Urinary stone component analysis method and apparatus, and electronic device and storage medium

Abstract

The present disclosure relates to the technical field of stone analysis, and particularly relates to a urinary stone component analysis method and apparatus, and an electronic device and a storage medium. The method comprises: classifying analyzed stone components of an original stone, so as to obtain a stone data set; performing data enhancement and re-balancing processing to obtain a processed data set; and training a deep neural network model to obtain a trained model, and deploying the trained model to a server to receive urinary stone components sent by the server, so as to obtain an analysis result.

Description

Method, device, electronic device and storage medium for analyzing urinary stone composition

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] The present disclosure claims priority to Chinese patent application number "202410165012.3" filed by Beijing Tsinghua Chang Gung Hospital on February 5, 2024, with the invention title "Method, device, electronic device and storage medium for analyzing the composition of urinary stones". Technical Field

[0003] The present disclosure relates to the field of stone analysis technology, and in particular to a "method, device, electronic device and storage medium for analyzing the composition of urinary stones." Background Art

[0004] In related technologies, infrared spectroscopy is used to analyze the composition of stones. The in vitro stone powder obtained after surgery is processed by the halide tablet method and then analyzed. The stone composition is determined based on the characteristics of its absorption peak in the infrared light region, or dual-energy CT is used to identify the main components of the stone. That is, the analysis workstation provided by the equipment is used to manually or automatically outline the area of ​​interest, calculate the average atomic number, electron density and other information in the area, and judge the stone type.

[0005] However, the infrared spectroscopy method used in related technologies to analyze stone composition can only be used for the analysis of postoperative in vitro samples, and requires professional personnel to operate, increasing labor costs. Local stone sampling can lead to misjudgment of stone composition and reduce the accuracy of stone composition analysis. In addition, the dual-energy CT method is limited by the equipment resolution and cannot analyze mixed-component stones, which reduces the applicability of stone composition analysis and urgently needs to be resolved.

[0006] Public content

[0007] This disclosure is based on the following issues and understandings:

[0008] Urinary tract stones are a common and frequently occurring disease in my country, with an incidence of approximately 2-3%. In high-incidence areas, they can account for over 40% of urology hospitalizations. Furthermore, urinary stone recurrence rates are high, reaching 50% within 5-10 years and 75% within 20 years. The primary hazard of urinary stone disease is renal impairment caused by stone obstruction. Endoscopic surgery, a milestone in the treatment of urinary stones, is now widely used in clinical practice and is a key component of minimally invasive urology. It primarily includes ureteroscopic and percutaneous nephrolithotomy.

[0009] Stone composition analysis is the foundation for developing treatment and prevention strategies for stone patients. From a surgical perspective, stones of different composition may require different surgical plans. Endoscopic laser lithotripsy also requires different laser lithotripsy parameters. From a postoperative prevention perspective, long-term prevention strategies for patients with stones of different composition also differ significantly. From the perspective of exploring the causes of stone formation, stone composition analysis combined with morphological structure helps to refine the distinction between stone types and lays the foundation for developing personalized stone prevention plans. Therefore, domestic and international guidelines recommend that every patient with urinary stones undergo stone composition analysis. Currently, infrared spectroscopy is the most widely used method for stone composition analysis in clinical practice.

[0010] Infrared spectroscopy is the most classic method for analyzing stone composition. The in vitro stone powder obtained after surgery is processed by the halide tablet method and then analyzed. The stone composition is determined based on the characteristics of its absorption peak in the infrared light region. It has high accuracy, but this technology still has the following shortcomings: first, it can only be used for the analysis of in vitro samples after surgery, and the instruments and equipment are relatively expensive and require professional operation; second, this method only uses trace samples the size of "millet grains" for analysis. For common mixed-component stones in clinical practice, it is very likely to lead to deviations in the analysis results caused by sampling errors. In addition, the morphological structure of the stone is very meaningful for exploring the cause of the disease. After surgical lithotripsy, the original morphological structure of the stone is destroyed, and the powdered sample of the infrared spectroscopy method cannot achieve component analysis including the stone structure.

[0011] In recent years, many studies have used dual-energy CT to identify the main components of stones. Early studies were conducted using commercial medical dual-energy CT. The analysis workstation provided by the device was used to manually or automatically outline the region of interest, calculate the average atomic number, electron density and other information within the region, and determine the stone type. In vivo and in vitro related work has demonstrated that this method performs well in distinguishing uric acid stones from non-uric acid stones, but it still faces challenges in distinguishing other component subtypes. After the introduction of machine learning methods, the detection effect of dual-energy CT in distinguishing different subtypes has been partially improved. However, due to the limitations of the equipment resolution, dual-energy CT is difficult to distinguish stones with mixed components, and it is difficult to achieve component analysis while taking into account the description of morphological characteristics, which urgently needs to be improved.

[0012] The present disclosure provides a method, device, electronic device and storage medium for analyzing the composition of urinary stones to solve the problem that the infrared spectroscopy method used in related technologies for analyzing the composition of stones can only be used for the analysis of postoperative ex vivo samples, and requires professional personnel to operate, which increases labor costs. Local stone sampling may lead to misjudgment of stone composition and reduce the accuracy of stone composition analysis. In addition, the dual-energy CT method is limited by the equipment resolution and cannot analyze stones with mixed components, which reduces the applicability of stone composition analysis.

[0013] The first aspect of the present disclosure provides a method for analyzing the composition of urinary stones, comprising the following steps: collecting a stone image of a target original stone, analyzing the stone composition of the target original stone based on the stone image, and classifying the stone composition to obtain a target stone dataset of the target original stone; performing data enhancement on the target stone dataset to obtain an enhanced stone dataset, and performing data rebalancing on the enhanced stone dataset to obtain a processed stone dataset; training a deep neural network model using the processed stone dataset to obtain a trained deep neural network model, and deploying the trained deep neural network model to a server to receive the urinary stone composition sent by the server to obtain the analysis results output by the trained deep neural network model.

[0014] In some embodiments, the collecting of the target original stone image and analyzing the stone composition of the target original stone based on the stone image include: collecting the stone image of the target original stone before laser lithotripsy and the stone image of the target original stone after laser lithotripsy; analyzing the stone image before laser lithotripsy and the stone image after laser lithotripsy to obtain the stone composition of the target original stone.

[0015] In some embodiments, the enhanced stone dataset is subjected to data rebalancing processing to obtain a balanced stone dataset, including: obtaining the classification scale of the stone images in the enhanced stone dataset, detecting whether the classification scale meets the preset balanced classification condition, upsampling the classification in the classification scale that does not meet the preset balanced classification condition, and obtaining the upsampled balanced classification, so as to obtain the balanced stone dataset based on the upsampled balanced classification.

[0016] In some embodiments, before deploying the trained deep neural network model to the server, it also includes: determining whether the trained deep neural network model meets preset compliance conditions; if the preset compliance conditions are met, deploying the trained deep neural network model to the server; if the preset compliance conditions are not met, rebuilding a new deep neural network model and training the new deep neural network model to obtain a new trained deep neural network model, until the new trained deep neural network model meets the preset compliance conditions, and then deploying the new trained deep neural network model to the server.

[0017] The second aspect of the present disclosure provides an apparatus for analyzing the composition of urinary stones, including: an acquisition module for acquiring a stone image of a target original stone, analyzing the stone composition of the target original stone based on the stone image, and classifying the stone composition to obtain a target stone data set of the target original stone; a first acquisition module for performing data enhancement on the target stone data set to obtain an enhanced stone data set, and performing data rebalancing processing on the enhanced stone data set to obtain a processed stone data set; a second acquisition module for training a deep neural network model using the processed stone data set to obtain a trained deep neural network model, and deploying the trained deep neural network model to a server to receive the urinary stone composition sent by the server to obtain the analysis results output by the trained deep neural network model.

[0018] In some embodiments, the acquisition module includes: an acquisition unit for collecting stone images before laser lithotripsy and stone images after laser lithotripsy of the target original stone; a determination unit for analyzing the stone images before laser lithotripsy and the stone images after laser lithotripsy to obtain the stone composition of the target original stone.

[0019] In some embodiments, the first acquisition module includes: an acquisition unit, used to obtain the classification scale of the stone images in the enhanced stone data set, detect whether the classification scale meets the preset balanced classification conditions, upsample the classifications in the classification scale that do not meet the preset balanced classification conditions, and obtain the upsampled balanced classification, so as to obtain the balanced stone data set based on the upsampled balanced classification.

[0020] In some embodiments, the apparatus of the embodiments of the present disclosure further includes: a judgment module for judging whether the trained deep neural network model meets preset compliance conditions; a first processing module for deploying the trained deep neural network model to the server if the preset compliance conditions are met before the trained deep neural network model is deployed to the server; a second processing module for rebuilding a new deep neural network model and training the new deep neural network model to obtain a new trained deep neural network model if the preset compliance conditions are not met before the trained deep neural network model is deployed to the server, until the new trained deep neural network model meets the preset compliance conditions, and then the new trained deep neural network model is deployed to the server.

[0021] The third aspect of the present disclosure provides an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for analyzing the composition of urinary stones as described in the above embodiment.

[0022] A fourth aspect of the present disclosure provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the above method for analyzing the composition of urinary stones.

[0023] The disclosed embodiment can analyze and classify the target original stones based on the collected target original stone images, obtain the target stone data set of the target original stones, then perform data enhancement to obtain the enhanced stone data set, perform data rebalancing on the enhanced stone data set to obtain the processed stone data set, thereby training a deep neural network model, obtaining the trained deep neural network model, and deploying it to a server to receive the urinary stone components sent by the server, so as to obtain the analysis results output by the trained deep neural network model, thereby effectively improving the accuracy and applicability of the stone composition analysis. Thus, it solves the problem that the infrared spectroscopy method in the related art is only used for the analysis of postoperative ex vivo samples, local stone sampling leads to misjudgment of stone composition, and reduces the accuracy of stone composition analysis. In addition, the dual-energy CT method is limited by the equipment resolution, which reduces the applicability of stone composition analysis.

[0024] Additional aspects and advantages of the present disclosure will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The above and / or additional aspects and advantages of the present disclosure will become apparent and readily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0026] FIG1 is a flow chart of a method for analyzing the composition of urinary stones according to an embodiment of the present disclosure;

[0027] FIG2 is a schematic structural diagram of a device for analyzing the composition of urinary stones according to an embodiment of the present disclosure;

[0028] FIG3 is a schematic structural diagram of an electronic device provided according to an embodiment of the present disclosure.

[0029] Explanation of reference numerals: 10 - device for analyzing the composition of urinary stones; 100 - acquisition module; 200 - first acquisition module; and 300 - second acquisition module; 301 - memory; 302 - processor; and 303 - communication interface. DETAILED DESCRIPTION

[0030] The following describes in detail embodiments of the present disclosure, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present disclosure, and should not be construed as limiting the present disclosure.

[0031] The following describes the method, device, electronic device and storage medium for analyzing the composition of urinary stones according to the embodiments of the present disclosure with reference to the accompanying drawings. In view of the problems that the infrared spectroscopy method is only used for the analysis of postoperative ex vivo samples in the related art mentioned in the background technology center, local stone sampling leads to misjudgment of stone composition, reducing the accuracy of stone composition analysis, and the dual-energy CT method is limited by the equipment resolution, which reduces the applicability of stone composition analysis, the present disclosure provides a method for analyzing the composition of urinary stones, in which the stone composition of the target original stone can be analyzed and classified based on the stone image of the target original stone that has been collected, to obtain a target stone data set of the target original stone, and then data enhancement is performed to obtain an enhanced stone data set, and the enhanced stone data set is subjected to data rebalancing processing to obtain a processed stone data set, thereby training a deep neural network model to obtain a trained deep neural network model, and deploying it to a server to receive the urinary stone composition sent by the server to obtain the analysis results output by the trained deep neural network model, thereby effectively improving the accuracy and applicability of stone composition analysis. This solves the problem that the infrared spectroscopy method in the related technology is only used for the analysis of postoperative in vitro samples, local stone sampling leads to misjudgment of stone composition, and reduces the accuracy of stone composition analysis. In addition, the dual-energy CT method is limited by the equipment resolution, which reduces the applicability of stone composition analysis.

[0032] Specifically, FIG1 is a flow chart of a method for analyzing the composition of urinary stones provided in an embodiment of the present disclosure.

[0033] As shown in FIG1 , the method for analyzing the composition of urinary stones includes the following steps:

[0034] In step S101, a stone image of a target original stone is collected, the stone components of the target original stone are analyzed according to the stone image, and the stone components are classified to obtain a target stone data set of the target original stone.

[0035] It can be understood that the embodiments of the present disclosure can collect stone images of the original stones in the following steps. For example, in clinical surgery, professional equipment such as endoscopes can be used to obtain stone images captured in real stone surgery videos, and the stone components of the original stones can be analyzed based on the stone images, and the stone components can be classified to obtain a stone data set of the original stones. Among them, the embodiments of the present disclosure can obtain a stone data set of a certain scale according to the analysis requirements, which is not specifically limited here, so as to improve the feasibility of stone analysis.

[0036] Among them, in some embodiments of the present disclosure, stone images of the target original stone are collected, and the stone composition of the target original stone is analyzed based on the stone images, including: collecting stone images of the target original stone before laser lithotripsy and stone images after laser lithotripsy; analyzing the stone images before laser lithotripsy and stone images after laser lithotripsy to obtain the stone composition of the target original stone.

[0037] During the actual implementation process, the disclosed embodiment can collect a large number of real stone surgery videos taken by professional equipment such as endoscopes during clinical operations, and capture stone images of the stone surface before laser lithotripsy and stone images of relatively complete cross-sections of the stone after laser lithotripsy in the surgery video. Then, relevant technical personnel can conduct a comprehensive analysis of the stone images before laser lithotripsy and the stone images after laser lithotripsy, and obtain the stone composition of the original stone based on postoperative infrared spectroscopy analysis, and then mark the stone composition, such as a binary group marked as <image-stone composition>, and save it to a local computer, and repeat the above process until a certain scale of data set is obtained. The specific scale is not specifically limited here, which effectively improves the robustness of the stone composition analysis.

[0038] In step S102, data enhancement is performed on the target stone dataset to obtain an enhanced stone dataset, and data rebalancing is performed on the enhanced stone dataset to obtain a processed stone dataset.

[0039] It can be understood that the embodiments of the present disclosure can perform data enhancement on the target stone dataset to obtain an enhanced stone dataset. For example, since the acquisition cost of stone surgery videos in clinical surgery is high and the scale is small, the stone images in the target stone dataset can be rotated, contrast transformed, brightness transformed, randomly cropped, etc., that is, multiple copies of different information based on the stone images are obtained, thereby increasing the scale of the target stone dataset. Then, the embodiments of the present disclosure can perform data rebalancing processing in the following steps on the enhanced stone dataset to obtain a processed stone dataset, which effectively improves the accuracy of stone composition analysis.

[0040] Among them, data enhancement can be rotation, contrast transformation, brightness transformation, random cropping, etc., to expand the training set and improve the generalization ability of the deep neural network model in the following steps.

[0041] Among them, in some embodiments of the present disclosure, the enhanced stone dataset is subjected to data rebalancing processing to obtain a balanced stone dataset, including: obtaining the classification scale of the stone images in the enhanced stone dataset, detecting whether the classification scale meets the preset balanced classification conditions, upsampling the classifications in the classification scale that do not meet the preset balanced classification conditions, and obtaining the balanced classification after upsampling, so as to obtain the balanced stone dataset according to the balanced classification after upsampling.

[0042] In the embodiment of the present disclosure, the preset balanced classification condition is a condition that the distribution of stone components in the classification scale is relatively balanced. When the preset balanced classification condition is met, stone data of different categories can be maintained at approximately the same level. The preset balanced classification condition is set by technical personnel in this field according to actual conditions and is not specifically limited here.

[0043] In some embodiments, the embodiments of the present disclosure can obtain the classification scale of the stone images in the enhanced stone data set. Then, the embodiments of the present disclosure can analyze the classification scale of the stone images in the enhanced stone data set to detect whether the classification scale meets the balanced classification condition. When the balanced classification condition is not met, that is, when the stone component distribution is extremely unbalanced, for example, when a low-frequency classification with a lower data scale is detected, the stone components of the low-frequency classification can be upsampled so that stone data of different categories remain at approximately the same level, effectively improving the balance of the stone component distribution.

[0044] In step S103, the processed stone data set is used to train the deep neural network model to obtain a trained deep neural network model, and the trained deep neural network model is deployed to the server to receive the urinary stone components sent by the server to obtain the analysis results output by the trained deep neural network model.

[0045] It can be understood that the embodiments of the present disclosure can pre-build a deep neural network model. For example, deep learning technology can be used to build a deep neural network model (based on ResNet-152-V2) that identifies stone components through stone images, and the processed stone data set is used to train the deep neural network model to obtain a trained deep neural network model. The trained deep neural network model is deployed to the server, and the user can upload the patient's clearer endoscopic digital image of the stone to the server, so that the embodiments of the present disclosure can receive the urinary stone components sent by the server to obtain the analysis results output by the trained deep neural network model, thereby effectively improving the efficiency and accuracy of stone component identification.

[0046] In some embodiments, the disclosed embodiments can use endoscopic stone images of patients to train deep neural network models and identify stone components, and can be changed to use one or more videos of patients undergoing endoscopic examinations to manually mark stone components for deep neural network model training and evaluation. Secondly, the input of endoscopic digital images of stones can be changed to endoscopic stone videos to analyze stone components, thereby improving the applicability of stone composition analysis.

[0047] In some cases, the use of patient endoscopic stone images to train deep neural network models and identify stone components in the embodiments of the present disclosure can be changed to using patient physical indicators, such as urine composition and blood indicators, combined with endoscopic stone images to train and evaluate deep neural network models. In clinical use, the accuracy of stone composition analysis is improved by inputting patient physical indicators and patient endoscopic stone images.

[0048] In some embodiments, before deploying the trained deep neural network model to the server, it also includes: determining whether the trained deep neural network model meets the preset compliance conditions; if the preset compliance conditions are met, deploying the trained deep neural network model to the server; if the preset compliance conditions are not met, rebuilding a new deep neural network model and training the new deep neural network model to obtain a new trained deep neural network model, until the new trained deep neural network model meets the preset compliance conditions, and then deploying the new trained deep neural network model to the server.

[0049] In some embodiments, the embodiments of the present disclosure can determine whether the trained deep neural network model meets the qualifying conditions. For example, the qualifying conditions can be set as the trained deep neural network model reaches more than 98% in both the training set and the test set. When the qualifying conditions are met, the trained deep neural network model is deployed to the server for stone composition analysis. When the pre-qualified conditions are not met, the trained deep neural network model is fine-tuned, or a new deep neural network model is reconstructed and trained to obtain a new trained deep neural network model, which is evaluated until the new trained deep neural network model reaches more than 98% in both the training set and the test set. The new trained deep neural network model is exported and saved. Then, the embodiments of the present disclosure can deploy the new trained deep neural network model to the server, effectively improving the accuracy of stone composition analysis.

[0050] According to the method for analyzing the composition of urinary stones proposed in the embodiment of the present disclosure, the composition of the target original stones can be analyzed and classified based on the stone images of the target original stones that have been collected, and the target stone data set of the target original stones can be obtained. Then, data enhancement can be performed to obtain an enhanced stone data set. The enhanced stone data set can be subjected to data rebalancing processing to obtain a processed stone data set, thereby training a deep neural network model to obtain a trained deep neural network model, and deploying it to a server to receive the urinary stone composition sent by the server to obtain the analysis results output by the trained deep neural network model, thereby effectively improving the accuracy and applicability of the stone composition analysis. Thus, the problem that the infrared spectroscopy method in the related art is only used for the analysis of postoperative ex vivo samples, local stone sampling leads to misjudgment of stone composition, and reduces the accuracy of stone composition analysis is solved. In addition, the dual-energy CT method is limited by the equipment resolution, which reduces the applicability of stone composition analysis.

[0051] Next, the device for analyzing the composition of urinary stones according to the embodiment of the present disclosure will be described with reference to the accompanying drawings.

[0052] FIG2 is a block diagram of an apparatus for analyzing the composition of urinary stones according to an embodiment of the present disclosure.

[0053] As shown in FIG2 , the device 10 for analyzing the composition of urinary stones includes: a collection module 100 , a first acquisition module 200 and a second acquisition module 300 .

[0054] Specifically, the acquisition module 100 is used to acquire a stone image of the target original stone, analyze the stone composition of the target original stone according to the stone image, and classify the stone composition to obtain a target stone data set of the target original stone.

[0055] The first acquisition module 200 is used to perform data enhancement on the target stone dataset to obtain an enhanced stone dataset, and perform data rebalancing processing on the enhanced stone dataset to obtain a processed stone dataset.

[0056] The second acquisition module 300 is used to train the deep neural network model using the processed stone data set to obtain the trained deep neural network model, and deploy the trained deep neural network model to the server to receive the urinary stone components sent by the server to obtain the analysis results output by the trained deep neural network model.

[0057] In some embodiments, the acquisition module 100 includes: an acquisition unit and a determination unit.

[0058] The acquisition unit is used to acquire images of the target original stone before and after laser lithotripsy.

[0059] The determination unit is used to analyze the stone images before and after laser lithotripsy to obtain the stone composition of the target original stone.

[0060] In some embodiments, the first acquisition module 200 includes: an acquisition unit.

[0061] Among them, the acquisition unit is used to obtain the classification scale of the stone images in the enhanced stone data set, detect whether the classification scale meets the preset balanced classification conditions, upsample the classification in the classification scale that does not meet the preset balanced classification conditions, and obtain the balanced classification after upsampling, so as to obtain the balanced stone data set according to the balanced classification after upsampling.

[0062] In some embodiments, the apparatus 10 of the embodiment of the present disclosure further includes: a judgment module, a first processing module, and a second processing module.

[0063] Among them, the judgment module is used to determine whether the trained deep neural network model meets the preset qualification conditions.

[0064] The first processing module is used to deploy the trained deep neural network model to the server if the preset qualification conditions are met before the trained deep neural network model is deployed to the server.

[0065] The second processing module is used to rebuild a new deep neural network model and train the new deep neural network model to obtain a new trained deep neural network model if the preset qualification conditions are not met before the trained deep neural network model is deployed to the server, until the new trained deep neural network model meets the preset qualification conditions and the new trained deep neural network model is deployed to the server.

[0066] It should be noted that the above explanation of the embodiment of the method for analyzing the components of urinary stones is also applicable to the device for analyzing the components of urinary stones in this embodiment, and will not be repeated here.

[0067] According to the device for analyzing the composition of urinary stones proposed in the embodiment of the present disclosure, the composition of the target original stones can be analyzed and classified based on the stone images of the target original stones that have been collected, and the target stone data set of the target original stones can be obtained. Then, data enhancement can be performed to obtain an enhanced stone data set. The enhanced stone data set can be subjected to data rebalancing processing to obtain a processed stone data set, thereby training a deep neural network model to obtain a trained deep neural network model, and deploying it to a server to receive the urinary stone composition sent by the server to obtain the analysis results output by the trained deep neural network model, thereby effectively improving the accuracy and applicability of the stone composition analysis. As a result, the problem that the infrared spectroscopy method in the related art is only used for the analysis of postoperative ex vivo samples, local stone sampling leads to misjudgment of stone composition, and reduces the accuracy of stone composition analysis is solved. In addition, the dual-energy CT method is limited by the equipment resolution, which reduces the applicability of stone composition analysis.

[0068] FIG3 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure. The electronic device may include:

[0069] Memory 301 , processor 302 , and computer programs stored in the memory 301 and executable on the processor 302 .

[0070] When the processor 302 executes the program, the method for analyzing the composition of urinary stones provided in the above embodiment is implemented.

[0071] In addition, electronic equipment also includes:

[0072] The communication interface 303 is used for communication between the memory 301 and the processor 302 .

[0073] The memory 301 is used to store computer programs that can be run on the processor 302 .

[0074] The memory 301 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0075] If memory 301, processor 302, and communication interface 303 are implemented independently, communication interface 303, memory 301, and processor 302 can be interconnected via a bus and communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Buses can be categorized as address buses, data buses, control buses, and the like. For ease of illustration, FIG3 shows only one thick line, but this does not imply that there is only one bus or only one type of bus.

[0076] Optionally, in a specific implementation, if the memory 301 , the processor 302 and the communication interface 303 are integrated on a chip, the memory 301 , the processor 302 and the communication interface 303 can communicate with each other through an internal interface.

[0077] The processor 302 may be a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present disclosure.

[0078] This embodiment also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-mentioned method for analyzing the composition of urinary stones.

[0079] In the description of this specification, the description with reference to the terms "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present disclosure. In this specification, the schematic expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and the features of different embodiments or examples, unless they are contradictory.

[0080] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. Throughout the present disclosure, "N" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0081] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or N executable instructions for implementing a custom logical function or step of a process, and the scope of the preferred embodiments of the present disclosure includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present disclosure pertain.

[0082] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or N wires (electronic devices), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program can be obtained electronically by optically scanning the paper or other medium and then editing, interpreting or processing it in other suitable ways as necessary, and then storing it in a computer memory.

[0083] It should be understood that various parts of the present disclosure can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiment, the N steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, it can be implemented using any one or a combination of the following technologies known in the art: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0084] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0085] In addition, the functional units in the various embodiments of the present disclosure may be integrated into a single processing module, or each unit may exist physically separately, or two or more units may be integrated into a single module. The aforementioned integrated modules may be implemented in the form of hardware or in the form of software functional modules. If the integrated modules are implemented in the form of software functional modules and sold or used as independent products, they may also be stored in a computer-readable storage medium.

[0086] The storage medium mentioned above may be a read-only memory, a magnetic disk, or an optical disk, etc. Although the embodiments of the present disclosure have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present disclosure. A person of ordinary skill in the art may make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present disclosure.

Claims

1. A method for analyzing the composition of urinary stones, wherein: The following steps are involved: Collecting a stone image of a target original stone, analyzing the stone composition of the target original stone according to the stone image, and classifying the stone composition to obtain a target stone dataset of the target original stone; performing data enhancement on the target stone dataset to obtain an enhanced stone dataset, and performing data rebalancing processing on the enhanced stone dataset to obtain a processed stone dataset; as well as The processed stone data set is used to train a deep neural network model to obtain a trained deep neural network model, and the trained deep neural network model is deployed to a server to receive the urinary stone components sent by the server to obtain the analysis results output by the trained deep neural network model.

2. The method according to claim 1, wherein The collecting of a target original stone image and analyzing the target original stone composition according to the stone image includes: Collecting images of the target original stone before and after laser lithotripsy; The stone images before and after laser lithotripsy are analyzed to obtain the stone composition of the target original stone.

3. The method according to claim 1, wherein The step of performing data rebalancing on the enhanced stone dataset to obtain a balanced stone dataset comprises: Obtain the classification scale of the stone images in the enhanced stone data set, detect whether the classification scale meets the preset balanced classification condition, upsample the classification in the classification scale that does not meet the preset balanced classification condition, and obtain the upsampled balanced classification, so as to obtain the balanced stone data set according to the upsampled balanced classification.

4. The method according to claim 1, wherein Before deploying the trained deep neural network model to the server, the method further includes: Determine whether the trained deep neural network model meets the preset qualification conditions; If the preset criteria are met, the trained deep neural network model is deployed to the server; If the preset qualifying conditions are not met, a new deep neural network model is reconstructed and trained to obtain a new trained deep neural network model, until the new trained deep neural network model meets the preset qualifying conditions, and then the new trained deep neural network model is deployed to the server.

5. A device for analyzing the composition of urinary stones, wherein: include: an acquisition module, configured to acquire a stone image of a target original stone, analyze the stone composition of the target original stone according to the stone image, and classify the stone composition to obtain a target stone dataset of the target original stone; a first acquisition module, configured to perform data enhancement on the target stone dataset to obtain an enhanced stone dataset, and perform data rebalancing processing on the enhanced stone dataset to obtain a processed stone dataset; as well as The second acquisition module is used to train a deep neural network model using the processed stone data set to obtain a trained deep neural network model, and deploy the trained deep neural network model to a server to receive the urinary stone components sent by the server to obtain the analysis results output by the trained deep neural network model.

6. The device according to claim 5, wherein The acquisition module includes: an acquisition unit, configured to acquire images of the target original stone before and after laser lithotripsy; The determination unit is used to analyze the stone image before laser lithotripsy and the stone image after laser lithotripsy to obtain the stone composition of the target original stone.

7. The device according to claim 5, wherein The first acquisition module includes: An acquisition unit is used to obtain the classification scale of the stone images in the enhanced stone data set, detect whether the classification scale meets the preset balanced classification condition, upsample the classification in the classification scale that does not meet the preset balanced classification condition, and obtain the balanced classification after upsampling, so as to obtain the balanced stone data set according to the balanced classification after upsampling.

8. The device according to claim 5, wherein Also includes: A judgment module is used to judge whether the trained deep neural network model meets the preset qualification conditions; A first processing module is configured to deploy the trained deep neural network model to the server if the preset compliance condition is met before the trained deep neural network model is deployed to the server; The second processing module is used to rebuild a new deep neural network model and train the new deep neural network model to obtain a new trained deep neural network model if the preset qualification condition is not met before the trained deep neural network model is deployed to the server, until the new trained deep neural network model meets the preset qualification condition, and then deploy the new trained deep neural network model to the server.

9. An electronic device, wherein: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the method for analyzing the composition of urinary stones as described in any one of claims 1 to 4.

10. A computer-readable storage medium having a computer program stored thereon, wherein: The program is executed by a processor to implement the method for analyzing the composition of urinary stones as described in any one of claims 1 to 4.

Images