines.

[0144] The memory 302 can be used to store computer programs and / or modules. The processor 301 implements various functions of the computer device by running or executing the computer programs and / or modules stored in the memory 302 and accessing the data stored in the memory 302. The memory 302 may mainly include a program storage area and a data storage area. The program storage area may store an operating system, at least one application required for a function, etc.; the data storage area may store data created based on the use of the processing device, etc. In addition, the memory may include high-speed random access memory and non-volatile memory, such as a hard disk, internal memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0145] When the processor 301 is used to execute the computer program stored in the memory 302, it can specifically implement the following functions:

[0146] In the medical image acquisition step, a first medical image is acquired, wherein the first medical image is specifically a magnetic resonance image;

[0147] In the medical image motion artifact processing step, motion artifact processing is performed on the first medical image using a preconfigured motion artifact processing model, wherein the motion artifact processing model is a machine learning model, and the motion artifact processing model is used to determine whether motion artifacts exist in the medical image input to the model, and if it is determined that motion artifacts exist, continue to perform motion artifact correction processing on the medical image input to the model;

[0148] In the medical image repeated processing phase, a motion artifact processing model is used to determine whether a second medical image after the correction of the first medical image still has motion artifacts. If motion artifacts still exist, the medical image acquisition phase and the medical image motion artifact processing phase are triggered to be re-executed until the preset conditions configured for the number of acquisitions are met, and a medical image set consisting of the corrected medical images is formed;

[0149] Display medical image collections.

[0150] Those skilled in the art will clearly understand that for the convenience and brevity of description, the specific working processes of the medical image processing apparatus, processing equipment and corresponding units described above can refer to the following. Figure 1 The description of the medical image processing method in the corresponding embodiment will not be repeated here.

[0151] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be accomplished by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0152] To this end, the present application provides a computer-readable storage medium, which stores a plurality of instructions, which can be loaded by a processor to execute the present application as follows: Figure 1 For the steps of the medical image processing method in the corresponding embodiment, the specific operations can be referred to as follows Figure 1 The description of the medical image processing method in the corresponding embodiment will not be repeated here.

[0153] The computer-readable storage medium may include a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0154] Due to the instructions stored in the computer readable storage medium, the present application can be executed as follows: Figure 1 Corresponding to the steps of the medical image processing method in the embodiment, the present application can be implemented as follows Figure 1The beneficial effects that can be achieved by the medical image processing method in the corresponding embodiment are detailed in the previous description and will not be repeated here.

[0155] The above is a detailed introduction to the medical image processing method, apparatus, processing equipment and computer-readable storage medium provided by the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the core idea of ​​the present application. At the same time, for those skilled in the art, based on the idea of ​​the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A medical image processing method, characterized in that: The method comprises: In the medical image acquisition step, a first medical image is acquired, wherein the first medical image is specifically a magnetic resonance image; In the medical image motion artifact processing step, motion artifact processing is performed on the first medical image using a preconfigured motion artifact processing model, wherein the motion artifact processing model is a machine learning model, and the motion artifact processing model is used to determine whether motion artifacts exist in the medical image input to the model, and if it is determined that motion artifacts exist, continue to perform motion artifact correction processing on the medical image input to the model; In the medical image reprocessing step, the motion artifact processing model is used to determine whether the second medical image after the correction of the first medical image still has the motion artifact. If the motion artifact still exists, the medical image acquisition step and the medical image motion artifact processing step are triggered to be re-executed until the preset conditions configured for the number of acquisitions are met, and a medical image set consisting of the corrected medical images is formed; The medical image collection is displayed.

2. The method according to claim 1, characterized in that The motion artifact processing model is also used to perform noise reduction processing on a specified noise signal of a medical image input to the model.

3. The method according to claim 1, characterized in that The displaying of the medical image collection includes: screening a target medical image set from the medical image set based on image quality scores of different medical images in the medical image set; The target medical image set is displayed.

4. The method according to claim 1, wherein The medical image set is specifically an image sequence, and displaying the medical image set includes: Based on the image quality scores of different medical images in the medical image set, and under a preset medical image sequence adjustment strategy, adjusting the sequence order of different medical images in the medical image set; The medical image set after sequence adjustment is displayed.

5. The method according to claim 1, characterized in that Displaying the medical image collection includes: With the goal of preserving image content with no motion artifacts or a lower degree of motion artifacts in the same image region as much as possible, the corrected medical images are fused to achieve the goal of updating the representation and content of the medical image set; The updated medical image collection is displayed.

6. The method according to claim 1, characterized in that The method further comprises: Extracting the recorded position of the lesion tissue corresponding to the subject from the system; Based on the medical image set, combining the content of motion artifact location, diseased tissue location and normal tissue location, and generating a corresponding motion artifact report; The motion artifact report is displayed.

7. The method according to claim 6, characterized in that Based on the medical image set, the content of the motion artifact position, the diseased tissue position and the normal tissue position is combined to generate a corresponding motion artifact report, including: Based on the medical image set, generating an identification suggestion of the positional relationship between the motion artifact position, the diseased tissue position, and the normal tissue position; Based on the medical image set, the corresponding motion artifact report is generated by combining the content of the motion artifact position, the diseased tissue position and the normal tissue position, and combining the identification suggestions of the positional relationship between the three.

8. A medical image processing device, characterized in that: The device comprises: an acquisition unit, configured to acquire a first medical image in a medical image acquisition step, wherein the first medical image is specifically a magnetic resonance image; a primary processing unit, configured to perform motion artifact processing on the first medical image using a preconfigured motion artifact processing model in a medical image motion artifact processing link, wherein the motion artifact processing model is a machine learning model, and the motion artifact processing model is configured to determine whether motion artifacts exist in the medical image input to the model, and to continue to perform motion artifact correction processing on the medical image input to the model if motion artifacts are determined to exist; a secondary processing unit configured to determine, during a repeated medical image processing step, whether the motion artifact still exists in a second medical image after the correction of the first medical image by using the motion artifact processing model; if the motion artifact still exists, triggering re-execution of the medical image acquisition step and the medical image motion artifact processing step until a preset condition configured for the number of acquisitions is met, thereby forming a medical image set consisting of the corrected medical images; A display unit is used to display the medical image set.

9. A processing device, characterized in that: The method comprises a processor and a memory, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the method according to any one of claims 1 to 7 is executed.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the method according to any one of claims 1 to 7.

Images