Method, system and device for determining the shooting part of medical images

Abstract

The embodiments of this specification provide a method, system, and apparatus for determining the imaging site of a medical image. The method comprises: obtaining a medical image and a site label of the medical image; determining a first candidate site of the medical image based on the site label; processing the medical image using a site classification model to obtain a site classification result; and determining the imaging site corresponding to the medical image based on the site classification result and the first candidate site.

Description

Technical Field

[0001] This specification relates to the field of medical devices, and in particular to a method, system, and device for determining the shooting location of a medical image. Background Art

[0002] When scanning and imaging with medical imaging equipment, a large number of images of varying types are typically captured. For example, X-ray imaging, a common medical examination method, can capture images of various body parts, and both body part and body position information are included in the captured medical images. However, due to issues with the equipment and user operation, the captured medical images are prone to missing or inconsistent information about the captured body part and body position.

[0003] Therefore, it is hoped to provide a method for determining the shooting site of a medical image, which can help screen out medical images that are consistent with the actual shooting site and improve screening efficiency. Summary of the Invention

[0004] In one aspect, the present disclosure provides a method for determining a medical image's imaging site. The method comprises: obtaining a medical image and a site label for the medical image; determining a first candidate site for the medical image based on the site label; processing the medical image using a site classification model to obtain a site classification result; and determining the imaging site corresponding to the medical image based on the site classification result and the first candidate site.

[0005] In some embodiments, the part classification result includes at least two candidate parts and a probability value corresponding to each candidate part, and determining the shooting part corresponding to the medical image based on the part classification result and the first candidate part further includes: determining a first probability value corresponding to the first candidate part based on the part classification result; determining the maximum probability value and the corresponding candidate part in the part classification result, determining the corresponding candidate part as the second candidate part, and determining the maximum probability value as the second probability value corresponding to the second candidate part; and determining the shooting part corresponding to the medical image based on the first probability value and the second probability value.

[0006] In some embodiments, determining the shooting part corresponding to the medical image based on the first probability value and the second probability value further includes: determining the ratio of the first probability value and the second probability value; comparing the ratio with a preset threshold to generate a comparison result; and based on the comparison result, selecting one from the first candidate part and the second candidate part as the shooting part of the medical image.

[0007] In some embodiments, the method further includes: obtaining a reference medical image and its reference part and true part; determining a first accuracy of the reference part and a second accuracy of the part classification model based on the true part of the reference medical image; and determining the preset threshold based on the first accuracy and the second accuracy.

[0008] In some embodiments, the method further includes: obtaining a user's historical film reading records; and determining the preset threshold based on the user's historical film reading records.

[0009] In some embodiments, the method further includes: acquiring an image to be processed; and pre-processing the image to be processed based on the part label to obtain the medical image.

[0010] In some embodiments, the method further includes: determining whether the medical image satisfies a preset condition based on a shooting site of the medical image; and designating the medical image as a target medical image in response to the medical image satisfying the preset condition.

[0011] In some embodiments, the method further includes: generating training samples based on the medical image and the photographed part thereof; and training an image processing model using the training samples.

[0012] Another aspect of this specification provides a system for determining the imaging site of a medical image. The system includes: an acquisition module for acquiring a medical image and a site label of the medical image; a first determination module for determining a first candidate site of the medical image based on the site label; a classification module for processing the medical image using a site classification model to obtain a site classification result; and a second determination module for determining the imaging site corresponding to the medical image based on the site classification result and the first candidate site.

[0013] Another aspect of the present specification provides an apparatus for determining a shooting part of a medical image, comprising at least one processor and at least one storage device, wherein the storage device is configured to store instructions. When the at least one processor executes the instructions, the aforementioned method for determining a shooting part is implemented.

[0014] Another aspect of the present specification provides a computer-readable storage medium, wherein the storage medium stores computer instructions. When a computer reads the computer instructions in the storage medium, the computer executes the above-mentioned method for determining the shooting part.

[0015] The imaging part determination method and / or system provided in embodiments of this specification determines a first candidate part based on the part label of a medical image, processes the medical image using a part classification model to obtain a part classification result, and then determines the corresponding imaging part of the medical image based on the part classification result and the first candidate part. This method / system fully utilizes the imaging part information and image information of the medical image, balancing the medical image part information with the image part classification result, thereby improving the screening success rate of medical images. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] This specification will be further described in the form of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting, and in these embodiments, like numbers represent like structures, wherein:

[0017] Figure 1 is a schematic diagram of an application scenario of an exemplary medical image capturing part determination system according to some embodiments of this specification;

[0018] Figure 2 is a module diagram of an exemplary medical imaging part determination system according to some embodiments of this specification;

[0019] Figure 3 is a flowchart of an exemplary method for determining a medical imaging location according to some embodiments of this specification;

[0020] Figure 4 is a schematic diagram of an exemplary method for determining a medical imaging location according to some embodiments of this specification;

[0021] Figure 5 is a flowchart of an exemplary determination of a preset threshold according to some embodiments of this specification;

[0022] Figure 6 It is an information diagram of exemplary medical images shown in some embodiments of this specification. DETAILED DESCRIPTION

[0023] To more clearly illustrate the technical solutions of the embodiments of this specification, the following briefly describes the drawings required for describing the embodiments. Obviously, the drawings described below are merely examples or embodiments of this specification. Those skilled in the art can apply this specification to other similar scenarios based on these drawings without inventive effort. Unless otherwise apparent from the context or otherwise noted, the same reference numerals in the figures represent the same structure or operation.

[0024] It should be understood that the terms "system," "device," "unit," and / or "module" used herein are a method for distinguishing different components, elements, parts, portions, or assemblies at different levels. However, if other terms can achieve the same purpose, the terms may be replaced by other expressions.

[0025] As used in this specification and claims, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not refer to the singular but also include the plural. Generally speaking, the terms "comprises" and "include" only indicate the inclusion of the steps and elements specifically identified, and these steps and elements do not constitute an exclusive list. A method or apparatus may also include other steps or elements.

[0026] Although this specification makes various references to certain modules or units in the system according to embodiments of the specification, any number of different modules or units can be used and run on the client and / or server. The modules are illustrative only, and different aspects of the system and method can use different modules.

[0027] Flowcharts are used throughout this specification to illustrate the operations performed by systems according to embodiments of this specification. It should be understood that preceding or following operations do not necessarily need to be performed in exact order. Instead, the steps may be processed in reverse order or simultaneously. Furthermore, other operations may be added to these processes, or one or more operations may be removed from these processes.

[0028] Medical images can be divided according to the parts of the body that were photographed, such as the hands and knee joints, and the part and position information are all present in the image information. However, due to equipment (e.g., X-Ray equipment) reasons, technician operation reasons, etc., the image information of medical images often contains missing information about the part of the body and the position of the body, or the information is inconsistent with the actual part of the body / position of the body. With the application of artificial intelligence in the field of medical imaging (e.g., X-Ray imaging) (e.g., XR-assisted diagnosis, intelligent quantitative evaluation of X-Ray image shooting quality, etc.), screening out medical images that are consistent with the actual part of the body can not only improve the user experience and reduce the occurrence of abnormal situations, but also has important significance for these artificial intelligence applications.

[0029] Generally, medical images may include basic information and image information. Basic information includes the scanned part, and image information is a digital image. In some embodiments, the image capture part can be determined based on the scanned part in the basic information of the medical image. However, due to factors such as equipment and shooting reasons, it is easy for the part and body position to be missing or inconsistent with the actual shooting. That is, the scanned part information carried by the medical image contains errors, which can easily lead to inaccurate determination of the captured part and image screening errors. In some embodiments, the captured part can be determined based on the image information carried by the medical image using a deep learning image classification method. However, the deep learning-based image classification method completely relies on the image information of the medical image. Once the classification result is wrong, it will lead to an incorrect screening result.

[0030] In an embodiment of the present specification, a method for determining the shooting part of a medical image is provided. A first candidate part is determined by combining the part label carried by the medical image, a part classification result is determined based on a part classification model, and the shooting part corresponding to the medical image is determined based on the part classification result and the first candidate part. This method can fully utilize the scanning part information and image information of the medical image, weigh the scanning part information of the medical image and the part classification result of the image, thereby improving the screening success rate of the medical image.

[0031] Figure 1 This is a schematic diagram of an application scenario of an exemplary medical image shooting part determination system according to some embodiments of this specification.

[0032] like Figure 1 As shown, the imaging part determination system 100 may include a medical device 110, a network 120, a terminal device 130, a processing device 140, and a storage device 150. The various components in the imaging part determination system 100 may be interconnected via the network 120. For example, the medical device 110 and the terminal device 130 may be connected or communicated with each other via the network 120. For another example, the medical device 110 and the processing device 140 may be connected or communicated with each other via the network 120.

[0033] The medical device 110 can be used to scan a scanned object within a detection area or a scanning area to obtain scan data of the scanned object. In some embodiments, the scanned object can include a biological object and / or a non-biological object. For example, the scanned object can be a living or non-living organic and / or inorganic substance.

[0034] In some embodiments, the medical device 110 can be a non-invasive imaging device for disease diagnosis or research purposes. For example, the medical device 110 can include a single-modality scanner and / or a multi-modality scanner. A single-modality scanner can include, for example, an ultrasound scanner, an X-ray scanner (X-Ray scanner), a computed tomography (CT) scanner, a magnetic resonance imaging (MRI) scanner, an ultrasound scanner, a positron emission tomography (PET) scanner, an optical coherence tomography (OCT) scanner, an ultrasound (US) scanner, an intravascular ultrasound (IVUS) scanner, a near-infrared spectroscopy (NIRS) scanner, a far-infrared (FIR) scanner, or the like, or any combination thereof. A multi-modality scanner can include, for example, an X-ray imaging-magnetic resonance imaging (X-ray-MRI) scanner, a positron emission tomography-X-ray imaging (PET-X-ray) scanner, a single-photon emission computed tomography-magnetic resonance imaging (SPECT-MRI) scanner, a positron emission tomography-computed tomography (PET-CT) scanner, a digital subtraction angiography-magnetic resonance imaging (DSA-MRI) scanner, or the like. The scanners provided above are for illustrative purposes only and are not intended to limit the scope of this description.

[0035] The network 120 may include any suitable network capable of facilitating information and / or data exchange of the imaging site determination system 100. In some embodiments, at least one component of the imaging site determination system 100 (e.g., the medical device 110, the terminal device 130, the processing device 140, the storage device 150) may exchange information and / or data with at least one other component of the imaging site determination system 100 via the network 120. For example, the processing device 140 may obtain the medical image of the scanned object and its site label from the medical device 110 via the network 120. The network 120 may include a public network (e.g., the Internet), a private network (e.g., a local area network (LAN)), a wired network, a wireless network (e.g., an 802.11 network, a Wi-Fi network), a frame relay network, a virtual private network (VPN), a satellite network, a telephone network, a router, a hub, a switch, a fiber optic network, a telecommunication network, an intranet, a wireless local area network (WLAN), a metropolitan area network (MAN), a public switched telephone network (PSTN), a Bluetooth TM Network, ZigBee TM In some embodiments, the network 120 may include at least one network access point. For example, the network 120 may include a wired and / or wireless network access point, such as a base station and / or an Internet exchange point. At least one component of the imaging part determination system 100 may connect to the network 120 via the access point to exchange data and / or information.

[0036] The terminal device 130 can communicate and / or connect with the medical device 110, the processing device 140, and / or the storage device 150. For example, a user can interact with the medical device 110 through the terminal device 130 to control one or more components of the medical device 110. In some embodiments, the terminal device 130 may include a mobile device 131, a tablet computer 132, a laptop computer 133, or any combination thereof. For example, the mobile device 131 may include a mobile control handle, a personal digital assistant (PDA), a smartphone, or any combination thereof.

[0037] The processing device 140 can process data and / or information obtained from the medical device 110, the at least one terminal device 130, the storage device 150, or other components of the imaging site determination system 100. For example, the processing device 140 can obtain medical images and their site labels from the medical device 110 and analyze and process them to determine the imaging site corresponding to the medical image. For another example, the processing device 140 can obtain reference medical images, their reference sites, and the actual sites from the storage device 150 to determine a preset threshold. In some embodiments, the processing device 140 can be a single server or a server group. The server group can be centralized or distributed. In some embodiments, the processing device 140 can be local or remote. For example, the processing device 140 can access information and / or data from the medical device 110, the at least one terminal device 130, and / or the storage device 150 via the network 120. For another example, the processing device 140 can directly connect to the medical device 110, the at least one terminal device 130, and / or the storage device 150 to access information and / or data. In some embodiments, the processing device 140 can be implemented on a cloud platform. For example, a cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an inter-cloud cloud, a multi-cloud, etc., or any combination thereof.

[0038] In some embodiments, the processing device 140 may include one or more processors (e.g., a single-chip processor or a multi-chip processor). By way of example only, the processing device 140 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), an application-specific instruction set processor (ASIP), a graphics processing unit (GPU), a physical processing unit (PPU), a digital signal processor (DSP), a field-programmable gate array (FPGA), a programmable logic device (PLD), a controller, a microcontroller unit, a reduced instruction set computer (RISC), a microprocessor, or any combination thereof. In some embodiments, the processing device 140 may be part of the medical device 110 or the terminal device 130. For example, the processing device 140 may be integrated into the medical device 110 to determine a preset threshold, the location of the medical image, and the like.

[0039] The storage device 150 can store data, instructions and / or any other information. For example, the storage device 150 can store scanned images and related information obtained by the medical device 110. In some embodiments, the storage device 150 can store data obtained from the medical device 110, at least one terminal device 130 and / or the processing device 140, such as scanned images, basic information, patient information and other data. In some embodiments, the storage device 150 can store data and / or instructions used by the processing device 140 to execute or use to complete the exemplary methods described in this specification. In some embodiments, the storage device 150 may include a large-capacity memory, a removable memory, a volatile read-write memory, a read-only memory (ROM), etc., or any combination thereof. In some embodiments, the storage device 150 can be implemented on a cloud platform.

[0040] In some embodiments, the storage device 150 may be connected to the network 120 to communicate with at least one other component in the imaging site determination system 100 (e.g., the medical device 110, at least one terminal device 130, and the processing device 140). At least one component in the imaging site determination system 100 may access data stored in the storage device 150 (e.g., scanned images of the scanned object, information data, etc.) via the network 120. In some embodiments, the storage device 150 may be part of the processing device 140.

[0041] It should be noted that the above description is provided for illustrative purposes only and is not intended to limit the scope of this specification. For those of ordinary skill in the art, under the guidance of the contents of this specification, various changes and modifications can be made. The features, structures, methods and other features of the exemplary embodiments described in this specification can be combined in various ways to obtain additional and / or alternative exemplary embodiments. For example, the storage device 150 can be a data storage device including a cloud computing platform (such as a public cloud, a private cloud, a community and a hybrid cloud, etc.). However, these changes and modifications will not deviate from the scope of this specification.

[0042] Figure 2 It is a module diagram of an exemplary medical image capturing part determination system according to some embodiments of this specification.

[0043] like Figure 2 As shown, in some embodiments, the imaging part determination system 200 may include an acquisition module 210, a first determination module 220, a classification module 230, and a second determination module 240. These modules may also be implemented as an application or a set of instructions read and executed by a processing engine. Furthermore, a module may be any combination of hardware circuitry and an application or set of instructions. For example, when a processing engine or processor executes an application or a set of instructions, the module may be part of the processing device 140.

[0044] The acquisition module 210 can be used to acquire medical images and part labels of the medical images. In some embodiments, the acquisition module 210 can be used to acquire images to be processed, pre-process the images to be processed based on the part labels, and obtain medical images.

[0045] In some embodiments, the acquisition module 210 may be used to acquire a reference medical image and its reference part and real part. In some embodiments, the acquisition module 210 may be used to acquire a user's historical film reading records.

[0046] The first determination module 220 can be used to determine a first candidate part of the medical image. In some embodiments, the first determination module 220 can determine the first candidate part based on a part label of the medical image.

[0047] The classification module 230 can be used to process the medical image using the part classification model to obtain a part classification result. In some embodiments, the part classification result can include at least two candidate parts and a probability value corresponding to each candidate part.

[0048] The second determination module 240 can be used to determine the imaging part corresponding to the medical image. In some embodiments, the second determination module 240 can be used to determine the imaging part corresponding to the medical image based on the part classification result and the first candidate part. In some embodiments, the second determination module 240 can be used to determine a first probability value corresponding to the first candidate part based on the part classification result. In some embodiments, the second determination module 240 can be used to determine the maximum probability value and the corresponding candidate part in the part classification result, determine the candidate part as the second candidate part, and determine the maximum probability value as the second probability value corresponding to the second candidate part. Further, the second determination module 240 can determine the imaging part corresponding to the medical image based on the first probability value and the second probability value.

[0049] In some embodiments, the second determination module 240 can be used to determine the ratio of the first probability value and the second probability value, compare the ratio with a preset threshold, and generate a comparison result; based on the comparison result, select one from the first candidate part and the second candidate part as the shooting part of the medical image.

[0050] In some embodiments, the second determination module 240 may be configured to determine a first accuracy of the reference part and a second accuracy of the part classification model based on the actual part of the reference medical image, and to determine a preset threshold based on the first accuracy and the second accuracy. In some embodiments, the second determination module 240 may be configured to determine the preset threshold based on the user's historical reading history.

[0051] In some embodiments, the imaging part determination system 200 may further include a screening module (not shown in the figure). The screening module may be configured to determine whether the medical image satisfies a preset condition based on the imaging part of the medical image, and designate the medical image as a target medical image in response to the medical image satisfying the preset condition.

[0052] In some embodiments, the imaging part determination system 200 may further include a training module (not shown). The training module may be used to generate training samples based on the medical images and the imaging parts thereof, and use the training samples to train the image processing model. In some embodiments, the training module may also be used to train a part classification model.

[0053] It should be noted that the above description of the imaging location determination system 200 and its modules is for convenience only and does not limit this specification to the illustrated embodiments. It is understood that, after understanding the principles of the system, those skilled in the art may arbitrarily combine the modules or form subsystems connected to other modules without departing from these principles. For example, in some embodiments, the acquisition module 210, the first determination module 220, the classification module 230, and the second determination module 240 may be different modules within a single system, or a single module may implement the functions of two or more of the aforementioned modules. For another example, in some embodiments, the acquisition module 210, the first determination module 220, the classification module 230, and the second determination module 240 may share a single storage module, or each module may have its own storage module. Such variations are within the scope of this specification.

[0054] Figure 3 is a flowchart of an exemplary method for determining a medical image capturing part according to some embodiments of this specification.

[0055] In some embodiments, the process 300 may be executed by a processor (eg, the processing device 140) or the imaging part determination system 200. The process 300 includes:

[0056] Step 310 , obtaining a medical image and its part labels. In some embodiments, step 310 may be performed by the processing device 140 or the acquisition module 210 .

[0057] In some embodiments, the medical image may include scanned image information (image information) and basic information. In some embodiments, the image information and / or basic information may be stored via a DICOM data structure.

[0058] Image information (imaging information) can refer to digital images. For example, a medical image / video of a scanned object can be obtained by scanning the object using medical device 110. In some embodiments, the digital image can include information about the body part and body position of the scanned object. For example, the body part can include the hand, abdomen, knee joint, head, neck, etc.; the body position can include lying flat, lying on the side, lying prone, etc. In some embodiments, the digital image can include 2D images, 3D images, 4D images, etc.

[0059] The basic information may reflect the diagnostic data of the scanned object. In some embodiments, the basic information may include at least one data field related to medicine, and each field contains information on a specific topic. For example, Figure 6 As shown in , the basic information of the DICOM data structure may include a tag field, a name field, a module field, a keyword field, a scanning part (photographing part) field, etc., and the column corresponding to each field contains corresponding descriptive information. Among them, a tag may refer to a label used to identify data, and each tag is a unique identifier for each data item. The name may reflect the description of the terminology involved in the diagnosis, such as end-to-end length, examination time, image type, etc. The module may be a component related to ultrasound and medical imaging communication, such as an image module, a study module, a series module, etc. The scanning part may reflect the target part of the patient that needs to be scanned. For example, if it is planned to scan the patient's knee joint, the field corresponding to the scanning part in the basic information will be recorded as "knee joint".

[0060] In some embodiments of the present specification, acquiring a medical image may refer to acquiring a digital image in the medical image, and acquiring a part tag of the medical image may refer to acquiring a tag corresponding to the digital image in the basic information.

[0061] In some embodiments, the captured part in the digital image and the scanned part in the basic information corresponding to the same medical image may be the same, different, or partially the same. For example, due to incorrect patient positioning by medical staff, incorrect operation of the equipment, or problems with the equipment itself, the obtained digital image may only capture part of the target part (for example, the basic information corresponds to the scanned part being the hand, but the digital image only captures half of the palm), or the scanned part recorded in the basic information may not be captured at all (for example, the basic information corresponds to the scanned part being the hand, but the digital image captures the scanned part being the arm).

[0062] In some embodiments, the medical image and the location label of the medical image can be obtained from a medical device (e.g., medical device 110). In some embodiments, the medical image and the location label of the medical image can be obtained from a storage device (e.g., storage device 150). In some embodiments, the medical image and the location label of the medical image can be obtained from a medical system such as an ultrasound system or a Picture Archiving and Communication System (PACS).

[0063] In some embodiments, an image to be processed may be acquired and pre-processed based on the body part labels to obtain a medical image. For example, the processing device 140 may acquire a medical image / video to be processed that includes two or more human body parts scanned by the medical device 110, segment the image / video to be processed based on the scanned parts corresponding to the body part labels, and obtain a medical image corresponding to the scanned parts.

[0064] Step 320 : Determine a first candidate part of the medical image based on the part label. In some embodiments, step 320 may be performed by the processing device 140 or the first determination module 220 .

[0065] The first candidate part can reflect the target part of the medical image, i.e., the part to be scanned. In some embodiments, the corresponding scanned part can be obtained based on the part tag and determined as the first candidate part of the medical image. For example, processing device 140 can parse the obtained part tag (0018, 0015), read the scanned part represented by the value corresponding to the tag, and determine it as the first candidate part of the corresponding medical image.

[0066] Step 330 : Process the medical image using the part classification model to obtain a part classification result. In some embodiments, step 330 may be performed by the processing device 140 or the classification module 230 .

[0067] The part classification model can be used to identify the imaging parts contained in a digital image. In some embodiments, the part classification model can be trained based on sample data. For example, a large number of digital images containing different imaging parts can be obtained as sample data, and the imaging parts in the sample data can be manually labeled. The data images are used as input, and the corresponding labels are used as a gold standard to train an initial model to obtain a part classification model. In some embodiments, the part classification model can be trained using any reasonable and feasible method, and this specification does not limit this.

[0068] In some embodiments, the part classification model may include but is not limited to a K-nearest neighbors (KNN) model, a logistic regression model, a linear discriminant analysis (LDA) model, a quadratic discriminant analysis (QDA), a LeNet-5 model, an AlexNet model, or the like, or any combination thereof.

[0069] In some embodiments, the input of the part classification model can be an acquired medical image (i.e., a digital image), and the output is a part classification result corresponding to the medical image. In some embodiments, the part classification result may include at least two candidate parts and a probability value corresponding to each candidate part. In some embodiments, at least two candidate parts of the part classification result may include specific human body part categories and / or other categories. For example, the part classification result may include at least one of the specific part categories such as hand, knee joint, ankle, neck, chest, head, and the category "other" that does not belong to a human body part. Exemplarily, the part classification result may include the category "hand" and the category "other", or the category "hand", "knee joint" and the category "other".

[0070] In some embodiments, at least two candidate parts of the part classification result may include parts that are easily confused and / or relatively close. For example, at least two candidate parts of the part classification result may include: knee joint, elbow joint, hand. In some embodiments, at least two candidate parts may include all parts of the human body. For example, the part classification result may include candidate parts such as hand, elbow joint, knee joint, ankle, sole, neck, chest, abdomen, waist, head, etc. In some embodiments, the probability value may include a value between 0-1, or a value between 0-100, which is not limited in this specification.

[0071] Step 340 : Determine the shooting part corresponding to the medical image based on the part classification result and the first candidate part. In some embodiments, step 340 may be performed by the processing device 140 or the second determination module 240 .

[0072] The shooting part corresponding to the medical image can reflect the part actually shot in the digital image.

[0073] In some embodiments, a first probability value corresponding to the first candidate part can be determined based on the part classification result. The first probability value can reflect the probability that the part captured in the medical image is the first candidate part. In some embodiments, an output value from the part classification result that matches the first candidate part can be obtained and determined as the first probability value. For example, if the first candidate part is determined to be the thorax, and the part classification model processes the medical image and outputs the following part classification results: thorax -85%, abdomen -80%, then the first probability value can be determined as 85%.

[0074] In some embodiments, the maximum probability value and the corresponding candidate part in the part classification result can be determined, the candidate part can be determined as the second candidate part, and the maximum probability value can be determined as the second probability value corresponding to the second candidate part. The second probability value can reflect the probability value of the captured part in the medical image being the second candidate part. In some embodiments, the maximum value among the output values ​​of the part classification result can be selected, the corresponding candidate part can be determined as the second candidate part, and the maximum value can be determined as the second probability value. For example, if the part classification model processes the medical image and outputs the part classification results as: chest -85%, abdomen -80%, waist -60%, then the second candidate part can be determined to be the chest, and its corresponding second probability value is 85%.

[0075] In some embodiments, the shooting part corresponding to the medical image may be determined based on the first probability value and the second probability value.

[0076] In some embodiments, a ratio of the first probability value to the second probability value can be determined, and the ratio can be compared with a preset threshold to generate a comparison result. In some embodiments, the ratio can be the first probability value / the second probability value, or the second probability value / the first probability value. The preset threshold can measure the confidence level between the scanned part recorded in the basic information and the captured part determined by the part classification model. For example, taking the ratio of the first probability value / the second probability value as an example, if the accuracy of the scanned part corresponding to the part label is higher, the preset threshold can be larger; if the accuracy of the part classification model is higher, the preset threshold can be smaller.

[0077] In some embodiments, the preset threshold can be determined based on historical data and / or user data. For example, a first accuracy of the reference part and a second accuracy of the part classification model can be determined based on the actual parts of the historical digital images, and the preset threshold can be determined based on the first accuracy and the second accuracy. For another example, a user preference can be determined based on the user's historical reading records, and the preset threshold can be determined based on the user preference. For more information on preset thresholds, please refer to Figure 5 The related descriptions will not be repeated here.

[0078] In some embodiments, the preset threshold can be adjusted as needed. For example, the preset threshold can be adjusted in real time based on the actual situation of the medical image currently being detected (for example, the presence of missing parts in the image). In another example, the preset threshold can be adjusted based on the parameters of the medical device (for example, the scanning field of view, lifespan, usage time, etc.). In another example, the user can manually set or adjust the preset threshold based on their personal preferences (such as preferring the scan part corresponding to the part label to be more accurate, or preferring the result of the imaged part obtained by the part classification model to be more accurate).

[0079] In some embodiments, based on the comparison results, one of the first and second candidate sites can be selected as the site to be captured in the medical image. For example, when the ratio is greater than or equal to a preset threshold, the first candidate site can be determined as the site to be captured in the medical image; when the ratio is less than the preset threshold, the second candidate site can be determined as the site to be captured in the medical image. For another example, when the ratio is less than the preset threshold, the first candidate site can be determined as the site to be captured in the medical image; when the ratio is greater than or equal to the preset threshold, the second candidate site can be determined as the site to be captured in the medical image.

[0080] In some embodiments, when the part classification result includes the category "Other," the imaging part of the medical image can be determined based on a preset value. For example, when the part classification result output by the part classification model includes a probability value for the category "Other," if the probability value is greater than a preset value, the imaging part of the medical image is deemed not to be any known type and is placed in a folder or database representing "Other." In some embodiments, when the part classification result includes the category "Other," the imaging part of the medical image can be determined based on the ratio of the output probability value for "Other" to a second probability value. In some embodiments, when the part classification result includes the category "Other," the imaging part of the medical image can be determined based on a preset value corresponding to the part classification model. The preset value corresponding to the part classification model can be used to determine the imaging part of the medical image obtained by the part classification model based on the part classification result. For example, the part category corresponding to the probability value in the part classification result greater than the preset value can be determined as the imaging part of the medical image, and this imaging part represents the imaging part of the medical image obtained by the part classification model.

[0081] In some embodiments, whether the medical image meets a preset condition may be determined based on the shooting site of the medical image, and in response to the medical image meeting the preset condition, the medical image may be designated as a target medical image.

[0082] The preset conditions may reflect the requirements for the captured part of the medical image. In some embodiments, the preset conditions may be determined based on the diagnostic requirements for the scanned object. For example, when diagnosing a patient's foot, the preset condition may be that the captured part of the medical image is the foot; when diagnosing a patient's hand, the preset condition may be that the captured part of the medical image is the hand. In some embodiments, the preset conditions may be determined based on the requirements for the medical image. For example, if it is necessary to filter medical images containing hands in the database, the preset condition may be that the captured part of the medical image is the hand. In some embodiments, the captured part in the preset conditions may include one or more. For example, the preset condition may be that the captured part of the medical image is the hand, or the foot, or the chest, etc.

[0083] In some embodiments, the target medical images can be used for medical diagnosis. For example, a diagnosis of a corresponding part of a patient can be made based on the selected target medical images. In some embodiments, the target medical images can be used for intelligent quantitative assessment of image quality. For example, the quality of the medical images of the corresponding part can be assessed to determine whether they meet the requirements.

[0084] In some embodiments, training samples can be generated based on medical images and their imaging sites, and the training samples can be used to train image processing models. For example, the imaging sites can be manually labeled as labels for the medical images, and the labeled medical images can be used as training samples to train the image processing model. In some embodiments, the image processing model can include a quality assessment model, a site classification model, a lesion recognition model, and the like. For example, the quality assessment model can be used to assess the imaging quality of medical images of specific imaging sites to determine whether the imaging sites in the medical images meet diagnostic requirements. For another example, the lesion recognition model can be used to identify the type and severity of lesions in specific sites in medical images.

[0085] It should be noted that the above description of process 300 is for illustrative purposes only and does not limit the scope of this specification. Those skilled in the art will be able to make various modifications and variations to process 300 under the guidance of this specification. For example, step 330 may be performed before step 320, or steps 320 and 330 may be performed simultaneously. However, such modifications and variations remain within the scope of this specification.

[0086] Figure 4 is a schematic diagram of an exemplary method for determining a medical image capturing site according to some embodiments of this specification.

[0087] like Figure 4 As shown in , in the embodiment of this specification, the process of determining the shooting part of the medical image is described by taking the scanning part corresponding to the part label as the knee joint and the part classification model as a binary classification model as an example.

[0088] In some embodiments, the processing device 140 can obtain a medical image 410 (here, a digital image) and a part tag 415 corresponding to the medical image from a DICOM data structure. Furthermore, or simultaneously, the processing device 140 can parse the part tag 415 to read the scan part represented by the tag value—the knee joint 420—and determine it as the first candidate part.

[0089] In some embodiments, processing device 140 may input medical image 410 into part classification model 433 for binary classification. Part classification model 433 processes medical image 410 and outputs prediction values ​​for the knee and hand in the medical image: knee joint output value 435 and hand output value 437.

[0090] In some embodiments, processing device 140 may determine knee joint output value 435 corresponding to knee joint 420 as a first probability value P1 for the first candidate part. Furthermore, processing device 140 may determine the maximum value between knee joint output value 435 and hand output value 437, determine the hand corresponding to the maximum value as the second candidate part, and determine the maximum value as a second probability value P2 corresponding to the second candidate part.

[0091] At 450, processing device 140 may calculate a ratio between first probability value 443 and second probability value 447, and compare this ratio with a preset threshold to generate a comparison result. When ratio P1 / P2 is less than the preset threshold, it indicates that the scanned part information carried by the basic information is inaccurate, and the candidate part corresponding to second probability value P2, the hand, is determined to be the captured part of medical image 410. When P1 / P2 is greater than or equal to the preset threshold, it indicates that the part information carried by the basic information is accurate, and the first candidate part, the knee joint, is determined to be the captured part of the medical image.

[0092] It should be noted that the above description of process 400 is merely for illustration and purpose, and does not limit the scope of application of this specification. For those skilled in the art, various modifications and changes can be made to process 400 under the guidance of this specification. For example, the part classification model can be a multi-classification model, which can output multiple prediction results (for example, knee joint output value, foot output value, elbow joint output value, and hand output value) after processing the medical image 410. However, these modifications and changes are still within the scope of this specification.

[0093] Figure 5 This is a flowchart of an exemplary method for determining a preset threshold value according to some embodiments of this specification. In some embodiments, process 500 may be executed by a processor (e.g., processing device 140) or imaging part determination system 200. It includes:

[0094] Step 510 , obtaining a reference medical image and its reference part and real part. In some embodiments, step 510 may be performed by the processing device 140 or the acquisition module 210 .

[0095] The reference medical image may refer to a digital image in which the imaging part has been determined. For example, the reference medical image may be a medical image in which the imaging part has been determined using process 300 .

[0096] The reference part may refer to part information obtained based on the part label of the reference medical image, for example, Figure 6 The scanning part is read from the Tag field shown in . The acquisition of the reference part is similar to that of the first candidate part. For more details, please refer to step 320 and its related description.

[0097] The actual location may refer to the actual location contained in the determined medical image, for example, the location determined based on process 300. In some embodiments, the actual location may include a location confirmed by the user, for example, the actual location determined by the user based on the medical image.

[0098] In some embodiments, the reference medical image and its reference part and the real part can be obtained from a database (eg, storage device 150). In some embodiments, the reference medical image and its reference part and the real part can be obtained from a medical system.

[0099] Step 520 : Determine a first accuracy of the reference part and a second accuracy of the part classification model. In some embodiments, step 510 may be performed by the processing device 140 or the second determination module 240 .

[0100] The first accuracy may reflect the accuracy of determining the shooting part in the digital image based on the part label, that is, the accuracy of the scanning part information carried in the basic information of the medical image.

[0101] The second accuracy can reflect the accuracy of determining the shooting part in the digital image based on the part classification model. For example, when the candidate part with the maximum probability value output by the part classification model is used as the shooting part of the medical image, the corresponding accuracy of the determination result.

[0102] In some embodiments, a first accuracy of the reference part and a second accuracy of the part classification model may be determined based on a real part of a reference medical image.

[0103] In some embodiments, the actual location of the reference medical image can be compared with the reference location to determine a first accuracy of the reference location. For example, the actual location of a large number of reference medical images can be compared with the reference location, and the proportion of reference medical images in which the reference location is consistent with the actual location can be recorded. The first accuracy can be determined based on this proportion.

[0104] In some embodiments, the actual part of the reference medical image can be compared with the part obtained using the part classification model to determine the second accuracy of the part classification model. For example, the reference medical image can be processed using the part classification model, and the candidate part with the highest probability value in the part classification result can be determined as the captured part of the reference medical image. The captured part is then compared with the actual part to determine whether it is consistent. The comparison results of a large number of reference medical images are then statistically analyzed to determine the second accuracy of the part classification model.

[0105] Step 530 : Obtain the user's historical film reading records. In some embodiments, step 530 may be performed by the processing device 140 or the acquisition module 210 .

[0106] A user's historical reading history may reflect the user's preferences, for example, the user's preference for trusting a region label or region classification model. In some embodiments, the historical reading history may include user feedback, selection information, or any combination thereof. For example, the user's historical reading history may include user feedback on the actual region of the medical image determined by the medical device or system, the actual region of the medical image selected by the user in the medical device or system, and the like.

[0107] In some embodiments, the user's historical film reading records can be obtained from a database (e.g., storage device 150). In some embodiments, the user's historical film reading records can be obtained from a terminal device (e.g., terminal device 130). In some embodiments, the user's historical film reading records can be obtained from a medical system.

[0108] Step 540 , determining a preset threshold. In some embodiments, step 540 may be performed by the processing device 140 or the second determination module 240 .

[0109] In some embodiments, a preset threshold value can be determined based on the first accuracy and the second accuracy. For example, taking the ratio of the first probability value to the second probability value as an example, a larger preset threshold value (e.g., 0.7, 0.8, 0.9, etc.) can be determined when the first accuracy is greater than the second accuracy, and a smaller preset threshold value (e.g., 0.3, 0.4, 0.5, etc.) can be determined when the first accuracy is less than the second accuracy. In some embodiments, a preset threshold value can be determined based on the first accuracy and the second accuracy, as well as a preset range. For example, a user can set an interval range that the preset threshold value needs to meet, and select an appropriate value within the interval range based on the first accuracy and the second accuracy as the preset threshold value.

[0110] In some embodiments, a preset threshold can be determined based on a user's historical film reading history. For example, taking the ratio of a first probability value to a second probability value as an example, if the historical film reading history indicates that the user tends to trust the part label, a larger preset threshold is determined; if the historical film reading history indicates that the user tends to trust the part classification model, a smaller preset threshold is determined. In some embodiments, the same or different preset thresholds can be determined for different users based on their historical film reading history.

[0111] In some embodiments, the preset threshold may be determined based on the first accuracy, the second accuracy, and the user's historical film reading records.

[0112] In some embodiments, a preset threshold can be determined based on hospital information. For example, hospital information can include the hospital's preferences for part labels and part classification models. In some embodiments, a preset threshold can be determined based on device parameters. For example, taking a ratio of a first probability value to a second probability value as an example, for devices that have been used for a long time, the accuracy of their various parameters may decrease, and the part in the captured image is more likely to deviate from the target part. Accordingly, a smaller preset threshold can be determined, i.e., the output result of the part classification model is more inclined.

[0113] It should be noted that the above description of process 500 is only for example and explanation, and does not limit the scope of application of this specification. For those skilled in the art, various modifications and changes can be made to process 500 under the guidance of this specification. For example, step 530 in process 500 can be omitted, and accordingly, in step 540, a preset threshold can be determined based on the first accuracy and the second accuracy. For another example, steps 510-520 can be omitted, and accordingly, in step 540, a preset threshold can be determined only based on the user's historical film reading record. For another example, step 527 can be further included: obtaining hospital information. However, these modifications and changes are still within the scope of this specification.

[0114] The beneficial effects that may be brought about by the embodiments of this specification include but are not limited to: (1) determining the shooting part of a medical image based on the scanning part information corresponding to the part label and the shooting part information in the image information can improve the accuracy of the determination result of the actual shooting part in the medical image; (2) by comparing the first probability value of the scanning part corresponding to the part label with the maximum value (second probability value) output by the part classification model and comparing with the preset threshold, the scanning part information of the medical image and the part classification result of the image can be weighed to obtain a more accurate shooting part; (3) determining the preset threshold for weighing basic information and the part classification model based on reference medical images and / or user historical film reading records can improve the flexibility and adaptability of the shooting part determination result and improve user experience; (4) using the method and / or system provided by the embodiments of this specification to determine the shooting part of a medical image can improve the efficiency and accuracy of screening medical images that meet the preset conditions; (5) using the method and / or system provided by the embodiments of this specification to determine the shooting part of a medical image and training an image processing model based on the medical image can improve the processing efficiency of artificial intelligence scenarios, making the medical diagnosis process more automated and intelligent. It should be noted that different embodiments may produce different beneficial effects. In different embodiments, the beneficial effects that may be produced may be any one or a combination of the above, or any other possible beneficial effects.

[0115] While the basic concepts have been described above, it will be apparent to those skilled in the art that the detailed disclosure is merely illustrative and does not limit this specification. Although not explicitly stated herein, various modifications, improvements, and revisions to this specification may be made by those skilled in the art. Such modifications, improvements, and revisions are suggested in this specification and remain within the spirit and scope of the exemplary embodiments of this specification.

[0116] This specification also uses specific terms to describe the embodiments of this specification. For example, "one embodiment," "an embodiment," and / or "some embodiments" refer to a feature, structure, or characteristic associated with at least one embodiment of this specification. Therefore, it should be emphasized and noted that references to "one embodiment," "an embodiment," or "an alternative embodiment" two or more times in different locations in this specification do not necessarily refer to the same embodiment. Furthermore, certain features, structures, or characteristics of one or more embodiments of this specification may be appropriately combined.

[0117] In addition, unless expressly stated in the claims, the order of the processing elements and sequences, the use of alphanumeric characters, or the use of other names described in this specification are not intended to limit the order of the processes and methods of this specification. Although the above disclosure discusses some of the invention embodiments currently considered useful through various examples, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments. On the contrary, the claims are intended to cover all modifications and equivalent combinations that are consistent with the spirit and scope of the embodiments of this specification. For example, although the system components described above can be implemented by hardware devices, they can also be implemented only by software solutions, such as installing the described system on an existing server or mobile device.

[0118] Similarly, it should be noted that, in order to simplify the presentation of this specification and thus facilitate understanding of one or more embodiments of the invention, the foregoing descriptions of the embodiments of this specification sometimes combine multiple features into a single embodiment, figure, or description thereof. However, this disclosure method does not imply that the subject matter of this specification requires more features than those recited in the claims. In fact, an embodiment may have fewer features than all of the features of a single disclosed embodiment.

[0119] In some embodiments, numbers are used to describe the quantity of components and attributes. It should be understood that such numbers used in the description of the embodiments are modified by the modifiers "about", "approximately" or "substantially" in some examples. Unless otherwise stated, "about", "approximately" or "substantially" indicate that the numbers are allowed to vary by ±20%. Accordingly, in some embodiments, the numerical parameters used in the description and claims are approximate values, which may change according to the required characteristics of individual embodiments. In some embodiments, the numerical parameters should take into account the specified significant digits and adopt the general method of retaining digits. Although the numerical domains and parameters used to confirm the breadth of their range in some embodiments of this specification are approximate values, in specific embodiments, the settings of such numerical values ​​are as accurate as possible within the feasible range.

[0120] Each patent, patent application, patent application publication, and other materials, such as articles, books, specifications, publications, and documents, cited in this specification is hereby incorporated by reference in its entirety. This includes application history documents that are inconsistent with or conflict with the content of this specification, as well as documents (currently or subsequently attached to this specification) that limit the broadest scope of the claims of this specification. It should be noted that if the descriptions, definitions, and / or terminology used in the accompanying materials are inconsistent or conflicting with the content of this specification, the descriptions, definitions, and / or terminology used in this specification will control.

[0121] Finally, it should be understood that the embodiments described in this specification are intended only to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, by way of example and not limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly described and illustrated in this specification.

Claims

1. A method for determining a shooting part of a medical image, characterized in that: include: Acquire a medical image and a part label of the medical image; Determining a first candidate part of the medical image based on the part label; Processing the medical image using a part classification model to obtain a part classification result, wherein the part classification result includes at least two candidate parts and a probability value corresponding to each candidate part; as well as Determining a shooting part corresponding to the medical image based on the part classification result and the first candidate part includes: Determining a first probability value corresponding to the first candidate part based on the part classification result; determining a maximum probability value and a corresponding candidate part in the part classification result, determining the corresponding candidate part as a second candidate part, and determining the maximum probability value as a second probability value corresponding to the second candidate part; and A shooting part corresponding to the medical image is determined based on the first probability value and the second probability value.

2. The method according to claim 1, characterized in that The determining, based on the first probability value and the second probability value, the shooting part corresponding to the medical image further includes: determining a ratio of the first probability value to the second probability value; Comparing the ratio with a preset threshold to generate a comparison result; and Based on the comparison result, one of the first candidate part and the second candidate part is selected as the shooting part of the medical image.

3. The method according to claim 2, characterized in that The preset threshold is determined based on historical data, including: Obtain reference medical images and their reference and true locations; determining a first accuracy of the reference part and a second accuracy of the part classification model based on the real part of the reference medical image; and The preset threshold is determined based on the first accuracy and the second accuracy.

4. The method according to claim 2, characterized in that The preset threshold is determined based on user data, including: Obtain the user's historical film viewing records; and The preset threshold is determined based on the user's historical film reading records.

5. The method according to claim 1, characterized in that The method further comprises: Obtaining images to be processed; and The image to be processed is preprocessed based on the part label to obtain the medical image.

6. The method according to any one of claims 1 to 5, characterized in that The method further comprises: Determining whether the medical image meets a preset condition based on the shooting location of the medical image; and In response to the medical image satisfying the preset condition, the medical image is designated as a target medical image.

7. The method according to any one of claims 1 to 5, characterized in that The method further comprises: generating training samples based on the medical image and the photographed part thereof; and The image processing model is trained using the training samples.

8. A system for determining the shooting part of a medical image, characterized in that: include: An acquisition module, configured to acquire medical images and part labels of the medical images; A first determining module is configured to determine a first candidate part of the medical image based on the part label; a classification module, configured to process the medical image using a part classification model to obtain a part classification result, wherein the part classification result includes at least two candidate parts and a probability value corresponding to each candidate part; as well as a second determining module, configured to determine a shooting part corresponding to the medical image based on the part classification result and the first candidate part, the second determining module further configured to: Determining a first probability value corresponding to the first candidate part based on the part classification result; Determining a maximum probability value and a corresponding candidate part in the part classification result, determining the corresponding candidate part as a second candidate part, and determining the maximum probability value as a second probability value corresponding to the second candidate part; as well as A shooting part corresponding to the medical image is determined based on the first probability value and the second probability value.

9. A device for determining the shooting part of a medical image, characterized in that: The apparatus includes at least one processor and at least one storage device, wherein the storage device is used to store instructions. When the at least one processor executes the instructions, the method according to any one of claims 1 to 7 is implemented.

Images