Acetabulum hammer based on AI intelligent algorithm

Abstract

The invention discloses an acetabulum hammer based on an AI intelligent algorithm, and belongs to the technical field of intelligent medical equipment. A multi-mode sensor set, an AI processing module, a man-machine interaction module, a dual-mode operation module and an anomaly detection module are arranged in the acetabular hammer body, real-time monitoring is conducted through the multi-mode sensor set, the risk of fracture or poor fixation of an implant caused by misoperation is reduced, the density change of bone tissue is judged through spectral analysis, and the accuracy of the bone tissue is improved. The AI processing module is used for carrying out deep feature extraction on parameter data in an operation based on a convolutional neural network, identifying different operation stages and operation states, analyzing change modes of Z-direction pressure and RX-direction torque in different frequency bands, and carrying out real-time recognition on the condition of excessive driving or insufficient driving so as to reduce driving position errors; the change trend of the intraoperative bone mineral density is judged in real time, complex intraoperative data is converted into interpretable operation state information, errors caused by experience-dependent judgment in the operation are reduced, and the operation qualification rate of green doctors is increased.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent medical devices, and particularly to an acetabular hammer based on an AI intelligent algorithm. Background Art

[0002] Traditional acetabular hammers rely on the operator's hand feeling and experience. During the operation, the control of pressure and torque is inaccurate, which easily leads to insufficient insertion of the acetabulum. There is a lack of real-time bone density feedback and motion trajectory monitoring, increasing the risk of intraoperative fractures and the probability of postoperative prosthesis loosening. Existing devices do not have intelligent decision-making support and cannot dynamically adjust operation parameters.

[0003] At the same time, the feedback information through a single sensor (such as pressure or acceleration) is limited, making it difficult to comprehensively evaluate the surgical state and unable to adapt to complex intraoperative scenarios. Summary of the Invention

[0004] The purpose of the present invention is to provide an acetabular hammer based on an AI intelligent algorithm to solve the problems raised in the above background art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An acetabular hammer based on an AI intelligent algorithm, including an acetabular hammer main body, on which a handheld rod is provided. An in-place degree indicator is provided on the handheld rod, and a multi-modal sensor group, an AI processing module, a human-computer interaction module, a dual-mode operation module, and an anomaly detection module are provided inside the handheld rod;

[0006] The multi-modal sensor group is configured to collect intraoperative parameter data; the AI processing module is configured to process the intraoperative parameter data in real time, extract the characteristic patterns of the intraoperative parameter data, and track the surgical process to predict the stress state and insertion degree of the bone tissue, and provide adaptive parameter adjustment suggestions; the human-computer interaction module is configured to communicate with an external display device and provide real-time feedback information to the operator; the dual-mode operation module is configured to adjust the operation mode of the acetabular hammer; the anomaly detection module is configured to automatically detect intraoperative abnormal situations and trigger a safety alarm when an anomaly occurs.

[0007] Further, the multi-modal sensor group includes a six-axis force sensor, a 9-axis acceleration sensor, and a bone conduction sound sensor;

[0008] Among them, the six-axis force sensor is used to detect the Z-direction pressure and the RX-direction torque; the 9-axis acceleration sensor is used to obtain the acceleration, angular velocity, and attitude information of the acetabular hammer in space; the bone conduction sound sensor is used to collect the bone conduction sound signals of the bone tissue in different states during the operation.

[0009] Furthermore, the AI processing module is equipped with an MCU containing a convolutional neural network unit to perform real-time processing on the original intraoperative parameter data collected by the multi-modal sensor group, removing outliers and normalizing the data, performing time alignment and interpolation processing on the intraoperative parameter data, and generating a standardized time data sequence with equal time intervals;

[0010] Extract the data features of the intraoperative parameter data through the convolutional neural network. The data features include mechanical waveform features, acceleration curve features, and bone conduction sound spectrum features;

[0011] During the extraction of mechanical waveform features, obtain the Z-direction pressure and RX-direction torque change patterns in different frequency bands through multi-scale convolutional kernels; during the extraction of acceleration curve features, adopt the time window segmentation method to identify the motion features in different operation stages; during the extraction of bone conduction sound spectrum features, combine the short-time Fourier transform to convert the time-domain signal into frequency-domain features;

[0012] Use a feature fusion network to perform multi-modal fusion of mechanical waveform features, acceleration curve features, and bone conduction sound spectrum features; map the data features extracted in real time to the preset surgical state categories, and the surgical state categories include in-place insertion, about-to-be-in-place insertion, and insufficient insertion.

[0013] Furthermore, the AI processing module tracks the surgical process through the time data sequence, predicts the force state and insertion degree of the bone tissue, specifically including:

[0014] Perform sequence modeling on the intraoperative parameter data at different time steps, generate a standardized time data sequence with equal time intervals, and capture the temporal pattern of changes in the intraoperative parameter data based on the time data sequence;

[0015] Combine the data of the six-axis force sensor and the 9-axis acceleration sensor to dynamically calculate the force change curve of the acetabular hammer, and determine whether the impact force applied by the surgeon meets the surgical standard; through the spectral analysis of bone conduction sound, real-time identify the change in bone tissue density, judge whether there is a risk of insufficient insertion or bone injury, automatically record and mark the key time points during the operation; if an abnormal operation is detected during the operation, provide a warning message to the surgeon in real time through the human-computer interaction module;

[0016] Generate an adaptive parameter adjustment suggestion based on the surgical process prediction result, and the adaptive parameter adjustment suggestion includes adjusting the impact force and optimizing the motion trajectory.

[0017] Furthermore, the AI processing module provides an adaptive parameter adjustment suggestion, specifically including:

[0018] Define the state space, which includes real-time intraoperative parameter data and surgical state categories; set the action space, which includes the adjustment ranges of the impact force, rotation angle, and impact frequency of the acetabular hammer;

[0019] During the operation, based on the current state space, calculate the adjustment strategy for the current surgical state. The adjustment strategy includes reducing the impact force, changing the rotation angle, and adjusting the impact frequency, and the adjustment range of the adjustment strategy is controlled within the action space;

[0020] During the operation, continuously evaluate the effect after the operation adjustment, and through the human-computer interaction module, provide real-time parameter adjustment suggestions to the surgeon. If it is detected that the surgeon's operation has not been adjusted, provide a sound feedback prompt;

[0021] After the operation, store the intraoperative parameter adjustment records to form a surgical optimization plan.

[0022] Further, the human-computer interaction module is connected to the in-place degree indicator light. The in-place degree indicator light is provided with red, green, and blue indicator lights, and during the operation, it is controlled based on the real-time surgical state through the in-place degree indicator light. Among them, the red indicator light indicates insufficient insertion, the blue indicator light indicates that the insertion is about to be in place, and the green indicator light indicates that the insertion is in place.

[0023] Further, the human-computer interaction module is electrically connected to the wireless data transmission interface, and the human-computer interaction module communicates with the external display device through the wireless data transmission interface.

[0024] Further, the operation mode includes a recording mode and a reasoning mode;

[0025] When the dual-mode operation module is in the recording mode, record and mark the changes in intraoperative parameter data during the operations of different patients; when the dual-mode operation module is in the reasoning mode, during the operation, based on the AI processing module, perform real-time analysis and navigation, and provide intraoperative early warnings and operation optimization suggestions.

[0026] Further, the abnormal detection module obtains the real-time data of the Z-direction pressure and RX-direction torque in the intraoperative parameter data, combines the surgical process data, partitions the mechanical data in different surgical stages, and distinguishes the force application modes in different surgical stages;

[0027] Set the bone tissue safe force threshold, and adjust the threshold based on the preoperative imaging individually;

[0028] During the operation, based on the force application modes in different surgical stages, combine the bone tissue safe force threshold to detect the Z-direction pressure and RX-direction torque. When the Z-direction pressure or RX-direction torque exceeds the bone tissue safe force threshold, calculate the overlimit degree;

[0029] Send a warning message to the surgeon through the human-computer interaction module, and at the same time activate the red flashing alarm of the in-place degree indicator light to prompt the surgeon to adjust the operation force.

[0030] Furthermore, the abnormal detection module acquires the bone conduction sound signal and the bone conduction sound spectrum characteristics; combines with the AI processing module to establish a sound feature library for bone tissue being inserted in place, about to be inserted in place, insufficient insertion, and bone mass injury;

[0031] Compare the real-time collected sound spectrum characteristics with the sound feature library to determine the current bone density change trend; when it is detected that the bone tissue spectrum characteristics deviate from the normal range and the high-frequency components are abnormally enhanced, it is determined that there is a risk of insufficient insertion and bone mass injury;

[0032] Provide real-time feedback to the surgeon through the human-computer interaction module. If it is determined that the insertion is insufficient, light up the red warning of the in-place degree indicator light; if it is determined that the insertion is about to be in place, light up the blue warning light and emit a prompt sound to remind the surgeon to adjust the operation force.

[0033] Compared with the prior art, the beneficial effects of the present invention are:

[0034] 1. The present invention collects intraoperative parameter data through a multi-modal sensor group, and real-time monitors the pressure in the Z direction and the torque change in the RX direction of the acetabular hammer, enabling the surgeon to accurately adjust the force application method, thereby reducing the risk of fractures or poor implant fixation caused by improper operation. The 9-axis acceleration sensor can provide the position, attitude, and motion characteristic data of the acetabular hammer in space, ensuring that the surgeon's operation on the acetabular hammer conforms to the established trajectory, improving the consistency and accuracy of the surgery. By analyzing the spectrum to judge the density change of bone tissue, it can be identified in real time whether there is a situation of insufficient insertion, enabling the AI processing module to monitor the surgical process in different dimensions, and providing more comprehensive analysis and optimization suggestions, reducing the occurrence of intraoperative emergencies, and reducing the risk of intraoperative fractures and insertion position errors.

[0035] 2. The AI processing module of the present invention performs deep feature extraction on intraoperative parameter data based on a convolutional neural network, and combines with a feature fusion network for multi-modal data analysis, enabling the system to accurately identify different surgical stages and operation states, analyze the change patterns of the pressure in the Z direction and the torque in the RX direction in different frequency bands, thereby accurately judging whether the impact force applied by the surgeon meets the standard, classifying the motion characteristics of different surgical stages, extracting the bone conduction sound spectrum characteristics, and based on the spectrum change of the bone tissue sound signal, judging the change trend of the bone density during the operation in real time, converting the complex intraoperative data into interpretable surgical state information, providing real-time and intuitive decision support for the surgeon, reducing the error of relying on experience judgment during the operation, improving the passing rate of novice doctors' operations, and improving the accuracy and consistency of the surgery. Brief Description of the Drawings

[0036] Figure 1 Schematic diagram of the main structure of the acetabular hammer of the present invention;

[0037] Figure 2 Schematic diagram of the acetabular hammer module of the AI intelligent algorithm of the present invention.

[0038] In the figure: 1. Main body of the acetabular hammer; 2. Handheld rod; 3. Indicator light for in-place degree; 4. Six-axis force sensor; 5. 9-axis acceleration sensor; 6. Bone conduction sound sensor; 7. MCU with convolutional neural network unit. Specific embodiments

[0039] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0040] Please refer to Figure 1-2 , the present invention provides the following technical solutions:

[0041] An acetabular hammer based on an AI intelligent algorithm includes a main body 1 of the acetabular hammer. A handheld rod 2 is provided on the main body 1 of the acetabular hammer. An in-place degree indicator light 3 is provided on the handheld rod 2. A multi-modal sensor group, an AI processing module, a human-computer interaction module, a dual-mode operation module, and an anomaly detection module are provided inside the handheld rod 2;

[0042] The multi-modal sensor group is configured to collect intraoperative parameter data; the AI processing module is configured to process the intraoperative parameter data in real time, extract the characteristic patterns of the intraoperative parameter data, and track the surgical process to predict the stress state and driving-in degree of the bone tissue, and provide suggestions for adaptive parameter adjustment; the human-computer interaction module is configured to communicate with an external display device and provide real-time feedback information to the surgeon; the dual-mode operation module is configured to adjust the operation mode of the acetabular hammer; the anomaly detection module is configured to automatically detect intraoperative abnormal situations and trigger a safety alarm when an anomaly occurs.

[0043] The multi-modal sensor group includes a six-axis force sensor 4, a 9-axis acceleration sensor 5, and a bone conduction sound sensor 6; wherein, the six-axis force sensor 4 is used to detect the Z-direction pressure and the RX-direction torque; the 9-axis acceleration sensor 5 is used to obtain the acceleration, angular velocity, and attitude information of the acetabular hammer in space; the bone conduction sound sensor 6 is used to collect the bone conduction sound signals of the bone tissue in different states during the operation.

[0044] In the above embodiment, intraoperative parameter data is collected through a multimodal sensor group and analyzed in real time by an AI processing module, which not only improves the accuracy of intraoperative data, but also enhances the controllability of the surgical process. The six-dimensional force sensor can monitor the pressure of the acetabular hammer in the Z direction and the torque changes in the RX direction in real time, so that the surgeon can accurately adjust the force application method, thereby reducing the risk of fractures or poor implant fixation caused by improper operation. The 9-axis acceleration sensor can provide the position, posture and motion characteristic data of the acetabular hammer in space, ensuring that the surgeon's operation of the acetabular hammer complies with the established trajectory and improves the consistency and accuracy of the operation.

[0045] In addition, the bone conduction sound sensor can capture the sound signals emitted by bone tissue during surgery, and determine the density changes of bone tissue through spectrum analysis, and identify in real time whether there is insufficient penetration. The combination of multimodal data enables the AI processing module to monitor the progress of the surgery in different dimensions and provide more comprehensive analysis and optimization suggestions, thereby effectively improving the surgeon's ability to control the details of the surgery, reducing the occurrence of emergencies during surgery, and reducing the risk of fractures and insertion position errors during surgery.

[0046] The AI processing module is equipped with an MCU7 containing a convolutional neural network unit, which processes the raw intraoperative parameter data collected by the multimodal sensor group in real time, removes outliers and normalizes them, performs time alignment and interpolation processing on the intraoperative parameter data, and generates a standardized time data sequence with equal time intervals.

[0047] Extracting data features of intraoperative parameter data through a convolutional neural network, wherein the data features include mechanical waveform features, acceleration curve features, and bone conduction sound spectrum features;

[0048] In the process of mechanical waveform feature extraction, the Z-axis pressure and RX-axis torque change patterns in different frequency bands are obtained through multi-scale convolution kernels; in the process of acceleration curve feature extraction, the time window segmentation method is used to identify the motion characteristics of different operation stages; in the process of bone conduction sound spectrum feature extraction, the time domain signal is converted into frequency domain features in combination with short-time Fourier transform;

[0049] A feature fusion network is used to multimodally fuse mechanical waveform features, acceleration curve features, and bone conduction sound spectrum features; the real-time extracted data features are mapped to preset surgical status categories, which include fully inserted, about to be inserted, and insufficiently inserted;

[0050] The AI processing module tracks the surgical progress through time data series and predicts the stress state and penetration degree of bone tissue, including:

[0051] Perform sequential modeling on the intraoperative parameter data at different time steps to generate a standardized time data series with equal time intervals, and capture the temporal patterns of changes in intraoperative parameter data based on the time data series.

[0052] Combine the data of the six-axis force sensor 4 and the nine-axis acceleration sensor 5 to dynamically calculate the force change curve of the acetabular hammer, and determine whether the impact force applied by the surgeon meets the surgical standards; through the spectral analysis of bone-conducted sound, real-time identify the changes in bone tissue density, judge whether there is a risk of insufficient insertion or bone injury, and automatically record and mark the key intraoperative time points; if an abnormal intraoperative operation is detected, provide a warning message to the surgeon in real time through the human-computer interaction module.

[0053] Generate adaptive parameter adjustment suggestions based on the surgical process prediction results. The adaptive parameter adjustment suggestions include adjusting the impact force and optimizing the movement trajectory to improve the surgical accuracy.

[0054] The AI processing module provides adaptive parameter adjustment suggestions, specifically including:

[0055] Define the state space, which includes real-time intraoperative parameter data and surgical state categories; set the action space, which includes the adjustment ranges of the impact force, rotation angle, and impact frequency of the acetabular hammer.

[0056] During the operation, calculate the adjustment strategy in the current surgical state according to the current state space. The adjustment strategy includes reducing the impact force, changing the rotation angle, and adjusting the impact frequency. The adjustment range of the adjustment strategy is controlled within the action space.

[0057] During the operation, continuously evaluate the effect after the operation adjustment, provide real-time parameter adjustment suggestions to the surgeon through the human-computer interaction module, and provide a sound feedback prompt if it is detected that the surgeon's operation has not been adjusted.

[0058] Store the intraoperative parameter adjustment records after the operation to form a surgical optimization plan.

[0059] In the above embodiment, the AI processing module performs deep feature extraction on the intraoperative parameter data based on the convolutional neural network (CNN), and combines the feature fusion network for multi-modal data analysis, enabling the system to accurately identify different surgical stages and operation states.

[0060] In the above embodiments, during the extraction of mechanical waveform features, multi-scale convolutional kernels are adopted, enabling the system to analyze the variation patterns of Z-axis pressure and RX-axis torque in different frequency bands, thereby accurately determining whether the impact force applied by the operator meets the standards. During the extraction of acceleration curve features, the AI processing module classifies the motion features at different surgical stages through time window segmentation, enabling the system to accurately identify the operator's operating habits at different stages and provide targeted adjustment suggestions. During the extraction of bone conduction sound spectrum features, combined with the short-time Fourier transform, the time-domain signal is converted into frequency-domain features, enabling the AI processing module to judge the changing trend of bone density during the operation in real time based on the spectrum changes of the bone tissue sound signal.

[0061] In the above embodiments, through these deep learning technologies, the system can convert complex intraoperative data into interpretable surgical state information, providing real-time and intuitive decision-making support for the operator, reducing the errors relying on empirical judgment during the operation, improving the passing rate of novice doctors' operations, and enhancing the accuracy and consistency of the operation.

[0062] The human-computer interaction module is connected to the in-place degree indicator 3. The in-place degree indicator 3 is provided with red, green, and blue indicator lights, and controls the in-place degree indicator 3 based on the real-time surgical state during the operation. Among them, the red indicator light indicates insufficient insertion, the blue indicator light indicates that the insertion is about to be in place, and the green indicator light indicates that the insertion is in place.

[0063] The human-computer interaction module is electrically connected to the wireless data transmission interface, and the human-computer interaction module communicates with the external display device through the wireless data transmission interface.

[0064] In the above embodiments, the addition of the human-computer interaction module enables the operator to obtain intuitive surgical state feedback at any time during the operation, thereby adjusting the operation strategy more efficiently. The in-place degree indicator uses red, green, and blue indicators, where the red indicator light indicates insufficient insertion, the blue indicator light indicates that the insertion is about to be in place, and the green indicator light indicates that the insertion is in place. Through this intuitive visual feedback, the operator can quickly judge whether the current operation meets the surgical requirements without frequently checking the external display device, improving the surgical efficiency.

[0065] In addition, the human-computer interaction module communicates with the external display device through the wireless data transmission interface, enabling the intraoperative parameter data to be synchronously displayed on the monitoring screen for the operator's reference. If it is detected that the operator's operation has not been adjusted, the system will also provide voice feedback prompts to enhance the operator's perception of the surgical state, improving the operator's perception of the intraoperative situation, reducing the surgical failure rate caused by improper operation, and providing a more efficient solution for precision medicine.

[0066] The operation modes include a recording mode and an inference mode;

[0067] When the dual-mode operation module is in the recording mode, it records and marks the changes in intraoperative parameter data during the surgical procedures of different patients; when the dual-mode operation module is in the inference mode, during the surgical procedure, it performs real-time analysis and navigation based on the AI processing module, providing intraoperative warnings and operation optimization suggestions.

[0068] In the above embodiment, the introduction of the dual-mode operation module enables the acetabular hammer to be used not only for real-time analysis and optimization during the surgical procedure but also for postoperative data recording and analysis. In the recording mode, the system automatically records the changes in intraoperative parameter data during the surgical procedures of different patients and stores the key time points and parameter adjustment records during the surgical procedure. These data can be used not only for postoperative review but also for the continuous optimization and training of the AI model, enabling the system to continuously improve the prediction accuracy of the surgical state. In the inference mode, the system provides intraoperative warnings and operation optimization suggestions based on the real-time analysis results of the AI processing module. Through this dual-mode operation, the surgeon can not only obtain intelligent assistance during the operation but also optimize their surgical skills through postoperative data analysis, thereby continuously improving the surgical quality and promoting the further development of intelligent surgery.

[0069] The anomaly detection module obtains the real-time data of the Z-direction pressure and RX-direction torque in the intraoperative parameter data, combines the surgical process data, partitions the mechanical data in different surgical stages, and distinguishes the force application modes in different surgical stages;

[0070] Set the safe force application threshold for bone tissue and adjust the threshold based on the preoperative images for personalization;

[0071] During the surgical procedure, based on the force application modes in different surgical stages, combine the safe force application threshold for bone tissue to detect the Z-direction pressure and RX-direction torque. When the Z-direction pressure or RX-direction torque exceeds the safe force application threshold for bone tissue, calculate the degree of overlimit;

[0072] Send a warning message to the surgeon through the human-machine interaction module, and at the same time activate the red flashing alarm of the in-place degree indicator 3 to prompt the surgeon to adjust the operation force;

[0073] The anomaly detection module obtains the bone conduction sound signal and the bone conduction sound spectrum characteristics; combines with the AI processing module to establish a sound feature library for bone tissue being driven in place, about to be driven in place, insufficiently driven in, and bone quality damage;

[0074] Compare the real-time collected sound spectrum characteristics with the sound feature library to determine the current bone density change trend; when it is detected that the bone tissue spectrum characteristics deviate from the normal range and the high-frequency components are abnormally enhanced, it is determined that there is a risk of insufficient driving in and bone quality damage;

[0075] Provide real-time feedback to the operator through the human-machine interaction module. If it is determined that the insertion is insufficient, light up the red warning of the in-place degree indicator 3; if it is determined that the insertion is about to be in place, light up the blue warning light and emit a prompt sound to remind the operator to adjust the operation force.

[0076] In the above embodiment, the application of the abnormal detection module enables the system to automatically identify abnormal situations during the operation and trigger a safety alarm in a timely manner to ensure the safety of the operation. Based on the six-axis force sensor, the real-time data of the Z-direction pressure and the RX-direction torque are obtained, and the force application modes in different surgical stages are distinguished by combining the surgical process data, so as to accurately detect whether the operator's operation exceeds the safe force threshold of the bone tissue. When it is detected that the Z-direction pressure or the RX-direction torque exceeds the threshold, the system will calculate the overrun degree and send a warning message to the operator through the human-machine interaction module, and at the same time activate the red flashing alarm of the in-place degree indicator to remind the operator to adjust the operation force.

[0077] In addition, abnormal detection is combined with bone conduction sound signals, and a sound feature library for bone tissue insertion in place, about to be in place, insufficient insertion, and bone damage is established based on the AI model. When the sound spectrum features collected in real time deviate from the normal range and the high-frequency components are abnormally enhanced, the system will determine that there is a risk of insufficient insertion and bone damage, and remind the operator to adjust the operation method through the in-place degree indicator and sound prompts, improving the operator's perception ability of abnormal situations and reducing the occurrence of intraoperative accidents.

[0078] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, according to the technical solution and inventive concept of the present invention, makes equivalent substitutions or changes, and should be covered by the protection scope of the present invention.

Claims

1. Acetabular mallet based on AI intelligent algorithm, comprising an acetabular mallet main body (1), characterized in that: A hand-held rod (2) is provided on the acetabular hammer main body (1). An in-place degree indicator light (3) is provided on the hand-held rod (2). Inside the hand-held rod (2), a multi-modal sensor group, an AI processing module, a human-computer interaction module, a dual-mode operation module, and an anomaly detection module are provided; The multi-modal sensor group is configured to collect intraoperative parameter data; the AI processing module is configured to process the intraoperative parameter data in real time, extract the characteristic patterns of the intraoperative parameter data, and track the surgical process to predict the stress state and driving degree of the bone tissue, and provide adaptive parameter adjustment suggestions; The human-computer interaction module is configured to communicate with an external display device for data and provide real-time feedback information to the surgeon; the dual-mode operation module is configured to adjust the operation mode of the acetabular hammer; The anomaly detection module is configured to automatically detect intraoperative abnormal conditions and trigger a safety alarm when an anomaly occurs.

2. The acetabular hammer based on the AI intelligent algorithm according to claim 1, wherein: The multi-modal sensor group includes a six-axis force sensor (4), a 9-axis acceleration sensor (5), and a bone conduction sound sensor (6); Among them, the six-axis force sensor (4) is used to detect the Z-direction pressure and the RX-direction torque; the 9-axis acceleration sensor (5) is used to obtain the acceleration, angular velocity, and attitude information of the acetabular hammer in space; the bone conduction sound sensor (6) is used to collect the bone conduction sound signals of the bone tissue in different states during the operation.

3. The acetabular hammer based on the AI intelligent algorithm according to claim 1, wherein: The AI processing module is equipped with an MCU (7) containing a convolutional neural network unit, and performs real-time processing on the original intraoperative parameter data collected by the multi-modal sensor group, eliminates outliers and normalizes the data, performs time alignment and interpolation processing on the intraoperative parameter data, and generates a standardized time data sequence with equal time intervals; Extract the data features of the intraoperative parameter data through a convolutional neural network. The data features include mechanical waveform features, acceleration curve features, and bone conduction sound spectrum features; In the process of extracting mechanical waveform features, obtain the Z-direction pressure and RX-direction torque change patterns in different frequency bands through multi-scale convolutional kernels; in the process of extracting acceleration curve features, adopt a time window segmentation method to identify the motion features in different operation stages; in the process of extracting bone conduction sound spectrum features, combine short-time Fourier transform to convert the time-domain signal into a frequency-domain feature; Adopt a feature fusion network to fuse the mechanical waveform features, acceleration curve features, and bone conduction sound spectrum features in a multi-modal manner; Map the data features extracted in real time to preset surgical state categories, and the surgical state categories include being driven in place, about to be driven in place, and insufficient driving.

4. The acetabular hammer based on the AI intelligent algorithm according to claim 3, wherein: The AI processing module tracks the surgical process through the time data sequence and predicts the stress state and driving degree of the bone tissue, specifically including: Perform sequence modeling on the intraoperative parameter data at different time steps to generate a standardized time data sequence with equal time intervals, and capture the temporal pattern of the change of the intraoperative parameter data based on the time data sequence; Combining the data of the six - axis force sensor (4) and the 9 - axis acceleration sensor (5), dynamically calculate the force change curve of the acetabular hammer, and determine whether the impact force applied by the surgeon meets the surgical standard; through the spectral analysis of bone - conducted sound, real - time identify the change of bone tissue density, judge whether there is a risk of insufficient insertion or bone injury, and automatically record and mark the key time points during the operation; if an abnormal operation is detected during the operation, provide a warning message to the surgeon in real - time through the human - machine interaction module; Generate an adaptive parameter adjustment suggestion based on the prediction result of the surgical process, and the adaptive parameter adjustment suggestion includes adjusting the impact force and optimizing the movement trajectory.

5. The acetabular hammer based on the AI intelligent algorithm according to claim 4, wherein: The AI processing module provides an adaptive parameter adjustment suggestion, specifically including: Define the state space, which includes real - time intraoperative parameter data and surgical state categories; set the action space, which includes the adjustment ranges of the impact force, rotation angle, and impact frequency of the acetabular hammer; During the operation, calculate the adjustment strategy in the current surgical state according to the current state space. The adjustment strategy includes reducing the impact force, changing the rotation angle, and adjusting the impact frequency, and the adjustment range of the adjustment strategy is controlled within the action space; During the operation, continuously evaluate the effect after the operation adjustment, provide real - time parameter adjustment suggestions to the surgeon through the human - machine interaction module. If it is detected that the surgeon's operation has not been adjusted, provide a sound feedback prompt; After the operation, store the intraoperative parameter adjustment record to form an optimized surgical plan.

6. The acetabular hammer based on the AI intelligent algorithm according to claim 1, wherein: The human - machine interaction module is connected to the in - place degree indicator light (3). The in - place degree indicator light (3) is provided with red, green, and blue indicator lights, and controls the in - place degree indicator light (3) based on the real - time surgical state during the operation. Among them, the red indicator light indicates insufficient insertion, the blue indicator light indicates that the insertion is about to be in place, and the green indicator light indicates that the insertion is in place.

7. The acetabular hammer based on the AI intelligent algorithm according to claim 6, wherein: The human - machine interaction module is electrically connected to the wireless data transmission interface, and the human - machine interaction module communicates with the external display device through the wireless data transmission interface.

8. The acetabular hammer based on the AI intelligent algorithm according to claim 1, characterized in that: The operation mode includes a recording mode and a reasoning mode; When the dual - mode operation module is in the recording mode, record and mark the changes of intraoperative parameter data during the operations of different patients; when the dual - mode operation module is in the reasoning mode, during the operation, based on the AI processing module, perform real - time analysis and navigation, and provide intraoperative early warning and operation optimization suggestions.

9. The acetabular hammer based on the AI intelligent algorithm according to claim 1, wherein: The abnormal detection module obtains the real - time data of the Z - direction pressure and RX - direction torque in the intraoperative parameter data, combines the surgical process data, partitions the mechanical data in different surgical stages, and distinguishes the force - bearing modes in different surgical stages; Set the safe force - bearing threshold of the bone tissue, and adjust the threshold based on the preoperative image personalizedly; During the operation, based on the force - bearing modes in different surgical stages, combine the safe force - bearing threshold of the bone tissue to detect the Z - direction pressure and RX - direction torque. When the Z - direction pressure or RX - direction torque exceeds the safe force - bearing threshold of the bone tissue, calculate the over - limit degree; Send a warning message to the surgeon through the human - machine interaction module, and at the same time activate the red flashing alarm of the in - place degree indicator light (3) to prompt the surgeon to adjust the operation force.

10. The acetabular hammer based on the AI intelligent algorithm according to claim 9, characterized in that: The abnormal detection module acquires bone conduction sound signals and bone conduction sound spectrum features; combines with the AI processing module to establish a sound feature library for bone tissue being inserted in place, about to be inserted in place, insufficient insertion, and bone mass injury; Compare the sound spectrum features collected in real time with the sound feature library to determine the current bone density change trend; when it is detected that the bone tissue spectrum features deviate from the normal range and the high-frequency components are abnormally enhanced, it is determined that there is a risk of insufficient insertion and bone mass injury; Provide real-time feedback to the surgeon through the human-computer interaction module. If it is determined that the insertion is insufficient, light up the red warning of the in-place degree indicator (3); if it is determined that the insertion is about to be in place, light up the blue warning light and emit a prompt sound to remind the surgeon to adjust the operation force.

Images