Special anesthesia and proton therapy integrated equipment for children and control method

Abstract

The invention discloses special anesthesia and proton treatment integrated equipment for children and a control method, relates to the field of intelligent medical treatment, and integrates five functional modules, namely anesthesia depth monitoring, respiratory gating synchronization, proton treatment positioning, intelligent feedback control and central control. The system collects multi-mode physiological signals such as electroencephalogram and myoelectricity in real time, and anesthesia state indexes and body movement early warning information are generated; a future displacement track is predicted in combination with the breathing phase, and accurate synchronization is achieved; a beam releasing and interrupting mechanism is linked through the feedback control module; the proton positioning module dynamically adjusts an irradiation path; the central module is scheduled in a unified mode, closed-loop control and safety protection are achieved, and the precision and safety of child proton treatment are remarkably improved.

Description

Technical Field

[0001] This specification relates to the field of smart healthcare, and more specifically, this application relates to an integrated device and control method for pediatric anesthesia and proton therapy. Background Technology

[0002] With the widespread application of proton therapy in pediatric cancer treatment, it has demonstrated unique advantages in reducing the radiation dose to normal tissues by utilizing the Bragg peak effect to achieve precise dose delivery to the tumor target area. However, existing proton therapy systems still face several key technical bottlenecks in practical clinical applications, especially for pediatric patients.

[0003] First, respiratory and body motion disturbances significantly affect treatment accuracy. Traditional systems rely on respiratory gating techniques such as 4D-CT to estimate target displacement, but pediatric patients experience increased respiratory synchronization prediction errors due to their high respiratory rate, unstable amplitude, and reduced spontaneous respiratory regulation under anesthesia. Furthermore, current technologies often rely on external markers to infer internal tumor location, lacking dynamic monitoring and control of the coupling relationship between respiratory phase and anesthesia, which can easily lead to deviations in beam release timing.

[0004] Secondly, anesthesia monitoring and proton therapy systems generally employ a separate architecture, lacking a data linkage mechanism. This prevents the system from dynamically adjusting the treatment plan or interrupting the beam based on real-time anesthesia depth (such as the bispectral index BIS). When insufficient anesthesia leads to body movement, or excessive anesthesia causes respiratory depression, the treatment system struggles to respond promptly.

[0005] Existing systems have limited signal fusion capabilities, relying solely on single physiological signals (such as chest and abdominal pressure or surface displacement) as control criteria. They fail to integrate multimodal monitoring information such as EEG, EMG, and heart rate, making it difficult to accurately identify complex changes in body movement or sedation states, thus affecting treatment safety and adaptability.

[0006] Furthermore, current proton therapy and anesthesia equipment is generally designed with adult parameters. Child patients differ in body size, metabolic rate, and physiological signal amplitude, making it difficult to accurately match existing sensors and fixation devices, potentially leading to discomfort, data discrepancies, or even secondary injuries. Traditional systems lack multi-source triggering and intelligent linkage capabilities in their safety control mechanisms, typically relying on a single threshold (such as displacement exceeding limits) to execute emergency closure. This fails to comprehensively consider collaborative risks such as anesthesia abnormalities, equipment delays, and signal loss, posing potential treatment safety hazards.

[0007] In summary, current proton therapy systems have not yet formed an integrated system deeply coupled with anesthesia monitoring. Breakthroughs are needed in dynamic synchronization, multimodal perception, closed-loop control, and physiological adaptation in children. More intelligent and highly integrated solutions are urgently needed. Summary of the Invention

[0008] The summary section introduces a series of simplified concepts, which will be further explained in detail in the detailed description section. This summary section is not intended to limit the key and essential technical features of the claimed technical solution, nor is it intended to determine the scope of protection of the claimed technical solution.

[0009] In a first aspect, the present invention proposes an integrated anesthesia and proton therapy device specifically for children, comprising:

[0010] Anesthesia depth monitoring module, respiratory gating synchronization module, proton therapy positioning module, intelligent feedback control module, and central control module;

[0011] The aforementioned anesthesia depth monitoring module is used to acquire multimodal physiological parameters, generate current anesthesia status indicators and body movement risk warning signals based on the aforementioned multimodal physiological parameters, and send the aforementioned current anesthesia status indicators and the aforementioned body movement risk warning signals to the aforementioned intelligent feedback control module;

[0012] The aforementioned respiratory gating synchronization module is used to predict future displacement trajectory information based on the patient's respiratory phase information, and to transmit the respiratory phase information and the future displacement trajectory information to the aforementioned intelligent feedback control module and the aforementioned proton therapy positioning module.

[0013] The aforementioned intelligent feedback control module is used to receive the aforementioned current anesthesia status indicators, the aforementioned body movement risk warning signals, the aforementioned respiratory phase information and the aforementioned future displacement trajectory information, assess the proton beam triggering conditions, and send the beam control signal for beam release or interruption to the aforementioned proton therapy positioning module.

[0014] The aforementioned proton therapy positioning module is used to receive the aforementioned respiratory phase information, the aforementioned future displacement trajectory information, and the aforementioned beam control signal to perform precise proton beam irradiation;

[0015] The aforementioned central control module is used to integrate the operating status and data signals of the above modules, realize system-level time synchronization and closed-loop control, and issue a safety interrupt command in abnormal conditions.

[0016] In one feasible implementation, the aforementioned anesthesia depth monitoring module includes a bispectral index sensor, a multi-parameter physiological monitoring unit, an electromyography signal monitoring unit, and a state assessment unit.

[0017] The aforementioned bifrequency index sensor is used to acquire brainwave signals in real time.

[0018] The aforementioned multi-parameter physiological monitoring unit integrates a heart rate variability analysis module, a blood oxygen saturation sensor, and an end-tidal carbon dioxide sensor.

[0019] The aforementioned electromyography (EMG) signal monitoring unit is deployed on the surface of the patient's limb muscle groups to monitor changes in muscle electrical activity in real time. When a rapid potential change is detected that exceeds a set threshold, a body movement risk warning signal is output to the status assessment unit.

[0020] The aforementioned status assessment unit is used to integrate the aforementioned EEG signals, heart rate variability, blood oxygen saturation, end-tidal carbon dioxide and electromyography signals, and generate current anesthesia status indicators and body movement risk warning information through a set dynamic weight algorithm. The aforementioned current anesthesia status indicators and the aforementioned body movement risk warning signals and levels are then transmitted to the intelligent feedback control module.

[0021] In one feasible implementation, the aforementioned bifrequency index sensor is positioned on the patient's forehead.

[0022] In one feasible implementation, the aforementioned multi-parameter physiological monitoring unit is embedded in an anesthesia mask as a flexible sensor.

[0023] In one feasible implementation, the above-mentioned respiratory gating synchronization device includes a 4D optical surface tracking unit, an intracavitary pressure sensor, and a dynamic compensation algorithm unit.

[0024] The aforementioned 4D optical surface tracking unit is used to acquire continuous three-dimensional displacement data of the patient's chest and abdominal surface through a TOF camera array, forming a respiratory motion time series;

[0025] The aforementioned intracavitary pressure sensor is used to collect the pressure change curve inside the airway and obtain the aforementioned respiratory phase information;

[0026] The aforementioned dynamic compensation algorithm unit is used to construct a time series prediction model based on the aforementioned respiratory motion time series and the aforementioned respiratory phase information, and output the aforementioned future displacement trajectory information.

[0027] In one feasible implementation, the proton therapy positioning module includes a CBCT image modeling unit, an MRI real-time imaging unit, a six-dimensional bed adjustment unit, and a pencil beam scanning unit.

[0028] The CBCT image modeling unit is used to acquire cone-beam CT images of the patient's target area before the start of treatment and to perform image registration with the target area anatomical model in the treatment planning system to determine the initial target area spatial location.

[0029] The MRI real-time imaging unit is used to dynamically acquire magnetic resonance images of the patient's target soft tissue during treatment, identify the positional shift of the tumor caused by body movement or respiration, and provide data support for updating the irradiation coordinates.

[0030] The six-dimensional bed adjustment unit is used to fine-tune the patient's position by translation and rotation based on the aforementioned future displacement trajectory information and real-time MRI image results, so as to keep the actual irradiation target area and the planned target area spatially consistent.

[0031] The pencil beam scanning unit is used to control the release of the proton beam according to the preset energy layer and scanning trajectory based on the above-mentioned beam control signal, the above-mentioned respiratory phase information and the above-mentioned bed adjustment results, so as to achieve high-precision dose delivery to the tumor target area.

[0032] In one feasible implementation, the aforementioned six-dimensional bed adjustment unit includes a piezoelectric ceramic micro-motion platform.

[0033] Secondly, the present invention provides a control method for the pediatric-specific integrated anesthesia and proton therapy device described in any one of the first aspects, comprising:

[0034] The above-mentioned anesthesia depth monitoring module acquires multimodal physiological parameters, generates current anesthesia status indicators and body movement risk warning signals based on the above-mentioned multimodal physiological parameters, and sends the above-mentioned current anesthesia status indicators and body movement risk warning signals to the above-mentioned intelligent feedback control module.

[0035] The respiratory gating synchronization module predicts future displacement trajectory information based on the patient's respiratory phase information and transmits the respiratory phase information and future displacement trajectory information to the intelligent feedback control module and the proton therapy positioning module.

[0036] The intelligent feedback control module receives the current anesthesia status indicators, the body movement risk warning signal, the respiratory phase information and the future displacement trajectory information, assesses the proton beam triggering conditions, and sends the beam control signal for beam release or interruption to the proton therapy positioning module.

[0037] The proton therapy positioning module receives the respiratory phase information, the future displacement trajectory information, and the beam control signal to perform precise proton beam irradiation.

[0038] The central control module integrates the operating status and data signals of the modules mentioned above to achieve system-level time synchronization and closed-loop control, and issues a safety interrupt command in abnormal conditions.

[0039] In one feasible implementation, the above method further includes:

[0040] The above multimodal physiological parameters are filtered, normalized, and feature extracted to form a sequence of feature vectors within a sliding time window;

[0041] The above feature vectors are input into the first long short-term memory network model, which outputs the body movement risk level within a future predetermined time window.

[0042] If the above-mentioned risk level of physical movement reaches medium risk or above, a pause or intervention command will be sent to the above-mentioned proton therapy positioning module.

[0043] In one feasible implementation, the above method further includes:

[0044] Acquire CBCT images before treatment and register them with images from the treatment planning system to establish initial spatial reference coordinates for the target area;

[0045] During treatment, MRI dynamic image frames are acquired in real time, and image preprocessing and edge enhancement are performed.

[0046] Based on a non-rigid registration method, MRI images are aligned with CBCT images, and the tumor center displacement vector is extracted.

[0047] If the displacement exceeds the preset tolerance, the irradiation path compensation is performed based on the displacement vector control six-dimensional bed adjustment unit or pencil beam scanning unit.

[0048] The registration offset results are synchronously sent to the central control module for closed-loop correction and risk recording. In summary, compared to traditional technologies, this invention utilizes multimodal monitoring signals such as BIS and electromyography, combined with an LSTM body motion prediction model, to achieve pre-movement warning and automatically control beam release or interruption, significantly improving treatment safety. Employing a 4D optical tracking and intracavitary pressure dual-channel respiratory synchronization system, combined with a prediction algorithm, the target area positioning error is reduced from the traditional ±3mm to ±0.7mm, enhancing irradiation accuracy. This invention integrates CBCT modeling and MRI dynamic image registration, combined with six-dimensional bed fine-tuning, to achieve dynamic target area correction, effectively addressing displacement issues caused by breathing or body movement in children. The central control module integrates 12 types of data, performs risk assessment based on Monte Carlo simulation, and possesses a millisecond-level automatic interruption and backup ventilation mechanism, significantly improving fault response speed. The sensors, mask, and support structure all meet the low-pressure, high-sensitivity requirements of children, and combined with immersive sedation technology, improve treatment comfort and compliance, reduce drug usage, and shorten postoperative recovery time.

[0049] Other advantages, objectives and features of this application will be apparent in part from the description which follows, and in part from what those skilled in the art will understand through study and practice of this application. Attached Figure Description

[0050] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit this specification. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0051] Figure 1 A structural schematic diagram of an integrated anesthesia and proton therapy device for children provided in this application embodiment;

[0052] Figure 2 This is a flowchart illustrating a control method for an integrated anesthesia and proton therapy device for children, provided as an embodiment of this application. Detailed Implementation

[0053] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus. The technical solutions of the embodiments of this application will now be clearly and completely described in conjunction with the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them.

[0054] Please refer to the figure. Figure 1 A structural schematic diagram of an integrated anesthesia and proton therapy device 10 for children provided in this application embodiment, specifically including:

[0055] Anesthesia depth monitoring module 101, respiratory gating synchronization module 102, proton therapy positioning module 103, intelligent feedback control module 104, and central control module 105;

[0056] The aforementioned anesthesia depth monitoring module 101 is used to acquire multimodal physiological parameters, generate current anesthesia status indicators and body movement risk warning signals based on the aforementioned multimodal physiological parameters, and send the aforementioned current anesthesia status indicators and the aforementioned body movement risk warning signals to the aforementioned intelligent feedback control module;

[0057] The aforementioned respiratory gating synchronization module 102 is used to predict future displacement trajectory information based on the patient's respiratory phase information, and to transmit the respiratory phase information and the future displacement trajectory information to the aforementioned intelligent feedback control module and the aforementioned proton therapy positioning module.

[0058] The aforementioned intelligent feedback control module 103 is used to receive the aforementioned current anesthesia status indicators, the aforementioned body movement risk warning signals, the aforementioned respiratory phase information and the aforementioned future displacement trajectory information, assess the proton beam triggering conditions, and send the beam control signal for beam release or interruption to the aforementioned proton therapy positioning module.

[0059] The aforementioned proton therapy positioning module 104 is used to receive the aforementioned respiratory phase information, the aforementioned future displacement trajectory information, and the aforementioned beam control signal to perform precise proton beam irradiation;

[0060] The aforementioned central control module 105 is used to integrate the operating status and data signals of the above modules, realize system-level time synchronization and closed-loop control, and issue a safety interrupt command in abnormal conditions.

[0061] For example, this embodiment provides an integrated anesthesia and proton therapy device 10 specifically for children, aiming to solve problems in traditional proton therapy such as physical isolation between the anesthesia and treatment systems, severe respiratory and body movement interference, insufficient utilization of multimodal signals, and poor device adaptability to children. The device consists of five core modules: an anesthesia depth monitoring module 101, a respiratory gating synchronization module 102, a proton therapy positioning module 103, an intelligent feedback control module 104, and a central control module 105. The modules are interconnected through efficient data channels, and a real-time closed-loop control system is constructed under the coordination of the central control module.

[0062] The anesthesia depth monitoring module 101 is used to acquire multimodal physiological parameters of the patient in real time, including bispectral index (BIS), heart rate variability (HRV), blood oxygen saturation (SpO2), end-tidal carbon dioxide (EtCO2), and electromyography signals. This module has a built-in state assessment unit that can fuse and analyze the acquired signals to generate anesthesia state indicators describing the current sedation level, and output a body movement risk warning signal based on an LSTM prediction model. These anesthesia indicators and warning signals are sent to the intelligent feedback control module in real time for decision support during treatment.

[0063] The respiratory gating synchronization module 102, through an integrated 4D optical surface tracking unit and intracavitary pressure sensor, collects the patient's chest and abdominal surface displacement sequence and airway pressure change data, respectively. It then constructs a time-series prediction model between respiratory phase and displacement using a dynamic compensation algorithm, outputting displacement trajectory information for the next 500ms. This information is used to predict the proton beam irradiation window and is also transmitted to the proton therapy positioning module for dynamic adjustment of the treatment coordinates.

[0064] The intelligent feedback control module 103, serving as the decision-making center of this system, receives anesthesia status indicators and body movement warning signals from the anesthesia depth monitoring module in real time, as well as respiratory phase and displacement prediction information from the respiratory gating module. Based on a comprehensive assessment of multiple conditions, it determines whether the triggering conditions for proton beam release are met. When the conditions are met, such as the patient being in the end-expiratory phase, low body movement risk, and displacement prediction fluctuations within a tolerable range, the module will output a beam release control signal. If any risk factor exceeds a set threshold, it will output an interruption or pause signal to ensure treatment safety.

[0065] After receiving the beam control signal, respiratory phase and displacement information, the proton therapy positioning module 104 combines CBCT image modeling and MRI real-time imaging technology to locate and confirm the patient's tumor target area and register the images. At the same time, it calls the six-dimensional bed adjustment platform to fine-tune the patient's position. Finally, it uses the pencil beam scanning unit to accurately release the proton beam according to the preset energy layer and trajectory, ensuring that the spatial accuracy of dose irradiation is controlled at the millimeter level.

[0066] The central control module 105, serving as the system's overall coordination and scheduling center, is responsible for integrating the operational status and data signals of all the aforementioned modules, ensuring timing consistency between modules through a unified timestamp mechanism. Furthermore, the central control module's built-in Monte Carlo simulation engine is used for multi-source signal fusion and risk assessment. When premonitory body movement, communication delays, or irradiation errors are detected, it can issue a safety interruption command within milliseconds, triggering rapid responses from all sub-modules to cut off the beam and activate backup ventilation channels, ensuring the safety and treatment accuracy of pediatric patients throughout the entire treatment cycle.

[0067] In summary, the integrated anesthesia and proton therapy device and its control method for children proposed in this invention address the technical shortcomings of existing proton therapy systems in pediatric tumor treatment by designing and deeply integrating several key modules, resulting in significant technical advantages and practical application effects. Firstly, regarding the linkage between anesthesia and beam control, this solution integrates an anesthesia depth monitoring module and an intelligent feedback control module, overcoming the problems of independent anesthesia monitoring and proton therapy equipment and response delays in traditional systems. This module integrates multimodal physiological parameters such as bispectral index (BIS), heart rate variability (HRV), oxygen saturation (SpO2), end-tidal carbon dioxide (EtCO2), and electromyographic signals. Utilizing a body movement prediction model based on long short-term memory networks (LSTM), it can predict body movement 500ms before it occurs and trigger the suspension of beam flow or adjustment of anesthetic drug dosage, achieving a closed-loop feedback between the anesthesia state and proton beam release conditions, effectively preventing target area mis-irradiation due to unstable sedation. In terms of respiratory synchronization control, this scheme employs a dual-channel structure of a 4D optical surface tracking system and an intracavitary pressure sensor, combined with a dynamic compensation algorithm. It predicts the chest and abdominal movement trajectory within the next 500ms based on the current respiratory phase and transmits this information synchronously to the proton therapy positioning module. By accurately identifying the end-expiratory apnea and triggering beam release, this system reduces the irradiation error of traditional equipment from approximately ±3mm to ±0.7mm, significantly improving dose concentration and irradiation safety in the tumor area. Regarding image guidance and positioning accuracy, this system integrates CBCT image modeling and real-time MRI imaging units. Before treatment, a three-dimensional structural model of the target area is acquired via CBCT and registered with the treatment planning system. During treatment, real-time dynamic MRI images are acquired to capture the tumor's positional shift caused by respiration or body movement. A non-rigid image registration algorithm extracts the displacement vector, driving a six-dimensional bed adjustment unit to correct the patient's position in real time, ensuring the proton beam always accurately hits the planned target area. Furthermore, pencil beam scanning technology (PBS) can flexibly adjust the energy layer and scanning path based on real-time feedback, further controlling the dose delivery error to within 2%, adapting to the treatment needs of complex sites. In terms of system-level control and safety assurance, this solution is equipped with a central control module that integrates and manages 12 types of data streams, including anesthesia, respiration, imaging, and feedback. Based on a Monte Carlo simulation evaluation mechanism, the central module analyzes the system status and potential risks in real time. In emergency situations such as pre-movement symptoms, communication anomalies, or sensor malfunctions, it issues an interrupt command within milliseconds and switches to the backup ventilation system. This mechanism significantly improves the system's responsiveness and fault tolerance. Finally, considering the physiological characteristics of pediatric patients, this solution has been specifically optimized in terms of hardware structure and sensing system. All sensors are deployed using low-pressure, flexible materials, and the anesthesia mask and bed support system incorporate memory silicone and low-pressure shaping structures to ensure high adaptability for children aged 0 to 12 years.Furthermore, it is equipped with an AR immersive visual guidance interface and alpha wave sound field technology to assist sedation, further reducing the amount of sedative drugs used, minimizing intraoperative fear response, and significantly shortening postoperative recovery time. In summary, the integrated system provided by this invention, in the context of proton therapy for pediatric tumors, can achieve dynamic linkage between anesthesia depth and treatment beam, high-precision synchronous control of respiratory phase and image positioning, and closed-loop safety assurance throughout the entire process. It overcomes the limitations of traditional solutions in multimodal signal fusion, real-time feedback control, and individualized adaptation, providing a systematic solution for improving the accuracy, safety, and comfort of pediatric proton radiotherapy.

[0068] In one feasible implementation, the aforementioned anesthesia depth monitoring module includes a bispectral index sensor, a multi-parameter physiological monitoring unit, an electromyography signal monitoring unit, and a state assessment unit.

[0069] The aforementioned bifrequency index sensor is used to acquire brainwave signals in real time.

[0070] The aforementioned multi-parameter physiological monitoring unit integrates a heart rate variability analysis module, a blood oxygen saturation sensor, and an end-tidal carbon dioxide sensor.

[0071] The aforementioned electromyography (EMG) signal monitoring unit is deployed on the surface of the patient's limb muscle groups to monitor changes in muscle electrical activity in real time. When a rapid potential change is detected that exceeds a set threshold, a body movement risk warning signal is output to the status assessment unit.

[0072] The aforementioned status assessment unit is used to integrate the aforementioned EEG signals, heart rate variability, blood oxygen saturation, end-tidal carbon dioxide and electromyography signals, and generate current anesthesia status indicators and body movement risk warning information through a set dynamic weight algorithm. The aforementioned current anesthesia status indicators and the aforementioned body movement risk warning signals and levels are then transmitted to the intelligent feedback control module.

[0073] For example, the anesthesia depth monitoring module is designed to achieve accurate and real-time assessment of the anesthetic status and movement risk of pediatric patients, thereby ensuring the stability and safety of proton therapy. This module consists of four key units: a bifrequency EEG sensor, a multi-parameter physiological monitoring unit, an electromyography (EMG) signal monitoring unit, and a status assessment unit.

[0074] A bispectral index (BIS) sensor is placed on the patient's forehead to collect brainwave signals in real time. By analyzing the frequency distribution and phase characteristics, it outputs a standardized anesthetic depth index (such as the BIS value, typically ranging from 0 to 100), with a target anesthesia range of 40–60, representing a moderate state of sedation. This index is a core basis for assessing the level of sedation and consciousness.

[0075] A multi-parameter physiological monitoring unit is integrated into the anesthesia mask using flexible electronics, enabling continuous acquisition of key vital signs data. These include: heart rate variability (HRV) to reflect fluctuations in autonomic nervous activity; oxygen saturation (SpO2) to reflect oxygenation status; and end-tidal carbon dioxide (EtCO2) to assess ventilation efficiency. These parameters constitute the basic monitoring of respiratory and circulatory status, helping to assess for potential abnormalities during anesthesia.

[0076] The electromyography (EMG) signal monitoring unit is deployed on the surface of the patient's limb muscles and detects muscle discharge activity in real time through a low-impedance electrode array. When the system detects a short-term, high-amplitude potential change (such as a transient fluctuation exceeding a preset threshold), it considers that there may be a precursor to body movement and immediately sends a body movement risk warning signal to the downstream status assessment unit.

[0077] The state assessment unit, as the core module of the fusion processing, is responsible for integrating and dynamically analyzing the collected multimodal physiological parameters. This unit deploys a dynamic weighting algorithm that assigns different weights to different physiological signals based on their current state, enhancing the robustness of the assessment results. For example, if HRV fluctuates drastically and is accompanied by a decrease in EtCO2, the weight for judging respiratory abnormalities is increased; if electromyography (EMG) mutations are significant, the determination of body movement risk level is strengthened. Based on the above fusion assessment, the state assessment unit outputs two main results: first, quantitative indicators reflecting the current anesthesia state; and second, graded body movement risk warning information (e.g., low, medium, high risk). These two types of data are synchronously transmitted to the intelligent feedback control module to drive proton beam triggering and treatment decisions.

[0078] The anesthesia depth monitoring module described in this embodiment enables intelligent monitoring and early warning of the anesthesia status and movement risks of pediatric patients without relying on manual observation, thereby improving the safety and treatment accuracy of the entire proton therapy process.

[0079] In one feasible implementation, the aforementioned bifrequency index sensor is positioned on the patient's forehead.

[0080] In one feasible implementation, the aforementioned multi-parameter physiological monitoring unit is embedded in an anesthesia mask as a flexible sensor.

[0081] For example, in order to achieve high-precision monitoring of the anesthesia status of pediatric patients and ensure their wearing comfort and data acquisition stability, both the bifrequency EEG sensor and the multi-parameter physiological monitoring unit adopt structural optimization and physiological adaptation design.

[0082] The bispectral index (BIS) sensor is positioned on the patient's forehead, specifically the central area of ​​the forehead. This location has minimal hair coverage, stable skin resistance, and is close to the frontal cortex of the brain, making it an ideal area for acquiring brain activity. The sensor adheres to the skin via medical-grade adhesive electrodes, enabling real-time capture of the high-frequency (Beta) and low-frequency (Delta) components of the EEG signal and calculating their ratio to output a BIS value, used to quantify the current depth of anesthesia. This setup not only ensures signal quality but also reduces physical interference and discomfort for children, making it suitable for continuous wear during prolonged treatments.

[0083] To simultaneously acquire critical vital signs such as respiratory and circulatory status, a multi-parameter physiological monitoring unit is integrated inside the anesthesia mask, employing a flexible sensor structure. This flexible sensor material boasts excellent biocompatibility and conformability, allowing it to closely fit the contours of a child's face without obstructing the ventilation path or causing pressure. The unit includes the following modules: a heart rate variability analysis module to analyze changes in electrocardiographic intervals to assess the state of the autonomic nervous system; a blood oxygen saturation sensor to measure microvascular reflection signals in the fingertips or facial area, providing real-time feedback on oxygenation levels; and an end-expiratory carbon dioxide sensor, connected to the anesthesia circuit, to detect CO2 concentration at the end of each expiration, used to assess ventilation patency.

[0084] Overall, this embodiment, by placing an EEG sensor in the forehead and embedding a flexible multi-parameter monitoring unit in the anesthesia mask, not only achieves high-precision, low-interference data acquisition but also fully considers the tolerance and suitability of pediatric patients during treatment. This integrated, non-invasive monitoring scheme significantly improves the real-time performance and safety of anesthesia depth assessment, providing a solid foundation for subsequent body movement risk assessment and proton beam control decisions.

[0085] In one feasible implementation, the above-mentioned respiratory gating synchronization device includes a 4D optical surface tracking unit, an intracavitary pressure sensor, and a dynamic compensation algorithm unit.

[0086] The aforementioned 4D optical surface tracking unit is used to acquire continuous three-dimensional displacement data of the patient's chest and abdominal surface through a TOF camera array, forming a respiratory motion time series;

[0087] The aforementioned intracavitary pressure sensor is used to collect the pressure change curve inside the airway and obtain the aforementioned respiratory phase information;

[0088] The aforementioned dynamic compensation algorithm unit is used to construct a time series prediction model based on the aforementioned respiratory motion time series and the aforementioned respiratory phase information, and output the aforementioned future displacement trajectory information.

[0089] For example, to improve the accuracy of target displacement compensation caused by breathing in pediatric patients during proton therapy, the respiratory gating synchronization device integrates a 4D optical surface tracking unit, an intracavitary pressure sensor, and a dynamic compensation algorithm unit to form a high spatiotemporal resolution respiratory synchronization prediction system, aiming to achieve sub-second proton beam triggering control.

[0090] The 4D optical surface tracking unit consists of a TOF (Time-of-Flight) camera array positioned above and to the sides of the treatment device. It captures the three-dimensional deformation trajectory of the patient's chest and abdominal surfaces during respiration without contact. This unit continuously acquires three-dimensional point cloud data at a frame rate of 30Hz or higher and constructs a complete respiratory motion time series based on spatial reference calibration. The system establishes a local ROI (Region of Interest) tracking template by setting reference regions (such as below the xiphoid process and the center of the abdomen), thereby ensuring that the recorded surface displacement changes are highly correlated with the fluctuations in the internal target area.

[0091] An intracavitary pressure sensor is integrated into the endotracheal tube or ventilation circuit to acquire real-time pressure change curves within the airway. These curves reflect fluctuations in intrapulmonary pressure differentials at different phases of inspiration and expiration. After filtering and phase decoding, precise respiratory phase information (e.g., currently in the ascending phase of inspiration, the descending phase of expiration, or the end-expiratory plateau) can be extracted. This sensor complements optical tracking's inaccessible information about internal pressure states, improving the model's robustness to abnormal breathing patterns (such as hiccups, breath-holding, and non-rhythmic ventilation).

[0092] The dynamic compensation algorithm unit is used to fuse the two types of data mentioned above to construct a respiratory prediction model. This module employs time series modeling methods (such as GRU with attention mechanism or Transformer), using the respiratory motion time series as the main input and the respiratory phase as an auxiliary constraint to establish a predictive model of target displacement response to time. Through training, the model can output predicted displacement trajectories within the next 500 milliseconds, including displacement vectors in three directions and their confidence levels, which guide the proton beam control logic to trigger irradiation only during the resting window at the end of expiration (typically accounting for 10%-20% of a respiratory cycle).

[0093] In summary, this embodiment effectively captures the complex and unstable respiratory movement patterns of pediatric patients through a multimodal synchronous sensing and prediction mechanism, enabling high-precision dynamic control of proton beam irradiation timing. This method significantly reduces the risk of irradiation deviation caused by target movement, and is particularly suitable for brain, abdominal, and thoracic tumor treatment scenarios where extremely high positioning accuracy is required.

[0094] In one feasible implementation, the proton therapy positioning module includes a CBCT image modeling unit, an MRI real-time imaging unit, a six-dimensional bed adjustment unit, and a pencil beam scanning unit.

[0095] The CBCT image modeling unit is used to acquire cone-beam CT images of the patient's target area before the start of treatment and to perform image registration with the target area anatomical model in the treatment planning system to determine the initial target area spatial location.

[0096] The MRI real-time imaging unit is used to dynamically acquire magnetic resonance images of the patient's target soft tissue during treatment, identify the positional shift of the tumor caused by body movement or respiration, and provide data support for updating the irradiation coordinates.

[0097] The six-dimensional bed adjustment unit is used to fine-tune the patient's position by translation and rotation based on the aforementioned future displacement trajectory information and real-time MRI image results, so as to keep the actual irradiation target area and the planned target area spatially consistent.

[0098] The pencil beam scanning unit is used to control the release of the proton beam according to the preset energy layer and scanning trajectory based on the above-mentioned beam control signal, the above-mentioned respiratory phase information and the above-mentioned bed adjustment results, so as to achieve high-precision dose delivery to the tumor target area.

[0099] For example, to achieve high-precision proton beam irradiation of tumor target areas in pediatric patients, the proton therapy positioning module integrates a CBCT image modeling unit, an MRI real-time imaging unit, a six-dimensional bed adjustment unit, and a pencil beam scanning unit, forming an integrated treatment positioning system with image guidance, adaptive adjustment, and dynamic compensation capabilities.

[0100] The CBCT image modeling unit is activated before each treatment begins to acquire cone-beam CT images of the patient in their current position. CBCT can provide high-resolution three-dimensional structural information within seconds. The system then registers this image with a pre-defined target anatomical model in the treatment planning system (TPS), using either rigid registration or a fusion-based feature-point-based non-rigid registration algorithm to determine the spatial offset between the actual and planned target areas. This process generates initial spatial reference coordinates for the target area, which serve as the baseline for subsequent dynamic image tracking and proton beam path control.

[0101] The real-time MRI imaging unit activated during treatment continuously acquires dynamic image frames of the target soft tissue, making it particularly suitable for soft tissue-rich areas such as the brain, abdomen, or pelvis. This unit uses high-speed MRI sequences (such as bSSFP or FLASH) to acquire dynamic slices and performs image preprocessing, noise suppression, and edge enhancement operations in real time. Based on a non-rigid image registration method, the system aligns the current MRI image with the CBCT baseline image, extracting displacement vectors and deformation information of the tumor region. This data is used to determine whether target displacement has occurred and whether dynamic adjustment of the irradiation coordinates is necessary.

[0102] The six-dimensional bed adjustment unit uses the extracted future displacement trajectory information and MRI image registration results as input to drive the bed to perform spatial fine-tuning. This bed employs a high-precision piezoelectric ceramic micro-motion platform, capable of millimeter-level or sub-millimeter-level adjustments in six degrees of freedom (X, Y, Z-axis translation and Pitch, Roll, Yaw rotation). The system analyzes and superimposes the estimated displacement vector with the actual image deviation, dynamically adjusting the bed's attitude to realign the tumor center with the planned irradiation path, ensuring the proton beam hits the expected area.

[0103] Based on the aforementioned beam control signals, respiratory phase information, and bed adjustment results, the pencil beam scanning unit controls the precise release of the proton beam according to a preset energy layer and two-dimensional scanning path. The system employs PBS (Pencil Beam Scanning) technology to control the proton beam energy layer by layer and achieve lateral scanning guided by a high-speed magnetic field. Irradiation occurs only during the end-expiratory rest phase, and combined with a real-time displacement correction strategy, ensures that each proton beam is accurately deposited on the target layer, with dose distribution error controlled within 2%.

[0104] In summary, this embodiment achieves continuous monitoring and real-time correction of the target area during proton therapy through the synergistic mechanism of multimodal image guidance and six-dimensional body positioning adjustment. This significantly reduces positioning deviations caused by variations in children's breathing, body movement, or fluctuations in anesthesia, thereby effectively improving the accuracy and safety of the treatment.

[0105] In one feasible implementation, the aforementioned six-dimensional bed adjustment unit includes a piezoelectric ceramic micro-motion platform.

[0106] For example, to achieve high-precision dynamic adjustment of the patient's position, the six-dimensional bed adjustment unit integrates a piezoelectric ceramic micro-motion platform as the core actuator. This platform has micron-level displacement accuracy and rapid response capability, and can perform real-time compensation based on the dynamic displacement of the target area during proton therapy.

[0107] Specifically, the piezoelectric ceramic micro-motion platform consists of multiple piezoelectric actuation units, each corresponding to a displacement adjustment of six degrees of freedom: translation in the three linear directions (X, Y, and Z) and rotation in the three angular directions (Pitch, Roll, and Yaw). High-precision position sensors (such as capacitive displacement sensors or optical encoders) are embedded within the platform to provide real-time feedback on the platform's attitude status, and a closed-loop control algorithm ensures the stability and accuracy of the adjustment process.

[0108] After receiving the target area offset vector information transmitted by the MRI real-time imaging unit, the central control module calculates the current offset and future displacement trend, and inputs the result to the bed controller. The piezoelectric platform performs fine-tuning actions in each axis according to the instructions to correct positional deviations caused by respiratory phase fluctuations, body movement, or changes in anesthesia status.

[0109] For example, when the tumor moves up and down due to respiratory movements, the platform can quickly adjust the Z-axis position and reserve a motion buffer according to the prediction model; if the patient is detected to be slightly flipped or swayed, the platform will correct it by adjusting the Roll or Yaw angle, so that the tumor is always kept in the center of the preset irradiation path.

[0110] Compared to traditional hydraulic or electric motor driven beds, this platform has advantages such as fast response, no mechanical lag, and high positioning accuracy (up to 1-10μm), making it particularly suitable for rapid response and compensation for small-amplitude positional deviations in pediatric proton therapy.

[0111] Through the aforementioned technical approach, the six-dimensional bed adjustment unit not only ensures the spatial accuracy of proton beam irradiation but also reduces the risk of dose deviation caused by body position drift, providing highly safe and precise human positioning support for proton therapy in children.

[0112] The second aspect, such as Figure 2 As shown, Figure 2 This is a schematic flowchart illustrating a control method for an integrated anesthesia and proton therapy device for children, provided as an embodiment of this application. The present invention proposes a control method for the integrated anesthesia and proton therapy device for children described in any of the first aspects, comprising:

[0113] S110. Obtain multimodal physiological parameters through the above-mentioned anesthesia depth monitoring module, generate current anesthesia status indicators and body movement risk warning signals based on the above-mentioned multimodal physiological parameters, and send the above-mentioned current anesthesia status indicators and body movement risk warning signals to the above-mentioned intelligent feedback control module.

[0114] S120. The respiratory gating synchronization module predicts the future displacement trajectory information based on the patient's respiratory phase information, and transmits the respiratory phase information and the future displacement trajectory information to the intelligent feedback control module and the proton therapy positioning module.

[0115] S130. Receive the current anesthesia status indicators, the body movement risk warning signal, the respiratory phase information and the future displacement trajectory information through the intelligent feedback control module, assess the proton beam triggering conditions, and send the beam control signal for beam release or interruption to the proton therapy positioning module.

[0116] S140. Receive the respiratory phase information, the future displacement trajectory information and the beam control signal through the proton therapy positioning module and perform precise proton beam irradiation.

[0117] S150. The central control module integrates the operating status and data signals of the above modules to achieve system-level time synchronization and closed-loop control, and issues a safety interrupt command in abnormal conditions.

[0118] For example, in step S110, the system collects the patient's multimodal physiological parameters through the anesthesia depth monitoring module. These parameters include, but are not limited to, bispectral index (BIS), heart rate variability (HRV), peripheral capillary oxygen saturation (SpO2), end-tidal carbon dioxide (EtCO2), and electromyographic signals of the limbs. The collected raw data is filtered, normalized, and feature extracted before being input into the state assessment unit. Based on a set dynamic weight fusion model, the assessment unit quantifies the current anesthesia depth, identifies the presence of abrupt changes in electromyographic signals or other premonitory body movements, generates corresponding anesthesia state indicators and body movement risk warning signals, and sends the results to the intelligent feedback control module.

[0119] In step S120, the system processes the patient's respiratory status information through a respiratory gating synchronization module. This module uses a 4D optical surface tracking device to capture continuous three-dimensional displacement data of the chest and abdominal surfaces, and an intracavitary pressure sensor to acquire airway pressure fluctuations in real time to identify respiratory phases (such as inspiration and expiration). By establishing a predictive model based on the respiratory displacement time series and phase information, the dynamic compensation algorithm unit outputs the predicted displacement trajectory for the next 500 milliseconds and synchronously transmits the respiratory phase information and future displacement trajectory information to the intelligent feedback control module and the proton therapy positioning module.

[0120] In step S130, the intelligent feedback control module receives the aforementioned multi-source information input, including anesthesia status indicators, body movement risk signals, respiratory phase, and predicted displacement trajectory. This module performs real-time evaluation based on the set proton beam triggering rules. If the irradiation safety conditions are met (e.g., end-expiratory apnea, low body movement risk level), a beam release signal is issued; otherwise, a beam interruption signal is issued. This beam control signal is then transmitted to the proton therapy positioning module.

[0121] In step S140, the proton therapy positioning module integrates the aforementioned input signals to perform precise irradiation. The CBCT (Cone-Beam Computed Tomography) image modeling unit within the module completes target area location modeling at the start of treatment, while the MRI (Magnetic Resonance Imaging) real-time imaging unit dynamically senses tumor position shifts during irradiation. Combining displacement prediction from the respiratory gating synchronization module and beam control signals from the intelligent feedback control module, the system controls the six-dimensional bed adjustment unit to fine-tune the patient's position, while simultaneously driving the pencil beam scanning unit to achieve dynamic energy layer irradiation of the target area, ensuring dose accuracy is controlled within millimeter-level error range.

[0122] In step S150, the central control module, acting as the core coordination unit, is responsible for integrating the operational status and data signals of the aforementioned modules and performing system-level time synchronization control. For example, it ensures strict alignment between MRI image frames, respiratory trajectory prediction, and beam release, guaranteeing closed-loop safety and high-precision execution throughout the treatment process. Simultaneously, when the system detects an abnormal state (such as a sharp increase in body movement level, communication delay exceeding limits, or image registration failure), the central control module will immediately issue an interrupt command, automatically suspending treatment and entering a safety mode to prevent any uncertainties from causing dosage errors in pediatric patients.

[0123] In summary, this embodiment constructs an integrated proton therapy platform for children with high reliability and precise control capabilities through modular collaborative control processes and information fusion mechanisms.

[0124] In one feasible implementation, the above method further includes:

[0125] The above multimodal physiological parameters are filtered, normalized, and feature extracted to form a sequence of feature vectors within a sliding time window;

[0126] The above feature vectors are input into the first long short-term memory network model, which outputs the body movement risk level within a future predetermined time window.

[0127] If the above-mentioned risk level of physical movement reaches medium risk or above, a pause or intervention command will be sent to the above-mentioned proton therapy positioning module.

[0128] For example, to further enhance the ability to predict body movement risks during anesthesia, the system introduces a deep learning-based body movement prediction mechanism. As part of the pre-procedure safety control for proton therapy, this mechanism relies on dynamic analysis and predictive modeling of multimodal physiological parameters to effectively reduce the risk of irradiation deviation caused by sudden body movement.

[0129] After acquiring multimodal physiological parameters such as electroencephalogram (BIS), electromyography (EMG), heart rate variability (HRV), oxygen saturation (SpO2), and end-tidal carbon dioxide (EtCO2), the system performs preprocessing operations for different signal sources. This process includes bandpass filtering to remove high-frequency interference and low-frequency drift, normalization to eliminate dimensional differences between different sensors, and feature extraction operations, such as EMG mutation slope, HRV frequency domain indices, and BIS fluctuation amplitude. The processing results are uniformly encoded into a feature vector sequence within a sliding time window.

[0130] These feature vectors are fed into a pre-trained First Long Short-Term Memory (LSTM) network model. This model learns from historical feature sequences to capture the evolution of a patient's physiological state over the past few seconds and predicts the risk level of bodily movement within the next 0.5 to 1 second. The output is presented as a tiered label, such as low risk (0), medium risk (1), and high risk (2), along with a confidence score.

[0131] Once the system determines that the current body movement risk level reaches medium risk or above (i.e., a significant possibility of involuntary body movement is predicted in the short term), the intervention process will be triggered immediately. At this time, the central control module will issue an irradiation pause command or adjust the beam release strategy to the proton therapy positioning module based on the output instructions of the LSTM model. If necessary, physical isolation will be achieved through bed adjustment or beam interruption, thereby avoiding dose release when the patient is about to move, ensuring the accuracy and safety of irradiation.

[0132] Overall, this embodiment integrates multimodal physiological data with a deep prediction model to construct an intelligent recognition mechanism for forward-looking safety control, effectively improving the system's sensitivity and response speed to children's physical movement warnings, and reducing the incidence of dose mis-dosing and radiotherapy adverse events.

[0133] In one feasible implementation, the above method further includes:

[0134] Acquire CBCT images before treatment and register them with images from the treatment planning system to establish initial spatial reference coordinates for the target area;

[0135] During treatment, MRI dynamic image frames are acquired in real time, and image preprocessing and edge enhancement are performed.

[0136] Based on a non-rigid registration method, MRI images are aligned with CBCT images, and the tumor center displacement vector is extracted.

[0137] If the displacement exceeds the preset tolerance, the irradiation path compensation is performed based on the displacement vector control six-dimensional bed adjustment unit or pencil beam scanning unit.

[0138] The registration offset results are simultaneously sent to the aforementioned central control module for closed-loop correction and risk recording.

[0139] For example, to ensure that the proton beam irradiation process is always accurately aligned with the tumor target area, the system is designed with a dynamic image registration and displacement compensation mechanism that combines CBCT and MRI images to solve the target area displacement problem caused by breathing and body movement. It is particularly suitable for application scenarios in pediatric tumor treatment where spatial accuracy is extremely important.

[0140] Before treatment begins, the system acquires a three-dimensional cone-beam CT image of the patient's target area using the CBCT image modeling unit, and then registers this image with a pre-defined target area anatomical model in the Treatment Planning System (TPS). This registration process employs rigid registration technology to establish an initial spatial reference coordinate system for subsequent image comparison and displacement correction during treatment. This initial reference coordinate system defines the spatial position of the target area center within the planned irradiation path.

[0141] During the treatment process, the real-time MRI imaging unit dynamically acquires soft tissue image frames of the patient's target area at a high frame rate. Due to MRI's significant advantage in soft tissue resolution, it can accurately present the tumor contour and the dynamic changes of surrounding tissues. After acquiring each new MRI image frame, the system immediately performs image preprocessing operations, including noise filtering, histogram equalization, and edge enhancement processing, to improve registration accuracy.

[0142] The processed MRI images will be non-rigidly registered with the initial CBCT images. Compared to rigid registration, non-rigid registration can capture local deformations and tissue drift caused by respiration or body movement, and extract the three-dimensional displacement vector of the tumor center point relative to the initial reference coordinates. This vector information can reflect the offset between the target area and the pre-set path of the proton beam in real time.

[0143] Once the displacement of the tumor center is detected to exceed a preset tolerance threshold (e.g., ±1.0 mm), the system will control the six-dimensional bed adjustment unit or pencil beam scanning unit in the proton therapy positioning module to perform irradiation path compensation based on the displacement vector. Specifically, the six-dimensional bed adjustment unit can perform fine adjustments in translation and rotation directions to re-align with the target area; while the pencil beam scanning unit can dynamically adjust the energy level and spot arrangement in the scanning path according to the displacement information to ensure accurate dose delivery to the target tissue.

[0144] Finally, the registered displacement vector and compensation record will be synchronously sent to the central control module, which will use them for closed-loop correction and risk tracking. On the one hand, this ensures time consistency and operational coordination among modules; on the other hand, it facilitates subsequent treatment review and attribution analysis of abnormal irradiation behaviors.

[0145] In summary, this embodiment constructs a high-precision, high-response dynamic positioning and control mechanism through a combination strategy of "CBCT-MRI non-rigid registration and real-time compensation control," which significantly improves the irradiation accuracy and safety assurance of proton therapy for pediatric tumors.

[0146] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application 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. Such 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 this application.

Claims

1. A pediatric-specific integrated anesthesia and proton therapy device, characterized in that, include: Anesthesia depth monitoring module, respiratory gating synchronization module, proton therapy positioning module, intelligent feedback control module, and central control module; The anesthesia depth monitoring module is used to acquire multimodal physiological parameters, generate current anesthesia status indicators and body movement risk warning signals based on the multimodal physiological parameters, and send the current anesthesia status indicators and body movement risk warning signals to the intelligent feedback control module. The respiratory gating synchronization module is used to predict future displacement trajectory information based on the patient's respiratory phase information, and transmit the respiratory phase information and the future displacement trajectory information to the intelligent feedback control module and the proton therapy positioning module. The intelligent feedback control module is used to receive the current anesthesia status index, the body movement risk warning signal, the respiratory phase information and the future displacement trajectory information, evaluate the proton beam triggering conditions, and send the beam control signal for beam release or interruption to the proton therapy positioning module. The proton therapy positioning module is used to receive the respiratory phase information, the future displacement trajectory information and the beam control signal to perform precise proton beam irradiation. The central control module is used to integrate the operating status and data signals of the above modules, realize system-level time synchronization and closed-loop control, and issue a safety interrupt command in abnormal conditions.

2. The integrated anesthesia and proton therapy device for children according to claim 1, characterized in that, The anesthesia depth monitoring module includes a bispectral index sensor for electroencephalography (EEG), a multi-parameter physiological monitoring unit, an electromyography (EMG) signal monitoring unit, and a state assessment unit. The bi-frequency index sensor is used to collect brainwave signals in real time. The multi-parameter physiological monitoring unit integrates a heart rate variability analysis module, a blood oxygen saturation sensor, and an end-tidal carbon dioxide sensor. The electromyography (EMG) signal monitoring unit is deployed on the surface of the patient's limb muscle groups to monitor changes in muscle electrical activity in real time. When a rapid potential change is detected that exceeds a set threshold, a body movement risk warning signal is output to the status assessment unit. The state assessment unit is used to integrate the multimodal physiological parameters of the electroencephalogram (EEG) signal, heart rate variability, blood oxygen saturation, end-tidal carbon dioxide, and electromyography (EMG) signal, and generate the current anesthesia state index and body movement risk warning information through a set dynamic weighting algorithm. The current anesthesia state index and the body movement risk warning signal and level are then transmitted to the intelligent feedback control module.

3. The pediatric-specific integrated anesthesia and proton therapy device according to claim 2, characterized in that, The bifrequency index sensor is positioned on the patient's forehead.

4. The integrated anesthesia and proton therapy device for children according to claim 2, characterized in that, The multi-parameter physiological monitoring unit is embedded in the anesthesia mask as a flexible sensor.

5. The integrated anesthesia and proton therapy device for children according to claim 1, characterized in that, The respiratory gating synchronization device includes a 4D optical surface tracking unit, an intracavitary pressure sensor, and a dynamic compensation algorithm unit. The 4D optical surface tracking unit is used to acquire continuous three-dimensional displacement data of the patient's chest and abdominal surface through a TOF camera array to form a respiratory motion time series. The intracavitary pressure sensor is used to collect the pressure change curve inside the airway and obtain the respiratory phase information; The dynamic compensation algorithm unit is used to construct a time series prediction model based on the respiratory motion time series and the respiratory phase information, and output the future displacement trajectory information.

6. The integrated anesthesia and proton therapy device for children according to claim 1, characterized in that, The proton therapy positioning module includes a CBCT image modeling unit, an MRI real-time imaging unit, a six-dimensional bed adjustment unit, and a pencil beam scanning unit. The CBCT image modeling unit is used to acquire cone-beam CT images of the patient's target area before the start of treatment and to perform image registration with the target area anatomical model in the treatment planning system to determine the initial target area spatial location. The MRI real-time imaging unit is used to dynamically acquire magnetic resonance images of the patient's target soft tissue during treatment, identify the positional shift of the tumor caused by body movement or respiration, and provide data support for updating the irradiation coordinates. The six-dimensional bed adjustment unit is used to fine-tune the patient's position by translation and rotation based on the future displacement trajectory information and real-time MRI image results, so as to keep the actual irradiation target area and the planned target area spatially consistent. The pencil beam scanning unit is used to control the release of the proton beam according to the preset energy layer and scanning trajectory based on the beam control signal, the breathing phase information and the bed adjustment result, so as to achieve high-precision dose delivery to the tumor target area.

7. The pediatric-specific integrated anesthesia and proton therapy device according to claim 6, characterized in that, The six-dimensional bed adjustment unit includes a piezoelectric ceramic micro-motion platform.

8. A control method for the integrated anesthesia and proton therapy device for children according to any one of claims 1 to 7, characterized in that, include: The anesthesia depth monitoring module acquires multimodal physiological parameters, generates current anesthesia status indicators and body movement risk warning signals based on the multimodal physiological parameters, and sends the current anesthesia status indicators and body movement risk warning signals to the intelligent feedback control module. The respiratory gating synchronization module predicts future displacement trajectory information based on the patient's respiratory phase information, and transmits the respiratory phase information and the future displacement trajectory information to the intelligent feedback control module and the proton therapy positioning module. The intelligent feedback control module receives the current anesthesia status index, the body movement risk warning signal, the respiratory phase information, and the future displacement trajectory information, evaluates the proton beam triggering conditions, and sends the beam control signal for beam release or interruption to the proton therapy positioning module. The proton therapy positioning module receives the respiratory phase information, the future displacement trajectory information, and the beam control signal to perform precise proton beam irradiation. The central control module integrates the operating status and data signals of the above modules to achieve system-level time synchronization and closed-loop control, and issues a safety interrupt command in abnormal conditions.

9. The control method according to claim 8, characterized in that, The method further includes: The multimodal physiological parameters are filtered, normalized, and feature extracted to form a feature vector sequence within a sliding time window; The feature vector is input into the first long short-term memory network model, and the body movement risk level within a future predetermined time window is output. If the risk level of the body movement reaches medium risk or above, a pause or intervention command is sent to the proton therapy positioning module.

10. The control method according to claim 8, characterized in that, The method further includes: Acquire CBCT images before treatment and register them with images from the treatment planning system to establish initial spatial reference coordinates for the target area; During treatment, MRI dynamic image frames are acquired in real time, and image preprocessing and edge enhancement are performed. Based on a non-rigid registration method, MRI images are aligned with CBCT images, and the tumor center displacement vector is extracted. If the displacement exceeds the preset tolerance, the irradiation path compensation is performed based on the displacement vector control six-dimensional bed adjustment unit or pencil beam scanning unit. The registration offset results are synchronously sent to the central control module for closed-loop correction and risk recording.

Images