Surgical robot and safety control method thereof

Abstract

The invention provides a surgical robot and a safety control method thereof. The method comprises the following steps: acquiring audio information in an operating room; analyzing the audio information to obtain an abnormal degree score of the audio information; and under the condition that the abnormal degree score exceeds an abnormal threshold value, the surgical robot is controlled to be shut down or an operation part of the surgical robot is controlled to leave the patient. According to the characteristic that the medical staff cannot autonomously make the sound under the abnormal condition, the abnormal condition is automatically, intelligently and accurately recognized by analyzing the audio information of the operating room environment, recognizing keywords, duplicated words, volume characteristics and voiceprint characteristics in the operating room environment, and due to the fact that semantics is not analyzed, the abnormal condition can be automatically, intelligently and accurately recognized. Even if a doctor cannot correctly send out a voice instruction due to emotion tension, the system can timely and correctly respond, automatically recognize abnormal conditions, control the surgical robot to be shut down or control the operation part of the surgical robot to leave the patient, and guarantee the safety of the patient.

Description

Technical Field

[0001] This invention relates to the field of medical device technology, and in particular to a surgical robot and its safety control method. Background Technology

[0002] The use of surgical robots is crucial to patient safety. During operation, abnormal situations may arise, requiring the robot to be shut down or its operating parts removed from the patient. In practice, surgeons may be focused on the procedure or have their hands occupied by surgical instruments, failing to shut down the equipment in time and increasing the risk of injury to the patient.

[0003] Existing technologies mitigate risks by training patients on risk awareness and safe operating procedures for medical equipment. However, these methods are slow to take effect and offer limited improvement in safety. Another existing solution involves issuing voice commands. This solution uses the surgical robot to recognize and understand the voice commands, controlling the robot to shut down or remove its operating parts from the patient. However, in emergency situations or when doctors are under pressure, there is a high probability that voice commands cannot be issued promptly and correctly, leading to a continued high safety risk.

[0004] To at least partially address the aforementioned deficiencies, the present invention provides a surgical robot and a safety control method thereof. Summary of the Invention

[0005] This invention provides a surgical robot and its safety control method to address the shortcomings of existing technologies where surgical robots pose a high risk of injury to patients under abnormal conditions.

[0006] In a first aspect, the present invention provides a safety control method for a surgical robot, comprising: Obtain audio information from inside the operating room; The audio information is analyzed to obtain an abnormality score. If the abnormality score exceeds the abnormality threshold, the surgical robot is shut down or the operating part of the surgical robot is moved away from the patient.

[0007] Optionally, the step of analyzing the audio information to obtain an anomaly score for the audio information includes: Each evaluation metric of the audio information is scored; wherein, the evaluation metrics include: keywords, repeated words, long sounds, and the volume of keywords, repeated words, or long sounds; The scores are weighted according to the weights of each evaluation indicator to obtain the anomaly score.

[0008] Furthermore, the weight corresponding to the volume is greater than the weight of the keyword, repeated word, and long tone.

[0009] Optionally, the evaluation index also includes voiceprint, wherein the voiceprint score under abnormal emotions is higher than the voiceprint score under normal conditions.

[0010] Furthermore, before assigning scores to each evaluation indicator of the audio information, the method further includes: identifying identity information based on the voiceprint and excluding audio information input by irrelevant personnel.

[0011] Optionally, before analyzing the audio information, the method further includes: The audio information is noise-reduced to remove background noise.

[0012] Optionally, after acquiring the audio information in the operating room, the method further includes: Analyze whether the audio information contains any device malfunction sounds; If included, control the shutdown of the surgical robot or control the operating part of the surgical robot to leave the patient.

[0013] Optionally, the analysis process uses an audio information scoring model to process the audio information and give a score. The audio information scoring model includes one of the following forms: pattern matching model, hidden Markov model, feature extraction model, or neural network model.

[0014] Furthermore, the anomaly threshold is set in the following manner: Using a set of labeled audio samples, the audio information scoring model and the undetermined anomaly threshold are jointly trained to obtain the trained audio information scoring model and the adjusted anomaly threshold; wherein, the set of labeled audio samples includes audio samples labeled as normal and audio samples labeled as abnormal.

[0015] Furthermore, the method also includes: Automatic labeling of audio samples collected in the operating room; The audio information scoring model and the corresponding anomaly threshold are trained a second time using automatically labeled audio samples.

[0016] Furthermore, the automatic annotation of audio samples collected in the operating room includes: Obtain the surgical robot's status information within a preset time period after collecting audio samples; If the surgical robot does not stop or the operating part does not leave the patient within the preset time, the audio sample will be automatically marked as normal. If the surgical robot stops or the operating unit leaves the patient within the preset time period, the audio sample will be automatically marked as abnormal.

[0017] Secondly, the present invention also provides a surgical robot, comprising: an audio information acquisition module, a processing module, and an execution module; The audio information acquisition module is used to acquire audio information in the operating room; The processing module is used to process the audio information according to the safety control method of the surgical robot described in any of the preceding claims, and to control the execution module according to the processing result.

[0018] The surgical robot and its safety control method provided by this invention have at least the following beneficial effects: 1. This invention can automatically identify abnormal situations and control the shutdown of the surgical robot or control the operating part of the surgical robot to leave the patient, thus ensuring patient safety.

[0019] 2. Based on the characteristics of the sounds that medical staff involuntarily make under abnormal circumstances, this invention analyzes the audio information of the operating room environment to identify keywords, repeated words, volume characteristics, and voiceprint characteristics, and automatically, intelligently, and accurately identifies abnormal situations. Since this invention does not analyze semantic commands, it can respond promptly and correctly even if doctors fail to issue voice commands correctly due to emotional tension.

[0020] 3. By recognizing voiceprints, interference from audio information input by unauthorized personnel can be avoided.

[0021] 4. By recognizing abnormal sounds from the equipment, it automatically determines abnormal conditions (such as the surgical robot encountering an obstacle or the tool at the end of the robot breaking), ensuring the safe operation of the equipment. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0023] Figure 1 This is one of the flowcharts illustrating a safety control method for a surgical robot provided by the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0025] Example 1 like Figure 1 As shown, a safety control method for a surgical robot includes the following steps: S1: Obtain audio information from the operating room; During surgery, audio information within the operating room, including the voices of medical staff and the sounds of equipment operation, can be collected. This audio information can be captured using the microphones / microphone arrays built into the surgical robot, or it can be collected in real time using microphone arrays placed in the operating room, such as those positioned on the walls or ceiling, to ensure comprehensive sound capture within the operating room.

[0026] S2: Analyze the audio information to obtain an abnormality score for the audio information; Conventional speech recognition converts audio information into natural language characters, using these characters as control commands. Its essence is semantic understanding, and these commands typically require specific keywords; different operations require different keywords. In practical applications, during emergencies, operators may mispronounce commands due to nervousness or fail to issue timely and correct semantic instructions.

[0027] The inventors discovered that in emergency situations, operators may emit sounds lacking semantic information, such as "Hey hey hey!", "Ah (long sound)," or "Quick! Quick!". This invention, combining the vocal characteristics of operators in emergency situations, specifically evaluates the degree of abnormality and provides an abnormality score. Specifically, in some emergency situations, operators may be unable to issue clear and unambiguous voice commands, only uttering ambiguous keywords lacking semantic information, such as "Ah," "Hey," and "Quick," or repeating these words. Furthermore, the volume of sounds emitted in emergency situations is usually louder than in normal situations, and the voiceprint (e.g., timbre) of abnormal sounds may also differ from that of normal sounds. Further, before analyzing the audio information, noise reduction processing can be performed on the collected audio information to remove background noise and improve the accuracy of subsequent analysis. Noise reduction processing can employ techniques such as frequency domain filtering and adaptive filtering to effectively remove background noise, low-frequency noise from instruments and equipment, and other interference signals.

[0028] S3: If the abnormality score exceeds the abnormality threshold, control the shutdown of the surgical robot or control the operating part of the surgical robot to leave the patient.

[0029] When the calculated anomaly score exceeds a preset anomaly threshold, the system immediately triggers a safety mechanism, choosing to either shut down the surgical robot or safely remove its operating parts from the patient, depending on the current surgical status. Shutting down the robot could involve cutting off its power system or locking the robotic arm; removing it could involve following a preset safety path to ensure the robotic arm moves smoothly and without collision away from the patient. Preferably, based on the robotic arm's already executed movement path, the system controls the robot to remove it along the reverse path to avoid secondary injury to the patient.

[0030] In this method, the audio information scoring model can include one of the following forms: pattern matching model, hidden Markov model, feature extraction model, or neural network model. Among them, the neural network model has better learning ability and adaptability, and can handle complex audio features.

[0031] The anomaly threshold is set as follows: Using a pre-labeled audio sample set, the audio information scoring model and the proposed anomaly threshold are jointly trained to obtain the trained audio information scoring model and the adjusted anomaly threshold. The pre-labeled audio sample set includes audio samples labeled as normal (such as routine surgical conversations) and audio samples labeled as abnormal (such as instructions or warnings in emergency situations). Through training with a large number of samples, the system can accurately distinguish between normal and abnormal situations.

[0032] Conventional voice control involves recognizing voice commands with specific semantics and controlling the device's state accordingly. As an alternative, this approach can be combined with conventional voice control, enabling both the recognition and execution of conventional voice control commands and the identification of abnormal voice signals for safety control. In some examples, voice signals captured in the operating room are processed sequentially using two methods (conventional voice control and the voice safety control method of this application). In other examples, voice signals captured in the operating room are executed in parallel using the two methods; that is, the surgical robot runs both methods simultaneously, and the safety control method of this application directly takes over control when an emergency is detected, overriding the conventional voice control commands. In some examples, priorities can be set for the execution results of the two methods, for example, prioritizing the safety control method of this application and only executing conventional voice control if no abnormality is detected.

[0033] Example 2 like Figure 1 As shown, a safety control method for a surgical robot includes the following steps: S1: Obtain audio information from the operating room; During surgery, a microphone array positioned within the operating room captures audio information in real time, including the voices of medical staff and the sounds of equipment operation. The microphone array can be placed on the walls or ceiling of the operating room to ensure comprehensive sound capture.

[0034] S2: Analyze the audio information to obtain an abnormality score for the audio information; Before analyzing the audio information, noise reduction processing is first performed on the acquired audio data to remove background noise and improve the accuracy of subsequent analysis. Noise reduction processing can employ techniques such as frequency domain filtering and adaptive filtering to effectively remove interference signals such as air conditioner noise and low-frequency noise from instruments and equipment.

[0035] S20: Identify identity information based on voiceprint and exclude audio information input by irrelevant personnel; The system uses a pre-established voiceprint database to identify the identities of the collected audio, including the voices of the surgeon, assistant surgeon, anesthesiologist, and other relevant medical personnel. At the same time, it excludes the voice interference from irrelevant personnel who may enter the operating room (such as interns and visitors), ensuring that the system only analyzes the audio information of key medical personnel.

[0036] S21: Give scores for each evaluation indicator of audio information; Evaluation metrics include: keywords, repeated words, long vowels, and the volume of keywords, repeated words, or long vowels, as well as timbre. The system scores each of these metrics separately. - Keywords: Detect words indicating urgency such as "ai," "ah," "kuai," and "ting." 5 points are added for each keyword detected. In addition, to eliminate interference from modal particles in regular sentences, the independence of these words can be checked to eliminate pseudo-keywords appearing in complete sentences.

[0037] - Repeated words: Detect words that medical staff use repeatedly in a short period of time, such as "Hey hey hey!" or "Quick! Quick!". Each set of repeated words detected adds 3 points. Similarly, in order to eliminate the interference of repeated words in regular sentences, such as "going to work every day" or "noisy" that operators may use in chat, the independence of these words can be tested to exclude keywords that appear in complete sentences. - Long vowel: Detects prolonged tones in speech, indicating surprise or nervousness. 2 points are added for each long vowel detected; - Volume: Detects the decibel value of the voice. When the decibel value exceeds 30% of the normal conversation volume, add 5 points for every 10% exceeding it; - Timbre: By analyzing the spectral characteristics of the audio (such as sharpness and vibrato frequency), the speaker's emotional state is identified. When a rapid, high-pitched voice is detected, the score is 8 points, while the score is 2 points under normal conditions.

[0038] Among them, the score for timbre under abnormal emotions is higher than the score for timbre under normal conditions. This is because changes in timbre can directly reflect the emotional state of medical staff and are an important indicator for judging abnormal surgical conditions.

[0039] S22: Weight the scores according to the weights of each evaluation indicator to obtain the abnormality score; The system assigns different weights to each evaluation indicator. For example, volume has a weight of 0.3, keywords have a weight of 0.2, repeated words have a weight of 0.2, sustained sounds have a weight of 0.2, and timbre has a weight of 0.1. Volume has a higher weight than keywords, repeated words, and sustained sounds because a sudden increase in volume is usually the most direct indication of an emergency. The final anomaly score is obtained through weighted calculation, using the following formula: Abnormality score = Volume score × 0.3 + Keyword score × 0.2 + Repeated word score × 0.2 + Long note score × 0.2 + Timbre score × 0.1 S3: If the abnormality score exceeds the abnormality threshold, control the shutdown of the surgical robot or control the operating part of the surgical robot to leave the patient.

[0040] When the calculated abnormality score exceeds a preset abnormality threshold (e.g., 75 points), the system will immediately trigger a safety mechanism, choosing to either shut down the surgical robot or safely evacuate the robot's operating unit from the patient, depending on the current surgical status. Shutting down the robot involves cutting off its power system and locking the robotic arm; evacuation involves smoothly withdrawing the robotic arm from the patient's body along a preset safety path to avoid secondary injury.

[0041] It is understandable that the above scoring, weighting, and anomaly threshold are just examples. The specific numerical settings can be adjusted accordingly, as long as they can identify abnormal situations.

[0042] S4: Analyze the audio information to see if it contains any device malfunction sounds; In addition to analyzing the operator's voice, the system also monitors for abnormal noises from equipment in the operating room, such as abnormal friction sounds from robot joints, abnormal sounds from the hydraulic system, and abnormal sounds from motors. If such abnormal sounds are detected, the system will immediately trigger a safety mechanism to shut down the surgical robot or control the operating part of the surgical robot to move away from the patient, without waiting for the calculation results of the abnormality severity score.

[0043] In this method, the audio information scoring model can include one of the following forms: pattern matching model, hidden Markov model, feature extraction model, or neural network model. Pattern matching refers to identifying specific patterns or structures in data. A pattern can be a simple sequence of characters or a complex expression containing wildcards or special rules. Specifically, it can be used to match speech templates and identify keywords and repeated words through Dynamic Time Warping (DTW) technology. A Hidden Markov Model (HMM) is a statistical model used to describe a Markov process with hidden unknown parameters. HMMs can be applied to speech processing and can also be combined with Gaussian Mixture Models to model acoustic features. Feature extraction models, for example, utilize techniques such as Mel-frequency cepstral coefficients (MFCC) and Linear Predictive Coding (LPC) to extract acoustic features. A neural network model is a complex network system formed by extensive interconnection of a large number of simple processing units (called neurons). It simulates many basic characteristics of human brain function and is a highly complex nonlinear dynamic learning system. Neural network models have better learning ability and adaptability, and can handle complex audio features. Specific examples include Recurrent Neural Networks (RNNs), Long Short-Term Memory Networks (LSTMs), and Transformers.

[0044] The anomaly threshold is set as follows: Using a pre-labeled audio sample set, the audio information scoring model and the proposed anomaly threshold are jointly trained to obtain the trained audio information scoring model and the adjusted anomaly threshold. The pre-labeled audio sample set includes audio samples labeled as normal (such as routine surgical communication) and audio samples labeled as abnormal (such as instructions or warnings in emergency situations). Through training with a large number of samples, the system can accurately distinguish between normal and abnormal situations. During training, audio samples are input into the audio information scoring model to obtain corresponding scores. These scores are then combined with the initial anomaly threshold to determine whether the audio information is abnormal. By comparing the audio information's label (normal / abnormal), the parameters of the audio information scoring model and the initial anomaly threshold are adjusted based on feedback until the trained audio information scoring model and anomaly threshold achieve a high recognition accuracy.

[0045] To continuously optimize system performance, this method also includes automatically labeling audio samples collected in the operating room and using the automatically labeled audio samples to retrain the audio information scoring model and corresponding anomaly thresholds. The automatic labeling process is as follows: The status information of the surgical robot within a preset time period (e.g., 5 seconds) after the audio sample is collected is obtained; if the surgical robot has not stopped or the operating part has not left the patient within the preset time period, the audio sample is automatically labeled as normal; if the surgical robot has stopped or the operating part has left the patient within the preset time period, the audio sample is automatically labeled as abnormal. This method enables automatic labeling of audio samples, saving labor costs. This adaptive learning mechanism also allows the system to continuously improve, increasing the accuracy of identifying anomalies in specific hospitals / medical teams.

[0046] Furthermore, in some preferred embodiments, the changes in the surgical robot's state information caused by "normal operation switching, normal surgical procedure completion" in the above-mentioned automatically labeled data are also filtered out, and the corresponding audio data is changed and labeled as normal.

[0047] Furthermore, the aforementioned robot status information can be obtained based on the robot's work logs, or from video data related to the operation of the surgical robot, such as video data of the surgical robot in operation, or video data of the operation process captured by the surgical robot's built-in camera (such as the end-effector camera). The robot status information can also be obtained from medical imaging data, such as observing the operation / stopping status of the robot's operating parts through ultrasound imaging.

[0048] Example 3 like Figure 1 As shown in the figure, this embodiment provides a surgical robot, including an audio information acquisition module, a processing module, and an execution module.

[0049] The audio information acquisition module is used to collect audio information within the operating room. This module can be a microphone array installed on the walls or ceiling of the operating room to ensure comprehensive sound acquisition, including the voices of medical staff and the sounds of equipment operation. The audio information acquisition module is connected to the processing module via wired or wireless means to transmit the acquired audio information to the processing module.

[0050] The processing module processes audio information according to the safety control method for the surgical robot described in Embodiment 1, and controls the execution module based on the processing results. The processing module can be a computing unit integrating audio signal processing algorithms, capable of performing functions such as audio information analysis and anomaly score calculation as described in Embodiment 1. First, the module performs noise reduction processing on the acquired audio information. This noise reduction employs frequency domain filtering technology, analyzing the frequency components of the audio signal to remove irrelevant noise in the low and high frequency bands, while retaining key signals such as human voices and equipment sounds. Then, for human voices, identity is identified based on voiceprints. Specifically, processing module 2 pre-stores a database of voiceprint features of surgical team members. By comparing the voiceprint features of the audio signal, it determines whether the audio signal originates from a member of the surgical team. If the audio signal originates from a member of the surgical team, further analysis continues; if the audio signal originates from unrelated personnel, it is filtered out to avoid interfering with the system's judgment. Next, various evaluation indicators of the human voice audio (including keywords, repeated words, long vowels, volume, and timbre) are scored, and finally, an anomaly score is calculated based on the weight of each indicator. For equipment sounds, pattern matching is used to determine whether they belong to abnormal equipment noises, such as the friction sound of mechanical parts or abnormal noises from motors. If abnormal equipment noises are detected, the processing module will also generate control commands and send them to the execution module to take timely measures to deal with equipment malfunctions.

[0051] The execution module is the actuator of the surgical robot system, responsible for executing control commands. When the processing module detects an abnormality (such as an abnormality score exceeding an abnormality threshold or detecting abnormal sounds from the equipment), the execution module will perform corresponding safety operations according to the instructions of the processing module, such as shutting down the surgical robot or controlling the operating part of the surgical robot to move away from the patient. The execution module includes components such as the robot's power system, control system, and robotic arm. Specifically, the execution module is connected to the surgical robot's power system through a drive circuit. When it receives a stop command, the execution module cuts off the power supply to the power system, causing the surgical robot to immediately stop all movements and maintain its posture. When it receives a command to move the operating part away from the patient, the execution module activates the surgical robot's obstacle avoidance algorithm, adjusts the posture of the robotic arm, and quickly removes its operating part from the patient's body surface.

[0052] In this surgical robot system, three modules work together to form a complete safety monitoring and response system. The audio information acquisition module continuously monitors the operating room environment, the processing module analyzes the audio data in real time and makes judgments, and the execution module takes appropriate safety measures based on the processing results, jointly ensuring the safety of the surgical procedure.

[0053] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0054] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A safety control method for a surgical robot, characterized in that, include: Obtain audio information from inside the operating room; The audio information is analyzed to obtain an abnormality score. If the abnormality score exceeds the abnormality threshold, the surgical robot is shut down or the operating part of the surgical robot is moved away from the patient.

2. The safety control method for the surgical robot according to claim 1, characterized in that, The analysis of the audio information to obtain an anomaly score includes: Each evaluation metric of the audio information is scored; wherein, the evaluation metrics include: keywords, repeated words, long sounds, and the volume of keywords, repeated words, or long sounds; The scores are weighted according to the weights of each evaluation indicator to obtain the anomaly score.

3. The safety control method for the surgical robot according to claim 2, characterized in that, The weight corresponding to the volume is greater than the weight of the keyword, repeated words, and long sounds.

4. The safety control method for the surgical robot according to claim 2, characterized in that, The evaluation criteria also include voiceprints, where the voiceprint score under abnormal emotions is higher than the voiceprint score under normal conditions.

5. The safety control method for the surgical robot according to claim 4, characterized in that, Before assigning scores to each evaluation metric of the audio information, the method further includes: identifying identity information based on the voiceprint and excluding audio information input by irrelevant personnel.

6. The safety control method for a surgical robot according to claim 2, characterized in that, Before analyzing the audio information, the method further includes: The audio information is noise-reduced to remove background noise.

7. The safety control method for a surgical robot according to claim 1, characterized in that, After acquiring the audio information from the operating room, the process also includes: Analyze whether the audio information contains any device malfunction sounds; If included, control the shutdown of the surgical robot or control the operating part of the surgical robot to leave the patient.

8. The safety control method for a surgical robot according to claim 1, characterized in that, The analysis process uses an audio information scoring model to process the audio information and give a score. The audio information scoring model includes one of the following forms: pattern matching model, hidden Markov model, feature extraction model, or neural network model.

9. The safety control method for a surgical robot according to claim 8, characterized in that, The anomaly threshold is set in the following manner: Using a set of labeled audio samples, the audio information scoring model and the undetermined anomaly threshold are jointly trained to obtain the trained audio information scoring model and the adjusted anomaly threshold; wherein, the set of labeled audio samples includes audio samples labeled as normal and audio samples labeled as abnormal.

10. The safety control method for a surgical robot according to claim 9, characterized in that, The method also includes: Automatic labeling of audio samples collected in the operating room; The audio information scoring model and the corresponding anomaly threshold are trained a second time using automatically labeled audio samples.

11. The safety control method for a surgical robot according to claim 10, characterized in that, The automatic annotation of audio samples collected in the operating room includes: Obtain the surgical robot's status information within a preset time period after collecting audio samples; If the surgical robot does not stop or the operating part does not leave the patient within the preset time, the audio sample will be automatically marked as normal. If the surgical robot stops or the operating unit leaves the patient within the preset time period, the audio sample will be automatically marked as abnormal.

12. A surgical robot, characterized in that, include: Audio information acquisition module, processing module, execution module; The audio information acquisition module is used to acquire audio information in the operating room; The processing module is used to process the audio information according to the safety control method of the surgical robot according to any one of claims 1-11, and to control the execution module according to the processing result.

Images