Orthopedic implant monitoring method, device and system and readable storage medium

Abstract

The embodiment of the invention provides an orthopedic implant monitoring method, device and system and a readable storage medium, and the method comprises the steps: obtaining a sound signal generated by a knocking operation in a process that an implant is embedded into a designated region in a skeleton structure; and acoustic characteristics of the sound signal are acquired, and according to the acoustic characteristics, fitting degree information of the implant and a designated area in the skeleton structure is determined.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of communication technologies, and particularly to a method, device, system and readable storage medium for monitoring orthopedic implants. Background Art

[0002] Arthroplasty in surgical operations is a common treatment method widely used in the orthopedic field to restore joint function, relieve pain and improve the quality of life of patients. In arthroplasty, an artificial joint is implanted into the patient's body to replace the damaged or diseased natural joint. This surgery involves the use of orthopedic implants, such as artificial joint components, to replace the skeletal structure of the joint. However, ensuring the tight fit and accurate placement of these implants with the patient's bones is a crucial step in the surgery.

[0003] For example, in hip arthroplasty surgery, the patient's natural hip joint may be replaced by an artificial hip joint. This involves implanting components such as an acetabular cup component and a femoral head component into the patient's hip joint area to reconstruct the function of the hip joint. To ensure the firm fixation of these implants, orthopedic surgeons need to properly insert or impact them into the patient's skeletal anatomy. However, this process is not an easy task because ensuring the implants are in the right position requires experience and skill.

[0004] Traditionally, orthopedic surgeons mainly rely on tactile and auditory feedback during surgery to judge whether the implants are properly placed. Surgeons may use orthopedic mallets, impactors or inserter tools and judge whether the implants are correctly positioned by the force felt during impact or insertion. However, this subjective sensory feedback may vary depending on the surgeon's experience and touch, resulting in under-impaction or over-impaction of the implants.

[0005] Over-impaction may cause the patient's bone fracture, while under-impaction may cause the implant to loosen, thus affecting the surgical outcome and the patient's recovery. In addition, the tactile and auditory feedback during surgery is affected by many factors, such as the noise in the surgical environment, the differences in surgical instruments and the subjective judgment of the surgeon, which may further increase the risk of misjudgment. Summary of the Invention

[0006] The embodiments of the present application aim to provide a method, device, system and readable storage medium for monitoring orthopedic implants to solve the problem of how to ensure the accurate fitting of implants with bones.

[0007] In a first aspect, a method for monitoring orthopedic implants is provided, including:

[0008] During the process of implanting the implant into a specified area of the bone structure, acquiring a sound signal generated by a knocking operation;

[0009] Obtain the acoustic characteristics of the sound signal, and determine the fitting degree information between the implant and the specified area in the bone structure according to the acoustic characteristics.

[0010] Optionally, obtaining the acoustic characteristics of the sound signal, and determining the fitting degree information between the implant and the specified area in the bone structure includes:

[0011] Segment the sound signal into separate tapping events, and obtain the acoustic characteristics of each tapping event;

[0012] Analyze and classify the acoustic characteristics of the tapping events to obtain the classification results of the tapping events, and determine the fitting degree information between the implant and the specified area in the bone structure according to the classification results.

[0013] Optionally, analyzing and classifying the acoustic characteristics of the tapping events to obtain the classification results of the tapping events includes:

[0014] Create a feature vector for each tapping event through the acoustic characteristics;

[0015] Classify the tapping events through a machine learning algorithm to obtain the classification results of the tapping events, and the classification results include good fit or poor fit.

[0016] Optionally, the method further includes:

[0017] Evaluate the fitting degree information between the implant and the bone structure to identify whether the implant is suitable and can fill the specified area in the bone structure.

[0018] Optionally, the method further includes:

[0019] Collect ambient sound signals;

[0020] Compare the acoustic characteristics of the collected ambient sound signals with a pre-established sound pattern library to determine whether there are abnormal sounds;

[0021] If abnormal sounds are detected, trigger an alarm process.

[0022] Optionally, the alarm process includes:

[0023] Display the abnormal detection results through the user interface;

[0024] Obtain first information through the user interface, and the first information is used to indicate the existence of an abnormal situation or the non-existence of an abnormal situation;

[0025] If the first information indicates the existence of an abnormal situation, send an alarm trigger signal.

[0026] Second aspect, a monitoring device for an orthopedic implant is provided, including:

[0027] A first acquisition module, configured to acquire a sound signal generated by a tapping operation during the process of embedding the implant into a specified area of a bone structure;

[0028] A first processing module, configured to acquire acoustic features of the sound signal, and determine the fitting degree information between the implant and the specified area in the bone structure according to the acoustic features.

[0029] Optionally, the first processing module is further configured to:

[0030] Segment the sound signal into separate tapping events, and acquire acoustic features of each tapping event;

[0031] Analyze and classify the acoustic features of the tapping events, obtain classification results of the tapping events, and determine the fitting degree information between the implant and the specified area in the bone structure according to the classification results.

[0032] Optionally, the first processing module is further configured to:

[0033] Create a feature vector for each tapping event through the acoustic features;

[0034] Classify the tapping events through a machine learning algorithm, obtain classification results of the tapping events, and the classification results include good fit or poor fit.

[0035] Optionally, the device further includes:

[0036] An evaluation module, configured to evaluate the fitting degree information between the implant and the bone structure to identify whether the implant is suitable and can fill the specified area in the bone structure.

[0037] Optionally, the device further includes:

[0038] A second acquisition module, configured to acquire an environmental sound signal;

[0039] A second processing module, configured to compare the acoustic features of the acquired environmental sound signal with a pre-established sound pattern library to determine whether there is an abnormal sound;

[0040] An alarm module, configured to trigger an alarm process if an abnormal sound is detected.

[0041] Optionally, the alarm process includes:

[0042] Displaying the abnormal detection result through a user interface;

[0043] Obtain first information through a user interface, where the first information is used to indicate the existence of an abnormal situation or indicate the non-existence of an abnormal situation;

[0044] If the first information indicates the existence of an abnormal situation, then send an alarm trigger signal.

[0045] In a third aspect, there is provided a surgical assistance system including the device as described in the second aspect.

[0046] In a fourth aspect, there is provided a readable storage medium, on which a program or instruction is stored, and when the program or instruction is executed by a processor, the steps of the method as described in the first aspect are implemented.

[0047] In a fifth aspect, there is provided a readable storage medium, on which a program or instruction is stored, and when the program or instruction is executed by a processor, the steps of the method as described in the first aspect are implemented.

[0048] In this application, during the process of implanting an implant into a specified area of a bone structure, a sound signal generated by a tapping operation is obtained; the acoustic characteristics of the sound signal are obtained, and based on the acoustic characteristics, the fitting degree information between the implant and the specified area in the bone structure is determined to ensure the accurate fitting of the implant to the bone. By providing quantitative and repeatable feedback, more accurate and repeatable implant positioning is achieved, the surgical quality is improved, and the patient's safety is enhanced. Description of the Drawings

[0049] By reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of this application. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0050] Figure 1 is a flowchart of a method for monitoring an orthopedic implant provided by an embodiment of this application;

[0051] Figure 2 is a schematic diagram of a surgical assistance system based on sound signals provided by an embodiment of this application;

[0052] Figure 3 is a schematic diagram of the process of sound acquisition and transmission provided by an embodiment of this application;

[0053] Figure 4 is a schematic diagram of the user interface of a surgical assistance system based on sound signals provided by an embodiment of this application;

[0054] Figure 5 is a schematic diagram of the sound processing process provided by an embodiment of this application;

[0055] Figure 6 It is a schematic diagram of the abnormal detection and alarm triggering process provided by an embodiment of the present application;

[0056] Figure 7 It is a schematic diagram of a monitoring device for orthopedic implants provided by an embodiment of the present application. Detailed implementation manners

[0057] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, rather than all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0058] The term "including" and any variation thereof in the specification and claims of the present application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products, or devices. In addition, the use of "and / or" in the specification and claims means at least one of the connected objects. For example, A and / or B means including three cases: A alone, B alone, and both A and B existing.

[0059] In the embodiments of the present application, words such as "exemplary" or "for example" are used to represent examples, illustrations, or explanations. Any embodiment or design solution described as "exemplary" or "for example" in the embodiments of the present application should not be construed as being more preferred or having more advantages than other embodiments or design solutions. Rather, the use of words such as "exemplary" or "for example" is intended to present relevant concepts in a specific manner.

[0060] Refer to Figure 1 , an embodiment of the present application provides a method for monitoring orthopedic implants, and the specific steps include: Step 101 and Step 102.

[0061] Step 101: During the process of implanting the implant into a specified area of the bone structure, obtain the sound signal generated by the tapping operation;

[0062] It can be understood that the tapping (or hammering) operation refers to the operation of implanting the implant into a specified area of the bone structure.

[0063] For example, the implant can be an artificial prosthesis in joint replacement surgery, such as an artificial femoral head prosthesis in hip replacement surgery, and the specified area can be the femoral cavity, but is not limited thereto.

[0064] Step 102: Obtain the acoustic features of the sound signal, and determine the fitting degree information between the implant and a specified area in the bone structure according to the acoustic features.

[0065] Optionally, the acoustic features of the sound signal include at least one of the following: information such as frequency, time, waveform, amplitude, power, and energy.

[0066] Optionally, the fitting degree information can be used to indicate whether the implant is accurately fitted to the bone, or to indicate the position where the implant impacts the bone.

[0067] Optionally, the fitting degree information includes but is not limited to at least one of the following: information such as the gap thickness, separation degree, dislocation amount, and adhesion state between the implant and a specified area in the bone structure.

[0068] In an embodiment of the present application, obtaining the acoustic features of the sound signal and determining the fitting degree information between the implant and a specified area in the bone structure according to the acoustic features includes:

[0069] Segment the sound signal into individual tapping events, and obtain the acoustic features of each tapping event;

[0070] Analyze and classify the acoustic features of the tapping events to obtain the classification results of the tapping events, and determine the fitting degree information between the implant and a specified area in the bone structure according to the classification results.

[0071] In an embodiment of the present application, analyzing and classifying the acoustic features of the tapping events to obtain the classification results of the tapping events includes

[0072] Create a feature vector for each tapping event through the acoustic features;

[0073] Classify the tapping events through a machine learning algorithm to obtain the classification results of the tapping events, and the classification results include good fitting or poor fitting.

[0074] In an embodiment of the present application, the method further includes:

[0075] Evaluate the fitting degree information between the implant and the bone structure to identify whether the implant is suitable and can fill a specified area in the bone structure.

[0076] In an embodiment of the present application, the method further includes:

[0077] Collect ambient sound signals;

[0078] Compare the acoustic features of the collected ambient sound signals with a pre-established sound pattern library to determine whether there are abnormal sounds;

[0079] If an abnormal sound is detected, an alarm process is triggered.

[0080] In an embodiment of the present application, the alarm process includes:

[0081] Displaying the abnormal detection result through the user interface;

[0082] Obtaining first information through the user interface, where the first information is used to indicate the existence of an abnormal situation or the non-existence of an abnormal situation;

[0083] If the first information indicates the existence of an abnormal situation, an alarm trigger signal is sent.

[0084] Optionally, in the present application, an environmental sound signal can be collected through a sound recognition alarm system. This system can efficiently and accurately identify specific sound patterns, such as bone fracture sounds, tearing sounds, etc., and trigger an alarm notification in a timely manner when an abnormal sound occurs. This system consists of multiple closely cooperating components, jointly realizing precise sound recognition and abnormal situation alarm.

[0085] Optionally, the sound recognition alarm system includes a microphone, a sound processing unit, an abnormal detection module, an alarm, and a user interface. The microphone is used to collect the sound signal in the environment. The position and arrangement of the microphone ensure that the sound generated during the operation can be effectively received, thereby providing reliable sound source data.

[0086] The sound processing unit plays an important role in real-time processing and feature extraction of the sound signal. By performing real-time sound processing on the captured sound signal, noise can be removed, acoustic features can be enhanced, and operations such as spectrum analysis can be carried out, thereby providing valuable information for subsequent sound pattern recognition.

[0087] The abnormal detection module is used to analyze the acoustic features processed by the sound processing unit in order to accurately detect sound abnormalities. This module uses a pre-established sound pattern library for comparison to determine whether the sound conforms to the normal operating conditions. Once an abnormal sound is detected, this module will immediately trigger the alarm process.

[0088] The alarm is used to emit an emergency alarm signal in case of a sound abnormality. The alarm signal can be in various forms such as a sound alarm, a light signal, a vibration signal, etc., to ensure that doctors and the surgical team can quickly become aware of the occurrence of an abnormal situation.

[0089] The user interface is a window for providing interactive operations and status displays for the system. Through the user interface, doctors can monitor the system status in real time, adjust settings, and obtain detailed sound analysis results, thereby making the operation more intuitive and controllable.

[0090] This embodiment realizes the complete process from sound capture to abnormal alarm, providing a reliable auxiliary tool for surgical operations.

[0091] In a specific embodiment, an exemplary surgical operation auxiliary system based on sound signals is provided, such as Figure 2 shown. The system includes one or more microphones, one or more displays, a processing unit, a controller, and an alarm. It should be noted that, depending on the types of components adopted, all or part of the microphones, displays, and control units can be integrated or fused into a multifunctional device, as is well-known to those skilled in the art. The systems and methods described in this application are not only applicable to hip replacement surgeries but can also be seamlessly applied to other types of orthopedic surgeries. For example, the designs of these components can be modified in different orthopedic surgical operations, so the details and designs of the system can be adjusted according to actual needs to be applicable to orthopedic surgeries not limited to hip replacement surgeries. In other words, the essence and scope of the present invention are constructed as a platform technology with scalability that can be applied to a variety of surgical procedures.

[0092] The microphone can be placed inside or outside the sterile area of the operating room to capture the sound information generated during the real-time surgical process. The position of the microphone is arranged such that it can capture the sound between the hammer and the broach introducer, the sound between the broach and the cancellous bone, and the sound between the broach and the cortical bone in the patient's body. There are no restrictions on the number and type of microphones, so standard microphones can be used, or high-fidelity microphones can be selected, and the specific choice depends on the actual situation, as is well-known to professionals in the field.

[0093] The display is used to display a user interface, and the user interface is used to receive and display audio and / or visual data related to the ongoing surgical operation. The display can be various computing devices, such as a monitor, a TV, a smartphone, a tablet, or any digital screen capable of presenting audio and visual data. In addition, the display can also be a multi-screen display system that displays the same or different images received from the control unit. On the other hand, the display can also act as a computing device and directly receive signals from the microphone component.

[0094] The controller usually consists of one or more computing devices. The controller is responsible for running the described algorithms and software. It receives sound signals from the microphone and processes the data by analyzing the acoustic features. The processed data can be output through a display to provide real-time sound feedback and fit information. The controller is connected to the microphone and the display through a communication network, receives the sound signal from the microphone, processes the data and sends the result to the display to achieve the visual display of the sound feedback. For example, after receiving the sound signal from the microphone, the controller can record and analyze the sound signal to extract the acoustic features in the prediction model for monitoring the broaching process. Algorithms residing on or run by the controller can be used to analyze and / or classify the acoustic features of the sound signal. The processed data can then be sent to a digital display for presenting or otherwise reporting the output of the processed information.

[0095] In a practical application scenario, the microphone can be positioned above the patient's surgical area. During total hip replacement surgery, the surgeon will first remove the patient's femoral head, and then gradually hollow out the femoral cavity using a broach that matches the size and shape of the implant. During this process, by tapping with a series of progressively larger broaches, the sound generated will be captured by the microphone and transmitted in real time to the control unit, as Figure 3 shown. Of course, these recordings can also be made directly on the controller. Next, in the controller, these audio can be segmented into individual tapping events, then the acoustic features can be extracted, and subsequently classified through machine learning techniques to determine whether it is a good or poor fit. The classification model for machine learning can be trained based on pre-collected patient data. The decision results can be weighted, combined, and then passed to the surgeon through a digital display and auditory feedback, thereby providing acoustic estimation information about fit and filling.

[0096] The software running in or operated by the system has a certain architecture. It optimizes the usability of the existing software platform and can expand the functions of the existing software platform. This software provides one or more applications, as Figure 4 shown, for users (such as patients, clinicians, etc.) to access and use for performing various functions. These applications can be used at the user's location or remotely, and each application has a graphical user interface (GUI) through which users can interact with the information in the system. The GUI can be customized for specific users, user groups, or user types, or can be consistent among all users or a specific subset of users. In addition, the system software can also provide a set of main GUIs through which users can select the GUIs of one or more other applications to interact, or access various information in the system simultaneously.

[0097] In terms of software presenting data, various optional formats and quantities can be adopted. For example, a multi-layer format can be used, where by viewing the information presented in the lower layers, users can obtain more details. These layers can be implemented through drop-down menus, tabbed mock folders, or other layering techniques, which are well-known to professionals in the field.

[0098] In addition, the software can also include standard reporting mechanisms, such as generating printable result reports, or generating electronic result reports that can be transmitted to computing devices connected to the system, in forms such as icons, text, email messages, or file attachments. In addition, specific results can also trigger alarm signals, such as generating alarm icons, emails, text messages, or phone calls to notify patients, doctors, nurses, first responders, or other healthcare providers about specific results.

[0099] In a specific embodiment, data can be transmitted from a microphone, a control unit, and a digital display through wired or wireless communication means. The wireless communication component is used to transmit information between systems and can be achieved through a wide area network, including but not limited to the Internet, an electronic network, an optical network, a wireless network, a physical security network, or a virtual private network, etc., which can all form part of a network system understood by those of ordinary skill in the art for transmitting data to additional computing devices. This extended network can also cover intermediate nodes, such as gateways, routers, bridges, ISP networks, public switched telephone networks, proxy servers, firewalls, etc., to transmit information items and other data throughout the system. Data transmission can be carried out through wireless communication and can adopt various wireless-based technologies, such as radio signals, near-field communication systems, hypersonic signals, infrared systems, cellular signals, GSM, etc. In some cases, data transmission can be carried out without the aid of a specific network or can directly enter and exit system components without a specific network.

[0100] For example, the system includes software specifically designed to collect audio data of the hammering sequence and segment the data into individual hammer strikes. This software is also capable of processing the data into acoustic features of interest and classifying the data as a good fit or a poor fit through machine learning algorithms, which can be trained as needed. This software also has the ability to weight and combine decisions, and then can pass the acoustic estimates of fit and fill to the surgeon. The system can perform the following program steps: First, the system collects sound signals or data. In one embodiment, the audio comes from a microphone in the operating room, recording the sound of the hammering broach during the operation. For example, the microphone can be fixed above the patient's surgical lamp but within the sterile area to obtain the best recording effect. In addition, the gain of the microphone can be optimized and adjusted to receive the sound containing the audio data, and can distinguish the relevant signal from the background noise. In one embodiment, the collected audio is continuously received from the microphone and streamed or otherwise transmitted to the control unit via a wireless transmission protocol (such as Bluetooth), as shown in Figure 5 shown. Another method can also record the audio through a wired microphone connected by a cable, or through a microphone connected to a wearable mobile device. The recording can also be recorded and then transmitted to the control unit at a time outside of real-time transmission. Alternative fixtures can be used to fix the microphone above the patient, and these fixtures can be temporary, room-specific, or custom-designed fixtures specifically for the microphone.

[0101] Next, the audio is segmented into individual hammering events. Typically, steps depicting the hit detection process are taken. For example, first, the audio is divided into a series of hammering events that correspond to the process of the surgeon hammering the broach into the femur and then repeatedly removing the broach and connecting it to the lock. There may be multiple hammering events for each broach size, depending on the surgeon's operation. For example, the segmentation of the hammering events can be performed within each 100-second window by identifying the regions where the sound signal exceeds 50% of the amplitude limit tuned to by the microphone. Each individual hammering event can be separated individually. For example, if the sequence signal is stereo, it can be combined into a mono mixed signal. Then, the signal is extracted by retaining only the maximum absolute amplitude points (i.e., the decimation factor) in every 100 points. The envelope of the signal is calculated through the total variation envelope. This method smooths the decaying components of the signal while retaining the sharp transitions at the moments of the hammering events. In addition, a smoothed version of the signal can be obtained by minimizing a specific optimization problem, which can be solved using convex optimization methods such as gradient descent, Newton's method, and interior point method.

[0102] During the audio data processing, in order to convert the data into useful acoustic features, a series of signal processing techniques and calculation formulas can be adopted. One important step is to extract the spectral features of the sound signal, such as spectral envelope and spectral centroid, etc. The spectral envelope represents the amplitude distribution of the signal at different frequencies, while the spectral centroid represents the central frequency of the signal energy. These features can be calculated by the following formulas:

[0103] 1. Spectral Envelope:

[0104] - Perform short-time Fourier transform (STFT) on the sound signal to obtain the spectrogram.

[0105] - In each time window, select the frequency with the largest amplitude as the estimated value of the spectral envelope.

[0106] 2. Spectral Centroid:

[0107] - Calculate the weighted average of the frequencies in each time window, with the weight being the amplitude of that frequency.

[0108] - This can be expressed as: \[f_c=\frac{\sum{f_i\cdot A_i}}{\sum{A_i}}\]

[0109] where \(f_c\) is the spectral centroid, \(f_i\) is the \(i\)-th frequency, and \(A_i\) is the amplitude of that frequency.

[0110] In addition, during the extraction of acoustic features, the time-domain and frequency-domain features of the signal can also be considered, such as mean, standard deviation, zero crossing rate, etc. These features can be calculated by the following formulas:

[0111] 3. Mean:

[0112] - Average the amplitudes of the sound signal.

[0113] 4. Standard Deviation:

[0114] - Calculate the standard deviation of the amplitudes of the sound signal.

[0115] 5. Zero Crossing Rate:

[0116] - Calculate the number of times the sound signal crosses the zero axis, which can be expressed as: \[ZCR=\frac{1}{N - 1}\sum_{i=1}^{N - 1}|sgn(x[i]) - sgn(x[i - 1])|\]

[0117] Among them, \(N\) is the number of sampling points of the signal, \(x[i]\) is the \(i\)-th sampling point, and \(sgn\) is the sign function.

[0118] By extracting these acoustic features, a feature vector can be created for each hammering event. Next, machine learning algorithms can be used to classify the hammering events to determine whether they are good fits or poor fits. A commonly used classifier is the support vector machine (SVM), which can learn the patterns of good fits and poor fits through training samples. After the classifier is trained, new acoustic feature vectors can be input into the classifier to obtain classification results.

[0119] In terms of decision weights and combinations, the system can use the method of weighted sum to determine the appropriate and filled acoustic estimates. The weighted sum can be expressed as: \[y=\sum_{i = 1}^{N}w_i\cdot x_i\]

[0120] Among them, \(y\) is the final acoustic estimate, \(N\) is the number of acoustic features, \(w_i\) is the weight, and \(x_i\) is the \(i\)-th acoustic feature.

[0121] In summary, through acoustic feature extraction and machine learning classification, the system can automatically judge the fitting degree of hammering events and transmit the appropriate and filled acoustic estimates to the surgeon. This process utilizes technologies such as signal processing, feature extraction, and machine learning, providing real-time analysis and feedback for surgical operations, thereby enhancing the success rate of the surgery and the safety of the patient.

[0122] In a specific embodiment, the following alternative method can be considered to achieve sequence segmentation. By identifying the key acoustic features of the release latch, the moments of hammering and releasing the latch can be effectively distinguished. These features are relatively easy to identify in the time-domain signal. Each identified pulse can be processed by applying the discrete Fourier transform (DFT), and the pulse can be classified with the help of key spectral features to determine whether it is a hammering or a release of the latch. Here, simple machine learning techniques such as support vector machine (SVM), kernel machine, or the like can be used to classify this difference through training to determine whether it is a hammering or a release of the latch. At the same time, the moving average of the last few hits can be calculated to determine when the broaching is completed and the release starts.

[0123] Another method for single hit segmentation is achieved by analyzing the signal energy within a given time window. As the signal decays, the power of the signal should decrease, so the first half of the signal should contain more power than the second half. The time window for segmentation can be longer or shorter than four seconds. These analyses can be performed on stereo mix data, not limited to mono. The dead zone limit and smoothing constant value can be adjusted according to the data collection situation to adapt to different scenarios. Other smoothing functions, such as quadratic smoothing, can also be used to find the envelope of the signal. The above optimization problems can be solved using various well-established methods, such as the interior point method, Newton's method, or gradient descent method.

[0124] When performing data conversion and feature extraction, a well-established fast Fourier transform (FFT) algorithm, such as the Cooley - Tukey, Bluestein, and Rader FFTs, can be used to perform the DFT. Acoustic features including the signal power in specific frequency bands can be extracted. For example, the signal power in the 1 - 2 kHz, 2 - 4 kHz, and 5 - 7 kHz frequency bands can be recorded. Thus, each sound of a hammer strike is transformed into a 3×1 vector. The power is calculated as the square of the integral (area under) of the DFT in a given frequency band. In another embodiment, additional acoustic features for classification include the power in lower frequency bands; the decay rate of the signal and specific harmonic regions; the zero crossing rate; and cepstrum analysis. Time domain analysis and wavelet transform analysis can also be performed. For example, Mel - Frequency Cepstral Coefficients (MFCCs) can also be used as a metric for fitting. The process of obtaining MFCCs may include: pre - emphasis using a first - order high - pass filter to emphasize the high - frequency components of the signal; taking the short - time Fourier transform of the windowed signal; converting the frequency scale to the Mel scale, where the spectral power is calculated over multiple frequency bands; taking the power logarithm of each band; and using the discrete cosine transform for decorrelation and compression. It should be further understood that variations of the above examples are equally feasible and not restricted, including alternative parameter selections, transforms, frequency scales, etc.

[0125] In another embodiment, this embodiment considers the following alternative method for sequence segmentation. The powers in the 1 - 2 kHz, 2 - 4 kHz, and 5 - 7 kHz frequency bands are regarded as data points in three - dimensional space. In this supervised learning problem, this embodiment uses machine learning methods to process. Among the known data, that is, the data before the final fit and the data known to be the final fit, an optimal separation plane is constructed. This separation plane is used to show the suitability for fitting through an X - ray display. Note that all data comes from the first five hits of each sequence.

[0126] This embodiment uses a soft - threshold support vector machine to construct this separation plane. The optimization problem can be formulated in the following form:

[0127] Minimize:

[0128] 1 / 2*||a||^2+∑_{i=1}^{n}u_i

[0129] Subject to:

[0130] y_i(a^T x_i - b)≥1 - u_i, i∈{1,2,...,n}

[0131] u_i≥0, i∈{1,2,...,n}

[0132] u_i≥1 - y_i(a^T x_i - b), i∈{1,2,...,n}

[0133] Wherein, xi is a known size vector representing the ith data point in three - dimensional space, yi is the known binary label of data point i (-1 if it is poorly - fitted data, 1 if it is best - fitted data), a and b form the optimal separation plane aTx = b, u is the penalty applied to each data point if the data point falls on the wrong side of the separation plane. μ is the penalty constant that controls the trade - off between margin separation and point misclassification, and the current value is 1.

[0134] When smoothing a signal, an optimization method can be used to minimize a specific optimization problem to obtain a smoothed version of the signal. Commonly used convex optimization methods include gradient descent, Newton's method, and interior - point method.

[0135] Gradient Descent:

[0136] Gradient descent is an iterative optimization method that reduces the value of the objective function by moving along the negative gradient direction at each step.

[0137] Example of an optimization problem: Minimize the quadratic error of a smoothed signal, where \(x\) is the smoothed signal, \(y\) is the original signal, and \(w\) is the weight.

[0138] Objective function: \[J(x)=\sum_{i=1}^{N}w_i\cdot(x_i - y_i)^2\]

[0139] Parameter update: \[x_i^{(k + 1)}=x_i^{(k)}-\alpha\cdot 2\cdot w_i\cdot(x_i^{(k)}-y_i)\]

[0140] Where \(k\) is the number of iterations and \(\alpha\) is the learning rate.

[0141] Newton's Method:

[0142] Newton's method uses the second derivative of the objective function to approximate the function, thus converging quickly.

[0143] Example of an optimization problem: Minimize the squared error of a smooth signal.

[0144] Objective function: \[J(x)=\sum_{i = 1}^{N}(x_i - y_i)^2\]

[0145] Parameter update: \[x_i^{(k + 1)} = x_i^{(k)}-\alpha\cdot(H^{-1}\cdot g)_i\]

[0146] where \(k\) is the number of iterations, \(\alpha\) is the learning rate, \(H\) is the Hessian matrix of the objective function, and \(g\) is the gradient of the objective function.

[0147] Interior Point Method:

[0148] The Interior Point Method is a convex optimization method that solves the problem by finding the optimal solution inside the constraints.

[0149] Example of an optimization problem: Minimize the regularized error of a smooth signal, where \(x\) is the smooth signal, \(y\) is the original signal, and \(\lambda\) is the regularization parameter.

[0150] Objective function: \[J(x)=\sum_{i = 1}^{N}(x_i - y_i)^2+\lambda\sum_{i = 1}^{N - 1}|x_{i + 1}-x_i|\]

[0151] The Interior Point Method transforms this problem into a constrained problem and solves it within the constraints.

[0152] Finally, the result of this problem will be the plane defined by \(a\) and \(b\), i.e., \(a^T x = b\). By solving this problem, an optimal separation plane can be obtained for classifying poorly fitting data and best-fitting data.

[0153] In the anomaly detection and alarm triggering phase, when a sound anomaly is detected, the system will immediately trigger an alarm to notify the medical team of possible anomalies, such as Figure 6As shown below. The steps of this process are as follows: First, in the sound processing unit, the system uses a specialized anomaly detection algorithm to analyze real-time sound signals. This algorithm compares the input sound pattern with a pre-defined normal pattern to identify whether there is an abnormal sound pattern. Once the anomaly detection algorithm detects a potential abnormal sound pattern, it generates corresponding anomaly detection results. These results are then passed to the result confirmation stage. The surgeon receives these anomaly detection results through the system's decision support interface. The interface presents the detected abnormal sound pattern and related information in a clear manner to help the surgeon accurately understand the situation. The surgeon will carefully review the anomaly detection results to confirm whether there is indeed an abnormal situation. They can further analyze the results to ensure that there are no false alarms. Once the surgeon confirms the accuracy of the anomaly detection results, the system generates an alarm trigger signal. This alarm trigger signal is passed to the alarm system, which is responsible for issuing an alarm. The alarm can be in the form of sound, vision, or other appropriate means to notify relevant medical personnel inside or outside the operating room.

[0154] In the stage of identifying abnormal sound signals, the system considers using the Support Vector Machine (SVM) as the core classification algorithm to achieve accurate segmentation and classification of the sound event sequence. SVM can effectively process samples in a high-dimensional feature space. By finding a hyperplane that maximally separates samples of different classes, classification is achieved.

[0155] For a linearly separable binary classification problem, the mathematical representation of SVM is as follows:

[0156] f(x)=\text{sign}\left(\sum_{i=1}^{n}\alpha_i y_i K(x_i,x)+b\right)

[0157] Where f(x) is the classification function, x is the input sample, x_i is the training sample, y_i is the corresponding class label, K(x_i,x) is the kernel function, \alpha_i is the corresponding Lagrange multiplier, and b is the bias term. The kernel function K(x_i,x) is used to map the input data to a high-dimensional space, so that linear separation can be achieved in the high-dimensional space.

[0158] SVM can also combine with kernel functions, such as the Radial Basis Function (RBF), to map the data to a higher-dimensional feature space. The calculation formula of the RBF kernel function is as follows:

[0159] K(x,x')=\exp\left(-\frac{\|x-x'\|^2}{2\sigma^2}\right)

[0160] Where x and x' are sample features, and \(\sigma\) is a parameter that controls the shape of the RBF kernel function.

[0161] In this embodiment, a large number of machine learning techniques are used, such as parametric methods, maximum likelihood estimation, hidden Markov models, kernel machines, k-nearest neighbor algorithms, and artificial neural networks, to solve this classification problem. Parametric methods can use prior distributions to provide specificity for a particular patient. For the training data, prior information can be obtained from the early broach sizes or sequences of broaches with the same size. The initial hit count can also be adjusted, or all hits or the final hit can be used as the basis for classification.

[0162] When more specific classification is needed, all features described herein can be considered. Dimensionality reduction techniques such as principal component analysis (PCA) or linear discriminant analysis (LDA) can be used to reduce the dimensionality of the features and pass them to the classification algorithm. Thus, the system can automatically identify the most discriminative features. For example, different frequency bands may provide information about the adaptation quality of different subjects, such as age, gender, and weight. SVM can also be used as a classification algorithm. Additionally, alternative implementations include SVM-based kernel extensions, linear discriminant analysis, and kernel discriminant analysis, as well as k-nearest neighbor classification, etc.

[0163] PCA reduces the dimension by projecting the original features onto a new coordinate system by calculating the eigenvectors of the covariance matrix of the data. The calculation process is as follows:

[0164] \text{Covariance Matrix:}\(\Sigma=\frac{1}{n}\sum_{i = 1}^{n}(x_i-\bar{x})(x_i-\bar{x})^T\)

[0165] \text{Eigenvalue Decomposition:}\(\Sigma = V\Lambda V^T\)

[0166] \text{Select Top Eigenvectors:}\(U = V[:, :k]\)

[0167] Where n is the number of samples, \(x_i\) is the sample feature, \(\bar{x}\) is the mean vector, V is the matrix of eigenvectors of the covariance matrix, \(\Lambda\) is the diagonal matrix of eigenvalues, and U is the feature matrix after dimensionality reduction.

[0168] Once the classifier is trained, the dimensionality-reduced projection matrix and the separating hyperplane can be used to classify the hit quality of new, compatible patients. Compatibility here refers to similarities in aspects such as gender, age, weight, etc. The classifier may return a confidence metric to indicate the degree of certainty of the classification result. This confidence can be combined with the classification result to provide a more concise fitting quality indicator for the surgeon.

[0169] The classification result can be displayed on the user interface, providing visual and auditory feedback. This interface can present different views for evaluating all available features. This helps the surgeon to perform accurate analysis and judgment. For example, two-dimensional data of the 2-1 kHz and 2-5 kHz components can be plotted in real time to help the surgeon make more informed clinical decisions. This interface can also display additional data as needed, such as spectrograms or image data of key acoustic features.

[0170] In summary, the system and method are applicable to real-time classification and evaluation for guiding the surgeon's surgical decision-making. The technology not only includes the processing and feature extraction of sound signals, but also the application of a variety of machine learning algorithms, as well as an interactive interface with the surgeon, enabling it to effectively utilize the feature information for patient operations.

[0171] The system and method of this application have multiple advantages. The following are several of the main advantages, including but not limited to:

[0172] 1. Precise positioning and filling of the femoral shaft position: Through sound analysis, this system can optimize the fit and filling of the femoral shaft in the femoral cavity. By analyzing acoustic features, the system can accurately judge the position that the actual implant rod will occupy, thus providing a true gauge for the broach of the femoral shaft position after implantation. This helps to ensure a tight fit between the implant and the bone structure and improve the success rate of the surgery.

[0173] 2. Non-invasive auxiliary tool: Acoustic analysis, as a non-invasive auxiliary tool, can supplement the surgeon's tactile and auditory information. Compared with traditional surgical methods, this system can provide more objective data and quantitative feedback, enabling the surgeon to more accurately judge the position and status of the implant to ensure the best surgical effect.

[0174] 3. Providing auditory feedback: The system provides auditory feedback to the surgeon, enabling them to be informed in a timely manner when there is an abnormal sound and indirectly confirm when the broach reaches the appropriate position in the femoral cavity, that is, the maximum contact between the broach and the compact bone. This real-time feedback helps the doctor to more precisely control the surgical operation, thereby optimizing the embedding depth of the implant to the greatest extent.

[0175] 4. Integration into existing surgical procedures: This system can be easily integrated into current surgical workflows and procedures, providing important feedback information to surgeons while minimizing additional surgical time. This means that doctors can benefit from the advantages of this innovative technology without changing their existing operating habits.

[0176] 5. Surgical training tool: This system is not only suitable for experienced surgeons but also can be used as a surgical training tool for new surgeons or students. Through real-time audio feedback and a visual interface, trainers can gain a deeper understanding of the interaction between implants and bones, enhancing the training effect.

[0177] 6. Integration of intraoperative information: This system has the ability to integrate intraoperative information collected by other means, thereby enhancing the three-dimensional aspects of the surgery. By integrating multi-source information such as audio analysis and imaging data, doctors can obtain more comprehensive and accurate surgical information, further improving the precision and success rate of the surgery.

[0178] See Figure 7 , an embodiment of the present application provides a monitoring device for orthopedic implants, including:

[0179] A first acquisition module 701, configured to obtain a sound signal generated by a tapping operation during the process of embedding an implant into a specified area of a bone structure;

[0180] A first processing module 702, configured to obtain an acoustic feature of the sound signal, and determine the fitting degree information between the implant and a specified area in the bone structure according to the acoustic feature.

[0181] In an implementation manner of the present application, the first processing module is further configured to:

[0182] Segment the sound signal into individual tapping events, and obtain an acoustic feature of each tapping event;

[0183] Analyze and classify the acoustic features of the tapping events to obtain a classification result of the tapping events, and determine the fitting degree information between the implant and a specified area in the bone structure according to the classification result.

[0184] In an implementation manner of the present application, the first processing module is further configured to:

[0185] Create a feature vector for each tapping event through the acoustic feature;

[0186] Classify the tapping events through a machine learning algorithm to obtain a classification result of the tapping events, where the classification result includes good fitting or poor fitting.

[0187] In one embodiment of the present application, the device further includes:

[0188] An evaluation module for evaluating the fitting degree information between the implant and the bone structure to identify whether the implant is suitable and can fill a specified area in the bone structure.

[0189] In one embodiment of the present application, the device further includes:

[0190] A second acquisition module for acquiring ambient sound signals;

[0191] A second processing module for comparing the acoustic features of the acquired ambient sound signals with a pre-established sound pattern library to determine whether there are abnormal sounds;

[0192] An alarm module for triggering an alarm process if abnormal sounds are detected.

[0193] In one embodiment of the present application, the alarm process includes:

[0194] Displaying the abnormal detection result through a user interface;

[0195] Obtaining first information through the user interface, where the first information is used to indicate the presence or absence of an abnormal situation;

[0196] If the first information indicates the presence of an abnormal situation, then send an alarm trigger signal.

[0197] The device provided by the embodiments of the present application can implement Figure 1 each process implemented by the method embodiments shown, and achieve the same technical effects. To avoid repetition, it will not be elaborated here.

[0198] The embodiments of the present application further provide a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, it implements each process of the method embodiments shown above Figure 1 and can achieve the same technical effects. To avoid repetition, it will not be elaborated here.

[0199] Wherein, the processor is the processor in the terminal described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs, etc.

[0200] The steps of the methods or algorithms described in connection with the disclosure of the present application may be implemented in hardware or by a processor executing software instructions. The software instructions may be composed of corresponding software modules, and the software modules may be stored in RAM, flash memory, ROM, EPROM, EEPROM, registers, hard disks, removable hard disks, CD-ROMs, or any other form of storage medium well-known in the art. An exemplary storage medium is coupled to the processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium may also be a component of the processor. The processor and the storage medium may be carried in an ASIC. Additionally, the ASIC may be carried in a core network interface device. Of course, the processor and the storage medium may also exist as discrete components in the core network interface device.

[0201] Those skilled in the art should be able to realize that in one or more of the above examples, the functions described in the present application can be implemented by hardware, software, firmware, or any combination thereof. When implemented using software, these functions can be stored in a computer-readable medium or transmitted as one or more instructions or codes on a computer-readable medium. The computer-readable medium includes computer storage media and communication media, where the communication media includes any medium that facilitates the transfer of a computer program from one place to another. The storage medium can be any available medium accessible by a general or special-purpose computer.

[0202] The specific embodiments described above further elaborate on the purpose, technical solutions, and beneficial effects of the present application. It should be understood that the above description is only the specific embodiments of the present application and is not used to limit the protection scope of the present application. Any modifications, equivalent replacements, improvements, etc. made on the basis of the technical solutions of the present application should be included within the protection scope of the present application.

[0203] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, system, or computer program product. Therefore, the embodiments of the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the embodiments of the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0204] Embodiments of the present application are described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing device generate a means for implementing the functions specified in one or more of the flows Figure 1 one or more flows and / or blocks Figure 1 or a means for implementing the functions specified in one or more of the blocks.

[0205] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction means that implements the functions specified in one or more of the flows Figure 1 one or more flows and / or blocks Figure 1 or a means for implementing the functions specified in one or more of the blocks.

[0206] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, so that the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more of the flows Figure 1 one or more flows and / or blocks Figure 1 or a means for implementing the functions specified in one or more of the blocks.

[0207] Obviously, those skilled in the art can make various modifications and variations to the embodiments of the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the embodiments of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

Claims

1. A monitoring method for orthopedic implants, characterized in that, Comprising: During the process of implanting an implant into a specified area of a bone structure, obtaining a sound signal generated by a tapping operation; Obtaining acoustic characteristics of the sound signal, and determining the fitting degree information between the implant and the specified area in the bone structure according to the acoustic characteristics.

2. The method according to claim 1, characterized in that, Obtaining acoustic characteristics of the sound signal, and determining the fitting degree information between the implant and the specified area in the bone structure according to the acoustic characteristics, including: Segmenting the sound signal into individual tapping events, and obtaining acoustic characteristics of each tapping event; Analyzing and classifying the acoustic characteristics of the tapping events to obtain classification results of the tapping events, and determining the fitting degree information between the implant and the specified area in the bone structure according to the classification results.

3. The method according to claim 2, characterized in that, Analyzing and classifying the acoustic characteristics of the tapping events to obtain classification results of the tapping events, including: Creating a feature vector for each tapping event through the acoustic characteristics; Classifying the tapping events through a machine learning algorithm to obtain classification results of the tapping events, where the classification results include good fit or poor fit.

4. The method according to claim 1, characterized in that, The method further comprises: Evaluating the fitting degree information between the implant and the bone structure to identify whether the implant is suitable and can fill the specified area in the bone structure.

5. The method according to claim 1, characterized in that, The method further comprises: Collecting ambient sound signals; Comparing the acoustic characteristics of the collected ambient sound signals with a pre-established sound pattern library to determine whether there are abnormal sounds; If abnormal sounds are detected, triggering an alarm process.

6. The method according to claim 5, characterized in that, The alarm process includes: Displaying the abnormal detection result through a user interface; Obtaining first information through the user interface, where the first information is used to indicate the existence of an abnormal situation or the non-existence of an abnormal situation; If the first information indicates the existence of an abnormal situation, sending an alarm trigger signal.

7. A monitoring device for orthopedic implants, characterized in that, Comprising: A first acquisition module, configured to obtain a sound signal generated by a tapping operation during the process of implanting an implant into a specified area of a bone structure; A first processing module, configured to obtain acoustic characteristics of the sound signal, and determine the fitting degree information between the implant and the specified area in the bone structure according to the acoustic characteristics.

8. The device according to claim 7, characterized in that, The first processing module is further configured to: Segment the sound signal into individual tapping events, and obtain acoustic characteristics of each tapping event; Analyze and classify the acoustic characteristics of the tapping events to obtain classification results of the tapping events, and determine the fitting degree information between the implant and the specified area in the bone structure according to the classification results.

9. The device according to claim 8, characterized in that, The first processing module is further configured to: Create a feature vector for each tapping event through the acoustic characteristics; Classify the tapping events through a machine learning algorithm to obtain classification results of the tapping events, where the classification results include good fit or poor fit.

10. The device according to claim 7, characterized in that, The device further comprises: An evaluation module, configured to evaluate the fitting degree information between the implant and the bone structure to identify whether the implant is suitable and can fill the specified area in the bone structure.

11. The device according to claim 7, wherein The device further comprises: A second acquisition module, configured to collect ambient sound signals; A second processing module, configured to compare the acoustic features of the collected environmental sound signals with a pre-established sound pattern library to determine whether there is an abnormal sound; An alarm module, configured to trigger an alarm process if an abnormal sound is detected.

12. The device according to claim 11, wherein The alarm process includes: Displaying the abnormal detection result through a user interface; Obtaining first information through the user interface, where the first information is used to indicate the existence of an abnormal situation or the non-existence of an abnormal situation; If the first information indicates the existence of an abnormal situation, then sending an alarm trigger signal.

13. A surgical assistance system, wherein Including the device according to any one of claims 7 to 12.

14. A readable storage medium, wherein A program or instruction is stored on the readable storage medium, and when the program or instruction is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images