Prompting system, method, equipment and medium for hip replacement
By analyzing hammer sounds using deep learning models in hip replacement, providing real-time feedback, it solves the problem that surgeons have difficulty controlling hammer force accurately, and improves the safety and accuracy of the surgery.
Patent Information
- Application Number
- CN202510063830.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-15
- Publication Date
- 2025-04-08
AI Technical Summary
During hip replacement, surgeons have difficulty controlling the hammer force accurately, resulting in unfixed prosthesis or bone damage, lacking objectivity and real-time feedback.
A prompt system is adopted, including a sound acquisition module, a data preprocessing module, an audio multi-feature analysis module and a user real-time feedback module, and provides accurate feedback through end-to-end deep learning models.
It improves the safety and accuracy of prosthesis installation, reduces the risk of postoperative complications, enhances the intelligence and real-time nature of the surgery, and improves the work efficiency of doctors and the medical experience of patients.
Smart Images
Figure CN119770240A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of medical devices, and in particular to a prompt system, method, device and medium for hip replacement. Background Art
[0002] Hip replacement surgery is a common orthopedic procedure that aims to replace a hip joint damaged by disease or injury. This surgery usually requires implanting a prosthesis (an artificial joint) into the patient's bones to restore their function and relieve pain. However, in traditional surgical methods, surgeons mainly rely on their own experience to judge whether the fit between the prosthesis and the bone is tight enough.
[0003] In order to ensure that the prosthesis can be firmly fixed on the bone, surgeons usually use a hammer to tap the prosthesis. This process requires great precision, because if the hammering force is too strong, it may cause the surrounding healthy bone structure to break; on the contrary, if the hammering force is insufficient, the prosthesis may not achieve the ideal stability, thus affecting the postoperative recovery effect and even increasing the risk of reoperation. Therefore, how to control the hammering force has become a key issue.
[0004] Currently, in most cases, surgeons can only rely on their own feel and years of clinical experience to make decisions. However, this method has certain limitations: first, it lacks objectivity - even experienced doctors may be affected by various factors (such as fatigue, mood swings, etc.); second, without real-time data support, it is difficult to accurately evaluate the effect of each tap; finally, for novice or less experienced surgeons, the learning curve is long and prone to errors. Summary of the invention
[0005] In order to overcome the defects of the above-mentioned prior art, the present application provides a prompt system, method, device and medium for hip replacement surgery, which can provide real-time monitoring and feedback to improve the safety and accuracy of prosthesis installation.
[0006] In order to solve the above technical problems, this application provides the following technical solutions: According to a first aspect of an embodiment of the present application, a prompt system for hip replacement is provided, comprising: A sound collection module, used for collecting sound signals generated during the hip replacement surgery; A data preprocessing module, used for preprocessing the sound signal to extract target sound information corresponding to hammering during the installation of the prosthesis in the hip replacement surgery; the target sound information carries the start time and end time corresponding to each hammering during the installation of the prosthesis; An audio multi-feature analysis module, for analyzing the target sound information based on an end-to-end deep learning model by comprehensively applying time domain, frequency domain and time-frequency domain feature analysis techniques to obtain time domain features, frequency domain features and time-frequency domain features; and for fusing the time domain features, the frequency domain features and the time-frequency domain features to obtain fused features; and for selecting and optimizing the fused features to obtain indicative acoustic features; The user real-time feedback module is used to analyze and distinguish the indicative acoustic features in real time based on end-to-end real-time deep learning reasoning technology to obtain audio analysis results and provide feedback.
[0007] In an exemplary embodiment, the sound collection module includes a microphone array, and the microphone array is used to collect the sound signal and to adjust the weight of each microphone in the microphone array in real time according to the collection situation.
[0008] In an exemplary embodiment, the data preprocessing module includes an adaptive filter and a voice activity detection submodule, and the adaptive filter and the voice activity detection submodule are electrically connected; The adaptive filter is used to perform automatic gain control, noise reduction and echo cancellation on the sound signal by dynamically adjusting parameters to maximize the distinction between the hammer sound signal and the background noise; The voice activity detection submodule is used to segment the hammering sound signal to analyze and obtain the start time and end time corresponding to each hammering when the prosthesis is installed; The target sound information includes the hammering sound signal, and the start time and end time corresponding to each hammering when the prosthesis is installed.
[0009] In an exemplary embodiment, the deep learning model adopts a hybrid neural network model, which combines a convolutional neural network for extracting local features of the hammer sound signal and a long short-term memory network or Transformer for processing time series dependencies in the hammer sound signal.
[0010] In an exemplary embodiment, the audio multi-feature analysis module also includes a self-supervised learning model, which is used to learn and extract features that meet preset conditions from a large amount of unlabeled environmental sound data to enhance the ability to recognize and analyze hammer sounds in hip replacement surgery under data scarcity conditions.
[0011] In an exemplary embodiment, the deep learning model is deployed on a low-resource computing platform to ensure the real-time performance and computing efficiency of the prompt system.
[0012] In an exemplary embodiment, the system further comprises: A transfer learning framework is provided for transferring sound features matching the scenario of hip replacement from other medical sound analysis tasks to solve the problem of small amount of data on hip replacement as a specific type of surgery.
[0013] According to a second aspect of an embodiment of the present application, a method for prompting hip replacement is provided, which is applied to any one of the above-mentioned prompting systems for hip replacement, and the method comprises: The sound collection module collects the sound signal generated during the hip replacement surgery; The sound collection module transmits the sound signal to the data preprocessing module; The data preprocessing module preprocesses the sound signal to extract the target sound information corresponding to the hammering during the installation of the prosthesis in the hip replacement surgery; The data preprocessing module transmits the target sound information to the audio multi-feature analysis module; The audio multi-feature analysis module is based on an end-to-end deep learning model and comprehensively uses time domain, frequency domain and time-frequency domain feature analysis techniques to analyze the target sound information to obtain the time domain features, the frequency domain features and the time-frequency domain features; The audio multi-feature analysis module fuses the time domain feature, the frequency domain feature and the time-frequency domain feature to obtain the fused feature; The audio multi-feature analysis module selects and optimizes the fusion feature to obtain the indicative acoustic feature; The audio multi-feature analysis module transmits the indicative acoustic feature to the user real-time feedback module; The user real-time feedback module analyzes and distinguishes the indicative acoustic features in real time based on end-to-end real-time deep learning reasoning technology to obtain the audio analysis results and provide feedback.
[0014] According to a third aspect of an embodiment of the present application, there is provided an electronic device comprising a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or the at least one program is loaded and executed by the processor to implement the above-mentioned prompt method for hip replacement.
[0015] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, wherein at least one instruction or at least one program is stored in the storage medium, and the at least one instruction or the at least one program is loaded and executed by a processor to implement the above-mentioned prompt method for hip replacement.
[0016] By adopting the above technical solution, the present application has the following beneficial effects: (1) Compared with the audio processing method of template matching in the prior art, this application adopts an advanced audio processing method, which has the following advantages: First, through the end-to-end deep learning model, information is directly extracted from multi-dimensional acoustic features, avoiding the limitations of traditional template matching methods and improving the recognition and analysis capabilities of hammer sounds during surgery. At the same time, the end-to-end deep learning model can automatically learn and adapt to various complex audio patterns, thereby improving the intelligence level of the prompt system; second, since the traditional template matching method relies on a pre-defined template library, it may not be able to effectively recognize unknown or mutated sound patterns, while the end-to-end deep learning model can adaptively learn and recognize new audio features, thereby avoiding the limitations of the traditional template matching method; third, through the deep learning model, the prompt system can more accurately recognize the hammer sounds during surgery, improve recognition accuracy, reduce the possibility of misjudgment, and thus improve surgical safety and efficiency; fourth, since the end-to-end deep learning model has strong generalization capabilities, it can cope with changes in different surgical scenarios and operating methods, and enhance the generalization capabilities of the prompt system, making the prompt system more flexible and reliable in practical applications; fifth, the application of advanced audio processing technology enables doctors to more intuitively understand the sound changes during surgery, so as to make more accurate judgments and decisions, which helps to improve doctors' work efficiency and satisfaction and achieve the effect of improving user experience.
[0017] (2) This application significantly improves surgical safety and the accuracy of prosthesis installation by adopting a real-time feedback mechanism, reduces the risk of postoperative complications, and the immediacy of the feedback allows doctors to immediately know whether their operations are accurate, helps doctors quickly locate problems, and then adjust operations in a timely manner to ensure the smooth progress of the operation, saves time in finding errors, improves surgical efficiency, and avoids surgical failures or complications caused by misoperation. In addition, the real-time feedback mechanism also helps to improve the overall quality of medical services and provide patients with a better medical experience.
[0018] (3) Through transfer learning, existing large-scale data sets can be used to pre-train the deep learning model, so that the deep learning model can complete the training process of the prompt system for hip replacement in a shorter time, helping the deep learning model to better generalize to the prompt system for hip replacement, thereby enhancing the generalization ability and learning efficiency of the deep learning model, improving the performance and accuracy of the deep learning model, reducing the data collection cost of hip replacement, and effectively solving the problem of the small number of data sets for hip replacement as a specific type of surgery. In addition, since transfer learning can transfer knowledge from one field to another, it realizes and promotes cross-domain knowledge sharing and application.
[0019] (4) The design of the prompt system of this application takes into account the complex environment of the operating room (such as electromagnetic interference, noise and other factors in the operating room) and adopts anti-interference measures to ensure the stability and reliability of the prompt system in different surgical scenarios to adapt to various emergencies, thereby improving the adaptability and operational flexibility of the prompt system. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.
[0021] Figure 1 A structural block diagram of a prompt system for hip replacement provided in an embodiment of the present application; Figure 2 A practical application structure diagram of a prompt system for hip replacement provided in an embodiment of the present application Figure 3 A schematic diagram of a flow chart of a prompting method for hip replacement provided in an embodiment of the present application; Figure 4 A hardware structure block diagram of an electronic device that runs a prompt method for hip replacement surgery provided in an embodiment of the present application. DETAILED DESCRIPTION
[0022] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0023] The term "one embodiment" or "embodiment" as referred to herein refers to a specific feature, structure or characteristic that may be included in at least one implementation of the present application. In the description of the embodiments of the present application, it should be understood that the orientation or positional relationship indicated by the terms "upper", "lower", "top", "bottom", etc. is based on the orientation or positional relationship shown in the accompanying drawings, which is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation of the present application. The terms "first" and "second" are used only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, the features defined as "first" and "second" may include one or more of the features explicitly or implicitly. Moreover, the terms "first", "second", etc. are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described here can be implemented in an order other than those illustrated or described here. Unless otherwise clearly specified and limited, the terms "installed", "connected" and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0024] See also Figure 1 , which is a structural block diagram of a prompt system for hip replacement provided by an embodiment of the present application, the prompt system for hip replacement includes: A sound collection module, used to collect sound signals generated during hip replacement surgery; A data preprocessing module is used to preprocess the sound signal to extract the target sound information corresponding to the hammering during the installation of the prosthesis in hip replacement surgery; the target sound information carries the start time and end time corresponding to each hammering during the installation of the prosthesis; An audio multi-feature analysis module is used to analyze the target sound information based on an end-to-end deep learning model by comprehensively applying time domain, frequency domain and time-frequency domain feature analysis techniques to obtain time domain features, frequency domain features and time-frequency domain features; and to fuse the time domain features, frequency domain features and time-frequency domain features to obtain fused features; and to select and optimize the fused features to obtain indicative acoustic features; preferably, the end-to-end deep learning model directly realizes automatic learning from the original audio signal to the final recognition and classification of the prosthesis installation state, using structures including but not limited to convolutional neural networks (CNNs) and recurrent neural networks (RNNs) and Transformer to make full use of the time and frequency information in the sound signal, without using performance cross methods such as template matching, thereby improving the overall performance of the system and reducing the deviation of the model; The user real-time feedback module is used to analyze and distinguish indicative acoustic features in real time based on end-to-end real-time deep learning reasoning technology, running on the edge computing platform, to obtain audio analysis results and provide feedback; different from the template matching method, it can analyze and distinguish signals in real time in various actual scenarios, and provide real-time operation feedback to the surgeon through the graphical user interface and sound signals, including confirmation signals for successful installation and warning signals for potential problems, and the real-time feedback technology is low-latency; The sound collection module, the data preprocessing module, the audio multi-feature analysis module and the user real-time feedback module are electrically connected in sequence; Among them, the deep learning model is deployed on a low-resource computing platform to ensure the real-time and computing efficiency of the prompt system.
[0025] In a specific embodiment, the prompt system for hip replacement also includes a transfer learning framework for migrating sound features that match the scene of hip replacement from other medical sound analysis tasks to solve the problem of small amount of data for hip replacement as a specific type of surgery; The sound collection module includes a microphone array, which is used to collect sound signals and adjust the weight of each microphone in the microphone array in real time according to the collection situation; wherein the microphone array is designed to automatically adjust the sensitivity and directionality of each microphone to adapt to different sound sources and noise levels in the operating room; Optionally, the sound collection module also includes a calculation submodule and an adjustment device, the microphone array is installed on the adjustment device, the calculation submodule can be built into the adjustment device, and can also be installed on the adjustment device, the calculation submodule is electrically connected to the microphone array and the adjustment device, respectively, and is used to calculate and adjust the adjustment device according to the calculation result when the weights of each microphone determined by the microphone array do not meet the preset rule, so that the relative angle between the surface where the microphone array is located and the hammer point changes, so that the weights of each microphone finally obtained by the microphone array meet the preset rule; Preferably, the preset rule may be that the microphone at the center of the microphone array has the highest weight, and the weights of the surrounding microphones decrease from near to far, that is, the hammer sound collected by the microphone at the center of the microphone array is the largest, and the adjustment device may be an electric universal adjustment stand, and the adjustment device may be a processor; Specifically, the microphone array is composed of a single or multiple microphones configured to optimize the capture of sound signals, especially in the complex acoustic environment of the operating room, and can effectively reduce environmental noise and instrument interference such as instruments, conversations, and other noises; The data preprocessing module includes an adaptive filter and a voice activity detection submodule, and the adaptive filter and the voice activity detection submodule are electrically connected; An adaptive filter is used to automatically perform gain control, noise reduction and echo cancellation on the sound signal by dynamically adjusting parameters to maximize the distinction between the hammer sound signal and the background noise; Optionally, the adaptive filter monitors the level of the input signal in real time and adjusts the gain of the amplifier as needed to keep the output signal at a relatively stable level to achieve automatic gain control, and then uses a digital signal processing algorithm to identify and suppress non-speech components to achieve noise reduction, and then estimates the echo path and generates a signal opposite to the echo to offset it to achieve echo cancellation, thereby achieving the effect of maximizing the distinction between the hammer sound signal and the background noise, and then uses the high-pass and low-pass filters in the adaptive filter to accurately extract the key frequency components of the hammer sound; The voice activity detection submodule is used to segment the hammer sound signal to analyze and obtain the start time and end time of each hammer during the installation of the prosthesis; the hammer point can be accurately located through voice activity detection technology, the start time of each hammer can be saved, and the audio waveform can be adaptively processed in real time as subsequent input; The target sound information includes the hammer sound signal, as well as the start time and end time corresponding to each hammer when the prosthesis is installed; The audio multi-feature analysis module includes a hybrid neural network model and a self-supervised learning model, and the hybrid neural network model and the self-supervised learning model are electrically connected; The hybrid neural network model combines a convolutional neural network for extracting local features of the hammer sound signal and a long short-term memory network or Transformer for processing time series dependencies in the hammer sound signal; The self-supervised learning model is used to learn and extract features that meet preset conditions from a large amount of unlabeled environmental sound data, so as to enhance the ability to recognize and analyze hammer sounds in hip replacement surgery under data scarce conditions.
[0026] In an optional embodiment, the end-to-end deep learning model can also use techniques such as quantization, pruning, and knowledge distillation to optimize the computational complexity and storage requirements of the model, so that the prompt system can operate stably in a surgical environment with high real-time requirements while maintaining high recognition accuracy.
[0027] like Figure 2 As shown, in an actual application scenario, the prompt system for hip replacement surgery may include a sound collection component, an audio multi-feature real-time analysis component, and a real-time feedback prompt component.
[0028] It should be noted that the system provided in the above embodiment, when implementing its functions, only uses the division of the above-mentioned functional modules as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.
[0029] It can be seen from the above technical solutions of the embodiments of the present application that the embodiments of the present application provide an intelligent and real-time prosthesis installation prompt system for artificial hip replacement by utilizing microphone arrays, data preprocessing and audio multi-feature analysis technology. The system accurately captures the hammering sound during surgery through the microphone array, analyzes the hammering audio data through advanced audio processing technology, and provides real-time feedback on the combination status of the prosthesis and the bone, thereby improving the accuracy and safety of the surgery and having broad application prospects.
[0030] The innovation of the technical solution in the embodiment of the present application is that a deep learning model is used to perform a comprehensive feature analysis of the audio signal, including the fusion and optimization of time domain, frequency domain and time-frequency domain features. The feature extraction part involves a variety of feature extraction methods, but the system can optimize the audio analysis process through an automated learning process to improve the recognition accuracy. Unlike traditional methods, the system as a whole does not rely on manually designed features, but automatically extracts and optimizes key features in the sound signal through a deep learning model, thereby achieving accurate monitoring and feedback of whether the bone handle size is appropriate and whether the bone handle fixation state is reliable during the prosthesis installation process. Compared with the traditional template matching method, the prompt system based on end-to-end deep learning in this technical solution has stronger adaptability, robustness and generalization ability, can handle complex and changeable sound data, and provide accurate feedback in real time. Its ability of automatic feature learning and extraction, multi-level feature fusion and intelligent decision-making feedback makes it show obvious advancement and advantages in improving surgical safety and accuracy.
[0031] Specifically: 1. Automatic feature learning and extraction vs. manual template design: The template matching method involved in the prior art relies on pre-designed templates or rules, usually based on manual experience and simple feature extraction methods (such as amplitude, frequency or time window, etc.). This method requires the system to define the template in advance and compare it with the audio signal to be detected. This makes the template matching method very sensitive to variant or uncertain signals (such as changes in sound intensity, noise interference, etc.), and is easily affected by noise and interference, making it unsuitable for use in operating rooms.
[0032] The end-to-end deep neural network in this technical solution can automatically learn useful features from raw data without manually designing templates. This makes the deep learning model more robust and adaptable when processing complex and changeable sound data. It can autonomously extract the most relevant features for analysis in a variety of different sound patterns, hammer strength, frequency changes, etc., greatly reducing dependence on manual intervention.
[0033] 2. Adaptability and generalization ability: Template matching methods used in existing technologies usually perform poorly when faced with new, unseen sound types or environmental noises because they rely only on specific templates or rules. When there are certain deviations in the sound signal (such as noise interference, changes in the severity of hammer sounds, etc.), template matching methods are difficult to deal with effectively.
[0034] The end-to-end method in this technical solution can automatically learn the diversity and complexity of audio signals through a large amount of training data, and has strong generalization ability. Even in different surgical environments, different doctors' operations or other uncertain factors, the deep learning model can still maintain high accuracy and robustness and adapt to new data.
[0035] 3. Multi-level feature fusion vs. single feature matching: Template matching methods in the prior art usually rely on direct comparison with a predetermined template, use limited features, and can usually only process simple, single-dimensional information (such as frequency, amplitude, etc.). If the target sound has a small difference from the template, the template matching method often cannot effectively identify it.
[0036] The end-to-end approach in this technical solution can comprehensively analyze the time domain, frequency domain and time-frequency domain features, and fuse and optimize these multi-dimensional features. This fusion can not only improve the model's ability to recognize complex sounds, but also improve the accuracy and sensitivity of the system. The deep learning model can learn and judge at the level of multiple features, thereby achieving higher-precision recognition and feedback.
[0037] 4. Real-time feedback capability and intelligent decision-making: Template matching methods in the prior art usually require matching through pre-designed rules during execution. Once the sound signal does not match the predetermined template, it may lead to wrong judgment or no feedback. The feedback mechanism of template matching methods often does not have the ability of real-time learning and dynamic adjustment.
[0038] The deep learning method in this technical solution can not only analyze audio signals in real time, but also optimize the analysis results through the training process, and provide intelligent decision support by combining historical data and real-time feedback. The model can dynamically adjust the feedback strategy according to the changes in the actual surgery (such as hammer sounds of different intensities and frequencies), thereby providing doctors with more accurate real-time suggestions.
[0039] 5. Reduce human errors and biases: The template matching methods involved in the prior art strongly rely on manually defined templates and rules. Improper manual design or incorrect template selection may lead to high false alarm and missed alarm rates. Especially when facing unknown or changing sound signals, the effectiveness of the template matching method will be greatly reduced.
[0040] The end-to-end method in this technical solution can be trained based on actual clinical in-situ data collection, effectively adjust model parameters, and effectively reduce errors caused by human intervention. Its data-driven characteristics enable the system to optimize based on actual sound signals, reducing the impact of manual definition deviations on system performance.
[0041] Corresponding to the prompt system for hip replacement provided in the above-mentioned embodiment, the embodiment of the present application also provides a prompt method for hip replacement. Since the prompt method for hip replacement provided in the embodiment of the present application corresponds to the prompt system for hip replacement provided in the above-mentioned embodiment, the implementation method of the aforementioned prompt system for hip replacement is also applicable to the prompt method for hip replacement provided in the present embodiment, and will not be described in detail in the present embodiment.
[0042] See also Figure 3 , which is a flow chart of a prompting method for hip replacement provided in an embodiment of the present application; the method comprises: Step S1: The sound collection module collects the sound signal generated during hip replacement surgery; Step S2: The sound collection module transmits the sound signal to the data preprocessing module; Step S3: the data preprocessing module preprocesses the sound signal to extract the target sound information corresponding to the hammering during the installation of the prosthesis in hip replacement surgery; Step S4: the data preprocessing module transmits the target sound information to the audio multi-feature analysis module; Step S5: The audio multi-feature analysis module analyzes the target sound information based on an end-to-end deep learning model by comprehensively using time domain, frequency domain and time-frequency domain feature analysis techniques to obtain time domain features, frequency domain features and time-frequency domain features; Step S6: the audio multi-feature analysis module fuses the time domain features, the frequency domain features and the time-frequency domain features to obtain a fusion feature; Step S7: the audio multi-feature analysis module selects and optimizes the fusion features to obtain indicative acoustic features; Step S8: the audio multi-feature analysis module transmits the indicative acoustic features to the user real-time feedback module; Step S9: The user real-time feedback module analyzes and distinguishes the indicative acoustic features in real time based on end-to-end real-time deep learning reasoning technology to obtain audio analysis results and provide feedback.
[0043] It should be noted that the method and system embodiment provided in the above embodiments belong to the same concept, and the specific application process is detailed in the system embodiment, which will not be repeated here.
[0044] An embodiment of the present application also provides an electronic device, including a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or at least one program is loaded and executed by the processor to implement a prompt method for hip replacement as provided in the above method embodiment.
[0045] The memory can be used to store software programs and modules. The processor executes various functional applications and realizes high-level autonomous driving by running the software programs and modules stored in the memory. The memory may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, application programs required for functions, etc.; the data storage area may store data created according to the use of the device, etc. In addition, the memory may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage device. Accordingly, the memory may also include a memory controller to provide the processor with access to the memory.
[0046] The method embodiments provided in the embodiments of the present application may be executed in a computer terminal, a server or a similar computing device, that is, the above-mentioned electronic device may include a computer terminal, a server or a similar computing device.
[0047] Figure 4 is a hardware structure block diagram of an electronic device for running a prompt method for hip replacement provided by an embodiment of the present application, such as Figure 4 As shown, the internal structure of the electronic device may include but is not limited to: a processor, a network interface and a memory. The processor, network interface and memory in the electronic device may be connected via a bus or other means. Figure 4 The example of connecting through bus is taken in the following.
[0048] Among them, the processor (or CPU (Central Processing Unit)) is the computing core and control core of the electronic device. The network interface may optionally include a standard wired interface, a wireless interface (such as WI-FI, a mobile communication interface, etc.). Memory is a memory device in an electronic device for storing programs and data. It is understandable that the memory here can be a high-speed RAM storage device or a non-volatile memory device, such as at least one disk storage device; optionally, it can also be at least one storage device located away from the aforementioned processor. The memory provides a storage space, which stores the operating system of the electronic device, which may include but is not limited to: Windows system (an operating system), Linux (an operating system), Android (Android, a mobile operating system) system, IOS (a mobile operating system) system, etc., and this application does not limit this; and, in the storage space, one or more instructions suitable for being loaded and executed by the processor are also stored, and these instructions can be one or more computer programs (including program codes). In the embodiment of this specification, the processor loads and executes one or more instructions stored in the memory to implement the prompt method for hip replacement provided in the above method embodiment.
[0049] An embodiment of the present application also provides a computer-readable storage medium, in which at least one instruction or at least one program is stored, and the at least one instruction or at least one program is loaded and executed by a processor to implement the prompt method for hip replacement provided in the method embodiment.
[0050] Optionally, in this embodiment, the above-mentioned storage medium may include but is not limited to: a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk, and other media that can store program codes.
[0051] It should be noted that the above-mentioned sequence of the embodiments of the present application is for description only and does not represent the advantages and disadvantages of the embodiments. And the above describes a specific embodiment of this specification. Other embodiments are within the scope of the attached claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multi-small sample image classification and parallel processing are also possible or may be advantageous.
[0052] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.
[0053] A person skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware or by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.
[0054] The above are only preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.
Claims
1. A prompt system for hip replacement, characterized in that: include: A sound collection module, used for collecting sound signals generated during the hip replacement surgery; A data preprocessing module, used for preprocessing the sound signal to extract target sound information corresponding to hammering during the installation of the prosthesis in the hip replacement surgery; The target sound information carries the start time and end time corresponding to each hammering when the prosthesis is installed; An audio multi-feature analysis module is used to analyze the target sound information based on an end-to-end deep learning model by comprehensively using time domain, frequency domain and time-frequency domain feature analysis techniques to obtain time domain features, frequency domain features and time-frequency domain features; and used to fuse the time domain feature, the frequency domain feature and the time-frequency domain feature to obtain a fused feature; and for selecting and optimizing the fused features to obtain indicative acoustic features; The user real-time feedback module is used to analyze and distinguish the indicative acoustic features in real time based on end-to-end real-time deep learning reasoning technology to obtain audio analysis results and provide feedback.
2. The prompt system for hip replacement according to claim 1, characterized in that: The sound collection module includes a microphone array, and the microphone array is used to collect the sound signal and to adjust the weight of each microphone in the microphone array in real time according to the collection situation.
3. The prompt system for hip replacement according to claim 2, characterized in that: The data preprocessing module includes an adaptive filter and a voice activity detection submodule, and the adaptive filter and the voice activity detection submodule are electrically connected; The adaptive filter is used to perform automatic gain control, noise reduction and echo cancellation on the sound signal by dynamically adjusting parameters to maximize the distinction between the hammer sound signal and the background noise; The voice activity detection submodule is used to segment the hammering sound signal to analyze and obtain the start time and end time corresponding to each hammering when the prosthesis is installed; The target sound information includes the hammering sound signal, and the start time and end time corresponding to each hammering when the prosthesis is installed.
4. The prompt system for hip replacement according to claim 3, characterized in that: The deep learning model adopts a hybrid neural network model, which combines a convolutional neural network for extracting local features of the hammer sound signal and a long short-term memory network or Transformer for processing time series dependencies in the hammer sound signal.
5. The prompt system for hip replacement according to claim 4, characterized in that: The audio multi-feature analysis module also includes a self-supervised learning model, which is used to learn and extract features that meet preset conditions from a large amount of unlabeled environmental sound data to enhance the ability to recognize and analyze hammer sounds in hip replacement surgery under data scarcity conditions.
6. The prompt system for hip replacement according to claim 1, characterized in that: The deep learning model is deployed on a low-resource computing platform to ensure the real-time performance and computing efficiency of the prompt system.
7. The prompt system for hip replacement according to any one of claims 1 to 6, characterized in that: The system further comprises: A transfer learning framework is provided for transferring sound features matching the scenario of hip replacement from other medical sound analysis tasks to solve the problem of small amount of data on hip replacement as a specific type of surgery.
8. A method for prompting hip replacement, applied to the prompting system for hip replacement as claimed in any one of claims 1 to 7, characterized in that: The method comprises: The sound collection module collects the sound signal generated during the hip replacement surgery; The sound collection module transmits the sound signal to the data preprocessing module; The data preprocessing module preprocesses the sound signal to extract the target sound information corresponding to the hammering during the installation of the prosthesis in the hip replacement surgery; The data preprocessing module transmits the target sound information to the audio multi-feature analysis module; The audio multi-feature analysis module is based on an end-to-end deep learning model and comprehensively uses time domain, frequency domain and time-frequency domain feature analysis techniques to analyze the target sound information to obtain the time domain features, the frequency domain features and the time-frequency domain features; The audio multi-feature analysis module fuses the time domain feature, the frequency domain feature and the time-frequency domain feature to obtain the fused feature; The audio multi-feature analysis module selects and optimizes the fusion feature to obtain the indicative acoustic feature; The audio multi-feature analysis module transmits the indicative acoustic feature to the user real-time feedback module; The user real-time feedback module analyzes and distinguishes the indicative acoustic features in real time based on end-to-end real-time deep learning reasoning technology to obtain the audio analysis results and provide feedback.
9. An electronic device, characterized in that: It comprises a processor and a memory, wherein the memory stores at least one instruction or at least one program, and the at least one instruction or the at least one program is loaded and executed by the processor to implement the prompt method for hip replacement as claimed in claim 8.
10. A computer-readable storage medium, wherein at least one instruction or at least one program is stored in the storage medium, and the at least one instruction or the at least one program is loaded and executed by a processor to implement the prompt method for hip replacement as described in claim 8.