Operating room medical waste intelligent classification treatment system and method thereof

Abstract

The invention relates to the technical field of medical information systems, in particular to an intelligent classification treatment system and method for medical waste in an operating room, and the system comprises a waste recognition module which is used for recognizing the medical waste through a computer vision algorithm; determining a waste category based on the recognition result; the electronic guiding module is in communication connection with the waste identification module and is used for receiving the waste category information sent by the waste identification module; based on the waste category information, generating an electronic guide for guiding the medical staff to store the waste; the waste weighing and recording module is in communication connection with the electronic guiding module and is used for calculating and recording the weight and the capacity of the waste; waste weight and volume data are generated; the RFID reading module is in communication connection with the waste weighing and recording module and is used for automatically reading RFID electronic tag information; the source tracing of the waste is realized based on the RFID electronic tag information; the system can automatically complete the whole process from waste identification, classification guidance, weight recording to information tracking.

Description

Technical Field

[0001] The invention relates to the technical field of medical information systems, in particular to an intelligent classification and processing system for medical waste in an operating room and a method thereof. Background Art

[0002] With the rapid development of modern medical technology and the increasing demand for medical services, the management of medical waste in operating rooms has become a key issue in hospital operations. Traditional medical waste treatment methods mainly rely on manual classification and management, which is not only inefficient but also has high operational risks.

[0003] At present, some hospitals have begun to try to introduce automated equipment to assist in the classification and treatment of medical waste. For example, special garbage bags with barcodes are used, or simple weight sensors are used to record the amount of waste. However, these methods still have many shortcomings. First, they cannot accurately identify and classify different types of medical waste, especially when dealing with mixed waste. Secondly, these systems lack real-time monitoring and tracking capabilities, making it difficult to effectively manage the entire waste treatment process. Furthermore, existing systems are often difficult to adapt to emergencies, such as emergency treatment needs during epidemic outbreaks.

[0004] In addition, existing medical waste treatment systems also have obvious deficiencies in data collection and analysis. Most systems can only provide basic statistical data and cannot conduct in-depth data mining and analysis, which limits the ability of hospital management to optimize waste treatment processes and formulate long-term strategies.

[0005] Finally, the existing systems generally lack intelligent and humanized design. Medical staff often feel cumbersome and confused when using these systems, which not only reduces work efficiency but also increases the risk of operational errors.

[0006] In view of the above problems, there is an urgent need for an intelligent classification and treatment system and method for medical waste in operating rooms that can perform intelligent identification, accurate classification, full-process tracking, real-time monitoring, data analysis, and emergency handling capabilities. Summary of the invention

[0007] The intelligent classification and treatment system and method for medical waste in operating rooms of the present invention are proposed to solve the above technical problems. The present invention integrates advanced computer vision technology, Internet of Things technology, artificial intelligence algorithm and big data analysis technology to build a comprehensive, efficient and intelligent medical waste management system.

[0008] The present invention proposes an intelligent classification and treatment system for medical waste in an operating room, comprising:

[0009] Waste identification module for:

[0010] Identify medical waste through computer vision algorithms;

[0011] Based on the identification results, determine the waste category;

[0012] The electronic guidance module is in communication with the waste identification module and is used to:

[0013] receiving waste category information sent by the waste identification module;

[0014] Based on the waste category information, generate electronic instructions for medical staff to place the waste;

[0015] The waste weighing recording module is in communication with the electronic guidance module and is used to:

[0016] Calculate and record waste weights and volumes;

[0017] Generate waste weight and volume data;

[0018] The RFID reading module is in communication with the waste weighing recording module and is used to:

[0019] Automatically read RFID electronic tag information;

[0020] Based on the RFID electronic tag information, the waste can be traced back to its source;

[0021] Among them, the system automatically performs classified recycling of various types of waste according to a preset emergency mode; the emergency mode includes an epidemic prevention emergency mode and a normal emergency mode.

[0022] Preferably, the electronic guidance module comprises:

[0023] A screen display unit, used for displaying waste classification guidance information;

[0024] Indicator lamp unit for:

[0025] When the capacity corresponding to the classification function is not full, instruct and guide the staff to put the waste to be classified into the classification system;

[0026] When the capacity corresponding to the classification function is full, the red light flashes to remind the staff that the classification is complete.

[0027] Preferably, the waste identification module uses a preset multi-type detection algorithm to perform multi-target identification analysis on the types of medical waste, including:

[0028] Extract feature pixels through RGB color space to classify medical waste;

[0029] Taking multiple sets of RGB as input, it outputs the identification box information corresponding to the medical waste category, which is used to guide the classification operation and calculate the waste capacity.

[0030] Preferably, the waste weighing and recording module comprises:

[0031] Electronic weighing unit to detect the quantity of waste delivered before sorting;

[0032] Label encoding unit, used to:

[0033] Generate electronic tags according to different wastes;

[0034] Place the classified waste into the corresponding waste disposal bin;

[0035] IoT interaction unit for:

[0036] Get the classification processing results;

[0037] RFID electronic tags are bound to classified waste to record the weight, collection time and processing process information of classified waste.

[0038] Preferably, the waste weighing and recording module is divided into three types of waste bags according to the waste treatment process:

[0039] Burnable waste bags with a first label code to record the weight of recyclable waste and the time of collection;

[0040] Non-burnable waste bags with a second label code to record the weight and collection time of non-recyclable waste;

[0041] Infectious waste is bagged with a third label coded to record infectious waste weight and collection time.

[0042] Preferably, the Internet of Things interaction unit includes:

[0043] IoT backend server for processing classified waste treatment process information;

[0044] Control chip, used to calculate the weight of classified waste and issue label codes;

[0045] Smart cloud terminal for real-time viewing of current waste weight and historical waste treatment results.

[0046] Preferably, the RFID reading module comprises:

[0047] RFID electronic tags are used to identify each category of medical waste;

[0048] RFID reader / writer unit for:

[0049] Read electronic tag information through RFID;

[0050] Automatically record waste treatment processes;

[0051] Wherein, the RFID reader unit includes:

[0052] A first RFID reading unit, used to obtain the recyclable waste processing result;

[0053] The second RFID reading unit is used to obtain the non-recyclable waste treatment result;

[0054] The third RFID reading unit is used to obtain the infectious waste treatment result.

[0055] Preferably, the RFID electronic tag is disposable and is scanned during recycling to trace the historical information of the waste after treatment.

[0056] Preferably, the operation process of the epidemic prevention emergency mode includes:

[0057] Step 1: If an abnormal epidemic situation is detected, activate the epidemic prevention emergency mode;

[0058] Step 2: Generate electronic instructions to classify various types of medical waste according to the preset emergency storage process;

[0059] Step 3: The classified wastes are put into the corresponding waste treatment cylinders through their respective channels; the classified waste treatment results are summarized to the smart cloud terminal through the IoT backend server;

[0060] Step 4: When the waste capacity of each type of waste treatment cylinder reaches the threshold or the epidemic ends, statistical analysis of each treatment result is performed and the normal emergency mode is returned.

[0061] The medical waste intelligent classification and treatment method based on the operating room medical waste intelligent classification and treatment system comprises:

[0062] Identify medical waste through computer vision algorithms;

[0063] Guide staff to perform classification operations through electronic guidance modules;

[0064] Calculate and record waste weight and waste volume through the waste weighing and recording module;

[0065] The RFID reader module automatically reads the RFID electronic tag information to achieve waste tracking and tracing;

[0066] Automatically sort and recycle various types of waste.

[0067] The beneficial effects of the present invention are mainly reflected in the following aspects:

[0068] First, at the macro level, the present invention realizes the intelligent management of the whole process of medical waste treatment. Through the coordinated work of the waste identification module, the electronic guidance module, the waste weighing and recording module, and the RFID reading module, the system can automatically complete the entire process from waste identification, classification guidance, weight recording to information tracking. This not only greatly improves the efficiency of waste treatment, but also significantly reduces the risk of human error.

[0069] Secondly, the present invention solves the shortcomings of traditional systems in dealing with emergencies. By setting up epidemic prevention emergency mode and general emergency mode, the system can flexibly respond to various special situations and ensure that efficient and safe waste treatment processes can be maintained in emergency situations. This flexibility and adaptability is a major breakthrough of the present invention compared to the prior art.

[0070] At the micro level, each module of the present invention has unique innovations. For example, the waste identification module uses an improved deep learning algorithm to accurately identify various types of medical waste, even in a complex operating room environment. The electronic guidance module greatly reduces the cognitive burden of operators and improves the accuracy of classification through multimodal guidance.

[0071] In addition, the present invention also has significant advantages in data management and analysis. Through the Internet of Things technology and big data analysis, the system can monitor the waste treatment status in real time, generate detailed analysis reports, and provide decision support for hospital management. This not only helps to optimize the daily waste treatment process, but also provides data support for long-term strategic planning.

[0072] Finally, the system design of the present invention fully considers the convenience of human-computer interaction. The intuitive user interface and intelligent operation guidance greatly reduce the use threshold of medical staff, so that even in a busy operating room environment, staff can easily complete waste sorting and processing tasks.

[0073] In summary, the operating room medical waste intelligent classification and treatment system and method of the present invention comprehensively solve various problems existing in the existing medical waste treatment system through innovative technical integration and intelligent design. It not only improves the efficiency and safety of waste treatment, but also makes important contributions to the environmental protection and sustainable development of hospitals. The promotion and application of this system will bring revolutionary changes to modern hospital management. BRIEF DESCRIPTION OF THE DRAWINGS

[0074] Figure 1 It is the overall logic block diagram of the system of the present invention. DETAILED DESCRIPTION

[0075] See also Figure 1The present invention provides an intelligent classification and treatment system for medical waste in operating rooms and a method thereof, so as to solve the problems of inaccurate classification and low treatment efficiency of medical waste in operating rooms in the prior art. The present invention is described in detail below in conjunction with specific embodiments.

[0076] like Figure 1 As shown, the intelligent classification and treatment system for medical waste in operating rooms of the present invention includes a waste identification module 1, an electronic guidance module 2, a waste weighing and recording module 3 and an RFID reading module 4.

[0077] The waste identification module 1 is used to identify medical waste through a computer vision algorithm. Specifically, the waste identification module 1 uses a deep learning method to extract features and classify medical waste images through a convolutional neural network. In a preferred embodiment of the present invention, the waste identification module 1 uses the YOLOv5 algorithm for target detection and classification. The core idea of ​​the algorithm is to divide the image into S×S grids, each grid is responsible for predicting B bounding boxes, and each bounding box contains 5 prediction values: x, y, w, h and confidence. Among them, x and y represent the center coordinates of the bounding box, w and h represent the width and height of the bounding box, and the confidence reflects whether the target is contained in the predicted box and the accuracy of the position.

[0078] For each grid, the probabilities of C categories need to be predicted. Therefore, the number of predicted values ​​for each grid is B×5+C. The size of the predicted output tensor for the entire image is S×S×(B×5+C). The prediction process can be expressed as:

[0079]

[0080] Among them, P(Object) represents the probability that the bounding box contains the object, Represents the intersection-over-union ratio of the predicted box and the real box, P(Class i |Object) represents the probability of belonging to the i-th category under the condition of having a target.

[0081] In this way, the waste identification module 1 can accurately identify various types of medical waste in the operating room, such as syringes, gauze, scalpels, etc., and determine their locations and categories.

[0082] The electronic guidance module 2 is connected to the waste identification module 1 for guiding the medical staff to store the waste according to the identification result. The electronic guidance module 2 includes a screen display unit and an indicator light unit. The screen display unit can be an LCD display screen for displaying the type and storage location of the waste. The indicator light unit is composed of a plurality of LED lights, and LED lights of different colors correspond to different types of waste.

[0083] In actual application, when the waste identification module 1 identifies a piece of medical waste, the electronic guidance module 2 will immediately display the type of the waste and the location where it should be stored on the screen, and the corresponding LED indicator will light up. For example, when a syringe is identified, the screen may display "Syringe-Please put it in the red trash can" and the red LED light will light up. This intuitive guidance method greatly improves the efficiency and accuracy of medical staff in sorting waste.

[0084] It is worth noting that when the storage container of a certain type of waste is full, the indicator light unit will remind the staff by flashing a red light. This function can effectively prevent waste from overflowing and keep the operating room clean and safe.

[0085] The waste weighing and recording module 3 is connected to the electronic guidance module 2 for calculating and recording the weight and volume of the waste. In one embodiment of the present invention, the waste weighing and recording module 3 uses a high-precision electronic scale and a volume sensor. The electronic scale has an accuracy of up to 0.1g and can accurately record the weight of each piece of medical waste. The volume sensor measures the volume of the waste through 3D scanning technology with an accuracy of up to 1cm. 3 .

[0086] The Waste Weighing Recording Module 3 not only records the weight and volume of individual waste, but also accumulates the total weight and volume of each type of waste. This data is very important for hospital management and can be used to optimize waste treatment processes, estimate treatment costs, and develop more effective waste reduction strategies.

[0087] The RFID reading module 4 is connected to the waste weighing and recording module 3 for automatically reading the RFID electronic tag information to trace the waste. In the present invention, each medical waste container is affixed with an RFID tag. These tags contain information such as the type of waste, the time of generation, and the processing status. The RFID reading module 4 adopts UHF (ultra-high frequency) RFID technology, with a reading distance of up to 1-10 meters, a fast reading speed, and can read multiple tags at the same time.

[0088] Through RFID technology, this system can realize the full tracking of medical waste. From waste generation, classification, storage to final treatment, every link can be recorded and traced. This not only improves the transparency and controllability of medical waste treatment, but also provides strong support for the hospital's quality management and risk control.

[0089] The system of the present invention also has an emergency mode function, including an epidemic prevention emergency mode and a normal emergency mode. When a special situation (such as an epidemic outbreak) is detected, the system will automatically switch to the corresponding emergency mode and adjust the waste classification and treatment process to deal with the emergency. For example, in the epidemic prevention emergency mode, the system will treat all wastes as high-risk infectious wastes and uniformly treat them strictly.

[0090] In general, the intelligent classification and treatment system for medical waste in the operating room of the present invention realizes the intelligent identification, accurate classification, full-process tracking and emergency treatment of medical waste by combining advanced technologies such as computer vision, Internet of Things, and artificial intelligence. This not only greatly improves the efficiency and accuracy of medical waste treatment, but also makes important contributions to the safety management and environmental protection of the hospital. In a preferred embodiment of the present invention, the electronic guidance module 2 includes a screen display unit 21 and an indicator light unit 22. The screen display unit 21 can be a high-resolution LCD touch screen with a size of 10 inches and a resolution of 1920×1080. Such a configuration can clearly display waste classification guidance information, including details such as waste pictures, category names, and treatment methods. Preferably, the screen display unit 21 adopts a waterproof and dustproof design to meet the strict hygiene requirements of the operating room.

[0091] The indicator light unit 22 is composed of multiple groups of LED lights, each group corresponding to a waste category. In one embodiment of the present invention, the indicator light unit 22 includes four colors of LED lights: red, yellow, green, and blue, corresponding to infectious waste, pathological waste, injury waste, and pharmaceutical waste, respectively. Each color of LED light is divided into two states: always on and flashing. When the capacity corresponding to the classification function is not full, the LED light of the corresponding color remains always on, instructing and guiding the staff to put the waste to be classified into the corresponding classification container.

[0092] The system of the present invention specifically takes into account the problem of overflowing containers. When the container capacity of a certain type of waste reaches a preset threshold (e.g., 90%), the LED light of the corresponding color in the indicator light unit 22 will begin to flash, reminding the staff that the classification of this category is about to be completed. Preferably, the flashing frequency can be set to 1 Hz, which can attract attention without causing visual fatigue. When the container is completely full (reaching 100% capacity), the LED light will speed up the flashing frequency to 2 Hz, accompanied by an alarm sound from a buzzer, to ensure that the staff can replace the container in time.

[0093] The waste identification module 1 of the present invention adopts an advanced multi-type detection algorithm, which can perform multi-target identification analysis on the types of medical waste. In a specific embodiment, the system uses an improved Faster R-CNN (Regions with Convolutional Neural Networks) algorithm. The core idea of ​​the algorithm is to extract features through a convolutional neural network, then use a region proposal network (RPN) to generate candidate regions, and finally classify and regress these candidate regions into bounding boxes.

[0094] Specifically, the system of the present invention first extracts feature pixels through the RGB color space. Preferably, the system uses the HSV (hue, saturation, value) color space for conversion, because the HSV space is closer to the way humans perceive color, which is conducive to improving recognition accuracy. The conversion formula is as follows:

[0095]

[0096] v=max

[0097] Where R, G, and B represent the values ​​of the red, green, and blue channels, respectively, and the range is [0,1]. H represents hue, S represents saturation, and V represents brightness. After feature extraction, the system of the present invention uses multiple sets of RGB values ​​as input and outputs the identification box information corresponding to the medical waste category through the trained neural network model. This information is not only used to guide the classification operation, but also used to calculate the waste capacity. The position and size of the identification box can be expressed by the following formula:

[0098] x=x c +w×δx

[0099] y=y c +h×δy

[0100] w′=w×e δw

[0101] h′=h×e δh

[0102] Among them, (x, y) is the center coordinate of the recognition box, w' and h' are the width and height of the recognition box, (x c ,y c ) is the preset anchor box center coordinate, w and h are the preset anchor box width and height, δx, δy, δw and δh are the adjustment parameters predicted by the network.

[0103] The waste weighing and recording module 3 of the present invention comprises an electronic weighing unit 31, a label encoding unit 32 and an Internet of Things interaction unit 33. The electronic weighing unit 31 adopts a high-precision strain gauge weighing sensor with an accuracy of up to 0.1g. This accuracy can accurately detect the delivery quantity of classified waste before classification, and even lightweight medical waste such as cotton swabs, gauze, etc. can be accurately measured.

[0104] The label encoding unit 32 is responsible for encoding electronic labels according to different waste generation. In one embodiment of the present invention, the encoding uses a 128-bit unique identifier, which includes information such as waste type, generation time, and source department. For example, a typical encoding may be: 0101-20230611-0930-SURG-I NJ-001. This encoding means: 01 (infectious waste)-20230611 (generation date)-0930 (generation time)-SURG (surgery)-INJ (syringe)-001 (serial number).

[0105] After the label coding unit 32 generates the code, it will instruct the robot arm or conveyor belt to put the classified waste into the corresponding waste treatment cylinder. This process is fully automatic, avoiding the risk of cross infection caused by manual operation.

[0106] The IoT interaction unit 33 is responsible for obtaining the classification processing results and binding them to the classified waste through RFID electronic tags. It records detailed information such as the weight, collection time, and processing flow of the classified waste. This information is uploaded to the central database in real time through the 5G network, supporting hospital managers to conduct real-time monitoring and data analysis.

[0107] The system of the present invention uses three types of waste bags in the waste treatment process, namely, burnable waste bags, non-burnable waste bags and infectious waste bags. Each waste bag has its own specific label code for recording the weight and collection time of the corresponding type of waste.

[0108] Burnable garbage bags are red and have a first label code. This type of waste bag is mainly used to collect burnable medical waste such as gauze and cotton swabs. The format of the first label code is "B-YYYY-MM-DD-HHNN-WWW", where B stands for burnable, YYYY-MM-DD is the date, HHNN is the time, and WWW is the weight (unit: grams).

[0109] Non-burnable garbage bags are yellow and have a second label code. This type of waste bag is used to collect non-burnable medical waste such as discarded metal instruments and glassware. The second label code format is "N-YYYY-MM-DD-HHNN-WWW", where N stands for non-burnable.

[0110] Infectious waste bags are black and have a third label code. This type of waste bag is used to collect infectious medical waste, such as disposable medical supplies used by infected patients. The third label code format is "I-YYYY-MM-DD-HHNN-WWW", where I stands for infectious.

[0111] Through this sophisticated classification and coding system, the present invention can achieve accurate tracking and management of medical waste, greatly improving the safety and efficiency of medical waste treatment. In a preferred embodiment of the present invention, the Internet of Things interaction unit 33 includes an Internet of Things background server 331, a control chip 332 and an intelligent cloud terminal 333. These three components work together to form an efficient and reliable medical waste management system.

[0112] The IoT backend server 331 is the core of the entire system and is responsible for processing the treatment process information of classified waste. The server adopts a distributed architecture to ensure the high availability and scalability of the system. In a specific embodiment, the server uses the Apache Hadoop framework for big data processing, which can efficiently process and analyze a large amount of waste treatment data. The server is also equipped with a special data encryption module, using the AES-256 encryption algorithm to ensure the security of sensitive medical data.

[0113] The control chip 332 is a multifunctional embedded system that is responsible for calculating the weight of classified waste and issuing label codes. The chip adopts the ARM Cortex-M4 architecture, with a main frequency of up to 180MHz and powerful computing power. In order to improve the weighing accuracy, the control chip 332 uses the Kalman filter algorithm to process the sensor data. Its mathematical model can be expressed as:

[0114] x k =Ax k-1 +Bu k +w k

[0115] z k =Hx k +v k

[0116] Among them, x k is the state vector at time k, u k is the control vector, z k is the observation vector, w k and $v_k v k They are process noise and observation noise, respectively, and they are all assumed to be Gaussian white noise. A, B, and H are the state transfer matrix, control matrix, and observation matrix, respectively.

[0117] The intelligent cloud terminal 333 provides an intuitive user interface through which medical staff can view the current waste weight and historical waste treatment results in real time. The terminal adopts a responsive design and can be well displayed on devices of different sizes. Preferably, the intelligent cloud terminal 333 also integrates an artificial intelligence assistant, which can answer users' queries through natural language processing technology and provide intelligent decision support.

[0118] The RFID reading module 4 of the present invention comprises an RFID electronic tag 41 and an RFID reader-writer unit 42. The RFID electronic tag 41 is used to identify each classification of medical waste, while the RFID reader-writer unit 42 is responsible for reading the information of these tags and automatically recording the waste treatment process.

[0119] RFID electronic tags 41 use ultra-high frequency (UHF) technology and operate at a frequency of 860-960MHz. This frequency range allows the tag to have a longer reading distance (usually up to 3-6 meters) and faster data transmission speed. Each tag contains a unique 96-bit EPC (Electronic Product Code) that can store a wealth of waste information. The tag has a storage capacity of 512 bits, which is enough to record key information such as the type, weight, generation time, and processing status of the waste.

[0120] The RFID reader / writer unit 42 includes three subunits: a first RFID reading unit 421, a second RFID reading unit 422 and a third RFID reading unit 423. The three reading units are responsible for obtaining the processing results of recyclable waste, non-recyclable waste and infectious waste respectively.

[0121] The first RFID reading unit 421 is specifically used to obtain the recyclable waste treatment results. It uses a directional antenna, and the reading range is precisely controlled within the recyclable waste treatment area, effectively avoiding confusion with other types of waste information. The reading sensitivity of this unit can reach -80dBm, and even tags in metal containers can be accurately read.

[0122] The second RFID reading unit 422 is responsible for obtaining the non-recyclable waste processing results. Considering that the non-recyclable waste may contain metal objects, the unit adopts a special anti-metal interference technology. Specifically, it uses a circularly polarized antenna, which can read tags at all angles, greatly improving the reading success rate.

[0123] The third RFID reading unit 423 is used to obtain the results of infectious waste treatment. In view of the particularity and danger of infectious waste, the unit is equipped with additional safety measures. For example, it adopts a waterproof and dustproof design (protection level up to IP67) and can work normally during the disinfection process. In addition, the unit also has an automatic alarm function, which will immediately alert the management personnel when untreated infectious waste is detected.

[0124] In one embodiment of the present invention, the RFID electronic tag 41 is designed to be disposable. This design has several significant advantages: first, it avoids the risk of cross-infection that may be caused by the reuse of tags; second, disposable tags are usually low-cost and suitable for large-scale use; finally, disposable tags can be finally disposed of together with waste, simplifying the management process.

[0125] When the waste is finally recycled, the RFID reader unit 42 will scan the tag for the last time. This scan not only records the final disposal time and method of the waste, but also traces all the historical information of the waste. For example, the system can query the time of generation of the waste, the source department, the intermediate processing process, etc. This full tracking capability is of great significance for the quality control of medical waste management and the retrospective investigation of potential problems.

[0126] Through the above design, the RFID reading module 4 of the present invention realizes the full life cycle management of medical waste, greatly improving the safety, traceability and management efficiency of medical waste treatment. The system of the present invention is designed with special consideration of the emergency treatment needs under special circumstances, so an epidemic prevention emergency mode is set. The operation process of this mode includes four key steps, which are designed to quickly and efficiently respond to possible emergencies such as epidemics.

[0127] The first step is to detect abnormal epidemic conditions and start the epidemic prevention emergency mode. The system of the present invention detects abnormal epidemic conditions by real-time monitoring of multiple data sources. These data sources include but are not limited to patient visit data within the hospital, epidemic warning information issued by the local CDC, and analysis results of social media data by intelligent algorithms. When the system detects a potential epidemic risk, it will automatically trigger the epidemic prevention emergency mode. Specifically, the system uses an anomaly detection algorithm based on machine learning. The algorithm uses a long short-term memory network (LSTM) to analyze time series data.

[0128] The second step is to generate electronic guidance and classify various types of medical waste according to the preset emergency storage process. In the epidemic prevention emergency mode, the system of the present invention will adjust the conventional waste classification standards and adopt a more stringent classification scheme. For example, all waste that may have been in contact with the source of infection will be classified as high-risk infectious waste and require special treatment. The system will send detailed classification instructions to medical staff through the electronic guidance module 2. These instructions will be displayed on the screen display unit 21 in the form of pictures and texts, and the indicator light unit 22 will use a specific light mode (such as rapid flashing red) to remind staff to pay attention to the current special situation.

[0129] The third step involves the classification and information collection of waste. In this step, the classified waste is put into the corresponding waste treatment cylinder through the corresponding dedicated channels. These dedicated channels are specially designed to minimize the risk of contact during the waste treatment process. Each waste treatment cylinder is equipped with a highly sensitive sensor that can monitor the weight and volume of the waste in real time. When the waste is put in, the processing results will be immediately summarized to the smart cloud terminal 333 through the IoT background server 331. This process is fully automatic and does not require human intervention, thereby minimizing the risk of infection for staff.

[0130] Preferably, the system of the present invention also uses an intelligent scheduling algorithm based on deep reinforcement learning in this step. The algorithm can dynamically adjust the allocation strategy of each treatment cylinder according to the current waste treatment situation to optimize the overall treatment efficiency. The core of the algorithm is a deep Q network (DQN), and its update rule can be expressed as:

[0131]

[0132] Among them, Q(s t ,a t ) is in state s t Take action a t The value function, r t is the immediate reward, γ is the discount factor, and α is the learning rate.

[0133] The fourth step is the statistical analysis of the treatment results and mode switching. When the waste capacity of each type of waste treatment barrel reaches the preset threshold (usually set to 90% of the capacity), or when the epidemic ends, the system will automatically perform statistical analysis on each treatment result. This analysis process includes but is not limited to the following aspects: the amount of various types of waste generated, the treatment efficiency, the frequency of abnormal situations, etc. The system will generate a detailed analysis report to provide decision-making support for hospital management. After the analysis is completed, if the epidemic has been brought under control, the system will automatically return to the normal emergency mode and resume the normal waste treatment process.

[0134] It is worth noting that during the entire operation of the epidemic prevention emergency mode, the system of the present invention always maintains real-time data interaction with other hospital systems. For example, it will synchronize with the hospital's patient management system to obtain the latest epidemic-related information in a timely manner; at the same time, it will also provide real-time data on waste treatment to the hospital's resource allocation system to assist the hospital in overall resource planning.

[0135] The present invention also provides a medical waste intelligent classification and treatment method based on the above-mentioned operating room medical waste intelligent classification and treatment system. The method includes five main steps, each of which fully utilizes the advanced functions of the system of the present invention to achieve full-process intelligent management of medical waste treatment.

[0136] First, the method uses computer vision algorithms to identify medical waste. Specifically, the system uses an improved YOLOv5 algorithm, which adds an attention mechanism on the basis of the standard YOLOv5, and can better handle small-target, multi-category recognition tasks such as medical waste. The core idea of ​​the algorithm is to add a spatial attention module and a channel attention module to the feature extraction network, which can be expressed as:

[0137]

[0138] Among them, F is the original feature map, M c and M s They are channel attention and spatial attention operations, This improvement improves the recognition accuracy of the algorithm by about 5% compared to the standard YOLOv5, and performs particularly well in special scenarios such as medical waste.

[0139] Secondly, the method guides the staff to perform the classification operation through the electronic guidance module. This step makes full use of the screen display unit 21 and the indicator light unit 22 in the system of the present invention. The system will display detailed classification guidance on the screen according to the recognition results, and provide intuitive visual guidance through LED lights of different colors and flashing modes. This multi-modal guidance method greatly reduces the probability of operation errors and improves classification efficiency.

[0140] The third step is to calculate and record the weight and volume of the waste through the waste weighing and recording module. This method uses innovative 3D reconstruction technology in this step, capturing stereoscopic images of the waste through cameras at multiple angles, and then using a deep learning algorithm for 3D reconstruction to more accurately estimate the volume of the waste. This method can better reflect the actual treatment needs of the waste than the traditional simple weighing method, especially for large but light waste (such as used protective clothing).

[0141] The fourth step is to automatically read the RFID electronic tag information through the RFID reading module to achieve waste tracking and tracing. This method realizes contactless information reading throughout the process. The system automatically reads the RFID tag at each stage of waste generation, classification, transportation, and treatment, and uploads the information to the central database in real time. This not only ensures the traceability of the waste treatment process, but also provides valuable data support for subsequent data analysis and process optimization.

[0142] Finally, the system automatically performs classified recycling of various types of waste. In this step, the method uses an intelligent scheduling algorithm that can dynamically adjust the recycling order and route according to the current treatment status of various types of waste to maximize treatment efficiency and minimize the risk of cross-infection. This algorithm is based on multi-agent reinforcement learning, and each waste treatment unit is regarded as an agent, which optimizes the overall recycling strategy through collaborative learning.

[0143] In general, the intelligent classification and treatment method for medical waste provided by the present invention realizes the intelligent management of the whole process of medical waste treatment by combining advanced computer vision technology, Internet of Things technology and artificial intelligence algorithm. This not only greatly improves the treatment efficiency and accuracy, but also significantly reduces the workload and infection risk of medical staff, making important contributions to the safe operation of hospitals and environmental protection.

[0144] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. Intelligent classification and treatment system for medical waste in operating room, characterized by: include: Waste identification module for: Identify medical waste through computer vision algorithms; Based on the identification results, determine the waste category; The electronic guidance module is in communication with the waste identification module and is used to: receiving waste category information sent by the waste identification module; Based on the waste category information, generate electronic instructions for medical staff to place the waste; The waste weighing recording module is in communication with the electronic guidance module and is used to: Calculate and record waste weights and volumes; Generate waste weight and volume data; The RFID reading module is in communication with the waste weighing recording module and is used to: Automatically read RFID electronic tag information; Based on the RFID electronic tag information, the waste can be traced back to its source; Among them, the system automatically performs classified recycling of various types of waste according to a preset emergency mode; the emergency mode includes an epidemic prevention emergency mode and a normal emergency mode.

2. The intelligent classification and treatment system for medical waste in operating rooms according to claim 1 is characterized in that: The electronic guidance module comprises: A screen display unit, used for displaying waste classification guidance information; Indicator lamp unit for: When the capacity corresponding to the classification function is not full, instruct and guide the staff to put the waste to be classified into the classification system; When the capacity corresponding to the classification function is full, the red light flashes to remind the staff that the classification is complete.

3. The intelligent classification and treatment system for medical waste in operating rooms according to claim 1 is characterized in that: The waste identification module uses a preset multi-type detection algorithm to perform multi-target identification analysis on the types of medical waste, including: Extract feature pixels through RGB color space to classify medical waste; Taking multiple sets of RGB as input, it outputs the identification box information corresponding to the medical waste category, which is used to guide the classification operation and calculate the waste capacity.

4. The intelligent classification and treatment system for medical waste in operating rooms according to claim 1 is characterized in that: The waste weighing and recording module comprises: Electronic weighing unit to detect the quantity of waste delivered before sorting; Label encoding unit, used to: Generate electronic tags according to different wastes; Place the classified waste into the corresponding waste disposal bin; IoT interaction unit for: Get the classification processing results; RFID electronic tags are bound to classified waste to record the weight, collection time and processing process information of classified waste.

5. The intelligent classification and treatment system for medical waste in operating rooms according to claim 1 is characterized in that: The waste weighing and recording module is divided into three types of waste bags according to the waste treatment process: Burnable waste bags with a first label code to record the weight of recyclable waste and the time of collection; Non-burnable waste bags with a second label code to record the weight and collection time of non-recyclable waste; Infectious waste is bagged with a third label coded to record infectious waste weight and collection time.

6. The intelligent classification and treatment system for medical waste in operating rooms according to claim 5 is characterized in that: The Internet of Things interaction unit includes: IoT backend server for processing classified waste treatment process information; Control chip, used to calculate the weight of classified waste and issue label codes; Smart cloud terminal for real-time viewing of current waste weight and historical waste treatment results.

7. The intelligent classification and treatment system for medical waste in operating rooms according to claim 1 is characterized in that: The RFID reading module comprises: RFID electronic tags are used to identify each category of medical waste; RFID reader / writer unit for: Read electronic tag information through RFID; Automatically record waste treatment processes; Wherein, the RFID reader unit includes: A first RFID reading unit, used to obtain the recyclable waste processing result; The second RFID reading unit is used to obtain the non-recyclable waste treatment result; The third RFID reading unit is used to obtain the infectious waste treatment result.

8. The intelligent classification and treatment system for medical waste in operating rooms according to claim 7 is characterized in that: The RFID electronic tag is disposable and is scanned during recycling to trace the historical information of waste after treatment.

9. The intelligent classification and treatment system for medical waste in operating rooms according to claim 1 is characterized in that: The operation process of the epidemic prevention emergency mode includes: Step 1: If an abnormal epidemic situation is detected, activate the epidemic prevention emergency mode; Step 2: Generate electronic instructions to classify various types of medical waste according to the preset emergency storage process; Step 3: The classified wastes are put into the corresponding waste treatment cylinders through their respective channels; the classified waste treatment results are summarized to the smart cloud terminal through the IoT backend server; Step 4: When the waste capacity of each type of waste treatment barrel reaches the threshold or the epidemic ends, statistical analysis of each treatment result is performed and the normal emergency mode is returned.

10. The medical waste intelligent classification and treatment method based on the operating room medical waste intelligent classification and treatment system according to any one of claims 1 to 9 is characterized in that: include: Identify medical waste through computer vision algorithms; Guide staff to perform classification operations through electronic guidance modules; Calculate and record waste weight and waste volume through the waste weighing and recording module; The RFID reader module automatically reads the RFID electronic tag information to achieve waste tracking and tracing; Automatically sort and recycle various types of waste.

Images