Medical instrument auxiliary cleaning and disinfection system and method based on AR

Abstract

The invention discloses an AR-based medical instrument auxiliary cleaning and disinfection system and method, and relates to the technical field of medical disinfection. A cleaning and disinfection report of a medical instrument is generated according to virtualization of AR cleaning equipment on a preliminary cleaning scheme; the final cleaning scheme of the medical apparatus is determined according to the cleaning and disinfection conditions of the plurality of cleaning and disinfection areas of the cleaning and disinfection report, the use schedule of the medical apparatus and the current time, and the accuracy of the final cleaning scheme of the medical apparatus is improved. Therefore, an auxiliary cleaning scheme of the AR cleaning equipment is determined according to the surface cleaning coefficient of each cleaning and disinfection area, the surface cleaning coefficient of the adjacent surface area and the use scene of the medical equipment; according to the multiple attitude parameters and the three-dimensional form of the medical instrument, the reinspection path of the AR cleaning equipment for the medical instrument is determined, the surface cleaning gradient map of the medical instrument is determined based on the reinspection path and the multiple corresponding surface cleaning coefficients, and the accuracy of the surface cleaning gradient map of the medical instrument is improved.

Description

Technical Field

[0001] This invention relates to the field of medical disinfection technology, and in particular to an AR-based medical device-assisted cleaning and disinfection system and method. Background Technology

[0002] With the development of technology, AR cleaning equipment, as an augmented reality medical device cleaning device, has an intelligent cleaning system that integrates AR technology, Internet of Things (IoT), and intelligent cleaning. Medical devices are placed in the cleaning space of the AR cleaning equipment and are dynamically cleaned by the AR cleaning equipment. In the existing technology, multiple images of the medical device are collected, and the corresponding cleaning and disinfection area is determined based on the recognition of multiple images of the medical device. Dynamic cleaning is then performed on the cleaning and disinfection area, which executes a single cleaning plan. It ignores auxiliary cleaning plans and re-inspection paths, which affects the fine cleaning of medical devices and cannot present a surface cleaning gradient map of the medical device. Summary of the Invention

[0003] The purpose of this invention is to overcome the shortcomings of the prior art. This invention provides an AR-based medical device assisted cleaning and disinfection system and method.

[0004] This invention provides an AR-based medical device assisted cleaning and disinfection system, comprising: determining multiple cleaning and disinfection areas based on online detection of the medical device using an AR cleaning device; determining a cleaning and disinfection path based on the multiple cleaning and disinfection areas and the three-dimensional shape of the medical device; predicting a preliminary cleaning plan for the medical device based on the starting position, path shape, and positions of the multiple cleaning and disinfection areas; generating a cleaning and disinfection report for the medical device based on the virtual representation of the preliminary cleaning plan by the AR cleaning device; determining a final cleaning plan for the medical device based on the cleaning and disinfection status of the multiple cleaning and disinfection areas in the cleaning and disinfection report, the usage schedule of the medical device, and the current time; cleaning and disinfecting the medical device along the final cleaning plan using the AR cleaning device; determining an auxiliary cleaning plan for the AR cleaning device based on the surface cleanliness coefficient of each cleaning and disinfection area, the surface cleanliness coefficient of adjacent surface areas, and the usage scenario of the medical device; determining a re-inspection path for the medical device using the AR cleaning device based on multiple posture parameters and the three-dimensional shape of the medical device; and determining a surface cleanliness gradient map of the medical device based on the re-inspection path and the corresponding multiple surface cleanliness coefficients.

[0005] This invention provides an AR-based auxiliary cleaning and disinfection method for medical devices, which uses the aforementioned AR-based auxiliary cleaning and disinfection system to disinfect medical devices.

[0006] Compared with the prior art, the beneficial effects of the present invention are: In this embodiment of the invention, a cleaning and disinfection report for the medical device is generated based on the virtualization of the preliminary cleaning plan by the AR cleaning device. The final cleaning plan for the medical device is determined based on the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report, the usage schedule of the medical device, and the current time. The virtualization of the preliminary cleaning plan is introduced, which takes into account the overall consideration of the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report, the usage schedule of the medical device, and the current time, thereby improving the accuracy of the final cleaning plan for the medical device.

[0007] Therefore, the AR cleaning equipment cleans and disinfects medical devices according to the final cleaning plan. Based on the surface cleanliness coefficients of each cleaning and disinfection area, the surface cleanliness coefficients of adjacent surface areas, and the usage scenario of the medical device, the AR cleaning equipment determines an auxiliary cleaning plan. Based on the multiple posture parameters and three-dimensional shape of the medical device, the AR cleaning equipment determines the re-inspection path of the medical device. Based on the re-inspection path and the corresponding multiple surface cleanliness coefficients, the surface cleanliness gradient map of the medical device is determined. The auxiliary cleaning plan is introduced, and the auxiliary cleaning plan, the final cleaning plan, and the re-inspection path are combined as a whole to achieve refined cleaning of the medical device and improve the accuracy of the surface cleanliness gradient map of the medical device. Attached Figure Description

[0008] Figure 1 This is a schematic diagram of the AR-based medical device assisted cleaning and disinfection system in an embodiment of the present invention; Figure 2 This is a flowchart illustrating step S11 in the AR-based medical device assisted cleaning and disinfection system of this invention. Figure 3 This is a flowchart illustrating step S12 in the AR-based medical device assisted cleaning and disinfection system of this invention. Figure 4 This is a flowchart illustrating step S13 in the AR-based medical device assisted cleaning and disinfection system of this invention. Figure 5 This is a flowchart illustrating step S14 in the AR-based medical device assisted cleaning and disinfection system of this invention. Figure 6 This is a flowchart illustrating step S15 in the AR-based medical device assisted cleaning and disinfection system of this invention. Figure 7 This is a schematic diagram of the structural composition of an AR-based medical device-assisted cleaning and disinfection system in an embodiment of the present invention. Detailed Implementation

[0009] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0010] Please see Figures 1 to 7 An AR-based medical device assisted cleaning and disinfection system is applied to medical device assisted cleaning and disinfection scenarios. The AR-based medical device assisted cleaning and disinfection system includes: Step S11: Based on the online detection of medical devices by AR cleaning equipment, multiple cleaning and disinfection areas are determined, and the cleaning and disinfection path is determined based on the multiple cleaning and disinfection areas and the three-dimensional shape of the medical devices. Step S12: Predict the initial cleaning plan for medical devices based on the starting point, path shape, and location of multiple cleaning and disinfection areas of the cleaning and disinfection path. Step S13: Generate a cleaning and disinfection report for the medical device based on the virtual representation of the preliminary cleaning plan by the AR cleaning equipment. Determine the final cleaning plan for the medical device based on the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report, the usage schedule of the medical device, and the current time. Step S14: The AR cleaning equipment cleans and disinfects the medical device according to the final cleaning plan. The auxiliary cleaning plan of the AR cleaning equipment is determined based on the surface cleanliness coefficient of each cleaning and disinfection area, the surface cleanliness coefficient of adjacent surface areas, and the usage scenario of the medical device. Step S15: Determine the re-inspection path of the medical device by the AR cleaning equipment based on multiple posture parameters and three-dimensional shape of the medical device, and determine the surface cleaning gradient map of the medical device based on the re-inspection path and the corresponding multiple surface cleaning coefficients.

[0011] refer to Figure 2 In step S11, the specific steps are as follows: S111: The medical device is placed in the cleaning space of the AR cleaning equipment. The AR cleaning equipment rotates the medical device and takes pictures of the medical device from different positions based on the built-in camera to collect images of the medical device at different positions. Based on multiple images and the corresponding stereo data of the medical device, a virtual model of the medical device is determined and the stereo shape of the medical device is marked.

[0012] S112: Multiple cleaning and disinfection areas are determined based on online detection of a virtual model of the medical device. These multiple cleaning and disinfection areas are distributed on the outer and inner surfaces of the medical device and are located at different positions on the medical device. The AR cleaning equipment determines the cleaning and disinfection path based on the location of the multiple cleaning and disinfection areas, the three-dimensional shape of the medical device, and usage information.

[0013] In the embodiments of this application, the medical device is placed in the cleaning space of the AR cleaning device. The medical device needs to be properly positioned, which can be done by simple placement, fixing with clamps, or, for small devices, by using the internal structure of the device for limiting. Good positioning can reduce image blurring or model deviation caused by device shaking. The size of the cleaning space must be able to accommodate the target medical device and leave enough space for the internal mechanical structure of the device (such as the rotating shaft and camera bracket) to operate, avoiding collisions.

[0014] The instrument is placed on it and rotated; or it is a multi-degree-of-freedom robotic arm that can more flexibly adjust the posture of the instrument (not just rotation, but also tilting and flipping); the rotation can be continuous or stepwise (for example, rotating a certain angle each time, such as 30 degrees or 45 degrees, and then stopping); stepwise stopping helps the camera to shoot in a stable state, improving image quality; the rotation action needs to be precisely synchronized with the camera's shooting action; for example, the device is designed to trigger the camera to shoot when it rotates to a specific angle or stops.

[0015] AR cleaning equipment is typically equipped with one or more cameras, including: high-resolution visible light cameras: used to capture details, colors, and visible stains on the instrument surface; depth cameras (such as structured light or TOF): used to directly acquire 3D point cloud data of the instrument surface to assist in modeling and detection; and special spectral cameras (optional): such as near-infrared or ultraviolet cameras, used to detect specific types of residues (such as proteins, blood).

[0016] The camera will take pictures from different angles and distances depending on the rotation of the instrument; it is necessary to ensure uniform lighting and avoid shadow interference during shooting; for the internal channels or cavities of the instrument, special endoscopic cameras or the instruments' own channels are required for observation; each shot will generate one or more images (if multiple cameras are used), and these images are transmitted to the device's processing unit in real time or in batches.

[0017] The system incorporates multiple images and corresponding stereo data for medical devices, fusing the acquired multi-view image data with input data or pre-stored stereo data for the medical devices in the database. The stereo data consists of a previously scanned 3D model of the device, providing its ideal geometry. Simultaneously, using computer vision techniques (such as feature point matching and stereo vision), the system extracts 3D information from 2D images to generate a point cloud or mesh model of the device. The reconstructed model from the images is then registered (aligned) with the pre-stored stereo data. The system searches for correspondences between the two, using algorithms such as Iterative Closest Point (ICP) to correct and supplement the ideal model with the real-world data. The resulting virtual model contains both the ideal geometry and surface details of the actual device (such as preliminary information on wear and existing stains). The generated virtual model is then optimized by removing noise, filling gaps, and ensuring the model's integrity and smoothness.

[0018] On the generated virtual model, the system will automatically or semi-automatically identify and mark the key three-dimensional morphological features of the medical device, including geometric features such as joints, hinges, connectors, handles, tips, teeth, screw holes, and internal cavity entrances. These markings are usually visualized in the AR interface, such as on the AR glasses worn by the operator or on the device screen, and overlaid on the virtual model of the device or the actual device for intuitive understanding.

[0019] Furthermore, the system first loads the virtual model of the medical device generated or invoked in S111. This model includes not only geometry but also material information, standard dimensions, etc. The system analyzes the virtual model to identify areas that require special attention for cleaning and disinfection. These algorithms include: Geometric analysis: identifying physical structures in the model such as narrow channels, complex curved surfaces, gaps, and dead corners, which are prone to accumulating dirt and are difficult to clean; for example, identifying holes, threads, hinge joints, etc. in the model; Material analysis: if the model contains material information, the system can identify which areas are made of special materials (such as porous materials or easily corroded materials) and require different cleaning agents or methods; Stain / residue simulation: more advanced systems will combine the device's usage information (such as the type of surgery and contact with bodily fluids) to simulate the distribution of stains or biofilms on the virtual model, thereby more accurately predicting high-risk areas; Threshold judgment: the system determines which geometric features or simulated stains have reached the "threshold" that requires special cleaning based on preset rules or learned patterns.

[0020] Based on the above analysis, the system divides the virtual model into multiple "cleaning and disinfection areas". These areas clearly indicate the scope that needs to be cleaned, covering the outer surface, inner surface (such as the inside of the lumen), specific functional areas or easily contaminated areas of the model. Each area can be assigned an ID or label and its three-dimensional coordinate range is recorded. These areas are clearly distributed on the inner and outer surfaces of the instrument and located at different positions on the instrument to ensure coverage of all potential contamination points.

[0021] The system integrates three key pieces of information: the location of the cleaning and disinfection area: precise 3D coordinate information from the previous step; the 3D morphology of the medical devices: complete geometric information provided by the virtual model, including surface curvature, edges, and obstacles (such as protruding parts of the devices themselves); and usage information: this is crucial contextual information, including: device type (e.g., surgical scissors, suction tubes, endoscopes); the type of surgery most recently performed (e.g., orthopedics, general surgery, neurosurgery); the type of bodily fluids involved (e.g., blood, fat, tissue debris); usage frequency and the time since the last cleaning; and the path planning algorithm used by the system (e.g., A*, RR). The system generates one or more cleaning and disinfection paths using tools such as T (or similar instruments). It ensures the path passes through each marked cleaning and disinfection area, avoids collisions between cleaning tools (such as nozzles and brushes) and non-cleaned areas or the instrument's structure, minimizes the total path length, and reduces cleaning time. The path is adjusted according to surface curvature; for example, spiral or reciprocating motion is needed in curved areas, while straight lines or pushes at specific angles are needed within lumens. Usage information can influence the priority or details of the path; for example, if the instrument has just been used in orthopedic surgery, the system will prioritize cleaning paths for areas contaminated with bone fragments, or select a path mode more suitable for cleaning bone fragments.

[0022] The final output is one or more specific cleaning and disinfection paths; each path consists of a series of waypoints, which define the position, attitude (angle) and actions (such as spraying water or rotating brushes) that the cleaning tool needs to reach. These path data are stored and prepared for execution in step S14.

[0023] Specifically, the system identified and marked the following cleaning and disinfection areas: Area A: Inner surface of the clamp arm (high risk, with grease); Area B: Clamping surface of the clamp arm (high risk, with tissue debris); Area C: Joint axis and surrounding gaps (high risk, with blood); Area D: Tip (medium risk, requires sterility); Area E: Connection between the handle and the connecting shaft (medium risk, with gaps). These areas are highlighted in red on the AR-displayed virtual model.

[0024] The system acquired the precise three-dimensional coordinates of the region AE, as well as the complete 3D model information of the needle holder; at the same time, it took into account the usage information of "intestinal anastomosis surgery" and knew that it was necessary to focus on removing fat and blood.

[0025] The path planning algorithm begins to work: Starting point: The starting point of the path is set near the handle of the needle holder, so that the cleaning tool (assuming it is a robotic arm end effector with a nozzle and a micro brush) can start working; Coverage Area A (Inner Surface of the Clamp Arm): The algorithm plans a path for the actuator to move slowly along the inner surface of the clamp arm, accompanied by slight rotation, to ensure that the nozzle can spray water onto the entire inner surface; since the presence of grease is known, the path will be designed to be denser; Coverage Area B (Matching Surface): The path planning allows the actuator to move to the mating surface to wipe or spray back and forth. Coverage area C (joint axis): The algorithm plans a finer path, allowing the actuator's micro-brush or nozzle to reach deep into the joint axis for cleaning, requiring multiple attempts with angle adjustments; Coverage area D (tip): Path planning ensures that the tip is also touched by the water flow or brush, but with gentler movements; Coverage area E (handle connection): Path planning allows the actuator to carefully clean the gaps at the connection. The system attempts to connect the cleaning actions from area A to area E to form a continuous or minimally mobile path sequence to improve efficiency. Ultimately, the system generates a detailed cleaning and disinfection path, which includes the number of waypoints the actuator needs to move to, and the actions to be performed at each point (such as spraying water, using a specific type of brush, and holding time).

[0026] refer to Figure 3 In step S12, the specific steps are as follows: S121: The starting position of the cleaning and disinfection path is determined based on the detection of the cleaning and disinfection path. The starting position of the cleaning and disinfection path is used as the starting point of the cleaning and disinfection of the medical device. The first cleaning and disinfection space of the medical device is determined according to the starting position of the cleaning and disinfection path and the virtual model of the medical device. S122: Determine the first cleaning plan for the medical device based on the spatial shape of the first cleaning and disinfection space and the path shape of the cleaning and disinfection path; determine the second cleaning plan for the medical device based on the path shape of the cleaning and disinfection path and the location of multiple cleaning and disinfection areas. S123: Based on the first cleaning plan, the second cleaning plan and the model of the medical device, a preliminary cleaning plan is predicted for the medical device, which indicates the cleaning medium, multiple cleaning parameters and the cleaning sequence of various parts in the medical device.

[0027] In the embodiments of this application, the system first needs to obtain the cleaning and disinfection path data generated in step S112. This is usually a sequence containing a series of coordinate points (waypoints) and accompanying action instructions (such as spraying, brushing), and mark the starting position of the cleaning and disinfection path, which serves as the starting point for cleaning and disinfection of the medical device.

[0028] Once the starting point (e.g., WaypointD) is determined, the system marks it as the starting point of the current cleaning and disinfection task. The starting point is usually not just a point; it represents an area or space that needs to be treated first. The system needs to accurately define the size and shape of this initial treatment area based on the starting point coordinates and the virtual model. With the introduction of a virtual model of the medical device, the system determines a three-dimensional spatial range based on the structure of the joint axis in the virtual model, such as "a cylindrical area with a radius of 5mm and a length of 15mm centered on the joint axis centerline." This definition takes into account the geometry of the joint axis gaps and the distribution range of dirt.

[0029] The first cleaning and disinfection space serves as the basis for formulating the initial cleaning strategy. Optionally, based on the virtual model and the coordinates of Waypoint D, the system determines the first cleaning and disinfection space as: "the joint axis gap area, including the inner and outer surfaces, with an axial length of 15mm and a radial range of 5mm." This information is recorded and used in the next step S122 to formulate the initial cleaning strategy for this area. For the laparoscopic needle holder, this "starting line" is at the joint axis gap, and the area to be "run" in the first step is this gap and a small space around it. This ensures that the cleaning process can start in an organized and focused manner.

[0030] Furthermore, a first cleaning and disinfection space for medical devices is introduced, and the corresponding spatial form is marked, presenting the specific geometry of the first cleaning and disinfection space. Based on the above analysis, the system generates a cleaning plan for this specific space and initial path. This plan specifies in detail: Cleaning medium: For crevices, high-pressure pulsed water flow or enzyme cleaner of a specific concentration is required to ensure effective flushing away of hidden dirt; Actuator action: Fine nozzles are required for targeted spraying, or flexible brush heads are required for internal brushing; Actions include specific rotation, oscillation, or pulse frequencies; Parameter settings: Spray pressure, flow rate, duration, brush head speed, etc.; Safety constraints: Ensure that the actions do not damage the device (e.g., avoid applying excessive pressure to vulnerable areas).

[0031] Specifically, for the first cleaning and disinfection space (joint axis gap area) of the laparoscopic needle holder, the system analyzes its spatial shape as a "narrow V-shaped channel"; the path shape requires the actuator to precisely rotate the instrument so that the gap is aligned with the nozzle; based on this, the system generates the first cleaning plan: cleaning medium: high-pressure pulse enzyme cleaner (pressure set at 200kPa, pulse frequency 5Hz); actuator action: using a 1mm diameter slender nozzle, the actuator maintains a stable posture and sprays at the center of the gap; parameter settings: spraying lasts for 15 seconds, during which the actuator swings slightly (<5 degrees) left and right to ensure coverage of both sides of the gap; safety constraints: the nozzle maintains a minimum distance of 2mm from the instrument surface to prevent collision.

[0032] The system is no longer limited to the starting point, but analyzes the macroscopic form of the entire cleaning and disinfection path. This includes the relative positional relationships between areas in the path, the connection methods (straight lines, curves, rotations, etc.), and the overall motion pattern that the actuators need to complete (e.g., whether to process one side first and then flip it over, or to process them alternately). The system combines the path form with the positional information of multiple cleaning and disinfection areas previously determined in S111 / S112. This helps to understand the logical flow and spatial layout of the entire cleaning process to generate a second cleaning plan. This second cleaning plan includes: transition motion planning: how to transition from one area to another... Smooth, efficient, and safe movement to the next area; for example, moving from the joint axis gap to the inner surface of the left clamp arm requires rotating the instrument 90 degrees; Sequence optimization: fine-tune the cleaning sequence based on the path shape and area location to reduce unnecessary movements or flips; for example, if the inner surfaces of the left and right clamp arms are adjacent, they can be processed consecutively; General strategy setting: set a general cleaning strategy framework for multiple areas with similar shapes or locations in the path; for example, all inner surface areas use a strategy of jet water flow + short soaking time; Resource allocation: estimate the time required to complete the entire path, the amount of cleaning agent used, etc.

[0033] Specifically, the system analyzes the entire cleaning and disinfection path of the laparoscopic needle holder: starting from the joint axis gap (D), it proceeds sequentially to the inner surface of the left clamp arm (A), the inner surface of the right clamp arm (B), the occlusal surface (C), the handle connection (F), the tip (E), and finally to the outer surface (G); the regional positions show that A and B are relatively symmetrical, while C, F, E, and G are relatively independent; based on this, the system generates a second cleaning plan.

[0034] Transitional motion planning: D>A: Rotate the instrument approximately 90 degrees so that the inner surface of the left jaw arm faces upward; A>B: Continue rotating approximately 180 degrees so that the inner surface of the right jaw arm faces upward; B>C: Fine-tune the angle so that the biting surface faces the nozzle; C>F: Rotate the instrument so that the handle faces upward; F>E: Rotate the instrument so that the tip faces upward; E>G: Rotate the instrument so that the entire outer surface faces upward; Sequence optimization: Maintain the order D>A>B because they are logically connected; the order of C, F, E, and G is also basically planned according to the path because their positions are scattered.

[0035] General strategy settings: For A and B (inner surfaces), use a similar water jet strategy as the first option, but with slightly lower pressure (150kPa) for 20 seconds; for C (interlocking surface), use water jet for 25 seconds; for F (handle connection), use a brushing strategy (using a small brush head) for 30 seconds; for E (tip), use a fine low-pressure water jet (100kPa) for 15 seconds; for G (outer surface), use a regular water jet for 40 seconds; Resource allocation: Estimated total cleaning time is approximately 3 minutes, and cleaning agent usage is approximately 50ml.

[0036] Therefore, based on the first cleaning plan, the second cleaning plan, and the model of the medical device, a preliminary cleaning plan for the medical device is predicted. This preliminary cleaning plan marks the cleaning medium, multiple cleaning parameters, and the cleaning sequence of various parts in the medical device.

[0037] At this point, the system first integrates the first cleaning plan (for the first area) and the second cleaning plan (for the overall path strategy) generated in S122. This is not just a simple splicing, but rather ensuring that the two are coordinated and consistent. For example, the fine water flow pressure set for the joint axis gap in the first plan cannot conflict with the general high-pressure strategy set for the entire instrument in the second plan, or adjustments need to be made to adapt to the overall strategy. The system retrieves the database information of this specific model of laparoscopic needle holder. This information includes: historical cleaning data: records of past cleaning of this model of instrument, common problem areas, average cleaning time, etc.; manufacturer recommendations: cleaning parameter ranges, prohibited areas or methods recommended by the instrument manual or manufacturer; material information: materials of different components (such as stainless steel, titanium alloy, plastic, etc.), which will affect the pressure, temperature and chemical reagent types that can be withstood; known differences: individual differences or special design features of this model.

[0038] The system combines the integrated scheme and model information to make predictions and decisions, generating a preliminary, relatively complete cleaning scheme. This scheme will: unify parameters: ensure that the parameters used in all areas (such as basic pressure and basic flow rate) are within the range allowed by the model, and make fine adjustments according to the characteristics of each area; optimize the sequence: further optimize the path sequence based on the model information, for example, if the model information shows that a certain area is particularly prone to stains, its cleaning steps will be advanced; supplement details: supplement overlooked details based on the model information, such as additional rinsing steps at specific joints; mark key information: clearly mark the type of cleaning medium to be used, the specific cleaning parameters (pressure, time, action type, etc.) for each area / step, and the final cleaning sequence list; specifically, the system integrates the first scheme for joint gaps and the second scheme covering areas such as the occlusal surface and inner surface, and combines the database information of "laparoscopic needle holder model A-100"; the occlusal surface (area A) of model A-100 is made of a softer material and has a maximum pressure resistance of 200 kPa; the joint gap (area D) is a historically high-incidence area for stains.

[0039] refer to Figure 4 In step S13, the specific steps are as follows: S131: Based on the traversal of the virtual model of the medical device, the various parts of the medical device are determined. At the same time, the AR cleaning device collects the preliminary cleaning plan and constructs the cleaning and disinfection video of each part of the medical device based on the preliminary cleaning plan and the virtual model of the medical device. S132: In the cleaning and disinfection videos of various parts, multiple cleaning and disinfection dynamic images are determined based on the dynamic recognition of the cleaning and disinfection videos. Based on the multiple cleaning and disinfection dynamic images and the corresponding cleaning and disinfection areas, cleaning and disinfection reports for each part are determined. The cleaning and disinfection report of the medical device is generated by synthesizing the cleaning and disinfection reports of each part, and the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report is marked. S133: Collect the usage plan of medical devices, determine the next usage time of medical devices based on the analysis of the usage plan, and determine the final cleaning plan of medical devices based on the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report, the next usage time of medical devices, and the current time. This final cleaning plan serves as a further optimization of the preliminary cleaning plan.

[0040] In the embodiments of this application, the various parts of the medical device are determined by traversing the virtual model of the medical device. At the same time, the AR cleaning device collects the preliminary cleaning plan and constructs cleaning and disinfection videos of various parts of the medical device based on the preliminary cleaning plan and the virtual model of the medical device. This takes into account the overall consideration of traversing the virtual model of the medical device and ensures the accuracy of various parts of the medical device.

[0041] At this point, a virtual model of the medical device is acquired. By traversing the virtual model, the system will identify all the main geometric parts that make up the device, such as: needle holder tip (D area): used to hold the end of the suture; jaws (A area): the part that holds the needle, usually with teeth; joint axis (B area): the movable joint that connects the tip and the handle, with a gap; operating lever (C area): the part that connects the joint axis and the handle; handle (F area): the part that the operator holds; cavity (E area): if the needle holder is designed with a cavity (e.g. for certain special operations), it needs to be identified; connection interface (G area): the part that connects to other devices (such as power cord, light source interface, etc.).

[0042] The system not only identifies these parts, but also associates and marks these geometric parts with specific cleaning and disinfection areas (D, A, B, C, F, E, G) based on the cleaning and disinfection areas (D, A, B, C, F, E, G) determined in S112. For example, the geometric part of the "jaw" is marked as belonging to the "Area A" cleaning and disinfection area. The purpose of this is to ensure that the subsequent simulated cleaning can accurately act on every physical part that needs to be cleaned.

[0043] After completing model traversal and part identification, the AR cleaning device (or its control software) accurately retrieves the "preliminary cleaning plan" for the laparoscopic needle holder from the storage area generated in S12. This plan contains detailed execution instructions. The system analyzes each item in the preliminary cleaning plan to ensure that it understands the meaning of each instruction. For example, it needs to specify: cleaning sequence: D>A>B>C>F>E>G; cleaning medium for each part: for example, for area D (joint axis gap), a specific enzyme cleaning solution is required; cleaning action for each part: for example, for area A (jaw), high-pressure water jet is required; for area B (joint axis), a small brush is required; cleaning parameters for each part: for example, jet pressure (100kPa), jet time (15 seconds), brush speed (300RPM), brushing time (10 seconds), etc.; and the action of changing the cleaning sequence: for example, when moving from area A to area B, the actuator needs to lift, rotate, and then reposition itself to area B.

[0044] The system places a 3D virtual model of the medical device within a virtual cleaning environment. This environment includes virtual cleaning media (such as water flow and bubbles) and virtual cleaning tools (such as nozzles and brushes). Following the instructions of the preliminary cleaning plan, the system "executes" the cleaning and disinfection process in the virtual environment. This includes: the virtual cleaning tools (such as nozzles or brushes) moving to their corresponding virtual locations according to a planned path (D>A>B>C>F>E>G); at each location, the virtual actuator performs specified actions, for example, simulating water jets in area D and brushing actions in area B; and simulating the cleaning media (such as water flow). The system simulates interactions with instrument surfaces (especially areas with simulated stains, if any), such as simulating the process of stains being rinsed, dissolved, or scraped off. The entire process is rendered in real-time as a dynamic video sequence. This video is not just a simple animation; it attempts to simulate the physical effects of a real cleaning process, such as the direction of water flow, brush friction, and stain removal. The system divides the entire simulated cleaning process according to the cleaning area, generating cleaning and disinfection video clips for each area. For example, it will generate a "D-zone joint axis cleaning simulation video," a "A-zone jaw cleaning simulation video," and so on.

[0045] Furthermore, in the cleaning and disinfection videos of various parts, multiple cleaning and disinfection animations are determined based on the dynamic recognition of the cleaning and disinfection videos. Based on the multiple cleaning and disinfection animations and the corresponding cleaning and disinfection areas, cleaning and disinfection reports for each part are determined. A cleaning and disinfection report for the medical device is generated by synthesizing the cleaning and disinfection reports of each part, and the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report is marked. This approach takes into account the overall consideration of multiple cleaning and disinfection animations and the corresponding cleaning and disinfection areas, ensuring the accuracy of the cleaning and disinfection reports for each part.

[0046] At this point, the system acquires simulated cleaning and disinfection video clips generated in S131 for each part (D, A, B, C, F, E, G) of the needle holder. First, the system processes these videos, extracting keyframes or capturing video frames at a fixed frame rate (e.g., a few frames per second). For each extracted frame, the system applies image recognition or computer vision algorithms for "dynamic recognition." Here, "dynamic" refers to changes that occur during the cleaning process, such as: the flow of cleaning agent / water: identifying whether the cleaning agent or water covers the target area as expected; changes in stains / residues: identifying whether the simulated stains (preset in the virtual model) that were originally present have been removed or reduced; the movement of the brush / nozzle: identifying whether the tools (virtual brush, nozzle) performing the cleaning action cover the entire target area and whether there are any omissions; and physical interaction effects: identifying whether the water flow or brush action has produced the expected physical effect (such as rinsing or friction) on the instrument surface.

[0047] Based on the recognition results, the system generates a series of "cleaning and disinfection dynamic images" for each part. These dynamic images can be: keyframe images: showing key states in the cleaning process, such as the start of cleaning, maximum removal of stains, and the end of cleaning; change comparison images: comparing the images before cleaning (preset state in the virtual model) and after cleaning (simulated end frame of cleaning) to highlight the areas of change; and overlay images: generating an overlay image based on the recognized tool movement trajectory to show the area actually acted upon by the virtual brush or nozzle throughout the cleaning process, allowing for a clear view of any omissions.

[0048] Specifically, for the D zone (joint axis gap) of the needle holder: the system extracted several frames of the D zone cleaning video; it identified: first frame: there are preset simulated stains in the gap, and the cleaning agent begins to spray; middle frame: the stain color lightens, and the cleaning agent covers the entire gap; last frame: the stain has basically disappeared, and the gap is clean; the system generated a dynamic map of the D zone, including: a comparison map of the first frame (initial state) and the last frame (post-cleaning state), as well as an overlay map showing that the cleaning agent spray covers the entire gap area.

[0049] The system analyzes the dynamic images of each part generated in step 1; focusing on: coverage integrity: whether the coverage image shows that the cleaning tool has been applied to the entire cleaning and disinfection area (the actual parts corresponding to D, A, B, C, F, E, G); cleaning effect: whether the comparison image shows that the preset simulated stains have been effectively removed; dynamic process: whether the keyframes show a reasonable and coherent cleaning process; report generation logic: based on the analysis results, the system generates a "cleaning and disinfection report" for each part; the report content includes: part identification: for example, "needle holder joint axis (area B)"; status assessment: "pass" or "fail"; specifically, if "pass": it means that the coverage is complete and the stain removal effect meets expectations; it will also include parameter information, such as "cleaning agent spraying time 15 seconds, meets the requirements"; if "fail": it needs to point out the specific reason; for example, "the coverage is incomplete, about 10% of the area at the top of the joint axis gap is not covered by the cleaning agent" or "the stain removal is not thorough, and there are still a small amount of residue".

[0050] The system integrates the cleaning and disinfection reports generated for each part (D, A, B, C, F, E, G) into a complete "Medical Device Cleaning and Disinfection Report." This report lists all parts of the device that require cleaning and disinfection, along with their corresponding evaluation results. In the final report, the system clearly marks the cleaning and disinfection status of each area (D, A, B, C, F, E, G), typically indicating whether each area is "cleaned," "not cleaned," or "partially cleaned." Using a virtual model or diagram of the device, different colors (e.g., green for pass, red for fail, yellow for warning) are used to mark the corresponding areas, visually demonstrating the overall cleaning status. The report concludes with an overall conclusion, such as "The overall cleaning and disinfection of the device has passed" or "Some areas of the device did not meet the cleaning standards, and the plan needs to be adjusted." This report not only provides detailed cleaning status information but also allows medical staff or operators to easily understand the cleaning status of the device through markings, providing crucial information for subsequent decisions (e.g., whether sterilization is possible, whether manual verification is required, etc.).

[0051] Therefore, the usage schedule of medical devices is collected, and the next usage time of the medical devices is determined based on the analysis of the usage schedule. The final cleaning plan for the medical devices is determined based on the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report, the next usage time, and the current time. This final cleaning plan serves as a further optimization of the preliminary cleaning plan, taking into account the overall consideration of the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report, the next usage time, and the current time, ensuring the accuracy of the final cleaning plan. Simultaneously, a virtual version of the preliminary cleaning plan is introduced, taking into account the overall consideration of the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report, the medical device usage schedule, and the current time, further improving the accuracy of the final cleaning plan.

[0052] At this point, the system collects the usage plan of the medical device. From the usage plan collected in the previous step, the system clearly extracts the planned usage date and time. For the needle holder 78901, the parsing result is: the next usage time is 8:30 AM on July 8, 2025. The system converts this time information into a unified time format within the system (such as Unix timestamp or ISO8601 format) for subsequent time calculation and comparison.

[0053] The system retrieves the cleaning and disinfection report generated by S132; the report clearly marks the cleaning and disinfection status (pass / fail / warning) of each cleaning and disinfection area (D, A, B, C, F, E, G); the system obtains the current system time (e.g., the time when the cleaning and disinfection process ended, assuming it is 3:00 PM on July 7, 2025); then it calculates the time difference from the current time to the next usage time; time difference = next usage time (08:30 on July 8, 2025) - current time (15:00 on July 7, 2025); time difference = 17 hours and 30 minutes.

[0054] Sufficient time difference (e.g., ≥24 hours): The initial protocol is usually sufficient; the final protocol can remain consistent with the initial protocol, or, depending on the characteristics of the instrument, consider whether certain time-consuming steps can be skipped (e.g., reducing the number of rinsing steps, if there is a strict sterilization step afterward); however, for safety reasons, skipping steps requires great caution and strict rules; Short time difference (e.g., <24 hours): The initial protocol is sufficient; the final protocol remains unchanged; a short time difference means that the instrument will be used soon, and drying and storage after cleaning become more important; Based on the above analysis, the system generates an optimized final cleaning protocol, which includes all necessary cleaning steps, especially adding remedial measures for areas that failed or were warned in the report, and taking time constraints into account.

[0055] The system will output the final cleaning plan, which includes more detailed information on the cleaning medium, parameters, and sequence, and especially adds remedial steps for addressing problems. This final plan is an evolution and refinement of the preliminary plan (S123 output). It not only includes all the contents of the preliminary plan but has also been adjusted based on the actual simulation results (cleaning report) and actual usage requirements (time plan) to make it more complete and fit the actual application scenario. At this point, this final plan will guide subsequent AR-assisted cleaning operations or serve as control instructions for automated cleaning equipment. For the needle holder 78901, if the report shows that all areas have passed and the time difference is 17.5 hours, the final plan is highly consistent with the preliminary plan. However, if there is a warning on the joint axis of area B, the final plan will add a brushing step for area B.

[0056] refer to Figure 5 In step S14, the specific steps are as follows: S141: Collect the final cleaning plan, and determine multiple cleaning and disinfection items based on the analysis of the final cleaning plan. Based on the item content, item order and medical device of the multiple cleaning and disinfection items, trigger the cleaning and disinfection of the medical device, and execute them in sequence according to the item content of the multiple cleaning and disinfection items. S142: Monitor each cleaning and disinfection area in real time and dynamically mark the surface cleanliness coefficient of each cleaning and disinfection area. At the same time, determine the adjacent surface areas based on the surrounding detection of each cleaning and disinfection area and mark the surface cleanliness coefficient of the adjacent surface areas. Determine the difference in surface cleanliness coefficients based on the surface cleanliness coefficients of each cleaning and disinfection area and the surface cleanliness coefficients of the adjacent surface areas. S143: Determine the auxiliary cleaning plan for the AR cleaning equipment based on the surface cleanliness coefficient, the difference in surface cleanliness coefficient, and the usage scenario of the medical device in each cleaning and disinfection area. At this time, the auxiliary cleaning plan and the final cleaning plan are carried out simultaneously.

[0057] In the embodiments of this application, the final cleaning plan is collected, parsed, and multiple cleaning and disinfection items are output. Simultaneously, each field is parsed according to the data format (e.g., JSON key-value pairs). It recognizes this as a list containing multiple "cleaning and disinfection items." The system traverses the item list, extracting key information for each item: Item number: determines the execution order; Target area: determines which physical part of the instrument needs to be operated on (e.g., area A, area B, area C); the system needs to map this logical area name to specific coordinates or feature points on the instrument's 3D model; Operation type: brushing, rinsing, ultrasonic, disinfectant soaking, etc., which determines which actuator or subsystem needs to be activated; Tool / Tool ID: the brush head, nozzle, or other tool to be used; the system needs to ensure the tool is installed or ready; Parameters: specific operational details, such as angle, force, time, pressure, temperature, disinfectant type / concentration, etc. These parameters need to be converted into executable machine language (e.g., motor speed, pressure valve opening, timer settings); Remarks: optional additional information for system or operator reference.

[0058] The parsed information is converted into an internal instruction format, ready to be passed to the execution layer (such as the robotic arm controller, water pump controller, sensor interface, etc.); for example, "angle 45 degrees" is converted into the target angle value of the robotic arm joint motor; a queue of items to be executed is formed in memory, and each queue item contains all the information required to execute the step.

[0059] The system sends a "start" signal to the robotic arm controller and each actuator controller, accompanied by a prompt tone or screen display. The system retrieves the project instructions one by one according to the order of the project queue. For the currently retrieved project: Positioning and attitude adjustment: The robotic arm controller plans and executes the movement of the robotic arm based on the target area coordinates and operating angle / attitude parameters, precisely moving the selected tool to the designated position on the instrument and adjusting it to the correct angle and attitude; for example, when executing project 1, the robotic arm will move to align the fine brush head with the gap in area B and adjust it to a 45-degree angle; Execution of operation: The actuator controller activates the corresponding actuator according to the operation type and parameters; for example, in project 1, the motor controller drives the fine brush head to rotate with moderate force, while the timer starts counting down for 10 seconds.

[0060] Furthermore, each cleaning and disinfection area is monitored in real time, and the surface cleanliness coefficient of each cleaning and disinfection area is dynamically marked. At the same time, the adjacent surface areas are determined based on the surrounding detection of each cleaning and disinfection area, and the surface cleanliness coefficient of the adjacent surface areas is marked. The difference in surface cleanliness coefficients is determined based on the surface cleanliness coefficients of each cleaning and disinfection area and the surface cleanliness coefficients of the adjacent surface areas. This takes into account the overall consideration of the surface cleanliness coefficients of each cleaning and disinfection area and the surface cleanliness coefficients of the adjacent surface areas, ensuring the accuracy of the difference in surface cleanliness coefficients.

[0061] At this time, each cleaning and disinfection area is monitored in real time, and the surface cleanliness coefficient of each cleaning and disinfection area is dynamically marked. At the same time, the adjacent surface areas are determined based on the surrounding detection of each cleaning and disinfection area, and the surface cleanliness coefficient of the adjacent surface areas is marked. The difference in surface cleanliness coefficients is determined based on the surface cleanliness coefficients of each cleaning and disinfection area and the surface cleanliness coefficients of the adjacent surface areas. During the cleaning process, the cleanliness status of the instrument surface is continuously observed, the degree of cleanliness is quantitatively evaluated, and the difference in cleanliness between adjacent areas is noted.

[0062] The sensor system in AR cleaning equipment typically includes visible light cameras, infrared cameras (for detecting temperature or certain residues), and spectral sensors (for detecting specific chemical residues). This sensor data is transmitted to the central processing unit in real time. Monitoring objects: Multiple cleaning and disinfection areas identified in S112 (such as the joint gap B area of ​​the needle holder, the inner surface A area, the toothed groove C area, etc.). Monitoring methods: Visual analysis: The camera continuously captures images of the cleaning area; the images are analyzed to look for stains, bloodstains, residues, or color changes; for example, a color model (such as HSV) is used to determine if there is still red (bloodstains) or brown (rust). Texture analysis: Changes in surface texture are analyzed; clean metal surfaces and surfaces with residues will have different textures. Spectral analysis: The reflected or transmitted spectra are analyzed to determine if there are specific organic or inorganic residues, continuously acquiring the current visual and physical state information of each cleaning and disinfection area.

[0063] Key features are extracted from monitoring data, such as percentage of soiled area, average color value, texture complexity, and spectral matching degree. Based on a pre-set model or machine learning algorithm, these features are converted into a quantitative value—the Surface Cleanliness Coefficient (SCC). This coefficient is usually a value between 0 and 1, where 1 represents completely clean and 0 represents very dirty. Because it is real-time monitoring, the SCC value changes dynamically as the cleaning operation proceeds. The SCC value is recalculated and updated each time the sensor collects new data, providing an immediate and quantitative indicator of the cleanliness level for each cleaning and disinfection area.

[0064] Specifically, before the scrubbing began, the SCC of area B was only 0.2 (very dirty); after 5 seconds of scrubbing, the image analysis showed that the stain area had decreased by 60% and the color had lightened; the calculated SCC was updated to 0.5; after 10 seconds of scrubbing (end of project 1), the analysis showed that the stains were basically removed, with only a small amount of residue; the SCC was updated to 0.8. This SCC value of 0.8 will be recorded and displayed on the AR interface (using color coding, such as green for meeting the standard, yellow for near meeting the standard, and red for not meeting the standard).

[0065] Based on the virtual model of the instrument and the location of the current cleaning area, determine which other surface areas are physically adjacent to the current cleaning area; for example, the adjacent areas of area B (joint axis gap) include the toothed area C above it, part of the inner surface area A below it, and the transition area connecting these areas; when the sensor monitors the current cleaning area, its field of view or detection range usually naturally covers the adjacent areas. Clearly identify these adjacent areas, separate them from the current overall field of view, and mark the surface cleanliness coefficient of the adjacent surface areas. At this time, the same method is used for these adjacent areas to calculate and mark their SCC values ​​using sensor data; understand whether the current cleaning operation is effective not only on the target area but also on the adjacent areas. This helps to determine whether there are cleaning dead spots or whether the cleaning operation is too vigorous, causing dirt to spread to the adjacent areas.

[0066] Specifically, when area B is halfway through the washing process: based on the 3D model of the needle holder A-100, it is identified that the adjacent area of ​​area B is part of area C (tooth groove) and area A; the camera's field of view not only covers area B, but also part of area C and area A; image areas of area C and adjacent area A are specifically extracted, and their SCC values ​​are calculated respectively; assuming that the SCC of area C is 0.6 at this time (it has been washed before, but not thoroughly), and the SCC of adjacent area A is 0.9 (it is relatively clean).

[0067] For the current cleaning area and each neighboring area, calculate the absolute or relative difference in their SCC values; for example, calculate |SCC_B - SCC_C| and |SCC_B - SCC_A_adjacent|. This Difference Quantity (DQ) reflects the degree of cleanliness imbalance between the current cleaning area and its neighboring areas. The larger the difference, the cleaner (or dirtier) the current area is relative to its neighboring areas. This imbalance means that: there are cleaning dead spots in the neighboring areas; the cleaning operation in the current area is not thorough enough; the cleaning medium (such as water flow) washes dirt to the neighboring areas, causing the SCC of the neighboring areas to decrease while the SCC of the current area increases, forming a difference. This quantitatively assesses the spatial uniformity of cleanliness distribution and identifies potential cleaning problem points.

[0068] Specifically, assuming that after scrubbing area B for 10 seconds: SCC of area B = 0.8; SCC of area C = 0.6; SCC of adjacent area A = 0.9; calculate the difference: DQ_BC = |0.8-0.6| = 0.2; DQ_BA = |0.8-0.9| = 0.1. These two differences (0.2 and 0.1) will be recorded. The larger DQ_BC (0.2) indicates that the difference in cleanliness between area B and area C is more obvious. This suggests that: area C is indeed quite dirty and needs to be cleaned more thoroughly; or, although scrubbing area B is effective, it does not completely remove all dirt; or, some of the debris generated by scrubbing has entered area C.

[0069] Therefore, an auxiliary cleaning scheme for the AR cleaning equipment is determined based on the surface cleanliness coefficient, the difference in surface cleanliness coefficient, and the usage scenario of the medical device in each cleaning and disinfection area. At this time, the auxiliary cleaning scheme and the final cleaning scheme are carried out simultaneously, taking into account the overall consideration of the surface cleanliness coefficient, the difference in surface cleanliness coefficient, and the usage scenario of the medical device in each cleaning and disinfection area, thus ensuring the accuracy of the auxiliary cleaning scheme for the AR cleaning equipment.

[0070] At this point, the system incorporates the surface cleanliness coefficient (SCC) of each cleaning and disinfection area, the difference in SCC, and the usage scenario of the medical devices. For each cleaning and disinfection area, the system compares the current SCC value with a preset cleanliness threshold. This threshold is a fixed value (e.g., 0.9 represents more than 95% cleanliness) and is also dynamically adjusted according to the device type and usage scenario. The system pays special attention to areas with SCC values ​​below the threshold, as well as areas with SCC values ​​above the threshold but close to the critical value. The decision criteria are as follows: if the SCC value of an area is far below the threshold, the system judges that the area is not thoroughly cleaned; if the SCC value of an area is close to the threshold, the system judges that the area poses a risk and requires closer attention; if the SCC values ​​of all areas are far above the threshold, the system judges that the current cleaning effect is good.

[0071] Regarding the difference in surface cleanliness coefficient (DQ), which is the difference between the surface cleanliness coefficient (SCC) value of a certain area and its neighboring areas, the system analyzes the magnitude of each DQ value. A large DQ value (whether positive or negative) indicates a significant imbalance in cleanliness between adjacent areas. Decision criteria: Positive difference (current area SCC > neighboring area SCC): indicates the existence of cleaning dead zones in the neighboring areas, or that the cleaning operation in the current area has transferred some of the dirt to the neighboring areas; Negative difference (current area SCC < neighboring area SCC): indicates that the cleaning operation in the current area is not thorough enough, or that the area itself is heavily contaminated. The system pays special attention to DQ values ​​with large absolute values, regarding them as potential cleaning problem points.

[0072] For medical devices used in specific scenarios, the system gathers device information before or during cleaning, including device type (laparoscopic needle holder), intended use (e.g., for suturing delicate blood vessels vs. for suturing muscle tissue), and estimated level of contamination after use. The system adjusts cleanliness requirements based on the usage scenario. For example, for devices used in highly sterile surgeries (such as neurosurgery and cardiac surgery), a higher cleanliness threshold and stricter DQ requirements are set, allowing no significant imbalances. For surgeries with higher levels of contamination (such as intestinal surgery), the system is more tolerant of minor SCC deficiencies but strengthens monitoring of specific areas (such as lumens). For devices with lower levels of contamination, the system allows for a more lenient threshold. The usage scenario determines the "bottom line" and "priority" of cleaning; the system prioritizes ensuring high-risk areas meet high cleanliness standards.

[0073] The system combines the analysis results of the surface cleanliness coefficient, the difference in surface cleanliness coefficient, and the usage scenario of medical devices in each cleaning and disinfection area to make a comprehensive judgment. Based on the judgment results, the system generates specific auxiliary cleaning instructions or suggestions. These instructions are fine-tuning for the operation that is currently being performed or is about to be performed. The auxiliary solutions include: adjusting the intensity: increasing or decreasing the pressure of brushing or spraying; adjusting the duration: extending the cleaning time for a certain area; adjusting the angle / position: fine-tuning the robotic arm or brush head to better reach the suspected area; changing tools: suggesting or automatically switching to different types or sizes of brush heads / nozzles; adding auxiliary steps: inserting a temporary rinsing or washing step into the original plan. Alerts / Prompts: If the problem is serious, the system will issue an alert to the operator, prompting manual intervention or specific checks. These auxiliary solutions do not interrupt the original plan, but are applied simultaneously with the original plan. For example, the original plan is a standard 15-second scrub, but based on real-time data, the system decides to increase the intensity by 10% in the last 5 seconds. The auxiliary cleaning solutions do not replace the final solution, but rather supplement and optimize it in real time. The robotic arm and cleaning tools will make minor, real-time adjustments based on the instructions of the auxiliary solutions while executing the steps of the final solution, ensuring that the final cleaning effect always moves towards optimization.

[0074] refer to Figure 6 In step S15, the specific steps are as follows: S151: After the medical device completes the auxiliary cleaning plan and the final cleaning plan, collect multiple posture parameters of the medical device, determine the first sub-re-inspection path based on the multiple posture parameters, three-dimensional shape and model of the medical device, and determine the re-inspection path of the AR cleaning equipment for the medical device based on the first sub-re-inspection path, the location of multiple cleaning and disinfection areas and the usage scenario of the medical device. S152: The AR cleaning equipment performs a final inspection of the medical device along the re-inspection path and marks the corresponding surface cleanliness coefficient in sequence during the inspection process. At this time, the re-inspection path has multiple re-inspection nodes, which are arranged along the direction of the re-inspection path and marked with the corresponding surface cleanliness coefficient. S153: Determine a first surface cleaning gradient map based on the positions of multiple re-inspection nodes and the corresponding surface cleaning coefficients; determine a second surface cleaning gradient map based on multiple surface cleaning coefficients and the three-dimensional shape of the medical device; and determine the surface cleaning gradient map of the medical device based on the synthesis of the first and second surface cleaning gradient maps.

[0075] In the embodiments of this application, after the medical device completes the auxiliary cleaning plan and the final cleaning plan, multiple posture parameters of the medical device are collected. Posture parameters: ensure that the path planning is based on the current actual state of the device; three-dimensional shape: the 3D model data of the device, including all its surfaces, edges, holes, gaps and other geometric information, which is the basis of path planning and determines where the inspection points should be; model of the medical device: model information is associated with the typical inspection difficulties or standard inspection areas of this type of device; for example, knowing that it is a "needle holder", the system knows that the jaws and joint gaps are usually the key points.

[0076] The system generates a series of virtual checkpoints, which are evenly distributed or distributed on the instrument surface according to geometric features (such as large curvature changes or gaps). Considering the current posture, the system calculates the positions and angles that the sensor (assuming it is a small, high-resolution vision or spectral sensor) needs to move to in order to "see" or "scan" these checkpoints. The system plans the movement trajectory of the sensor relative to the instrument (usually along a certain scanning pattern on the instrument surface, such as spiral or linear scanning). This draft path is still relatively "idealized" and does not fully consider the key areas cleaned previously. The system outputs a preliminary inspection path (first sub-re-inspection path) covering the main surface of the instrument, represented by a series of positions and postures (poses) that the sensor needs to perform.

[0077] Specifically, based on the previously acquired posture (horizontal placement, jaws facing right and rear) and the 3D model of the needle holder A-100 (knowing it has jaws, jaw bars, joints, handles, etc.), the system generates the first sub-examination path: the starting point of the path is set near the jaws, as this is the most critical part; then the sensor is planned to move along the jaw bars from the tip towards the handle while scanning, covering the inner and outer surfaces of the jaw bars; in the joint area, the path will plan denser checkpoints or smaller movement steps to capture the situation in the gaps; after reaching the handle, the path will cover all sides of the handle; finally, a reverse scan or a path covering the other side of the jaw bars will be planned to ensure that nothing is missed. This path is a preliminary draft to ensure that the sensor can scan all major surface areas of the needle holder from the current posture.

[0078] Based on the first sub-path, optimizations and adjustments are made to generate a final, smarter, and more efficient re-examination path. This optimization considers key cleaning areas (areas that were not cleaned properly before or are prone to getting dirty) and the actual usage of the instruments (which areas come into contact with the patient and which areas are key functional points); the location of multiple cleaning and disinfection areas: those areas that require special attention to cleaning as identified in S112 (such as the joint gaps of the needle holder and the occlusal surfaces of the clamp jaws), these areas also need to be focused on during re-examination; the usage scenario of the medical devices: understanding how the instruments are used in surgery; for example, the needle holder is mainly used for clamping and suturing, and its clamp jaws and occlusal surfaces are in direct contact with tissues and sutures, these areas must be absolutely free of residue during re-examination, and the cleanliness requirements are the highest.

[0079] The first sub-path passes through the "cleaning and disinfection area," which is then reinforced. This involves increasing the density of checkpoints, extending the dwell time in that area, or using a higher resolution detection mode. For critical contact areas clearly defined in the "usage scenario" (such as the jaws and occlusal surfaces of the needle holder), even if they are covered in the first sub-path, they will be given the highest priority in the final path to ensure that these areas are thoroughly inspected. While ensuring the inspection quality of key areas, the trajectory of sensor movement is optimized to reduce unnecessary movement and improve re-inspection efficiency. For example, adjacent checkpoints are merged to plan a smoother trajectory. The optimized path ensures that the sensor will not collide with instruments or the internal structure of the cleaning equipment. Output: The final, optimized re-inspection path, which will be the actual path followed by the AR cleaning equipment when performing the final inspection.

[0080] Specifically, in the needle holder A-100, the system now has a first sub-path and knows: cleaning and disinfection area: joint gap, jaw occlusal surface; usage scenario: mainly used for clamping sutures, the jaw occlusal surface and tip are the key contact points.

[0081] At this point, in the section of the first sub-path passing through the joint gap, the system increases the number of checkpoints and slows down the sensor's movement speed in that area to ensure that conditions deep within the gap can be captured; in the section of the first sub-path passing through the jaw occlusal surface, the system not only increases the density of checkpoints and slows down the speed, but also switches to a detection mode that can better identify minute residues (such as light of a specific wavelength), and this part of the path is given the highest priority; for the jaw tip, although it is covered in the first sub-path, considering that it is in direct contact with the suture, the system also ensures its examination quality by adding a small number of checkpoints; the system optimizes the movement trajectory from the joint area to the jaw area, making it more direct and reducing ineffective movement in non-critical areas; the final determined re-examination path is an intelligent path that combines comprehensive coverage, focused reinforcement, and efficiency optimization; it ensures that the AR device can efficiently and thoroughly examine all critical areas of the needle holder A-100, especially those areas that are given special attention during cleaning and those that directly contact the patient during use.

[0082] Furthermore, the AR cleaning equipment performs a final inspection of the medical device along the re-inspection path and marks the corresponding surface cleaning coefficient in sequence during the inspection process. At this time, the re-inspection path has multiple re-inspection nodes, which are arranged along the direction of the re-inspection path and marked with the corresponding surface cleaning coefficient, thus introducing the marking of the corresponding surface cleaning coefficient.

[0083] At this point, the AR cleaning equipment performs a final inspection of the medical device along the re-inspection path. The equipment's robotic arm moves precisely along the re-inspection path determined in S151. The path has a defined start point, end point, and key points (i.e., re-inspection nodes) along the way. The equipment needs to precisely control its movement speed, acceleration, and attitude to ensure that the sensor can stably and evenly cover every target area on the path. When the sensor reaches a re-inspection node or its coverage area, it will start the inspection program.

[0084] During the detection process, the corresponding surface cleanliness coefficients are sequentially marked. The raw data acquired by the sensor (such as image pixel values, fluorescence intensity, resistance values, etc.) needs to be converted into standardized "surface cleanliness coefficients" through a preset algorithm or model. This coefficient is usually a value between 0 and 1, or a percentage between 0 and 100, representing the cleanliness level of that point. For example: SCC=1.0: indicates that the point is completely clean and no residue was detected; SCC=0.0: indicates that a large amount of residue was detected at the point, which is very dirty; SCC=0.8: indicates that the point has slight residue or is close to clean. The system will associate the calculated SCC value with the corresponding detection point coordinates and store it in the database or transmit it to the control system in real time. This can be regarded as putting a "cleanliness label" on each detected point on the digital twin model of the instrument. The system will complete the detection and marking one point after another according to the order of the robotic arm moving along the re-inspection path to ensure the continuity and traceability of the data.

[0085] When planning the S151 route, a series of "re-inspection nodes" were set on the route. These nodes are key checkpoints on the route, representing: the center point of areas with particularly high cleanliness requirements (such as the center of the jaw occlusion surface); areas with complex structures (such as the corners of joint gaps); representative sampling points of large areas; and turning points on the route, ensuring that no area is missed.

[0086] These re-inspection nodes are arranged in an orderly manner along the physical direction of the re-inspection path; the first node is a checkpoint near the starting point of the path, the last node is a checkpoint near the ending point of the path, and the intermediate nodes are arranged according to the order in which the robotic arm moves. This arrangement ensures the logic and systematic nature of the inspection. During the inspection process of S152, when the robotic arm arrives at and completes the inspection of a certain re-inspection node, the node not only records its spatial coordinates, but is also assigned a corresponding SCC value. In this way, the re-inspection path is not just a geometric trajectory, but a "data chain" with quality information.

[0087] Therefore, a first surface cleaning gradient map is determined based on the location of multiple re-inspection nodes and their corresponding surface cleaning coefficients. A second surface cleaning gradient map is determined based on multiple surface cleaning coefficients and the three-dimensional shape of the medical device. The surface cleaning gradient map of the medical device is then determined by synthesizing the first and second surface cleaning gradient maps. This approach takes into account the overall consideration of synthesizing the first and second surface cleaning gradient maps, ensuring the accuracy of the surface cleaning gradient map of the medical device. At the same time, an auxiliary cleaning scheme is introduced, and the auxiliary cleaning scheme, the final cleaning scheme, and the re-inspection path are integrated as a whole to achieve refined cleaning of the medical device and improve the accuracy of the surface cleaning gradient map of the medical device.

[0088] At this point, each re-inspection node has a clear three-dimensional coordinate (location) and the surface cleanliness coefficient (SCC) detected at that location. The "first surface cleanliness gradient map" can be understood as a visualization based on discrete point data; it is similar to a map, or a prototype of a scatter plot / heat map. At the same time, the three-dimensional coordinates of each re-inspection node are projected onto a two-dimensional plane (e.g., an unfolded view or main view of the instrument), and its SCC value is marked with color or numerical value at the corresponding location. The color usually transitions from low (e.g., red, indicating poor cleanliness) to high (e.g., green, indicating good cleanliness). If the re-inspection nodes are densely distributed, simple interpolation (e.g., linear interpolation) can be performed between the nodes to preliminarily estimate the cleanliness of the area between the nodes, making the graph look more continuous. This graph particularly emphasizes the locations defined as re-inspection nodes, as they are usually key areas or areas that were previously cleaned.

[0089] Specifically, the system collects the coordinates and SCC values ​​of nodes 1 to 7 (assuming, as previously stated: node 1: 0.95, node 2: 0.96, node 3: 0.97, node 4: 0.85, node 5: 0.99, node 6: 0.98, node 7: 0.97); the system plots these points on a view of the 2D unfolded schematic diagram or 3D model of the needle holder A-100; for example, green (0.95) is displayed in the handle area (node ​​1); green is displayed in the connecting axis area (node ​​2). (0.96); Green is displayed on the outer side of the joint (node ​​3) (0.97); Yellow or orange is displayed deep in the joint gap (node ​​4) (0.85), clearly distinguishing it from the surrounding area; Dark green is displayed in the center of the jaws (node ​​5) (0.99); Green is displayed on the edge of the jaws (node ​​6) (0.98); Green is displayed at the tip (node ​​7) (0.97). The first image generated in this way intuitively shows the cleanliness distribution of these key nodes, especially the relatively low cleanliness of node 4, which is obvious at a glance.

[0090] Introducing multiple surface cleanliness coefficients and the three-dimensional morphology of medical devices, the "second surface cleanliness gradient map" focuses more on continuity and overall distribution, attempting to display the changes in cleanliness across the entire three-dimensional surface of the device. Here, the 3D model surface of the medical device is divided into fine grid cells; the SCC values ​​of all detection points are correlated with their corresponding positions on the 3D model surface; for grid cells not directly detected, more complex interpolation (such as Kriging interpolation, radial basis function interpolation, etc.) is performed using the SCC values ​​of neighboring detection points to estimate their cleanliness; the estimated SCC value of each grid cell is mapped to color and then rendered on the 3D model, forming a cleanliness map covering the entire device surface with continuously changing colors.

[0091] Specifically, the system not only has data for nodes 1-7, but also detection data for other non-node locations along the re-inspection path (such as SCC values ​​for hundreds of points on the inner wall of the joint gap and the sides of the jaws); the system loads a precise 3D model of the needle holder A-100, dividing its surface into tens of thousands of tiny triangular patches; the system "projects" the SCC values ​​of all detection points onto the corresponding positions on the 3D model; then, for each small patch, the system calculates an estimated SCC value based on the positions and SCC values ​​of all surrounding detection points; for example, the area inside the joint gap appears yellow because the SCC of node 4 is low and other surrounding detection points are also low; while the jaw area appears dark green because the SCC of nodes 5, 6, and 7 is high; the final generated second image is a rotatable and scalable 3D model with a smooth color transition on its surface, intuitively showing the distribution of overall cleanliness from the handle to the tip, and from the joint gap to the outer surface; it not only shows the problem at node 4, but also the slight unevenness in the area surrounding the joint gap.

[0092] The "node emphasis" information of the first surface cleaning gradient map and the "overall continuity" information of the second surface cleaning gradient map are combined. For example, on the 3D model of the second map, special markers (such as circles, asterisks, and borders of different colors) or brighter colors can be used to highlight the re-inspection nodes in the first map whose SCC values ​​are below a certain threshold. Different weights are assigned to certain features of the first and second maps, and weighted fusion is performed. The accurate node data of the first map is used to correct the deviation of the interpolation in the second map, and the continuity of the second map is used to fill the gaps in the area between nodes in the first map.

[0093] Specifically, the system uses the generated rotatable second surface cleaning gradient map as a base; then, the system finds those nodes with SCC values ​​below a preset threshold (e.g., below 0.90) for re-inspection (in this example, only node 4, SCC 0.85); on the final composite map, the system adds a prominent red circle mark at node 4 (deep in the joint gap) in the 3D model, or adjusts the color of this area to a brighter yellow, with the text prompt "Cleanliness Concern". In this way, the final surface cleaning gradient map shows that the needle holder A-100 is generally very clean (most areas are green), and also clearly points out that the deep joint gap (node ​​4) is the only area with slightly lower cleanliness that requires special attention.

[0094] Please see Figure 7 , Figure 7 This is a schematic diagram of the structural composition of the AR-based medical device assisted cleaning and disinfection system according to an embodiment of the present invention; the AR-based medical device assisted cleaning and disinfection system includes: The cleaning and disinfection path module 21 is used to determine multiple cleaning and disinfection areas based on the online detection of medical devices by AR cleaning equipment, and to determine the cleaning and disinfection path based on the multiple cleaning and disinfection areas and the three-dimensional shape of the medical devices. The preliminary cleaning plan module 22 is used to predict the preliminary cleaning plan for medical devices based on the starting position, path shape, and location of multiple cleaning and disinfection areas of the cleaning and disinfection path. The final cleaning plan module 23 is used to generate a cleaning and disinfection report for the medical device based on the virtual representation of the preliminary cleaning plan by the AR cleaning device, and to determine the final cleaning plan for the medical device based on the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report, the usage schedule of the medical device, and the current time. The auxiliary cleaning scheme module 24 is used by the AR cleaning equipment to clean and disinfect the medical device according to the final cleaning scheme. The auxiliary cleaning scheme of the AR cleaning equipment is determined based on the surface cleanliness coefficient of each cleaning and disinfection area, the surface cleanliness coefficient of the adjacent surface area, and the usage scenario of the medical device. The surface cleaning gradient map module 25 is used to determine the re-inspection path of the medical device by the AR cleaning equipment based on multiple posture parameters and three-dimensional shape of the medical device, and to determine the surface cleaning gradient map of the medical device based on the re-inspection path and the corresponding multiple surface cleaning coefficients.

[0095] This invention provides an AR-based auxiliary cleaning and disinfection method for medical devices, which uses the aforementioned AR-based auxiliary cleaning and disinfection system to disinfect medical devices.

[0096] The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. An AR-based medical device assisted cleaning and disinfection system, characterized in that, include: The cleaning and disinfection path module is used to determine multiple cleaning and disinfection areas based on the online detection of medical devices by AR cleaning equipment, and to determine the cleaning and disinfection path based on the multiple cleaning and disinfection areas and the three-dimensional shape of the medical devices. The preliminary cleaning plan module is used to predict the preliminary cleaning plan for medical devices based on the starting point, path shape, and location of multiple cleaning and disinfection areas of the cleaning and disinfection path. The final cleaning plan module is used to generate a cleaning and disinfection report for the medical device based on the virtual preliminary cleaning plan provided by the AR cleaning device. The final cleaning plan for the medical device is determined based on the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report, the usage schedule of the medical device, and the current time. The auxiliary cleaning scheme module is used by the AR cleaning equipment to clean and disinfect medical devices according to the final cleaning scheme. The auxiliary cleaning scheme of the AR cleaning equipment is determined based on the surface cleanliness coefficient of each cleaning and disinfection area, the surface cleanliness coefficient of adjacent surface areas, and the usage scenario of the medical device. The surface cleaning gradient map module is used to determine the re-inspection path of the medical device by the AR cleaning equipment based on multiple posture parameters and three-dimensional shape of the medical device, and to determine the surface cleaning gradient map of the medical device based on the re-inspection path and the corresponding multiple surface cleaning coefficients.

2. The AR-based medical device assisted cleaning and disinfection system according to claim 1, characterized in that, The method involves determining multiple cleaning and disinfection zones based on online detection of medical devices using AR cleaning equipment, and determining the cleaning and disinfection path based on these zones and the three-dimensional shape of the medical devices, including: The medical device is placed in the cleaning space of the AR cleaning equipment. The AR cleaning equipment rotates the medical device and takes pictures of the medical device from different positions based on the built-in camera to collect images of the medical device at different positions. Based on multiple images and the corresponding three-dimensional data of the medical device, a virtual model of the medical device is determined and the three-dimensional shape of the medical device is marked. Multiple cleaning and disinfection areas are determined based on online detection of a virtual model of the medical device. These areas are distributed on the outer and inner surfaces of the medical device and are located at different positions on the device. The AR cleaning equipment determines the cleaning and disinfection path based on the location of these multiple cleaning and disinfection areas, the three-dimensional shape of the medical device, and usage information.

3. The AR-based medical device assisted cleaning and disinfection system according to claim 1, characterized in that, The preliminary cleaning plan for medical devices, based on the starting point, path shape, and locations of multiple cleaning and disinfection areas, includes: The starting position of the cleaning and disinfection path is determined based on the detection of the cleaning and disinfection path. This starting position of the cleaning and disinfection path serves as the starting point for the cleaning and disinfection of the medical device. The first cleaning and disinfection space of the medical device is determined based on the starting position of the cleaning and disinfection path and the virtual model of the medical device. The first cleaning plan for the medical device is determined based on the spatial shape of the first cleaning and disinfection space and the path shape of the cleaning and disinfection path; the second cleaning plan for the medical device is determined based on the path shape of the cleaning and disinfection path and the location of multiple cleaning and disinfection areas. Based on the first cleaning plan, the second cleaning plan, and the model of the medical device, a preliminary cleaning plan for the medical device is predicted. This preliminary cleaning plan marks the cleaning medium, multiple cleaning parameters, and the cleaning sequence of various parts in the medical device.

4. The AR-based medical device assisted cleaning and disinfection system according to claim 1, characterized in that, The process involves generating a cleaning and disinfection report for the medical device based on the virtual simulation of the preliminary cleaning plan using AR cleaning equipment. The final cleaning plan for the medical device is then determined based on the cleaning and disinfection status of multiple cleaning and disinfection areas in the report, the medical device's usage schedule, and the current time. This includes: The various parts of the medical device are determined by traversing the virtual model of the medical device. At the same time, the AR cleaning equipment collects the preliminary cleaning plan and constructs the cleaning and disinfection video of each part of the medical device based on the preliminary cleaning plan and the virtual model of the medical device. In the cleaning and disinfection videos of various parts, multiple cleaning and disinfection dynamic images are determined based on the dynamic recognition of the cleaning and disinfection videos. Based on the multiple cleaning and disinfection dynamic images and the corresponding cleaning and disinfection areas, cleaning and disinfection reports for each part are determined. The cleaning and disinfection report of the medical device is generated by synthesizing the cleaning and disinfection reports of each part, and the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report is marked.

5. The AR-based medical device assisted cleaning and disinfection system according to claim 4, characterized in that, The process of generating a cleaning and disinfection report for the medical device based on the virtual preliminary cleaning plan created by the AR cleaning device, and determining the final cleaning plan for the medical device based on the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report, the medical device's usage schedule, and the current time, further includes: The usage schedule of medical devices is collected, and the next usage time of the medical devices is determined based on the analysis of the usage schedule. The final cleaning plan of the medical devices is determined based on the cleaning and disinfection status of multiple cleaning and disinfection areas in the cleaning and disinfection report, the next usage time of the medical devices, and the current time. This final cleaning plan serves as a further optimization of the preliminary cleaning plan.

6. The AR-based medical device assisted cleaning and disinfection system according to claim 1, characterized in that, The AR cleaning equipment cleans and disinfects the medical device according to the final cleaning plan. Based on the surface cleanliness coefficients of each cleaning and disinfection area, the surface cleanliness coefficients of adjacent surface areas, and the usage scenario of the medical device, the AR cleaning equipment determines an auxiliary cleaning plan, including: The final cleaning plan is collected, and multiple cleaning and disinfection items are determined based on the analysis of the final cleaning plan. The cleaning and disinfection of medical devices is triggered according to the content, order, and medical device of the multiple cleaning and disinfection items, and is executed in sequence according to the content of the multiple cleaning and disinfection items.

7. The AR-based medical device assisted cleaning and disinfection system according to claim 6, characterized in that, The AR cleaning equipment cleans and disinfects the medical device according to the final cleaning plan. Based on the surface cleanliness coefficients of each cleaning and disinfection area, the surface cleanliness coefficients of adjacent surface areas, and the usage scenario of the medical device, an auxiliary cleaning plan for the AR cleaning equipment is determined, which also includes: Real-time monitoring of each cleaning and disinfection area, and dynamic marking of the surface cleanliness coefficient of each cleaning and disinfection area. At the same time, based on the surrounding detection of each cleaning and disinfection area, the adjacent surface area is determined and the surface cleanliness coefficient of the adjacent surface area is marked. The difference in surface cleanliness coefficient between each cleaning and disinfection area and the adjacent surface area is determined. The auxiliary cleaning scheme for the AR cleaning equipment is determined based on the surface cleanliness coefficient of each cleaning and disinfection area, the difference in surface cleanliness coefficient, and the usage scenario of the medical device. At this time, the auxiliary cleaning scheme and the final cleaning scheme are carried out simultaneously.

8. The AR-based medical device assisted cleaning and disinfection system according to claim 1, characterized in that, The process of determining the re-inspection path of the medical device using AR cleaning equipment based on multiple posture parameters and three-dimensional shape of the medical device, and determining the surface cleaning gradient map of the medical device based on the re-inspection path and corresponding multiple surface cleaning coefficients, includes: After the medical device completes the auxiliary cleaning plan and the final cleaning plan, multiple posture parameters of the medical device are collected. Based on the multiple posture parameters, three-dimensional shape and model of the medical device, the first sub-re-inspection path is determined. Based on the first sub-re-inspection path, the location of multiple cleaning and disinfection areas and the usage scenario of the medical device, the re-inspection path of the AR cleaning equipment for the medical device is determined.

9. The AR-based medical device assisted cleaning and disinfection system according to claim 8, characterized in that, The method of determining the re-inspection path of the medical device by the AR cleaning equipment based on multiple posture parameters and three-dimensional shape of the medical device, and determining the surface cleaning gradient map of the medical device based on the re-inspection path and the corresponding multiple surface cleaning coefficients, further includes: The AR cleaning equipment performs a final inspection of the medical device along the re-inspection path and marks the corresponding surface cleanliness coefficient in sequence during the inspection process. At this time, the re-inspection path has multiple re-inspection nodes, which are arranged along the direction of the re-inspection path and marked with the corresponding surface cleanliness coefficient. A first surface cleaning gradient map is determined based on the location of multiple re-inspection nodes and the corresponding surface cleaning coefficients. A second surface cleaning gradient map is determined based on multiple surface cleaning coefficients and the three-dimensional shape of the medical device. The surface cleaning gradient map of the medical device is determined by combining the first surface cleaning gradient map and the second surface cleaning gradient map.

10. An AR-based method for assisting in the cleaning and disinfection of medical devices, characterized in that, The medical device is disinfected using the AR-based medical device auxiliary cleaning and disinfection system described in any one of claims 1-9.

Images