Medical instrument irradiation sterilization apparatus

Abstract

The present application relates to the technical field of medical instrument irradiation sterilization, and discloses a medical instrument irradiation sterilization device, which comprises an irradiation cavity, a roller conveyor, a carrier body and an electron beam generating device, the carrier is provided with an RFID module, and the cavity is divided into multiple independent control sections.Through scanning and identification before loading, a three-dimensional density material map is constructed, particle transport simulation outputs a predicted dose field, an energy spectrum requirement map and a radiation chemical risk distribution; during irradiation, multiple beam energy and incident angle are jointly modulated in time sequence, a gas phase sensor on-line monitors by-products and self-adaptsly puts in a scavenger, a photoluminescence dosimeter reconstructs a three-dimensional dose field in real time; a controller compares a prediction and a measured deviation, links and adjusts a beam current and a scavenging strategy, and iteratively corrects simulation parameters across batches.

Description

Technical Field

[0001] This invention relates to the field of medical device irradiation sterilization technology, and specifically to a medical device irradiation sterilization device. Background Technology

[0002] Medical devices require end-point sterilization before leaving the factory. Conventional irradiation equipment uses fixed beam parameters to uniformly pass the medical device, loaded into a carrier, through the irradiation chamber and irradiate it at a fixed angle. The only control target is the absorbed radiation dose. Since medical devices are composed of various polymer materials and metal parts, with complex structures and often multiple independent packaging, the fixed energy beam penetration depth cannot eliminate the deep under-dose blind zone inside high-density, large-size devices, nor can it avoid surface over-dose damage to lightweight, thin-walled devices. There is a lack of physical simulation and pre-irradiation of the three-dimensional dose field before irradiation, and parameter scheduling relies on geometric experience. The accuracy of parameters for the first batch of irradiation of new devices is insufficient, and the process verification cycle is lengthy. During irradiation, gaseous radiation chemical byproducts such as aldehydes, hydrogen peroxide, and ozone generated in real time cannot be quantitatively monitored online. The dosage of scavenging agents is only mapped based on experience, and the process control of material chemical damage lacks quantitative basis. The spatial perception of the three-dimensional dose field in the chamber relies on sparse discrete point measurements, which cannot reconstruct the continuous three-dimensional dose distribution in real time, resulting in insufficient perception basis for precise closed-loop control. The above four defects are progressive and mutually restrictive, resulting in the existing device being unable to simultaneously ensure deep sterilization compliance, preservation of material chemical integrity, and adaptability to multiple batches. The core control logic is crude and lacks the ability to address the diversity and structural complexity of medical device materials.

[0003] Existing improvement schemes have made some progress in dose uniformity, material preservation, and multi-variety adaptability through multimodal pre-sensing, beam direction coordinated scheduling, zoned adjustable microenvironment cavities, and digital twin closed-loop control. However, there are still shortcomings such as the lack of three-dimensional dose field physical simulation before irradiation, the reliance on fixed data for parameter scheduling, the inability to quantitatively monitor radiation chemical byproducts online, the lack of real-time quantitative basis for chemical damage control, and the reliance on sparse discrete points for three-dimensional dose field sensing, which cannot reconstruct continuous distribution in real time. The root cause of these defects is that the functional modules are independent of each other and lack full connectivity and coordination. The four dimensions of simulation pre-sensing, energy modulation, chemical control, and dose sensing cannot form an overall linkage. There is an urgent need for a medical device irradiation sterilization device. Summary of the Invention

[0004] This invention provides a medical device irradiation sterilization device to solve the problems of deep underdose blind spots caused by the lack of full connectivity and collaboration, insufficient accuracy of initial irradiation parameters, lack of quantitative basis for chemical damage control, and weak real-time perception of three-dimensional dose field.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0006] In a first aspect, a medical device irradiation sterilization device includes an irradiation chamber, a roller conveyor, a carrier body placed on the roller conveyor, an electron beam generator disposed within the irradiation chamber, and a controller disposed on the outer wall of the irradiation chamber. The carrier body is equipped with an RFID module, and air curtains are symmetrically arranged within the irradiation chamber to divide the irradiation chamber into multiple independent control sections. The device also includes:

[0007] The dose prediction module, installed on the roller conveyor, acquires the three-dimensional density material spectrum of the medical device and performs particle transport simulation processing. The expected dose values ​​at each location obtained from the simulation are compared with the sterilization threshold to construct a unified simulation benchmark dataset containing the predicted three-dimensional dose field, energy spectrum requirement spectrum, and radiation chemical risk distribution map. From the simulation benchmark dataset, the energy level switching and beam inrush angle requirements at each spatial location, as well as the radiation chemical risk level of each independent control section, are extracted.

[0008] A beam modulation mechanism is located at the outlet of the electron beam generator and includes a switching component and an adjustment component. The switching timing of the switching component and the angle adjustment timing of the adjustment component are coordinated and arranged according to the energy level switching and angle requirements, and synchronized with the conveying speed of the roller conveyor to form a combined modulation timing.

[0009] The radiation removal module is located in each independent control section of the irradiation cavity. It includes a gas phase sensor array and a multi-component removal agent injection mechanism. According to the joint modulation timing, at each time point in the irradiation process, it collects multi-channel online detection data of multiple radiochemical byproducts in each independent control section by the gas phase sensor array and real-time dose readings of the optically stimulated luminescence dosimeter sensing element. The real-time dose readings are chained together in chronological order to construct a multi-source response parameter set containing real-time byproduct spectra and real-time dose readings.

[0010] The three-dimensional dose field construction module extracts real-time byproduct spectra from the multi-source response parameter set to adjust the flow rate of each component scavenger, and extracts real-time readings of each dose sensing node and combines them with three-dimensional spatial coordinate information. Using the predicted three-dimensional dose field as a priori constraint, it performs interpolation reconstruction processing to divide the irradiation cavity into multiple spatial monitoring partitions with different dose deviation characteristics, thereby obtaining a real-time three-dimensional dose field.

[0011] The controller performs position-by-position comparison between the real-time measured values ​​of the three-dimensional dose field of each spatial monitoring zone and the predicted three-dimensional dose field in the simulation benchmark dataset, calculates the dose deviation at each position, and obtains the spatial deviation field. The spatial deviation field is fed back to the beam modulation mechanism and the radiation removal module for linkage adjustment. At the same time, the deviation data is fed back to the dose pre-simulation module for simulation parameter correction, and iteratively generates an optimized joint modulation timing sequence. The iteration process continues until the global standard deviation of the spatial deviation field is lower than a preset threshold and the local deviation peak of each zone converges within the tolerance band. Finally, a dynamic modulation command sequence verified by closed-loop is output to drive the synchronous update of beam parameters and removal agent injection strategy.

[0012] Inter-module collaboration relationships:

[0013] The energy level switching and angle compensation requirements at each position in the energy spectrum demand map output by the dose pre-simulation module drive the switching timing of the energy reduction foil wheel and the beam injection angle timing of the beam modulation mechanism, transforming the required depth of penetration at each position from empirical judgment into a precise matching rule driven by the simulation prediction field.

[0014] The risk level of each segment in the radiation chemical risk distribution map output by the dose simulation module drives the pre-loaded formulation and initial flow rate baseline of the scavenger in each segment of the radiation removal module, so as to deploy scavenging capabilities that match the risk before irradiation begins.

[0015] The predicted 3D dose field output by the dose pre-exercise module serves as a priori constraint for the interpolation reconstruction of the 3D dose field construction module, significantly improving the reconstruction accuracy under sparse node reading conditions. The systematic deviation between the reconstructed real-time 3D dose field and the predicted field is fed back to the dose pre-exercise module, driving cross-batch iterative correction of simulation parameters.

[0016] Each time the beam modulation mechanism switches energy levels, it synchronously transmits the switching signal and the corresponding change in the byproduct generation rate to the radiation removal module, thereby achieving synchronous feedforward compensation of the removal agent flow rate and eliminating the response lag of pure feedback control during energy switching.

[0017] The timing sequence of the execution parameters of the beam modulation mechanism is associated with the dose increment of each node in the real-time three-dimensional dose field and stored to form a causal data pair of energy angle execution parameters and local dose field response, which is used for fine correction of the Monte Carlo simulation parameters at different energy levels.

[0018] The real-time byproduct spectra of each section of the radiation removal module are correlated and stored with the cumulative dose readings at the corresponding positions of the optically stimulated luminescence dosimeter sensing elements. A ternary correlation database of the damage risk of material generated by dose deposition radiochemical byproducts for each product is constructed, which provides a basis for predicting the material damage risk of subsequent batches of the same product before irradiation and optimizing the removal agent formulation parameters in advance.

[0019] The above-described solution of the present invention has at least the following beneficial effects:

[0020] By fully connecting and coordinating the dose pre-simulation module, the multi-level beam energy modulation module, the radiation chemical byproduct adaptive removal module, and the wireless three-dimensional dose field real-time reconstruction module, the four modules form a bidirectional driving relationship between each other. The measured feedback of any module can correct the operating parameters of the other modules in real time, thereby constructing an irradiation sterilization operation that is driven by simulation pre-simulation, corrected by actual measurement, decoupled by energy and chemical feedforward, and closed-loop feedback of three-dimensional dose field. Attached Figure Description

[0021] Figure 1 This is a three-dimensional structural diagram of a medical device irradiation sterilization device according to an embodiment of the present invention;

[0022] Figure 2 This is a cross-sectional plan view of a medical device irradiation sterilization device according to an embodiment of the present invention;

[0023] Figure 3 This is a three-dimensional transparent view of the vehicle body according to an embodiment of the present invention;

[0024] Figure 4 This is a perspective view of the combination of the liquid storage tank, metering pump, and atomizing nozzle according to an embodiment of the present invention;

[0025] Figure 5 This is a schematic diagram of the electron beam generator and foil wheel assembly according to an embodiment of the present invention;

[0026] Figure 6 This is a perspective view of the foil wheel according to an embodiment of the present invention;

[0027] Figure 7 This is a flowchart of an embodiment of the present invention;

[0028] Figure 8 This is a block diagram of an embodiment of the present invention.

[0029] Explanation of reference numerals in the attached figures:

[0030] In the diagram: 1. Irradiation chamber; 2. Roller conveyor; 3. Carrier body; 4. RFID module; 5. Controller; 6. Air curtain; 7. Electron beam generator; 8. Optically stimulated luminescence dosimeter sensor element; 9. Ultra-wideband wireless transceiver module; 10. Photoionization detector; 11. Electrochemical dual-channel sensor; 12. Low-dose industrial CT; 13. Support frame; 14. Near-infrared light source; 15. Infrared camera; 16. Rotating rod; 17. T-column; 18. Electric rod for adjusting conveying direction; 19. Rotating rod; 20. Connecting column; 21. Electric rod for adjusting carrier width; 22. Motor; 23. Foil wheel; 24. Energy-reducing foil sector; 25. Blank sector; 26. Position sensor; 27. Storage tank; 28. Metering pump; 29. ​​Piping; 30. Atomizing nozzle; 31. Receiving antenna array; 32. Differential optical absorption spectrometer. Detailed Implementation

[0031] Exemplary embodiments of the invention will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this invention will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.

[0032] like Figures 1 to 8 As shown, an embodiment of the medical device irradiation sterilization equipment of the present invention includes an irradiation chamber 1 and a roller conveyor 2 continuously circulating through the inlet and outlet of the irradiation chamber 1. A carrier body 3 is placed on the roller conveyor 2 for holding medical devices, and an RFID module 4 is fixed on the carrier body 3. A dose pre-playing module is arranged on the outside of the roller conveyor 2. Multiple sets of air curtain machines 6 are arranged vertically and horizontally along the conveying direction inside the irradiation chamber 1. The air curtain formed by the multiple sets of air curtain machines 6 divides the space inside the irradiation chamber into multiple independent control sections, and each section is equipped with a radiation-enhancing device. The irradiation chamber 1 includes a real-time monitoring mechanism for chemical byproducts, a gas phase sensor array, and a multi-component scavenger injection mechanism. Multiple electron beam generators 7 are symmetrically arranged vertically within the irradiation chamber 1, each located within an independent control section. A beam modulation mechanism, including a switching component and an adjustment component, is installed at the outlet of the electron beam generator. Optically stimulated luminescence dosimeter sensing elements 8 are embedded in a three-dimensional array within the carrier body 3. A receiving antenna array 31 is arrayed on the inner wall of the irradiation chamber 1. A controller 5 is located on the outer wall of the irradiation chamber 1 and is electrically connected to each module, performing full-process collaborative control.

[0033] One embodiment of the present invention provides a dose prediction module.

[0034] (1) Input data settings

[0035] A scanning unit is installed at the loading station, located outside the roller conveyor 2. An identification unit is mounted on one side of the scanning unit via a support frame 13. The identification unit includes an infrared camera 15 and a near-infrared light source 14 arranged side-by-side. The scanning unit uses a low-dose rotating cone-beam scanning method to non-destructively reconstruct the three-dimensional geometry, internal cavity, loading density, and packaging layers of the medical device within the carrier body 3. The detection results are processed into a three-dimensional voxel map of geometric density with a spatial resolution of no more than 2 mm. Each voxel stores the equivalent water density value at that location. (Unit: g / cm³). The identification unit detects wavelengths from 800nm ​​to 2500nm, detects the infrared characteristic absorption peaks of the instrument, and identifies material types such as polyethylene, polypropylene, polycarbonate, polyvinyl chloride, polyester, silicone, polylactic acid, and metals according to a preset spectral database. The detection results are processed into a material type distribution map, and the material type identifier m(x,y,z) and corresponding radiation tolerance level at each location are output.

[0036] Controller 5 registers and fuses the geometric density 3D voxel map and the material type distribution map in a unified 3D coordinate system: using the CT scan coordinate system as the main coordinate system, the material type identifier output by the hyperspectral recognition unit is mapped to the same grid position in the CT coordinate system through a spatial transformation matrix. Each voxel ultimately carries... Binary properties are used to form a three-dimensional density material map, which serves as the sole geometric input for particle transport simulation.

[0037] (2) Algorithm structure and simulation logic

[0038] The controller 5 incorporates an electron-photon coupling particle transport simulation module, the core algorithm logic of which is as follows:

[0039] Particle source definition: Based on the actual geometric parameters of the current beam generator, the electron beam is modeled as a quasi-single-energy term with an energy distribution following a Gaussian distribution and a full width at half maximum (FWHM) of no more than 5%. The initial particle coordinates are set at the exit section of the beam generator, the initial momentum direction is the beam axis direction, and the directional dispersion angle is no greater than 1°. The initial particle history number for each simulation is no less than 10. 7 A measure is taken to ensure that the relative statistical uncertainty of the monomer dose estimate is no greater than 2%.

[0040] Material cross-section data: For each material type in the 3D density material map, the corresponding mass collision stopping ability is read from the pre-installed NIST material cross-section database. and mass scattering cross section For regions of abrupt density gradient change (adjacent voxels) If the difference exceeds 0.3 g / cm³, an additional scattering interface layer is inserted at the interface, and the boundary scattering is handled by the weighted average of the material cross sections on both sides of the interface.

[0041] Transport process modeling: The energy deposition of electrons in each voxel is calculated using the Continuous Deceleration Approximation (CSDA), and the deposited energy... ,in Let be the path length of the electron within that voxel; the path length for each voxel. The calculation is obtained by geometric intersection of particle momentum direction and voxel geometry; for secondary electrons with energy below the cutoff energy (50keV), the deposition energy is deposited in situ in the current voxel and no longer tracked, so as to balance simulation accuracy and computational efficiency.

[0042] Secondary particle flux output: The average energy of secondary electrons within each voxel is simultaneously calculated during the transport process. and flux density (Number of particles / cm² / initial particles), used as input for the calculation of the radiochemical risk map.

[0043] Dosage conversion: converting the energy deposition values ​​of each voxel Divide by the mass of the voxel ( ) It is the three-dimensional voxel geometric volume in cm. 3 The predicted absorbed dose of the voxel was obtained. Normalized to the initial beam intensity (charge / cm²), a predicted three-dimensional dose field is formed.

[0044] (3) Logical derivation of the three outputs

[0045] Output 1 (Predicted 3D Dose Field): Absorbed dose of each voxel The expected absorbed dose at each location is obtained by multiplying the actual beam intensity parameter by the unit kGy·cm² / mC directly from the above simulation calculations.

[0046] Output 2 (Energy Spectrum Demand Map): For The fundamental reason for insufficient dose of voxels (sterilization threshold, typically 25 kGy) is the insufficient effective penetration depth of the beam at that location. Based on the empirical relationship between range and energy in an equivalent water medium (range R is approximately proportional to energy E, with the proportionality coefficient calibrated by NIST data), calculations are performed to... achieve Minimum required beam energy ;Will Mapped to the system's preset 3MeV, 5MeV, and 8MeV levels (taking the nearest level upwards), each voxel outputs an energy level with a switching level number (1~3). The voxel markings are summarized to form an energy spectrum requirement map. This map directly gives the minimum energy level required for each spatial location and is a spatial index table for driving the energy reduction foil wheel to switch.

[0047] Output 3 (Radiochemical Risk Distribution Map): For each independent control segment z, the secondary electron flux of all voxels within that segment. The average secondary electron flux of segment z is obtained by weighting the flux by material type. For each material type m, the unit flux aldehyde formation rate of the polymer material under secondary electron irradiation is read from a pre-stored database of radiochemical reaction cross-sections. ,

[0048] Hydrogen peroxide generation rate and ozone formation rate ;by Calculate the estimated cumulative amount of byproducts generated in each section, and classify the risk level into three categories: high (greater than 150% of the baseline value), medium (50% to 150%), and low (<50%), to form a radiation chemical risk distribution map. The physical meaning of this map is: high-risk sections have materials with higher secondary electron flux or higher radiation chemical reactivity, which generate more harmful byproducts during the process of reaching the sterilization dose, and require stronger pre-deployment of removal capabilities.

[0049] The above three outputs are generated by controller 5 before loading and serve as direct inputs for subsequent beam joint timing modulation and scavenger preloading.

[0050] One embodiment of the present invention provides a beam modulation mechanism.

[0051] The switching components include:

[0052] A motor 22 is fixed to one side of the electron beam generator 7. The drive end of the motor 22 is connected to a foil wheel 23 to drive the foil wheel 23 to rotate. Three energy-reducing foil sectors 24 and one blank sector 25 are evenly distributed and penetrated on the foil wheel 23. A positioning sensor 26 is embedded and fixed on the foil wheel 23, and the positioning sensor 26 corresponds to the position of the three energy-reducing foil sectors 24 and the blank sector 25. The equivalent aluminum thickness of the foil in each of the energy-reducing foil sectors 24 is different, and each equivalent aluminum thickness corresponds to a preset beam energy level. The blank sector 25 corresponds to the full-energy beam pass-through. The foil wheel 23 rotates around its central axis under the drive of the motor 22. Three energy-reducing foil sectors 24 and one blank sector 25 are evenly distributed around the circumference of the foil wheel 23. The three energy-reducing foil sectors 24 are embedded with foils of different aluminum equivalent thicknesses, which correspond to attenuating the full-energy electron beam to three preset energy levels. The blank sector 25 corresponds to full-energy direct flow. Reflective marks are set on the edges of each sector. The position sensor 26 feeds back the position signal when each sector enters the beam axis to the controller 5 to complete the level confirmation. The level switching response time is no more than 200ms.

[0053] The adjustment assembly includes: a connecting column 20 fixedly connected to the inner wall of the irradiation chamber 1; a rotating rod 16 passing through the connecting column 20; a T-column 17 rotatably connected to the outer side of the rotating rod 16; and a rotating rod 19 passing through the inner wall of the T-column 17. A carrier width direction adjustment electric rod 21 is rotatably connected to the side of the T-column 1 closest to the inner wall of the irradiation chamber 1, and one end of the carrier width direction adjustment electric rod 21 is rotatably connected to the irradiation chamber 1. One end of the electron beam generator's roller conveyor 2 is connected to a conveying direction adjustment electric rod 18, and the fixed end of the conveying direction adjustment electric rod 18 is rotatably connected to the T-column 17. When the carrier width direction adjustment electric rod 21 extends or retracts, it drives the T-column 17 to rotate around the rotating rod 16, adjusting the incident angle of the beam in the carrier width direction (adjustment range ±30°). When the conveying direction adjustment electric rod 18 extends or retracts, it drives the electron beam generator to rotate around the rotating rod 19, adjusting the incident angle of the beam in the conveying direction (adjustment range ±30°).

[0054] According to the energy level switching and angle requirements of each position in the energy spectrum demand diagram, the controller 5 arranges the rotation command of the foil wheel 23, the extension amount of the electric rod 18 for conveying direction adjustment, and the extension amount of the electric rod 21 for carrier width adjustment into a unified timing command package, and synchronizes it with the conveying speed of the roller conveyor 2, so that each position of the carrier body 3 is irradiated with the corresponding energy level and incident angle combination in sequence.

[0055] One embodiment of the present invention provides a radiation removal module.

[0056] Each independent control section is equipped with a gas phase sensor array, which includes: a photoionization detector 10 installed on the inner wall of the irradiation chamber 1, which continuously detects the total amount of volatile organic compounds in the section; an electrochemical dual-channel sensor 11 installed on one side of the photoionization detector 10 and fixedly connected to the inner wall of the irradiation chamber 1, which independently detects the concentrations of hydrogen peroxide and ozone; and a differential optical absorption spectrometer 32 installed on one side of the electrochemical dual-channel sensor 1 and connected to the inner wall of the irradiation chamber 1, which performs online detection of the concentrations of formaldehyde and acetaldehyde. The data from each sensor are combined into a real-time byproduct spectrum with a period of no more than 5 seconds and uploaded to the controller 5.

[0057] Each independent control section is equipped with a multi-component scavenger injection mechanism, which includes four storage tanks 27 fixed above the irradiation chamber 1 of each independent control section. The four storage tanks 27 respectively store aldehyde scavenger (storing sodium bisulfite and glycine compound formulation), hydrogen peroxide scavenger, ozone reducing agent (storing sodium ascorbate solution), and general antioxidant free radical scavenger. The outlet of each storage tank 27 is connected to a metering pump 28, and the outlet of the metering pump 28 is connected to a manifold 29. The manifold 29 is embedded inside the irradiation chamber 1, and the lower end of the manifold 29 is connected to an atomizing nozzle 30. The atomizing nozzle 30 penetrates the inner wall of the irradiation chamber 1 and extends into the corresponding independent control section. The atomization particle size of the atomizing nozzle 30 is not greater than 10 μm. The controller 5 independently adjusts the flow rate of each metering pump 28 according to the concentration of each component in the real-time by-product spectrum.

[0058] Before the carrier body 3 enters the irradiation chamber 1, the controller 5 presets the initial flow rate benchmark for each section according to the radiation chemical risk distribution map: the section with high risk level is preset with a larger initial benchmark flow rate, the section with medium risk level is preset with a medium benchmark flow rate, and the initial flow rate of the section with low risk level is set to zero; during the irradiation process, the controller adjusts the flow rate based on the preset benchmark value and the real-time byproduct spectrum superimposed in a closed loop to achieve a combination of preventive preloading and real-time precise compensation.

[0059] Each time the energy level is switched, the controller 5 calculates the feedforward compensation amount of the cleaning agent flow rate according to the correspondence between the current level and the reference generation rate of by-products in each material section. This compensation amount is then superimposed with the real-time by-product spectrum feedback amount to obtain the comprehensive target flow rate of each metering pump 28, thus eliminating the response lag of pure feedback control during energy switching.

[0060] One embodiment of the present invention provides a three-dimensional dose field construction module.

[0061] (1) Hardware settings and input data format

[0062] The inner wall of the carrier body 3 is arranged in a three-dimensional array of M×N×P (M columns in the left-right direction, N rows in the front-back direction, and P layers in the height direction) with optically stimulated luminescence dosimeter sensing elements 8 embedded in preset slots. Each node includes an optically stimulated luminescence dosimeter sensing element 8 ( The chip (with a measurement range of 1 mGy to 100 kGy, and an energy response deviation of no more than 3% between irradiation sensitivity and incident electron energy in the range of 3 to 10 MeV) and the ultra-wideband wireless transceiver module 9, with node three-dimensional coordinates It is calibrated and stored in controller 5 at the factory.

[0063] The receiving antenna array 31 receives data from each node sequentially, completing a round of dose reading acquisition for all nodes every 500ms. The input data format for each round is: a set of coordinate-dose binary pairs for N nodes. ,in This represents the cumulative dose reading increment at this node during this round of data acquisition (current round reading minus previous round reading, in kGy).

[0064] (2) Real-time reconstruction algorithm for three-dimensional dose field, including input, processing logic, output, and causal relationship.

[0065] Step 1, Residual Field Calculation (Basic Transformation to Eliminate MC System Errors):

[0066] For each node i, read the node coordinates from the predicted 3D dose field. Corresponding simulation prediction value Calculate the difference between the measured value and the predicted value of the node. The residual set { is obtained. }, Physical implications: What is captured is the systematic bias in MC simulation caused by material density uncertainty, cross-sectional parameter errors, and geometric approximations, rather than spatial noise; therefore, for Field interpolation comparison Direct field interpolation has smoother and more physically reasonable spatial variation characteristics.

[0067] Step 2, Residual field spatial interpolation (algorithm adaptation based on the physical characteristics of the irradiation dose field):

[0068] For any position in the 3D mesh Its residual interpolation The following domain-adapted weighted logic calculations are applied:

[0069] The three-dimensional Euclidean distance between each node i and the target point p As the basis for weighting, weight The feature association length L is taken as half the distance between adjacent nodes (typical value 25mm); the value of the feature association length L is based on: , MeV is the transverse scattering radius of the electron beam dose field in a homogeneous medium, which is approximately equal to the beam energy and much smaller than the node spacing of 50 mm. Therefore, the spatial variation of the dose field within the node spacing scale is mainly determined by the macroscopic structure of the MC prediction field. The residual δ reflects the slowly varying systematic error, and its spatial correlation length should be on the order of the node spacing.

[0070] For cases where the material type of the node is different from that of the target point, the corresponding weights will be adjusted. Additional multiplication by material compatibility degradation factor (Use 1.0 for the same material type and 0.3 for different material types) to avoid incorrect extrapolation of residuals between different materials;

[0071] After normalization, we get .

[0072] Step 3, Reconstructing Dose Field Synthesis: ;

[0073] Output meaning: Using the physically correct spatial gradient structure provided by MC simulation as the framework, and interpolating the measured residuals at the nodes to correct system deviations, the output continuous real-time three-dimensional dose field has both a physically reasonable spatial distribution and point-to-point accuracy assurance of measured data.

[0074] Step 4, Spatial Deviation Field and Compensation Decision:

[0075] Will and Subtracting at each grid location yields the spatial bias field. ;right At locations where the negative deviation exceeds the threshold by 10%, a compensation requirement marker is added. The coordinates of this location and the current dose deficit are output to the beam modulation mechanism, driving the switch to a higher energy level or adjustment of the beam angle for compensation irradiation at that location in the next transmission cycle. For locations with a positive deviation exceeding 10%, notify the corresponding independent control section to increase the dosage of free radical scavenger to address the additional free radical generation caused by excessive irradiation.

[0076] The intrinsic mechanism of causal relationship between input and output: The MC prediction field provides a physically correct spatial gradient structure, ensuring that the extrapolation of the reconstructed field between nodes follows the physical laws of particle transport (rather than the inverse square law of geometric distance); the measured values ​​of sparse nodes provide anchorage of absolute values, correcting the systematic overestimation or underestimation of MC caused by the uncertainty of material parameters; the superposition of the two makes the accuracy of the reconstructed field better than the scheme of simply using MC prediction (without measured correction) or simply using geometric interpolation (without physical constraints). The output real-time three-dimensional dose field directly reflects the actual absorbed dose state of each position of the instrument during irradiation, and is the physical basis for driving the compensation irradiation command.

[0077] (3) Cross-batch MC model parameter iterative correction algorithm, model construction end link

[0078] Input data: After each batch of irradiation is completed, extract the total cumulative dose of this batch at each node location from the node readings. Read the predicted values ​​of the corresponding coordinates from the MC prediction field. Read the actual energy level timing and incident angle timing of this batch from the associated storage of controller 5.

[0079] Grouping by material type: Group the N nodes according to the material type m of their coordinates, and each group yields several ( , Data pairs.

[0080] The logic for separating and extracting the three types of correction coefficients:

[0081] Density correction factor In homogeneous material regions (not near material interfaces, interface distance > 5 mm), the main cause of dose deviation is inaccurate material density input in the MC; data for this material type fits a linear relationship. , where the slope This is the density correction factor (physical meaning: This indicates that the density used in MC was lower than the actual density, leading to an overestimation of the simulated dose. Multiply by the input density of this material type in the next batch of simulations. Correct the material density parameters.

[0082] Beam energy correction factor Calculate the dose deviation for different energy levels E, i.e., decompose the node readings to the contribution of each energy level according to the execution parameter sequence of this batch; for each energy level E, the ratio of the expected value of the node reading to the predicted MC value within the corresponding time window is... ; In the next batch of simulations, the energy deposition response coefficient of the corresponding material at this energy level is multiplied to correct the energy dependence error of the cross-sectional data.

[0083] Geometric scattering correction factor The dose bias at nodes near the material interface (interface distance ≤ 5 mm) is mainly caused by interface scattering model errors; for these nodes... For data pairs, calculate the interfacial distance dependence of the bias and fit the interfacial scattering enhancement / attenuation factor. The next batch of simulations will multiply the energy deposition at the interface voxel by [percentage missing]. Make corrections.

[0084] Iterative update mechanism: After the three types of correction coefficients are extracted, they are fed back to the simulation parameter tables (density table, cross-sectional response table, interface correction table) of the corresponding material type in the dose pre-simulation module; new ones are added in each batch. The data pairs allow the sample size for fitting various correction coefficients to increase with each batch, and the fitting accuracy to converge with each batch. For the same SKU, after 5 batches of iterations, the systematic deviation between the predicted field and the measured field converges to within ±3%. The first batch of new SKUs uses the initial default parameters, and its correction coefficients are updated independently from the second batch onwards.

[0085] Technical effects and principles: The separation and extraction of three types of correction coefficients attribute and correct prediction errors caused by different physical causes separately, avoiding the overfitting problem caused by mixing all deviations into a single correction coefficient; the fine correction grouped by material type and energy level ensures that the correction coefficients are accurately matched with their corresponding physical mechanisms; the accumulation of data across batches increases the statistical confidence of the correction coefficients with each batch, essentially realizing online adaptive calibration of the MC simulation model.

[0086] In this invention, the optically stimulated luminescence dosimeter sensing element 8 achieves real-time online dose readout, employing one of the following two methods:

[0087] Option 1 (Radioluminescence Online Readout): Using The chip synchronously generates radioluminescence (RL) signals during high-energy electron beam irradiation. The RL signal is continuously acquired via an onboard photodetector guided through an optical fiber, directly obtaining the light intensity signal proportional to the instantaneous dose rate. After integration, the cumulative dose reading is output. This method eliminates the need for excitation light, does not consume the thermoluminescent information stored in the chip, and does not affect subsequent offline readout verification. In the high-energy electron beam irradiation environment, the readout laser path of the RL signal is spatially perpendicular to the irradiation beam axis. The photodetector is placed behind a lead shielding layer, the shielding thickness of which is determined based on the current energy level, ensuring that the interference of scattered electrons on the photodetector does not exceed 1% of the RL signal intensity.

[0088] Option 2 (Reusable Readout Optically Stimulated Luminescence Material System): Using BeO or Optically stimulated light materials with repeatable optical readout characteristics are used. Wavelength-selective excitation ensures that each readout consumes less than 1% of the energy stored in the wafer. During irradiation, short pulse excitation light (pulse width ≤ 10 ms, duty cycle ≤ 2%) is applied with a period of no more than 500 ms. During the application of the excitation light, the irradiation beam is paused through the shutter for a pause duration of no more than 15 ms, and the impact on the total dose is no more than 0.1%. The emitted fluorescence signal is collected and integrated successively to obtain the real-time cumulative dose reading.

[0089] The dose readings generated by both schemes are packaged into data frames containing node numbers via an ultra-wideband (UWB) wireless transceiver module and transmitted to the controller 5 via a receiving antenna array 31 symmetrically arranged on both sides of the inner wall of the irradiation cavity 1. The center frequency of the UWB wireless communication is 6.5 GHz, the pulse duration is 2 ns, the antenna spacing is 50 mm, and the communication packet error rate does not exceed 0.1% in the metal shielding environment of the high-energy electron beam irradiation cavity 1.

[0090] Step 1, Position-Time Mapping: Using the real-time position encoder output of the roller conveyor 2 as a reference, establish the position coordinates of the carrier body 3 within the irradiation chamber 1. The transport direction coordinates of each voxel in the energy spectrum demand map. Mapped to the moment when the corresponding voxel enters the beam irradiation region. ,in The current transmission speed, The coordinates are the entrance coordinates of the beam irradiation area.

[0091] Step 2, Switching Lead Calculation: Switching Response Time of Switching Components Effective beam width in the transmission direction not exceeding 200ms Approximately 20mm; when At a speed of 0.5 m / min, the time window for irradiating the same column of voxels is... far greater than The position deviation caused by the switching response time does not exceed 1.7mm, which is less than 2mm of the voxel resolution, meeting the accuracy requirements. Controller 5 anticipates each gear shift command in advance. The timing output ensures that the foil wheel switches into position before the target voxel enters the irradiation zone.

[0092] Step 3, Unified Timing Instruction Packet Generation: Controller 5 presses... The timing sequence is arranged for the energy level switching of each voxel (target position of foil wheel rotation), the incident angle of the conveying direction (extension and retraction of electric rod 18 in conveying direction adjustment) and the incident angle of the width direction (extension and retraction of electric rod 21 in width direction adjustment of the carrier), generating a unified timing instruction package, which is output to each actuator with a minimum scheduling granularity of 100ms. It is synchronized with the position encoder signal of roller conveyor 2 by hardware interruption to ensure that the timing deviation does not exceed 50ms.

[0093] The method for determining the feature association length L is as follows: The initial value of L is taken as 1 / 2 of the distance between adjacent nodes (typical value 25mm); if there is a region with uneven distance between adjacent nodes in the carrier body 3, the L in this region is taken as 1 / 2 of the actual nearest neighbor node distance; if the node distance in a certain direction exceeds 100mm (for example, due to the limitation of the instrument shape, the number of nodes is reduced), then L in that direction is changed to 100mm, and the gradient of the predicted three-dimensional dose field is used to replace the residual interpolation in the corresponding direction until the remaining node readings in that direction converge the gradient correction amount to within ±5% of the predicted value.

[0094] Material compatibility attenuation factor The physical basis is that at the interface of different materials, the energy deposition of the electron beam is affected by the abrupt change in the scattering cross section. The dose gradient on both sides of the interface is determined by the ratio of material density and the ratio of scattering cross section. The dose correlation between two materials is lower than that within the same material. The value corresponds to a typical polymer-metal interface pair with a density ratio between 0.5 and 3.0 (Monte Carlo simulations confirm that the effect of interface scattering on the dose is reduced by 65% ​​to 80% compared to the same material within this range, with the median value corresponding to 0.3); for extreme interfaces with a density ratio exceeding the above range (such as polyethylene-lead). It is further reduced to 0.1, which is automatically identified and assigned by controller 5 during registration and fusion processing.

[0095] Handling extreme cases with sparse nodes: If the number of effective nodes in a certain spatial direction decreases to 1 (i.e., interpolation cannot be formed in that direction), then the residual in that direction... Take the unique node residual value No weighted interpolation is performed; only... The predicted field is uniformly biased and corrected; at the same time, the controller 5 outputs a sparse node warning to the operator, indicating that the reconstruction accuracy in this direction only reflects the single-point correction rather than the spatial interpolation result.

[0096] All components of the scavenger used in this invention have been verified for radiation stability and meet the following conditions under high-energy electron beams within the working energy range (3MeV to 8MeV):

[0097] (1) Aldehyde scavenger (sodium bisulfite-glycine composite aqueous solution, mass concentration 0.5% to 2%): the degradation rate does not exceed 15% at a cumulative dose of 10 kGy. The degradation products are sodium sulfate and glycine oxidation products, both of which are non-toxic and low-volatile compounds and do not produce new volatile organic compounds.

[0098] (2) Hydrogen peroxide scavenger (catalase aqueous solution, activity units 100U / mL to 500U / mL): the enzyme activity retention rate is not less than 60% at a cumulative dose of 5kGy, and the protein degradation products after inactivation are amino acids, which do not affect the safety of medical devices; alternative formulations can use manganese dioxide particles fixed bed, and the chemical stability is not affected by radiation.

[0099] (3) Ozone reducing agent (sodium L-ascorbate aqueous solution, mass concentration 1% to 3%): at a cumulative dose of 10 kGy, it is converted into dehydroascorbic acid, which is then hydrolyzed into 2,3-diketogulonic acid at a chamber temperature (25°C to 40°C). Both are non-volatile and non-toxic compounds.

[0100] (4) General antioxidant free radical scavenger (N-acetylcysteine ​​aqueous solution, mass concentration 0.1% to 0.5%): oxidized to cystine at a cumulative dose of 5 kGy, which is a non-volatile and non-toxic compound.

[0101] The interactions between the four components under coexisting conditions (pH 6 to 8, temperature 20°C to 40°C) have been experimentally verified to be negligible: sodium bisulfite-glycine and sodium ascorbate are stable in coexistence under weakly acidic to neutral conditions; catalase does not exhibit catalytic activity towards N-acetylcysteine; and none of the components undergo detectable side reactions with the hydrolysis products of the ozone reducing agent. The component concentration ranges corresponding to the above verification data cover the full operating range of each metering pump 28 of this invention.

[0102] Convergence Criterion for Correction Coefficients: For a certain type of correction coefficient κ_n (the value of the nth batch) of material type m, the convergence condition is defined as follows: for K consecutive batches (K≥3), the following condition is met. ,in The convergence threshold is set to 3% by default, and can be adjusted by the user within the range of 1% to 10% according to accuracy requirements. Once the convergence condition is met, controller 5 marks the correction coefficient for that material type as converged. Subsequent batches will continue to update the coefficient, but using an exponentially weighted moving average (EWMA). To prevent overfitting.

[0103] New SKU initial processing: When irradiating a new SKU for the first time, if there are converged SKUs in the database with a density difference of no more than 10% and the same main material type, their correction coefficients are used as the initial values; otherwise, the standard default parameters are used, and their correction coefficients are updated independently from the second batch onwards.

[0104] Convergence failure handling: If the correction coefficient for the same material type still fails to meet the convergence condition after 10 iterations (the coefficient continues to change monotonically or the oscillation amplitude exceeds the limit), If the controller 5 determines that convergence has failed, it will trigger the following degraded measures: issue a warning to the operator that the uncertainty of the simulation prediction field of the current batch is high; automatically compress the compensation irradiation tolerance of the region corresponding to the material type by 50% from the default value (i.e., trigger compensation irradiation more aggressively) to make up for the insufficient simulation accuracy through feedback compensation of the measured dose field; and submit the complete measured data of the batch for manual review to analyze the physical reasons for the non-convergence of the correction coefficient.

[0105] Example workflow:

[0106] The staff loads the medical device to be irradiated into the carrier body 3; the roller conveyor 2 transports the carrier body 3 towards the low-dose industrial CT12, so that the carrier body 3 and the medical device inside are scanned by the scanning unit and the identification unit. The controller 5 completes the construction of the three-dimensional density material spectrum, calls the particle transport simulation program to pre-simulate the three-dimensional dose field, and outputs the predicted three-dimensional dose field, energy spectrum requirement spectrum and radiation chemical risk distribution map; the roller conveyor 2 sends the carrier into the irradiation chamber 1, the beam modulation mechanism starts the energy angle joint timing modulation according to the energy spectrum requirement spectrum, the scavenger system of each section preloads the scavenger according to the radiation chemical risk distribution map and starts the real-time by-product spectrum closed-loop control, the optically stimulated luminescence dosimeter sensing element 8 uploads the dose reading in real time, and the controller 5 continuously reconstructs the real-time three-dimensional dose field; the controller 5 issues a compensation irradiation command for the deviation exceeding the standard area, adjusts the scavenger flow rate for the by-product exceeding the standard area, and synchronously triggers feedforward compensation for energy switching events; after the batch is completed, the controller 5 extracts the system deviation correction Monte Carlo model parameters and adds the measured data to the ternary correlation database for subsequent batches of the same product.

[0107] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A medical device irradiation sterilization device, comprising an irradiation chamber, a roller conveyor, a carrier body placed on the roller conveyor, an electron beam generator disposed within the irradiation chamber, and a controller disposed on the outer wall of the irradiation chamber, wherein the carrier body is provided with an RFID module, and air curtains are symmetrically arranged within the irradiation chamber to divide the irradiation chamber into multiple independent control sections, characterized in that, Also includes: The dose prediction module, installed on the roller conveyor, acquires the three-dimensional density material spectrum of the medical device and performs particle transport simulation processing. The expected dose values ​​at each location obtained from the simulation are compared with the sterilization threshold to construct a unified simulation benchmark dataset containing the predicted three-dimensional dose field, energy spectrum requirement spectrum, and radiation chemical risk distribution map. From the simulation benchmark dataset, the energy level switching and beam inrush angle requirements at each spatial location, as well as the radiation chemical risk level of each independent control section, are extracted. A beam modulation mechanism is located at the outlet of the electron beam generator and includes a switching component and an adjustment component. The switching timing of the switching component and the angle adjustment timing of the adjustment component are coordinated and arranged according to the energy level switching and angle requirements, and synchronized with the conveying speed of the roller conveyor to form a combined modulation timing. The radiation removal module is located in each independent control section of the irradiation cavity. It includes a gas phase sensor array and a multi-component removal agent injection mechanism. According to the joint modulation timing, at each time point in the irradiation process, it collects multi-channel online detection data of multiple radiochemical byproducts in each independent control section by the gas phase sensor array and real-time dose readings of the optically stimulated luminescence dosimeter sensing element. The real-time dose readings are chained together in chronological order to construct a multi-source response parameter set containing real-time byproduct spectra and real-time dose readings. The three-dimensional dose field construction module extracts real-time byproduct spectra from the multi-source response parameter set to adjust the flow rate of each component scavenger, and extracts real-time readings of each dose sensing node and combines them with three-dimensional spatial coordinate information. Using the predicted three-dimensional dose field as a priori constraint, it performs interpolation reconstruction processing to divide the irradiation cavity into multiple spatial monitoring partitions with different dose deviation characteristics, thereby obtaining a real-time three-dimensional dose field. The controller performs position-by-position comparison between the real-time measured values ​​of the three-dimensional dose field of each spatial monitoring zone and the predicted three-dimensional dose field in the simulation benchmark dataset, calculates the dose deviation at each position, and obtains the spatial deviation field. The spatial deviation field is fed back to the beam modulation mechanism and the radiation removal module for linkage adjustment. At the same time, the deviation data is fed back to the dose pre-simulation module for simulation parameter correction, and iteratively generates an optimized joint modulation timing sequence. The iteration process continues until the global standard deviation of the spatial deviation field is lower than a preset threshold and the local deviation peak of each zone converges within the tolerance band. Finally, a dynamic modulation command sequence verified by closed-loop is output to drive the synchronous update of beam parameters and removal agent injection strategy.

2. The medical device irradiation sterilization equipment according to claim 1, characterized in that, Also includes: The dose prediction module includes a scanning unit and an identification unit. The scanning unit scans the three-dimensional geometric structure of the medical device inside the carrier body using a low-dose rotating cone-beam scanning method to obtain a three-dimensional voxel map of geometric density. Each voxel stores the equivalent water density value at that location. The identification unit detects the infrared characteristic absorption peaks of the medical device and its packaging material inside the carrier body using a preset detection band, identifies the material type at each location according to a preset spectral database, and obtains a material type distribution map and the radiation tolerance level of each material. The controller registers and fuses the geometric density three-dimensional voxel map and the material type distribution map in a unified three-dimensional coordinate system, so that each voxel carries a binary attribute of equivalent water density value and material type identifier, thereby obtaining the three-dimensional density material map. The controller then writes the corresponding SKU-specific formula information into the vehicle body before loading through the RFID module.

3. The medical device irradiation sterilization equipment according to claim 2, characterized in that, Also includes: The dose prediction module uses the density value and material type at each position in the three-dimensional density material map as boundary conditions to simulate the particle transport process under a given beam energy, obtain the expected absorbed dose value at each grid position, and simultaneously calculate the secondary particle flux density within each voxel, summarizing them to form the secondary particle flux distribution of each segment, which together form the predicted three-dimensional dose field. The positions in the predicted three-dimensional dose field where the expected dose value is lower than the sterilization threshold are matched with the correspondence between the electron beam range and energy. The nearest upward level is selected to obtain the minimum beam energy level required for each position, forming the energy spectrum demand map. The energy spectrum demand map serves as a spatial index for driving the level switching of the switching component. The secondary particle flux distribution of each section is correlated with the radiochemical reaction characteristics of each material read from the pre-stored radiochemical reaction section database to obtain the estimated unit dose generation rate of aldehydes, hydrogen peroxide and ozone in each independent control section. The risk levels are divided into high, medium and low to form the risk distribution map. The risk distribution map serves as the basis for setting the preload flow rate benchmark of the removal agent in each section of the radiation removal module.

4. The medical device irradiation sterilization equipment according to claim 1, characterized in that, The switching component includes: A motor fixed on the outlet side of the electron beam generator drives a foil wheel located at the motor drive end to rotate; The foil wheel is evenly distributed with multiple energy-reducing foil sectors and a single blank sector. Each energy-reducing foil sector is embedded with aluminum equivalent foils of different thicknesses. Each aluminum equivalent foil of different thicknesses corresponds to a different preset beam energy level. The blank sector corresponds to full-energy direct beam flow. The foil wheel is embedded with a positioning sensor, and the positioning sensor corresponds to the position of each sector. The positioning sensor feeds back the positioning signal of each sector when it enters the beam path to the controller, and the controller confirms the current energy level. The controller switches the energy level corresponding to each position in the energy spectrum demand map, arranges the rotation commands of the motor, and synchronizes the timing with the conveying speed of the roller conveyor, so that each position of the carrier body is irradiated by the required energy level beam in sequence.

5. The medical device irradiation sterilization equipment according to claim 1, characterized in that, The adjustment component includes: A connecting column is fixedly connected to the inner wall of the irradiation chamber and symmetrically arranged, and a rotating rod is provided through the inside of the connecting column; a T-column is rotatably connected to the outside of the rotating rod, and a rotating rod is provided through the inside of the T-column; The T-pillar is rotatably connected to a carrier width adjustment electric rod on the side near the inner wall of the irradiation chamber, and the end of the carrier width adjustment electric rod near the inner wall of the irradiation chamber is rotatably connected to the irradiation chamber. The end of the electron beam generator away from the roller conveyor is connected to a conveying direction adjusting electric rod, and the fixed end of the conveying direction adjusting electric rod is rotatably connected to the T-post; when the carrier width direction adjusting electric rod extends or retracts, it drives the T-post to rotate around the rotating rod, and independently adjusts the incident angle of the beam in the carrier width direction according to a preset adjustment range; When the electric rod for adjusting the transmission direction extends or retracts, it drives the electron beam generator to rotate around the rotating rod, and independently adjusts the incident angle of the beam in the transmission direction according to a preset adjustment range; the controller arranges the gear switching command of the switching component and the extension / retraction commands of the electric rod for adjusting the transmission direction and the electric rod for adjusting the width of the vehicle into a unified timing command according to the energy spectrum demand map, and performs joint energy angle matching irradiation on each position of the vehicle body.

6. The medical device irradiation sterilization equipment according to claim 1, characterized in that, Also includes: Radiation chemical byproducts include aldehydes, hydrogen peroxide, ozone, and volatile organic compounds; The gas phase sensor array includes a photoionization detector, an electrochemical dual-channel sensor, and a differential optical absorption spectrometer; The photoionization detector continuously detects the total amount of volatile organic compounds within the section to obtain the concentration value of volatile organic compounds; The electrochemical dual-channel sensor independently detects the hydrogen peroxide concentration and ozone concentration within the section, obtaining hydrogen peroxide concentration values ​​and ozone concentration values ​​respectively; the differential optical absorption spectrometer detects the formaldehyde concentration and acetaldehyde concentration within the section online, obtaining aldehyde concentration values; the controller combines the above concentration values ​​within a preset period to obtain the real-time byproduct spectrum. The multi-component scavenger injection mechanism includes independent storage tanks for storing aldehyde scavenger, hydrogen peroxide scavenger, ozone reducing agent and general antioxidant free radical scavenger, respectively. Each storage tank is connected to a metering pump, and the outlet of each metering pump is connected to an atomizing nozzle in the section, with an atomization particle size of no more than 10 μm. The aldehyde scavenger, hydrogen peroxide scavenger, ozone reducing agent, and general antioxidant free radical scavenger are all irradiation-stable aqueous solutions within the working energy range of the irradiation chamber. The components do not react with each other under coexisting conditions and do not produce new radiochemical byproducts that are harmful to medical devices under irradiation conditions. The controller independently adjusts the flow rate of each metering pump according to the concentration of each component in the real-time byproduct spectrum, and sprays the corresponding component scavenger into the section through the atomizing nozzle.

7. The medical device irradiation sterilization equipment according to claim 6, characterized in that, Also includes: The controller divides the risk level of each independent control segment in the risk distribution map into three levels: high, medium, and low. Before the vehicle body enters the irradiation chamber, the initial flow rate of each metering pump is preset to a reference value for the high-risk segment, the initial flow rate of each metering pump is preset to a medium reference value for the medium-risk segment, and the initial flow rate of each metering pump is preset to zero for the low-risk segment. During the irradiation process, the controller adjusts the flow rate of each metering pump based on a preset reference value and the amount of each component concentration in the real-time byproduct spectrum that exceeds the safety threshold, thereby combining preventive preloading of the scavenger with real-time precise compensation.

8. The medical device irradiation sterilization equipment according to claim 1, characterized in that, The three-dimensional dose field construction module also includes: The device consists of multiple wireless dose sensing nodes embedded in the carrier body in a three-dimensional array and a receiving antenna array disposed on the inner wall of the irradiation cavity. The optically stimulated dosimeter sensing element adopts a radioluminescence online reading method or a re-readable optically stimulated dosimeter material system, and packages and sends the detected dose readings and node numbers to the receiving antenna array. The receiving antenna array is symmetrically arranged on both sides of the transmission direction along the inner wall of the irradiation cavity, and transmits the received node data to the controller; the controller performs interpolation reconstruction processing on the dose readings of each node with the predicted three-dimensional dose field as a priori constraint to obtain the real-time three-dimensional dose field; The controller performs difference processing on the measured values ​​at each position in the real-time three-dimensional dose field and the corresponding values ​​in the predicted three-dimensional dose field to obtain the spatial deviation field. Positions with negative deviations exceeding the preset sterilization threshold tolerance value are marked with compensation requirement marks. The compensation requirement marks, along with the current position coordinates and the current dose gap, are output to the beam modulation mechanism to drive the current position to implement a higher energy level or adjust the incident angle for compensation irradiation in the next transmission cycle. For locations where the positive deviation exceeds the preset tolerance, the dosage of free radical scavenger is increased by an equal amount according to the preset dosage in the corresponding independent control section to deal with the additional free radical generation caused by excessive irradiation.

9. The medical device irradiation sterilization equipment according to claim 8, characterized in that, Also includes: After each batch of irradiation is completed, the controller performs deviation analysis between the real-time three-dimensional dose field and the predicted three-dimensional dose field, and extracts multiple correction coefficients by grouping them according to material type: Specifically, density correction coefficients are fitted to nodes within a uniform material region to correct the deviation between the simulated input density and the actual density; energy deposition response deviations are calculated for different energy levels, and beam energy correction coefficients are extracted to correct the energy dependence error of the cross-sectional data; geometric scattering correction coefficients are extracted for nodes near the material interface to correct the interface scattering model error. The correction coefficients are fed back to the dose pre-simulation module to correct the simulation input parameters for the corresponding material types. The controller associates and stores the execution parameter timing of the beam modulation mechanism in each batch with the dose increment of each node in the real-time three-dimensional dose field to obtain causal data pairs of the dose field response at each position for each energy level and angle combination. The correction coefficients are then refined and updated by energy level grouping. The controller monitors the change of each correction coefficient between consecutive batches. When the change of the correction coefficient for the same material type in adjacent batches does not exceed the preset convergence threshold, it is determined that the correction coefficient of the material type has converged. For material types that have not converged, data is accumulated and iteratively updated to achieve cross-batch iteration of simulation accuracy.

10. The medical device irradiation sterilization equipment according to claim 1, characterized in that, Also includes: When the switching component switches energy levels each time, the controller transmits the level switching signal to the radiation removal module, and performs feedforward adjustment on the flow rate of each component removal agent in each independent control section according to the correspondence between the current energy level and the reference generation rate of by-products in each material section, so as to obtain the feedforward compensation amount of the removal agent flow rate in each section under the current level. The controller superimposes the cleaning agent flow feedforward compensation amount with the real-time by-product profile feedback amount to obtain the comprehensive target flow of each metering pump, and outputs the comprehensive target flow to the corresponding metering pump for execution, thereby realizing synchronous compensation for cleaning agent dosing when the by-product profile changes due to energy level switching, and eliminating the response lag of pure feedback control during energy switching.

Images