Bone tumor intelligent diagnosis method, device and equipment
By employing an intelligent diagnostic method for bone tumors, which utilizes feature extraction and fusion networks to process knee joint images, the problem of low efficiency in existing bone tumor diagnosis has been solved, achieving rapid and accurate diagnostic results.
Patent Information
- Application Number
- CN202411163266.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2024-11-29
AI Technical Summary
Current technologies for diagnosing bone tumors are inefficient, requiring attending physicians to spend a significant amount of time analyzing multiple two-dimensional MRI images to determine the extent of the lesion, and lack rapid-response examination mechanisms.
An intelligent diagnostic method for bone tumors is adopted, which uses feature extraction network and feature fusion network to process knee joint images, including operations such as convolution, depthwise separable convolution, dilated convolution, and efficient convolution. The label smoothing algorithm is used to prevent overfitting. Combined with preprocessing such as denoising, image enhancement and normalization, the model is trained using Adam optimizer and DICE loss.
It enables rapid and accurate diagnosis of knee joint bone tumors, improving diagnostic efficiency and reducing doctors' analysis time.
Smart Images

Figure CN119049692A_ABST
Abstract
Description
Technical Field
[0001] The present application belongs to the field of intelligent diagnosis of bone tumors, and in particular, relates to an intelligent diagnosis method, device, equipment and computer-readable storage medium for bone tumors. Background Art
[0002] Bone tumors are tumors (cysts) that occur in bones or their appendages. Bone tumors can be benign or malignant. Benign bone tumors are easily cured and have a good prognosis, while malignant bone tumors develop rapidly, have a poor prognosis, and a high mortality rate. Malignant bone tumors are divided into primary and secondary. Malignant tumors from other tissues or organs in the body that metastasize to the bones through the blood circulation and lymphatic system are secondary malignant bone tumors. There is also a type of lesion called tumor-like lesions. The tissues of tumor-like lesions do not have the characteristics of tumor cell morphology, but their ecology and behavior are destructive to tumors. They are generally more limited and easier to cure.
[0003] Clinically, this type of tumor can be treated with intracystic curettage and chemical or physical inactivation of the tumor cavity to achieve satisfactory local control, thereby avoiding excessive functional loss caused by large-scale resection or tumor segment resection.
[0004] Clinically, magnetic resonance imaging (MRI) or CT is generally used as the main examination method for diagnosing bone tumor protrusions. This method can analyze and determine the disease extent and location of the specific structure of the bone tumor by detecting abnormal signals in the cystic part of the bone tumor. Although this method can analyze the MRI images of the bone tumor, it mainly uses two-dimensional sagittal MRI images to identify abnormal structures in the cystic part of the bone tumor, etc. It is necessary to combine sagittal MRI images at multiple angles to comprehensively analyze the degree of bone tumor lesions.
[0005] Therefore, the entire process requires the attending physician to consume most of his time to analyze and confirm the site of the patient's bone tumor lesions and determine the corresponding bone tumor treatment plan, which is relatively inefficient.
[0006] There is no rapid response inspection mechanism for the location of bone tumors and the examination of the extent of lesions in clinical practice. The attending physician needs to check the MRI images for manual identification, which is inefficient.
[0007] Therefore, how to diagnose knee bone tumors quickly and accurately is a technical problem that technical personnel in this field urgently need to solve. Summary of the invention
[0008] The embodiments of the present application provide a method, device, equipment and computer-readable storage medium for intelligent diagnosis of bone tumors, which can diagnose knee bone tumors quickly and accurately.
[0009] In a first aspect, an embodiment of the present application provides a bone tumor intelligent diagnosis method, comprising:
[0010] Obtain images of the patient's knee joint;
[0011] Input the knee joint image into the preset bone tumor diagnosis model and output the bone tumor diagnosis result;
[0012] Among them, the knee bone tumor diagnosis model includes a feature extraction network and a feature fusion network; the feature extraction network is used to extract feature maps of each level of the knee joint image; the feature fusion network is used to merge the feature maps of each level to enhance the representation ability of the model.
[0013] Optionally, a feature extraction network is used to extract feature maps of various levels of the knee joint image, including:
[0014] Convolve the knee joint image to obtain the first feature map;
[0015] Perform three depth-wise separable convolutions and dilated convolutions on the first feature map to obtain a second feature map.
[0016] The second feature map is subjected to three efficient convolutions in sequence to obtain the third feature maps after each fused convolution.
[0017] Optional, feature fusion network, used to merge feature maps at each level, including:
[0018] Attention is added, up-sampled and convolved on the first third feature map to obtain the fourth feature map;
[0019] The second third feature map and the fourth feature map are concatenated to obtain a fifth feature map;
[0020] Perform attention addition, upsampling and convolution on the fifth feature map to obtain the sixth feature map;
[0021] Convolve the third feature map to obtain the seventh feature map;
[0022] The sixth feature map and the seventh feature map are combined to obtain an eighth feature map.
[0023] Optionally, use a label smoothing algorithm to prevent overfitting in multi-label classification:
[0024]
[0025] in, represents the probability value after label smoothing; k is the total number of categories, pi is the true probability of the i-th category, qi is the network's predicted probability of the i-th category; zi is the output value of the last fully connected layer of the network; when the provided label is a category rather than a category probability, if i is equal to the true category, then pi = 1, otherwise pi = 0.
[0026] Optionally, after obtaining the patient's knee joint image, the following steps are also included:
[0027] The knee joint image is preprocessed; wherein the preprocessing includes: denoising, image enhancement, image smoothing, and normalization processing.
[0028] Optionally, preprocess the knee joint image, including:
[0029] Using Gaussian filtering algorithm to filter the knee joint image to obtain a first knee joint image;
[0030] Using histogram equalization to perform image enhancement on the first knee joint image to obtain a second knee joint image;
[0031] The second knee joint image is smoothed and normalized to obtain a preprocessed knee joint image.
[0032] Optionally, during model training, set the training batch_size to 32;
[0033] Set the initial learning rate to 1e-4, and add a learning rate decay strategy. Every 5000 iterations, the learning rate decays to 0.9 of the previous learning rate.
[0034] Set the optimizer to Adam optimizer;
[0035] Set the loss function to DICE loss;
[0036] The training set and validation set were validated once every 1000 iterations. The early stopping method was used to determine the stopping time of network training and obtain a knee bone tumor diagnosis model.
[0037] In a second aspect, an embodiment of the present application provides an intelligent bone tumor diagnosis device, comprising:
[0038] An image acquisition module, used for acquiring a knee joint image of a patient;
[0039] A bone tumor diagnosis module is used to input the knee joint image into a preset bone tumor diagnosis model and output the bone tumor diagnosis result;
[0040] Among them, the knee bone tumor diagnosis model includes a feature extraction network and a feature fusion network; the feature extraction network is used to extract feature maps of each level of the knee joint image; the feature fusion network is used to merge the feature maps of each level to enhance the representation ability of the model.
[0041] In a third aspect, an embodiment of the present application provides an electronic device, the electronic device comprising: a processor and a memory storing computer program instructions;
[0042] The processor implements a bone tumor intelligent diagnosis method when executing the computer program instructions.
[0043] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium having computer program instructions stored thereon, and the computer program instructions, when executed by a processor, implement an intelligent diagnosis method for bone tumors.
[0044] The intelligent bone tumor diagnosis method, device, equipment and computer-readable storage medium of the embodiments of the present application can diagnose knee bone tumors quickly and accurately.
[0045] The intelligent diagnosis method for bone tumors includes:
[0046] Obtain images of the patient's knee joint;
[0047] Input the knee joint image into the preset bone tumor diagnosis model and output the bone tumor diagnosis result;
[0048] Among them, the knee bone tumor diagnosis model includes a feature extraction network and a feature fusion network; the feature extraction network is used to extract feature maps of each level of the knee joint image; the feature fusion network is used to merge the feature maps of each level to enhance the representation ability of the model. BRIEF DESCRIPTION OF THE DRAWINGS
[0049] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0050] Figure 1 It is a flowchart of a bone tumor intelligent diagnosis method provided by an embodiment of the present application;
[0051] Figure 2 This is a schematic diagram of the structure of a knee bone tumor diagnosis model provided by an embodiment of the present application;
[0052] Figure 3 It is a structural schematic diagram of a bone tumor intelligent diagnosis device provided by an embodiment of the present application;
[0053] Figure 4 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present application. DETAILED DESCRIPTION
[0054] The features and exemplary embodiments of various aspects of the present application will be described in detail below. In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only intended to explain the present application, rather than to limit the present application. For those skilled in the art, the present application can be implemented without the need for some of these specific details. The following description of the embodiments is only to provide a better understanding of the present application by illustrating the examples of the present application.
[0055] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.
[0056] In order to solve the problems of the prior art, the embodiments of the present application provide a bone tumor intelligent diagnosis method, device, equipment and computer readable storage medium. The following first introduces the bone tumor intelligent diagnosis method provided by the embodiments of the present application.
[0057] Figure 1 It is a flowchart of a bone tumor intelligent diagnosis method provided by an embodiment of the present application; Figure 2 This is a schematic diagram of the structure of a knee bone tumor diagnosis model provided by an embodiment of the present application.
[0058] The intelligent diagnosis method for bone tumors comprises:
[0059] S101, obtaining a knee joint image of a patient;
[0060] S102, inputting the knee joint image into a preset bone tumor diagnosis model, and outputting a bone tumor diagnosis result;
[0061] Among them, the knee bone tumor diagnosis model includes a feature extraction network and a feature fusion network; the feature extraction network is used to extract feature maps of each level of the knee joint image; the feature fusion network is used to merge the feature maps of each level to enhance the representation ability of the model.
[0062] In one embodiment, a feature extraction network is used to extract feature maps of various levels of a knee joint image, including:
[0063] Convolve the knee joint image to obtain the first feature map;
[0064] Perform three depth-wise separable convolutions and dilated convolutions on the first feature map to obtain a second feature map.
[0065] The second feature map is subjected to three efficient convolutions in sequence to obtain the third feature maps after each fused convolution.
[0066] In one embodiment, a feature fusion network is used to merge feature maps at each level, including:
[0067] Attention is added, up-sampled and convolved on the first third feature map to obtain the fourth feature map;
[0068] The second third feature map and the fourth feature map are concatenated to obtain a fifth feature map;
[0069] Perform attention addition, upsampling and convolution on the fifth feature map to obtain the sixth feature map;
[0070] Convolve the third feature map to obtain the seventh feature map;
[0071] The sixth feature map and the seventh feature map are combined to obtain an eighth feature map.
[0072] In one embodiment, a label smoothing algorithm is used to prevent overfitting in multi-label classification:
[0073]
[0074] in, represents the probability value after label smoothing; k is the total number of categories, pi is the true probability of the i-th category, qi is the network's predicted probability of the i-th category; zi is the output value of the last fully connected layer of the network; when the provided label is a category rather than a category probability, if i is equal to the true category, then pi = 1, otherwise pi = 0.
[0075] In one embodiment, after acquiring the knee joint image of the patient, the method further includes:
[0076] The knee joint image is preprocessed; wherein the preprocessing includes: denoising, image enhancement, image smoothing, and normalization processing.
[0077] In one embodiment, the knee joint image is preprocessed, including:
[0078] Using Gaussian filtering algorithm to filter the knee joint image to obtain a first knee joint image;
[0079] Using histogram equalization to perform image enhancement on the first knee joint image to obtain a second knee joint image;
[0080] The second knee joint image is smoothed and normalized to obtain a preprocessed knee joint image.
[0081] In one embodiment, during the model training process, the batch_size of the training is set to 32;
[0082] Set the initial learning rate to 1e-4, and add a learning rate decay strategy. Every 5000 iterations, the learning rate decays to 0.9 of the previous learning rate.
[0083] Set the optimizer to Adam optimizer;
[0084] Set the loss function to DICE loss;
[0085] The training set and validation set were validated once every 1000 iterations. The early stopping method was used to determine the stopping time of network training and obtain a knee bone tumor diagnosis model.
[0086] Figure 3 1 is a schematic diagram of the structure of an intelligent bone tumor diagnosis device provided by an embodiment of the present application. The intelligent bone tumor diagnosis device comprises:
[0087] An image acquisition module 301 is used to acquire a knee joint image of a patient;
[0088] The bone tumor diagnosis module 302 is used to input the knee joint image into a preset bone tumor diagnosis model and output the bone tumor diagnosis result;
[0089] Among them, the knee bone tumor diagnosis model includes a feature extraction network and a feature fusion network; the feature extraction network is used to extract feature maps of each level of the knee joint image; the feature fusion network is used to merge the feature maps of each level to enhance the representation ability of the model.
[0090] Figure 4 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application is shown.
[0091] The electronic device may include a processor 401 and a memory 402 storing computer program instructions.
[0092] Specifically, the processor 401 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.
[0093] Memory 402 may include a large capacity memory for data or instructions. By way of example and not limitation, memory 402 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, memory 402 may include a removable or non-removable (or fixed) medium. In appropriate cases, memory 402 may be inside or outside of an electronic device. In a particular embodiment, memory 402 may be a non-volatile solid-state memory.
[0094] In one embodiment, the memory 402 may be a read-only memory (ROM). In one embodiment, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or a flash memory, or a combination of two or more of these.
[0095] The processor 401 implements any one of the intelligent bone tumor diagnosis methods in the above embodiments by reading and executing computer program instructions stored in the memory 402 .
[0096] In one example, the electronic device may further include a communication interface 403 and a bus 410. Figure 4 As shown, the processor 401, the memory 402, and the communication interface 403 are connected via a bus 410 and communicate with each other.
[0097] The communication interface 403 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.
[0098] Bus 410 includes hardware, software or both, and the parts of electronic equipment are coupled to each other. For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front side bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 410 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the application considers any suitable bus or interconnection.
[0099] In addition, in combination with the bone tumor intelligent diagnosis method in the above embodiment, the embodiment of the present application can provide a computer-readable storage medium for implementation. The computer-readable storage medium stores computer program instructions; when the computer program instructions are executed by the processor, any one of the bone tumor intelligent diagnosis methods in the above embodiment is implemented.
[0100] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.
[0101] The functional modules shown in the above-described block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. Programs or code segments can be stored in machine-readable media, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable media" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.
[0102] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.
[0103] The above reference is according to the method of the embodiment of the present application, the flow chart of the device (system) and the computer program product and / or the block diagram described various aspects of the present application.It should be understood that each square box in the flow chart and / or the block diagram and the combination of each square box in the flow chart and / or the block diagram can be realized by computer program instructions.These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the realization of the function / action specified in one or more square boxes of the flow chart and / or the block diagram.Such a processor can be but is not limited to a general-purpose processor, a special-purpose processor, a special application processor or a field programmable logic circuit.It can also be understood that each square box in the block diagram and / or the flow chart and the combination of the square boxes in the block diagram and / or the flow chart can also be realized by the dedicated hardware that performs the specified function or action, or can be realized by the combination of dedicated hardware and computer instructions.
[0104] The above is only a specific implementation of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.
Claims
1. A bone tumor intelligent diagnosis method, characterized in that: include: Obtain an image of the patient's knee joint; Input the knee joint image into the preset bone tumor diagnosis model and output the bone tumor diagnosis result; Among them, the knee bone tumor diagnosis model includes a feature extraction network and a feature fusion network; the feature extraction network is used to extract feature maps of each level of the knee joint image; The feature fusion network is used to merge feature maps at each level to enhance the representation ability of the model.
2. The intelligent diagnosis method for bone tumors according to claim 1, characterized in that: Feature extraction network, used to extract feature maps at various levels of knee joint images, including: Convolve the knee joint image to obtain the first feature map; Perform three depth-wise separable convolutions and dilated convolutions on the first feature map to obtain a second feature map. The second feature map is subjected to three efficient convolutions in sequence to obtain the third feature maps after each fused convolution.
3. The intelligent diagnosis method for bone tumors according to claim 2, characterized in that: Feature fusion network, used to merge feature maps at each level, including: Attention is added, up-sampled and convolved on the first third feature map to obtain the fourth feature map; The second third feature map and the fourth feature map are concatenated to obtain a fifth feature map; Perform attention addition, upsampling and convolution on the fifth feature map to obtain the sixth feature map; Convolve the third feature map to obtain the seventh feature map; The sixth feature map and the seventh feature map are combined to obtain an eighth feature map.
4. The intelligent diagnosis method for bone tumor according to claim 1, characterized in that: In order to prevent overfitting in multi-label classification, label smoothing algorithm is used: Among them, l(p,q) represents the probability value after label smoothing; k is the total number of categories, pi is the true probability of the i-th category, qi is the network's predicted probability of the i-th category; zi is the output value of the last fully connected layer of the network; when the provided label is a category rather than a category probability, if i is equal to the true category, then pi = 1, otherwise pi = 0.
5. The intelligent diagnosis method for bone tumor according to claim 1, characterized in that: After acquiring images of the patient's knee, it also includes: The knee joint image is preprocessed; wherein the preprocessing includes: denoising, image enhancement, image smoothing, and normalization processing.
6. The intelligent diagnosis method for bone tumor according to claim 1, characterized in that: Preprocess the knee joint image, including: Using Gaussian filtering algorithm to filter the knee joint image to obtain a first knee joint image; Using histogram equalization to perform image enhancement on the first knee joint image to obtain a second knee joint image; The second knee joint image is smoothed and normalized to obtain a preprocessed knee joint image.
7. The intelligent diagnosis method for bone tumor according to claim 1, characterized in that: During model training, set the training batch_size to 32; Set the initial learning rate to 1e-4, and add a learning rate decay strategy. Every 5000 iterations, the learning rate decays to 0.9 of the previous learning rate. Set the optimizer to Adam optimizer; Set the loss function to DICEloss; The training set and validation set were validated once every 1000 iterations. The early stopping method was used to determine the stopping time of network training and obtain a knee bone tumor diagnosis model.
8. An intelligent bone tumor diagnosis device, characterized in that: The device comprises: An image acquisition module, used for acquiring a knee joint image of a patient; A bone tumor diagnosis module is used to input the knee joint image into a preset bone tumor diagnosis model and output the bone tumor diagnosis result; Among them, the knee bone tumor diagnosis model includes a feature extraction network and a feature fusion network; the feature extraction network is used to extract feature maps of each level of the knee joint image; the feature fusion network is used to merge the feature maps of each level to enhance the representation ability of the model.
9. An electronic device, characterized in that: The electronic device comprises: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the intelligent diagnosis method for bone tumors according to any one of claims 1 to 7 is implemented.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by the processor, the intelligent diagnosis method for bone tumors according to any one of claims 1 to 7 is implemented.
Citation Information
Cited By
Image enhancement method and system for bone tumor
CN120525724A
Bone tumor image enhancement method and system
CN120525724B
Bone lesion classification method and device based on whole-body bone scanning image and clinical information, equipment and medium
CN120808043A
