An automatic separation, identification and counting system and method for dirty medical instruments in an orthopedic operating room

Abstract

The application provides a system and method for automatic separation, identification and counting of contaminated medical devices in an orthopedic operating room, belonging to the technical field of medical device management. The system includes a conveying mechanism, an RFID identification module, a sterile separation module, a visual identification module, an actuating mechanism, a storage module, a sterilization module, a control unit and a display module. The method includes the steps of device delivery and initial transmission, RFID preliminary identification and classification, multi-level separation processing, visual accurate identification and counting, large item grabbing and classified storage, small item collection and confirmation, data display and abnormal alarm. Through the cooperative work of multiple modules, the application integrates RFID preliminary classification, multi-level sterile separation and enhanced YOLOv5s visual identification technology, realizes the automatic, sterile and accurate separation, identification and counting of contaminated and overlapped orthopedic devices, the processing time is less than 1 minute, the identification accuracy is greater than 98%, effectively reduces the labor intensity and the risk of needle stick injury of medical staff, prevents RSI medical accidents, improves the operation efficiency and clinical safety.

Description

Technical Field

[0001] This invention belongs to the field of medical service technology, specifically relating to an automatic separation, identification and inventory system and method for soiled medical devices in orthopedic operating rooms. Background Technology

[0002] In orthopedic operating rooms, the management and disposal of medical instruments are crucial for ensuring surgical safety and efficiency. According to the Association of Peri-Operative Registered Nurses (AORN) 2025-2026 guidelines, orthopedic operating room environments are highly standardized, typically ranging from 15 to 25 square meters, and equipped with shadowless lamps, Mayo Clinics, back trays, and a Sterile Processing Department (SPD) transport system. Instruments are stored in sterile containers and transported from the SPD to the operating room via blue wrapping or hard cases to prevent contamination.

[0003] Orthopedic surgeries typically involve 100-200 instruments and can last 2-6 hours, increasing the risk of fatigue for medical staff. During surgery, instruments are placed on the Mayo table and transferred to the waste tray after use, often resulting in stacking, such as blood-stained titanium screws pressing on bone forceps or similarly shaped bone pins overlapping. After surgery, instruments remain mixed in the tray, potentially contaminated with blood, bone fragments, and other debris, requiring nurses to perform a rapid count before suturing. AORN requires a "triple count" procedure, but this process is often rushed, especially in emergency fracture repair surgeries, and prone to errors.

[0004] The unique challenges of orthopedic surgery stem from the diversity and similarity of instruments, as well as the complexity of the surgical environment. Implants such as titanium screws and bone pins are often small (<1cm), numerous, and similar in shape, making them difficult to identify manually when covered in contaminants, leading to counting difficulties. Sharp objects, such as titanium screws, can easily cause puncture wounds to medical staff during cleaning; tiny items, such as joint prostheses, are difficult to count due to obscuring contaminants; suture-related items, such as retractors and needle holders, are easily left in the patient's body, forming retained surgical items (RSI). According to data from the National Health Commission (NHC), orthopedic RSI accounts for approximately 20-30% of all RSIs. Common remnants include sponges (70%), instrument fragments, or small needles, resulting in a reoperation rate of up to 50% and an average extended hospital stay of 7 days. High BMI patients or complex anatomical structures further increase the risk of instrument entrapment, while prolonged surgery and contamination reducing visibility exacerbate these problems.

[0005] Current technologies primarily rely on manual counting and traditional auxiliary methods, but these have significant limitations. Manual counting is labor-intensive and prone to errors, especially under fatigue. Industrial sorting systems (such as coin sorters) are unsuitable for orthopedic instruments because they are irregularly shaped, variable, require a strictly sterile environment, and are susceptible to contamination from vibration or mechanical contact. Furthermore, they cannot achieve real-time (<1 minute) and high accuracy (over 99%) requirements to prevent medical accidents. Radio-frequency identification (RFID) systems (such as US patent US7518502B2) can be used to track instruments, but are only suitable for large, coded items and cannot effectively handle small items with overlapping contaminants or those requiring purely visual counting. Other patents, such as US10187742B2 (real-time tracking system) and US12082893B2 (robot positioning system), introduce navigation or robotics technologies, but lack AI optimization for overlapping contaminant images, resulting in slow clinical adoption and failing to comprehensively address the visual complexity and high-risk issues of orthopedic instrument cleaning.

[0006] Therefore, existing instrument cleaning methods are insufficient to meet the sterile, real-time, and accurate requirements of orthopedic operating rooms. There is an urgent need for a new system and method that can combine multiple modules to work together to efficiently process contaminated instruments, thereby reducing the risk of medical staff injury, improving counting accuracy, and preventing RSI medical accidents. Summary of the Invention

[0007] The purpose of this invention is to provide an automated separation, identification, and counting system and method for orthopedic surgical medical devices, addressing the problems of low efficiency, poor accuracy, easy injury to medical staff, and high risk of RSI (Recovery-Solving Injury) accidents in the postoperative medical device cleaning process of orthopedic surgery in existing technologies. By integrating multiple modules working collaboratively, it achieves real-time, high-accuracy separation, identification, and counting of contaminated and overlapping instruments, meeting operating room aseptic standards, improving surgical efficiency, reducing fatigue-induced counting errors, and enhancing overall clinical safety.

[0008] To achieve the above-mentioned objectives, the present invention provides an automatic separation, identification, and inventory system for soiled medical instruments in orthopedic operating rooms, comprising: A conveyor system used to transport soiled medical instruments after orthopedic surgery. The RFID identification module includes a high-frequency RFID reader array integrated below the conveyor mechanism, with a scanning range covering the entire strip surface, used to identify instruments with pre-coded tags in real time and to initially distinguish between large and small instruments. The aseptic separation module includes a medical-grade electromagnetic vibrator, a pulsed aseptic water atomizing nozzle array, and an aseptic airflow separation mechanism. It uses controllable low-frequency vibration, water flow, and airflow to break down the contaminants and physical overlap between instruments. The visual recognition module uses a multi-axis movable depth camera with a CoreXY structure to acquire images of instruments from multiple angles, and combines them with an enhanced YOLOv5s algorithm processing unit to identify the type, size and quantity of instruments in a state of overlapping dirt. The actuator, including electrically servo-driven parallel grippers, is used to grasp large instruments identified by RFID. The storage module includes a large item storage area and a small item storage area; the large item storage area is located on the side of the actuator and is used for the dedicated temporary storage and inventory of large instruments; the small item storage area is located at the end of the conveyor belt and is used for the collection and final processing of small instruments. The disinfection module includes ultraviolet lamps installed in the conveyor mechanism, the large item storage area, and the small item storage area; The control unit is used to receive detection data from the RFID identification module and the vision identification module, coordinate and control the operation of the transmission mechanism, the aseptic separation module and the actuator, and synchronously output instrument type, count and abnormal alarm information; The display module is an integrated LCD screen used to display the instrument type, count, and abnormal alarms in real time.

[0009] Preferably, the conveying mechanism uses a medical-grade stainless steel conveyor belt coated with a chemically resistant polytetrafluoroethylene coating; the operating speed of the conveying mechanism is adjustable between 0.1-0.5 m / s, and the edges are equipped with protective railings and inclined guide structures.

[0010] Preferably, the medical-grade electromagnetic vibrator has a frequency of 10-50Hz, an amplitude of 0.5-2mm, and a power of <30W. It is elastically connected to the stainless steel belt of the conveying mechanism through a silicone shock-absorbing pad and is located in the middle and rear section of the conveying mechanism. It breaks the dirt adhesion and physical overlap between small instruments through controllable low-frequency vibration, and provides a uniform spreading state to assist the YOLOv5s algorithm in accurately identifying small objects. The pulsed sterile water atomizing nozzle array is made of PTFE material, with a pressure of 0.1-0.3MPa and a pulse duration of 2-5s. It uses sterile pure water filtered through a 0.22-0.25μm filter and is located downstream of the vibration section of the conveying mechanism. The pulsed sterile water is used to rinse and dissolve blood clots / bone fragments, and the shear force of the water flow is used to assist in the separation of small instruments, while also initially cleaning surface dirt.

[0011] Preferably, the aseptic airflow separation mechanism includes a medical blower and an airflow guide rail disposed on both sides of the conveying mechanism. The airflow guide rail is made of medical-grade ABS plastic and designed as a gradient cross-section channel. The air outlet adopts an alternating activation mode, and the wind speed adjustment range is 5-10m / s. The airflow passes through a built-in HEPA filter to ensure sterile output and is monitored in real time by an integrated wind speed sensor, thereby directionally blowing away the dirt and adhesion of small instruments, improving the separation uniformity and counting accuracy of subsequent visual recognition, and avoiding instrument scattering or secondary contamination.

[0012] Preferably, the alternating activation mode switches between the left and right sides every 5-10 seconds; The size of the small instrument is <1cm.

[0013] Preferably, the depth camera is driven by a stepper motor to link the X and Y axes and acquire instrument images from multiple angles. The enhanced YOLOv5s algorithm adds a small object detection layer and a CBAM attention mechanism, and is trained on a blood-stained orthopedic instrument dataset to determine the type, size and quantity of instruments under smudged conditions.

[0014] Preferably, the parallel grippers have a clamping force of 0.5-5N, a response time of <0.5s, and the ends of the grippers are covered with flexible medical silicone pads, supporting 360° rotation and a 10cm lifting stroke.

[0015] Preferably, the large item storage area is a multi-layer stainless steel mesh frame structure, with a pressure sensor integrated at the bottom of the tank with an accuracy of 0.1g to automatically record weight changes to verify counting accuracy; a constant-on ultraviolet light is installed at the top, and a multi-axis movable depth camera with a CoreXY structure extends to cover the large item storage area, and combined with the YOLOv5s algorithm to process possible residual dirt images, accurately recording the size, quantity and type of large instruments, and updating the data to the display screen in real time, so as to facilitate quick verification by medical staff.

[0016] Preferably, the small item storage area is an inclined collection trough made of transparent medical-grade polycarbonate material, with a sliding angle of 15°-20°. The bottom of the trough is equipped with drainage holes and a fine filter with a pore size of 0.5mm. The small item storage area is connected to the conveyor belt frame of the conveying mechanism via a T-shaped connecting plate and is reinforced at the bottom with corner brackets. A multi-axis movable depth camera with a CoreXY structure extends to cover the small item storage area and, combined with the enhanced YOLOv5s algorithm, processes blood-stained overlapping images to accurately record the size, quantity, and type of small instruments. The data is simultaneously displayed on the display screen, and an alarm is triggered for potential omissions. A constantly lit ultraviolet light is installed on the top. An integrated vibration sensor monitors the sliding dynamics to prevent blockage or accumulation.

[0017] Another aspect of the present invention provides an automatic separation, identification, and counting method for orthopedic surgical medical devices. The method is based on the automatic separation, identification, and counting system and includes the following steps: S1. Instrument placement and initial transfer: Place the soiled medical instruments from orthopedic surgery at the starting end of the transfer mechanism. The control unit starts the transfer mechanism to transfer the instruments smoothly at a certain speed. The ultraviolet lamp of the disinfection module is turned on simultaneously for continuous disinfection. S2. Preliminary RFID Identification and Classification: During the operation of the conveying mechanism, the RFID identification module scans the instruments in real time through a high-frequency reader array, identifies large instruments with pre-coded tags, and transmits the identification results to the control unit to initially distinguish between large and small instruments. S3. Multi-level separation process: Based on RFID identification results and preliminary scanning data from the depth camera, the control unit activates the aseptic separation module in stages. S31. Vibration separation: Activate the medical-grade electromagnetic vibrator to break up the dirt adhesion and physical overlap between small instruments; S32. Water rinsing and separation: Activate the pulsed sterile water atomizing nozzle array to dissolve blood clots and bone fragments, preliminarily clean the surface of the instruments, and recover the waste liquid through the inclined drainage tank and fine filter screen below. S33. Airflow dispersion and separation: Activate the sterile airflow separation mechanism, and blow small instruments at a certain wind speed through the alternating air outlets of the airflow guide rail to further disperse overlapping instruments; S4. Precise Visual Recognition and Counting: The CoreXY structure of the visual recognition module has a multi-axis movable depth camera that moves along the XY axis to acquire images of the separated instruments from multiple angles. The image data is processed by the enhanced YOLOv5s algorithm to accurately identify the type, size and quantity of the instruments and transmit the results to the control unit. S5. Large item grasping and classification storage: The control unit triggers the parallel gripper of the actuator to locate and grasp large instruments based on RFID identification results and visual recognition data, and transfers them to the designated grid slots in the large item storage area. The pressure sensor in the large item storage area records weight changes to verify the counting accuracy, and the depth camera scans simultaneously to confirm the placement status. S6. Small item collection and confirmation: Small items are carried to the end of the conveyor and slide naturally into the small item storage area at an angle of 15°-20°. Vibration sensors monitor the sliding dynamics to prevent blockage or accumulation. A depth camera scans the small item storage area and combines it with a visual recognition algorithm to complete the final counting and confirmation. S7. Data Display and Anomaly Alarm: The display module shows the instrument type and counting results in real time. The control unit compares the RFID identification data with the visual identification data. If there is a discrepancy or anomaly in the counting, an alarm is triggered immediately.

[0018] The present invention has the following beneficial effects: 1. This invention integrates multiple modules to work together, realizing fully automated processing from instrument transmission, separation, identification to storage, without the need for a large amount of manual intervention, with a processing time of less than 1 minute, significantly improving processing efficiency and reducing errors caused by the labor intensity and fatigue of medical staff.

[0019] 2. This invention uses a combination of RFID preliminary classification and enhanced YOLOv5s visual recognition algorithm to accurately identify instruments in a state of overlapping contamination, with an accuracy rate of >98%, effectively preventing RSI medical accidents and reducing reoperation rate and hospitalization time.

[0020] 3. All modules of this invention are made of medical-grade materials and are equipped with full-process UV-C ultraviolet disinfection and HEPA-filtered sterile airflow design, which strictly meet the sterile standards of the operating room, avoid cross-contamination, and reduce the chance of medical staff coming into contact with sharp instruments, thus reducing the risk of puncture wounds.

[0021] 4. This invention is designed to address the diversity and irregularity of orthopedic instruments, and can effectively handle instruments of different sizes and shapes, especially small (<1cm) and similar-shaped instruments, solving the problems of difficulty in identifying and counting tiny instruments in the prior art.

[0022] 5. The overall frame of this invention is constructed with aluminum alloy profiles and reinforced with L-shaped and T-shaped connecting plates and 2028-type corner fittings. It has good shock resistance and ensures that instruments are placed stably and operated smoothly during high-intensity clinical use, reducing instrument displacement or identification errors caused by vibration.

[0023] 6. The control unit of this invention can store relevant data for each inventory count, supporting subsequent queries and traceability, which facilitates medical quality control. Attached Figure Description

[0024] The present invention will be further described below with reference to the accompanying drawings and embodiments. The accompanying drawings, which constitute a part of this application, are used to provide a further understanding of the present invention. The illustrative embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention.

[0025] Figure 1 System framework diagram of the present invention.

[0026] Figure 2 The system layout diagram of this invention.

[0027] Figure 3 Top view of the multi-axis movable depth camera with CoreXY structure of this invention.

[0028] Figure 4 Front view of the parallel gripper driven by the electric servo of the present invention.

[0029] Figure 5 Flowchart of the method of this invention. Detailed Implementation

[0030] The embodiments of the present invention will be further described below with reference to the accompanying drawings.

[0031] Example 1: like Figure 1-4 As shown, an automated separation, identification, and counting system for orthopedic surgical medical devices includes a conveying mechanism, an RFID identification module, a sterile separation module, a visual identification module, an actuator, a storage module, a disinfection module, a control unit, and a display module. This system aims to achieve real-time, high-accuracy separation, identification, and counting of contaminated and overlapping instruments by integrating an airflow separation mechanism, an enhanced YOLOv5s visual recognition algorithm, and RFID-assisted technology. This reduces the risk of injury to medical staff when handling sharp objects (such as titanium screws), simplifies the counting process of small items (such as joint prostheses), and effectively prevents medical accidents caused by suture-related items (such as retractors and needle holders) remaining in the patient's body. A further objective of this invention is to meet operating room sterility standards, improve surgical efficiency, reduce fatigue-induced counting errors, and enhance overall clinical safety, overcoming the limitations of existing technologies.

[0032] Furthermore, the conveying mechanism uses a medical-grade 304 stainless steel belt coated with a chemically resistant polytetrafluoroethylene coating. The belt is approximately 1.2m long and 30cm wide. The operating speed can be adjusted between 0.1-0.5m / s by the control unit. The belt edge is equipped with a guardrail 10 and an inclined guide structure 11 to prevent the instruments from slipping or overflowing.

[0033] Furthermore, the RFID identification module includes a high-frequency RFID reader array integrated below the conveyor mechanism, with a frequency of 13.56MHz and a 4×4 antenna configuration. The scanning range completely covers the surface of the conveyor mechanism, enabling real-time identification of large instruments with pre-coded tags (volume > 5cm), a multi-tag reading accuracy of ≥ 99%, and transmitting the identification results to the control unit to achieve preliminary differentiation between large and small instruments.

[0034] Furthermore, the aseptic separation module includes a medical-grade electromagnetic vibrator, a pulsed aseptic water atomizing nozzle array, and an aseptic airflow separation mechanism. The medical-grade electromagnetic vibrator has a frequency of 10-50Hz, an amplitude of 0.5-2mm, and a power of <30W. It is elastically connected to the conveyor belt via silicone shock-absorbing pads and is located in the middle to rear section of the conveyor mechanism, with a length of 40-60cm. The pulsed sterile water atomizing nozzle array is made of PTFE material, with a pressure of 0.1-0.3MPa and a pulse duration of 2-5s. It uses sterile pure water filtered through 0.22μm and is located downstream of the vibration section, with a length of 30-40cm. It has an inclined drainage channel and a fine filter screen below it for rapid recovery of waste liquid. Downstream, there is also a warm airflow drying section to ensure that the residual moisture of the instruments is <5%. The sterile airflow separation mechanism includes a medical blower and an airflow guide rail. The airflow guide rail is made of medical-grade ABS plastic and has a gradient cross-section channel. The air outlet adopts an alternating activation mode, switching between the left and right sides every 5-10 seconds. The wind speed is adjustable from 5-10m / s. The airflow is filtered by a built-in HEPA filter and monitored in real time by an integrated wind speed sensor.

[0035] Furthermore, the visual recognition module includes a multi-axis movable depth camera with a CoreXY structure, a resolution of 1920×1080, a depth accuracy of ±1mm, and an enhanced YOLOv5s algorithm processing unit. The depth camera is driven by a stepper motor to link the X and Y axes at a speed of 100mm / s, and can acquire instrument images from multiple angles. The enhanced YOLOv5s algorithm adds a small object detection layer and a CBAM attention mechanism. Trained on a dedicated dataset of blood-stained orthopedic instruments, it can accurately identify the type, size, and quantity of instruments under occlusion conditions, with a recognition accuracy of >98%.

[0036] Furthermore, the actuator includes an electrically servo-driven parallel gripper with an adjustable gripping force of 0.5-5N and a response time of <0.5s. The gripper tip is covered with a flexible medical silicone pad, supporting 360° rotation and a 10cm lifting stroke. It is located in the gripping area, which is approximately 30cm×30cm in size. The gripping area is seamlessly connected to the conveying mechanism via a T-shaped connecting plate. An infrared sensor array with an accuracy of ±1mm is installed in the area to monitor the gripping path and prevent collisions or mis-griping.

[0037] Furthermore, the storage module includes a large item storage area and a small item storage area. The large item storage area is a multi-layer stainless steel mesh rack structure, measuring approximately 40cm × 30cm × 20cm, with each mesh layer spaced 5cm apart. The bottom of the rack integrates a pressure sensor 12 with an accuracy of 0.1g, and the edges are equipped with protective covers to prevent secondary contamination of instruments. It also supports a manual removal interface. The small item storage area 62 is an inclined collection tank, measuring approximately 40cm × 25cm × 15cm, made of transparent medical-grade polycarbonate material, with a sliding angle of 15°-20°. The bottom of the tank is equipped with drainage holes and a fine filter with a 0.5mm aperture, and integrates a vibration sensor 13 to monitor the sliding dynamics and prevent blockage or accumulation.

[0038] Furthermore, the disinfection module includes UV-C ultraviolet lamps installed in the conveying mechanism, gripping area, large item storage area, and small item storage area. The wavelength is 253.7nm, and the power is 20W, 15W, 25W, and 20W respectively. All of them are in constant-on mode to ensure sterile disinfection throughout the process.

[0039] Furthermore, the system's conveying mechanism, gripping area, and storage module frame are constructed using aluminum alloy profiles and are securely fixed by L-shaped and T-shaped connecting plates. In critical areas with significant vibration, additional 2028-type corner brackets are used for reinforcement to improve structural stability and seismic performance.

[0040] Furthermore, the control unit is used to receive detection data from the RFID identification module and the visual identification module, coordinate and control the operation of the transmission mechanism, the aseptic separation module, and the actuator, synchronously output instrument type, count and abnormal alarm information, and store relevant data to support subsequent query and traceability.

[0041] Furthermore, the display module is an integrated LCD screen with a resolution of 1024×600, used to display the instrument type, count, and abnormal alarms in real time.

[0042] Example 2: like Figure 5 As shown, the automatic separation, identification, and counting method for orthopedic surgical medical devices of the present invention includes the following steps: S1. Instrument Placement and Initial Transfer: The contaminated medical instruments after orthopedic surgery are placed at the starting end of the transfer mechanism. Medical staff issue a start command through the display module, and the control unit starts the transfer mechanism. The transfer speed is adjusted to 0.1-0.5m / s according to the batch size of instruments to ensure smooth transfer. At the same time, the UV-C ultraviolet light of the disinfection module is turned on for continuous disinfection.

[0043] S2. RFID Preliminary Identification and Classification: During the operation of the conveying mechanism, the high-frequency reader array of the RFID identification module scans the instruments in real time, identifies large instruments (volume > 5cm) with pre-coded tags, and transmits the identification results (including the quantity, type and location information of large instruments) to the control unit. The control unit preliminarily distinguishes between large and small instruments based on this information.

[0044] S3. Multi-level separation process: Based on RFID identification results and preliminary scanning data from the depth camera, the control unit determines the degree of contamination and overlap of the instruments, adaptively adjusts the operating parameters of the aseptic separation module, and activates each separation component in stages: S31. Vibration separation: Activate the medical-grade electromagnetic vibrator and adjust the vibration duration to 5-15s according to the degree of instrument overlap. Vibrate at a frequency of 10-50Hz and an amplitude of 0.5-2mm to break up the dirt adhesion and physical overlap between small instruments, laying the foundation for subsequent separation.

[0045] S32. Water Flow Rinsing and Separation: Activate the pulsed sterile water atomizing nozzle array, adjust the pulse frequency according to the degree of instrument contamination, and pulse spray at a pressure of 0.1-0.3MPa for 2-5 seconds to dissolve blood clots and bone fragments, and initially clean the surface of the instruments. The inclined drainage tank and fine filter below quickly recover the waste liquid, which is then discharged after centralized treatment. The downstream warm airflow drying section is activated to dry the rinsed instruments, ensuring that the residual moisture content of the instruments is <5%.

[0046] S33. Airflow separation: Start the medical blower of the sterile airflow separation mechanism to generate sterile airflow filtered by HEPA. Adjust the wind speed to 5-10m / s according to the size and overlap of the instruments. The airflow blows small instruments in a directional manner through the alternating air outlets of the airflow guide rail (switching every 5-10 seconds) to further disperse the overlapping instruments and improve the uniformity of instrument separation.

[0047] S4. Precise Visual Recognition and Counting: The CoreXY structure depth camera of the visual recognition module moves along the XY axis to acquire images of the separated instruments from multiple angles. The image data is transmitted to the enhanced YOLOv5s algorithm processing unit. The algorithm first performs preprocessing on the image, such as denoising and enhancing contrast. Then, it accurately identifies the category, size and quantity of instruments through the small object detection layer and CBAM attention mechanism, especially focusing on the identification of tiny instruments <1cm. The recognition results are transmitted to the control unit, with a recognition accuracy of >98%.

[0048] S5. Large Item Grabbing and Classification Storage: Based on RFID identification results and visual recognition data, the control unit determines the accurate location and type of large instruments, triggering the parallel grippers of the actuator. The grippers adaptively adjust the clamping force (0.5-5N) according to the size and shape of the large instruments to avoid damaging the instrument surface. Under the real-time monitoring of the infrared sensor array, the parallel grippers accurately locate and grasp the large instruments, transferring them to the designated grid slots in the large item storage area. The pressure sensors in the large item storage area record weight changes, and the control unit compares the weight data with the identification data to verify the counting accuracy. The depth camera simultaneously scans the large item storage area to confirm the placement status of the large instruments.

[0049] S6. Small Item Collection and Confirmation: Small items are carried to the end of the conveyor mechanism and slide naturally into the small item storage area at a 15°-20° tilt angle. Vibration sensors monitor the sliding dynamics of the small items in real time. If blockage or accumulation occurs, a signal is sent to the control unit in a timely manner. The control unit triggers an alarm and adjusts the running speed of the conveyor mechanism appropriately. The depth camera scans the small item storage area and uses the enhanced YOLOv5s algorithm to perform a final count and confirmation of the small items to ensure that nothing is missed.

[0050] S7. Data Display and Anomaly Alarm: The display module shows the instrument type and counting results in real time (including the classification and total count of large and small instruments). The control unit compares the RFID identification data with the visual identification data. If there is a discrepancy in the count, an abnormal instrument position, or an unidentified instrument, an alarm is immediately triggered (including an audible alarm and a visual alarm on the display module), and the abnormal information is marked on the display module to remind medical staff to conduct manual verification.

[0051] S8. Data storage: The control unit stores data such as the type and quantity of instruments counted, count time, and abnormal records, which can be queried and traced by medical staff through the display module or by connecting to external devices.

[0052] Example 3: The system and method of this invention are not only applicable to rapid instrument verification in orthopedic operating rooms, but can also be widely extended to scenarios such as sterilization supply centers (SPD), instrument storage management, and surgical preparation rooms. When processing instruments in batches, it can significantly improve sorting and organization efficiency, reduce the risk of human contact contamination and puncture wounds, effectively prevent RSI accidents, reduce the incidence of medical errors, and shorten surgical turnaround time.

[0053] The system utilizes a depth camera to acquire real-time images of medical devices and employs enhanced YOLOv5s visual recognition technology for high-precision classification and counting. The recognition module analyzes the type, size, shape, and location of the devices. After receiving the recognition information, the control unit drives a servo motor to control the CoreXY slide rail system, precisely moving the gripping actuator above the device. Based on the actual situation, an appropriate gripping method (silicone grippers or vacuum suction cups) is selected. The actuator accurately transfers the device to a designated section of the rotating stage, achieving automatic classification and orderly placement of the devices.

[0054] In the batch processing scenario of the sterilization supply center, the system can efficiently handle the need for pre-cleaning inventory of large quantities of instruments. Through automated and intelligent operation, it significantly improves sorting and organization efficiency, reduces the risk of contamination and punctures from manual contact, and provides medical staff with a fast and reliable instrument inventory assistance solution in the end-of-line verification application in the operating room. It effectively prevents accidents involving retained surgical instruments (RSI), reduces the incidence of medical errors, and shortens surgical turnaround time.

[0055] It should be noted that the above examples are merely specific embodiments of the present invention, and the present invention is obviously not limited to the above embodiments, with many similar variations. All modifications that can be directly derived or conceived by those skilled in the art from the content disclosed in this invention should fall within the protection scope of this invention.

Claims

1. An automatic separation, identification, and inventory system for soiled medical devices in orthopedic operating rooms, characterized in that, include: A conveyor system used to transport soiled medical instruments after orthopedic surgery. The RFID identification module includes a high-frequency RFID reader array integrated below the conveyor mechanism, with a scanning range covering the entire strip surface, used to identify instruments with pre-coded tags in real time and to initially distinguish between large and small instruments. The aseptic separation module includes a medical-grade electromagnetic vibrator, a pulsed aseptic water atomizing nozzle array, and an aseptic airflow separation mechanism. It uses controllable low-frequency vibration, water flow, and airflow to break down the contaminants and physical overlap between instruments. The visual recognition module uses a multi-axis movable depth camera with a CoreXY structure to acquire images of instruments from multiple angles, and combines them with an enhanced YOLOv5s algorithm processing unit to identify the type, size and quantity of instruments in a state of overlapping dirt. The actuator, including electrically servo-driven parallel grippers, is used to grasp large instruments identified by RFID. The storage module includes a large item storage area and a small item storage area; the large item storage area is located on the side of the actuator and is used for the dedicated temporary storage and inventory of large instruments; the small item storage area is located at the end of the conveyor belt and is used for the collection and final processing of small instruments. The disinfection module includes ultraviolet lamps installed in the conveyor mechanism, the large item storage area, and the small item storage area; The control unit is used to receive detection data from the RFID identification module and the vision identification module, coordinate and control the operation of the transmission mechanism, the aseptic separation module and the actuator, and synchronously output instrument type, count and abnormal alarm information; The display module is an integrated LCD screen used to display the instrument type, count, and abnormal alarms in real time.

2. The automatic separation, identification, and inventory system for soiled medical devices in orthopedic operating rooms according to claim 1, characterized in that, The conveying mechanism uses a medical-grade stainless steel conveyor belt coated with a chemically resistant polytetrafluoroethylene coating; the operating speed of the conveying mechanism is adjustable between 0.1-0.5 m / s, and the edges are equipped with protective railings and inclined guide structures.

3. The automatic separation, identification, and inventory system for soiled medical devices in orthopedic operating rooms according to claim 1, characterized in that, The medical-grade electromagnetic vibrator has a frequency of 10-50Hz, an amplitude of 0.5-2mm, and a power of <30W. It is elastically connected to the stainless steel belt of the transmission mechanism through silicone shock-absorbing pads and is located in the middle and rear section of the transmission mechanism. It breaks the dirt adhesion and physical overlap between small instruments through controllable low-frequency vibration, and provides a uniform spreading state to assist the YOLOv5s algorithm in accurately identifying small objects. The pulsed sterile water atomizing nozzle array is made of PTFE material, with a pressure of 0.1-0.3MPa and a pulse duration of 2-5s. It uses sterile pure water filtered through a 0.22-0.25μm filter and is located downstream of the vibration section of the conveying mechanism. The pulsed sterile water is used to rinse and dissolve blood clots / bone fragments, and the shear force of the water flow is used to assist in the separation of small instruments, while also initially cleaning surface dirt.

4. The automatic separation, identification, and inventory system for soiled medical devices in orthopedic operating rooms according to claim 3, characterized in that, The sterile airflow separation mechanism includes a medical blower and an airflow guide rail located on both sides of the conveying mechanism. The airflow guide rail is made of medical-grade ABS plastic and designed with a gradient cross-section channel. The air outlet adopts an alternating activation mode, and the wind speed adjustment range is 5-10m / s. The airflow passes through a built-in HEPA filter to ensure sterile output and is monitored in real time by an integrated wind speed sensor. This directionally blows away the dirt and dirt adhering to and overlapping of small instruments, improving the separation uniformity and counting accuracy of subsequent visual recognition, while avoiding instrument scattering or secondary contamination.

5. The automatic separation, identification, and inventory system for soiled medical devices in orthopedic operating rooms according to claim 4, characterized in that, The alternating activation mode switches between the left and right sides every 5-10 seconds. The size of the small instrument is <1cm.

6. The automatic separation, identification, and inventory system for soiled medical devices in orthopedic operating rooms according to claim 1, characterized in that, The depth camera is driven by a stepper motor to link the X and Y axes and acquire instrument images from multiple angles. The enhanced YOLOv5s algorithm adds a small object detection layer and a CBAM attention mechanism. It is trained on a dedicated dataset of blood-stained orthopedic instruments and is used to determine the type, size and quantity of instruments under occlusion.

7. The automatic separation, identification, and inventory system for soiled medical devices in orthopedic operating rooms according to claim 1, characterized in that, The parallel grippers have a clamping force of 0.5-5N and a response time of <0.5s. The ends of the grippers are covered with flexible medical silicone pads, supporting 360° rotation and a 10cm lifting stroke.

8. The automatic separation, identification, and inventory system for soiled medical devices in orthopedic operating rooms according to claim 1, characterized in that, The large item storage area features a multi-layer stainless steel mesh frame structure, with a pressure sensor integrated at the bottom of the tank with an accuracy of 0.1g to automatically record weight changes and verify counting accuracy. A constantly lit ultraviolet light is installed at the top, and a multi-axis movable depth camera with a CoreXY structure extends to cover the large item storage area. Combined with the YOLOv5s algorithm, it processes possible residual dirt images, accurately records the size, quantity, and type of large instruments, and updates the data to the display screen in real time for easy verification by medical staff.

9. The automatic separation, identification, and inventory system for soiled medical devices in orthopedic operating rooms according to claim 1, characterized in that, The small item storage area is an inclined collection trough made of transparent medical-grade polycarbonate material, with a sliding angle of 15°-20°. The bottom of the trough is equipped with drainage holes and a fine filter with a pore size of 0.5mm. The small item storage area is connected to the conveyor belt frame of the conveying mechanism via a T-shaped connecting plate and is reinforced at the bottom with corner brackets. A multi-axis movable depth camera with a CoreXY structure extends to cover the small item storage area and, combined with the enhanced YOLOv5s algorithm, processes blood-stained overlapping images to accurately record the size, quantity, and type of small instruments. The data is simultaneously displayed on the screen and an alarm is triggered for potential omissions. A constantly lit ultraviolet light is installed on the top. An integrated vibration sensor monitors the sliding dynamics to prevent blockage or accumulation.

10. A method for automatic separation, identification, and counting of orthopedic surgical medical devices, characterized in that, The method is implemented based on the automatic separation, identification, and counting system according to any one of claims 1-9, and includes the following steps: S1. Instrument placement and initial transfer: Place the soiled medical instruments from orthopedic surgery at the starting end of the transfer mechanism. The control unit starts the transfer mechanism to transfer the instruments smoothly at a certain speed. The ultraviolet lamp of the disinfection module is turned on simultaneously for continuous disinfection. S2. Preliminary RFID Identification and Classification: During the operation of the conveying mechanism, the RFID identification module scans the instruments in real time through a high-frequency reader array, identifies large instruments with pre-coded tags, and transmits the identification results to the control unit to initially distinguish between large and small instruments. S3. Multi-level separation process: Based on RFID identification results and preliminary scanning data from the depth camera, the control unit activates the aseptic separation module in stages. S31. Vibration separation: Activate the medical-grade electromagnetic vibrator to break up the dirt adhesion and physical overlap between small instruments; S32. Water rinsing and separation: Activate the pulsed sterile water atomizing nozzle array to dissolve blood clots and bone fragments, preliminarily clean the surface of the instruments, and recover the waste liquid through the inclined drainage tank and fine filter screen below. S33. Airflow dispersion and separation: Activate the sterile airflow separation mechanism, and blow small instruments at a certain wind speed through the alternating air outlets of the airflow guide rail to further disperse overlapping instruments; S4. Precise Visual Recognition and Counting: The CoreXY structure of the visual recognition module has a multi-axis movable depth camera that moves along the XY axis to acquire images of the separated instruments from multiple angles. The image data is processed by the enhanced YOLOv5s algorithm to accurately identify the type, size and quantity of the instruments and transmit the results to the control unit. S5. Large item grasping and classification storage: The control unit triggers the parallel gripper of the actuator to locate and grasp large instruments based on RFID identification results and visual recognition data, and transfers them to the designated grid slots in the large item storage area. The pressure sensor in the large item storage area records weight changes to verify the counting accuracy, and the depth camera scans simultaneously to confirm the placement status. S6. Small item collection and confirmation: Small items are carried to the end of the conveyor and slide naturally into the small item storage area at an angle of 15°-20°. Vibration sensors monitor the sliding dynamics to prevent blockage or accumulation. A depth camera scans the small item storage area and combines it with a visual recognition algorithm to complete the final counting and confirmation. S7. Data Display and Anomaly Alarm: The display module shows the instrument type and counting results in real time. The control unit compares the RFID identification data with the visual identification data. If there is a discrepancy or anomaly in the counting, an alarm is triggered immediately.

Images