Foreign surgical instrument management system based on intelligent visual identification and information tracing

Abstract

The application discloses a kind of based on intelligent visual identification and information tracing extraneous surgical instrument management system, it is related to medical instrument management technical field, to solve the technical problem of current medical instrument non-systematic recovery distribution process, including visual acquisition module, intelligent identification algorithm module, dynamic risk assessment module and information tracing module, and output corresponding instrument type data and pollution state data, to establish dynamic tracing data chain in instrument processing process.The application utilizes intelligent identification algorithm module based on deep learning identification model to automatically identify instrument structure features, pollution area features and instrument state features in instrument image, to establish dynamic tracing data chain in instrument processing process, realizes dynamic risk control and information tracing in the whole process from recovery to distribution of extraneous surgical instrument, improves the security, intelligent degree and management accuracy in instrument processing process.

Description

Technical Field

[0001] This invention relates to the field of medical device management technology, and more specifically, to a management system for external surgical instruments based on intelligent visual recognition and information traceability. Background Technology

[0002] Currently, in surgical procedures such as orthopedics, neurosurgery, and laparoscopic surgery in hospitals, external surgical instruments provided by instrument suppliers are often used for auxiliary operations. Due to the wide variety, complex structure, and significant differences in specifications of these external surgical instruments, hospital sterilization supply centers typically still rely on manual verification, registration, and parameter selection for the collection, cleaning, sterilization, assembly, and distribution of these instruments. In existing technology, staff mainly rely on instrument labels, instrument numbers, or manual experience to classify and identify instruments, and select corresponding cleaning times, sterilization temperatures, and sterilization pressure parameters based on experience. This method suffers from low instrument identification efficiency, large errors in judging instrument status, and difficulty in tracing the instrument processing process.

[0003] Especially when imported surgical instruments have movable joints, slit structures, deep cavities, and multiple connection points, blood residue and contaminants easily adhere to their surfaces. Furthermore, different materials exhibit significant differences in their tolerance to sterilization temperatures and cleaning intensities. Using fixed parameters for cleaning and sterilization can easily lead to incomplete sterilization, instrument damage, or reduced instrument lifespan. Simultaneously, most existing management systems rely on fixed rule bases or static scoring methods for risk assessment, failing to dynamically adjust risk levels based on historical anomalies, contamination abnormalities, and sterilization anomalies. This results in a lack of adaptive risk control capabilities, hindering the automatic adjustment of processing procedures for different instrument types, and consequently affecting the safety, accuracy, and traceability reliability of imported surgical instrument management. Therefore, we propose an imported surgical instrument management system based on intelligent visual recognition and information traceability. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art, adapt to the needs of reality, and provide a management system for external surgical instruments based on intelligent visual recognition and information traceability, so as to solve the technical problem of the lack of a systematic recycling and distribution process for medical devices.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a management system for imported surgical instruments based on intelligent visual recognition and information traceability, comprising a visual acquisition module, an intelligent recognition algorithm module, a dynamic risk assessment module, and an information traceability module; The visual acquisition module is used to acquire image data of external surgical instruments; The intelligent recognition algorithm module is used to identify the structural features, contaminated area features and instrument status features in instrument images based on a deep learning recognition model. It also performs region filling, contour completion and gap closing processing on non-high reflective hollow structures in the instrument to eliminate the influence of hollow holes and gaps on structural segmentation, realize the accurate recognition of the overall shape and key parts of the instrument, and output the corresponding instrument type data and contamination status data. The dynamic risk assessment module is used to establish a device risk level model based on the device's structural complexity, contamination level, material sensitivity, and historical abnormal feedback data, and to generate device risk values ​​through a dynamic weight update algorithm. The dynamic weight update algorithm adaptively corrects the weight parameters in the risk level model based on the pollution anomaly, sterilization anomaly, and structural anomaly data generated during the historical processing of the medical devices, so that different types of medical devices correspond to different risk levels and processing processes in subsequent processing. The information traceability module is used to record the node association of instrument identification results, risk level results, and process control results to establish a dynamic traceability data chain in the instrument processing process.

[0006] Preferably, the visual acquisition module includes a top visual camera, a side visual camera, a ring-shaped supplementary lighting unit, and an image synchronous acquisition unit; The top vision camera is used to acquire a planar contour image of the instrument; The lateral vision camera is used to acquire three-dimensional structural images of the instrument; The ring-shaped supplementary lighting unit is used to reduce the interference of the reflective metal area of ​​the instrument on image acquisition; The image synchronization acquisition unit is used to acquire images from different angles in a time-synchronized manner.

[0007] Preferably, the intelligent recognition algorithm module includes an image preprocessing unit, a structural feature extraction unit, a pollution recognition unit, and a classification recognition unit; The image preprocessing unit is used to perform grayscale normalization, edge enhancement, and background separation on the instrument image. The structural feature extraction unit is used to extract the device contour node features, movable joint features, connection structure features, and cavity region features; The contamination identification unit is used to identify areas of residual blood and highly contaminated areas on the surface of the instrument. The classification and recognition unit is used to establish instrument classification results based on the instrument's structural features.

[0008] Preferably, the intelligent recognition algorithm module adopts a deep convolutional neural network model based on an attention mechanism; The deep convolutional neural network model extracts local structural features of the device through convolutional layers, reduces the feature dimension through pooling layers, and assigns high weights to the device's active connection region, slit region, and cavity region through an attention mechanism to improve the accuracy of complex device recognition. The intelligent recognition algorithm module is also equipped with a hollow structure processing unit. This unit uses a contour interpolation and completion algorithm to fill holes, complete contours, and close gaps in non-highly reflective hollow structures in instrument images, eliminating interference from the hollow structures on feature extraction and classification.

[0009] Preferably, the dynamic risk assessment module includes a complexity calculation unit, a pollution level calculation unit, a material sensitivity calculation unit, an anomaly feedback analysis unit, and a risk value generation unit; The complexity calculation unit is used to calculate the structural complexity of the device based on the number of device outline nodes, the number of movable joints, the number of connection parts, and the number of slit areas. The pollution level calculation unit is used to calculate the pollution level based on the area of ​​the polluted area and the pollution density. The material sensitivity calculation unit is used to generate material sensitivity based on the high temperature resistance and corrosion resistance of the instrument material. The anomaly feedback analysis unit is used to record historical anomaly data of the device. The risk value generation unit is used to generate device risk values ​​based on the calculation results.

[0010] Preferably, the risk value generation unit generates the device risk value based on a device risk level algorithm, which satisfies the following relationship:

[0011] in, This refers to the risk level value of the medical device. This represents the structural complexity coefficient of the instrument. This is the pollution level coefficient. The sensitivity coefficient of the instrument material. This is the coefficient for the number of historical abnormalities of the instrument. This is the abnormal time decay coefficient. The time interval between anomalies. , , as well as These are the corresponding weight coefficients. The abnormality rate per unit period, This represents the abnormal impact coefficient.

[0012] Preferably, the dynamic weight update algorithm periodically adjusts the weight coefficients α, β, γ, and δ based on the historical abnormality rate of the device, and its basic formula is as follows:

[0013] in, The current anomaly rate, The average abnormality rate, To learn the correction factor; When the abnormality rate of the instrument exceeds the preset threshold in the corresponding processing stage, the system automatically increases the weight value of the corresponding influencing factor. When the abnormality rate of the instrument does not exceed the preset threshold for multiple consecutive processing cycles, the system automatically reduces the weight value of the corresponding influencing factor.

[0014] Preferably, the system further includes a process control module, which is used to adaptively adjust the cleaning and sterilization parameters of the corresponding instruments according to the instrument risk value. The process control module includes a parameter matching unit, an equipment control unit, and an anomaly adjustment unit. The parameter matching unit is used to match the corresponding cleaning time, ultrasonic power, sterilization temperature and sterilization pressure parameters according to the instrument risk value. The equipment control unit is used to send corresponding control parameters to the cleaning equipment and sterilization equipment; The abnormality adjustment unit is used to automatically extend the sterilization time and increase the sterilization level when the instrument risk value is higher than a preset threshold.

[0015] Preferably, the information traceability module includes a timestamp generation unit, a node recording unit, a status association unit, and an anomaly location unit; The timestamp generation unit is used to record the time information of the instrument entering each processing node; The node recording unit is used to record the status data of the instrument during the stages of retrieval, cleaning, sterilization, assembly, and distribution; The state association unit is used to establish a dynamic association data chain for the data of each node in chronological order. The anomaly location unit is used to automatically locate abnormal nodes and generate early warning information when node data is abnormal.

[0016] Preferably, the system further includes a central management platform and a database module, wherein the central management platform is used for centralized management of the instrument processing status; The database module is used to store medical device standard atlas data, medical device processing parameter data, historical anomaly data, and medical device traceability data. The central management platform establishes data communication connections with the hospital's disinfection supply system, operating room management system, and medical device supplier database.

[0017] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention utilizes an intelligent recognition algorithm module based on a deep learning recognition model to automatically identify the structural features, contaminated area features, and state features of instruments in instrument images. A dynamic risk assessment module establishes a dynamic risk level model based on instrument structural complexity, contamination level, material sensitivity, and historical anomaly feedback data. Simultaneously, a dynamic weight update algorithm adaptively corrects the weight parameters in the risk level model, enabling the system to dynamically adjust instrument risk values ​​based on contamination anomalies, sterilization anomalies, and structural anomalies generated during historical instrument processing. Furthermore, a process control module adaptively adjusts cleaning time, ultrasonic power, sterilization temperature, and sterilization pressure parameters based on the instrument risk value, thus avoiding the problems of incomplete sterilization, instrument damage, and manual parameter misselection that occur with traditional fixed sterilization parameters in processing complex instruments. Simultaneously, an information traceability module records the instrument identification results, risk level results, and process control results as nodes, establishing a dynamic traceability data chain throughout the instrument processing process. This achieves dynamic risk control and information traceability of imported surgical instruments from collection to distribution, improving the safety, intelligence, and management accuracy of the instrument processing process.

[0018] 2. This invention also enables the visual acquisition module to acquire images of external surgical instruments from multiple angles through the cooperation of the top visual camera, the side visual camera, the ring-shaped supplementary lighting unit, and the image synchronous acquisition unit. The image preprocessing unit performs grayscale normalization, edge enhancement, and background separation on the instrument images. At the same time, it combines a deep convolutional neural network model based on the attention mechanism to assign high weights to the active connection area, slit area, and cavity area of ​​the instrument, so as to improve the recognition accuracy of complex instruments under occlusion, reflection, and contamination conditions. This solves the problem that the accuracy of instrument recognition decreases due to the complexity of the instrument structure, the reflection of the instrument surface, and the overlap of instruments, which in turn affects the accuracy of dynamic risk assessment results.

[0019] 3. This invention also uses the contour interpolation and completion algorithm in the intelligent recognition algorithm module to perform contour completion processing on non-highly reflective hollow structures. This effectively solves the problem of feature breakage, contour discontinuity, and decreased recognition accuracy caused by large overlapping areas of hollow instruments when the deep convolutional neural network model based on the attention mechanism focuses on active connection areas, slit areas, and cavity areas. This allows for dynamic adaptation to abnormal risk changes of different types of instruments during long-term use, avoiding the problem of inaccurate sterilization parameter adjustment and mismatch of processing technology caused by the combination of fixed risk model and hollow structure recognition defects. This further enhances the system's adaptive management capability for complex hollow foreign surgical instruments.

[0020] 4. This invention also enables the system to establish a dynamic data chain in chronological order based on the data generated by the device during the stages of retrieval, cleaning, sterilization, assembly, and distribution, through the cooperation between the timestamp generation unit, node recording unit, status association unit, and anomaly location unit in the information traceability module. Combined with the risk level algorithm, the system can dynamically record and trace the risk changes during the device handling process. When node data is abnormal, the system can automatically locate the abnormal node and generate early warning information. This solves the problem of lack of full-process correlation tracking after the dynamic risk parameters change in real time, and avoids the problem of difficulty in locating the source of anomalies and tracing the responsibility for device handling due to the dispersion of historical risk data of the device. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the system flow of the present invention; Figure 2 This is a schematic diagram of the system module connections of the present invention. Detailed Implementation

[0022] Example 1: As Figure 1-2 As shown, the present invention relates to a management system for imported surgical instruments based on intelligent visual recognition and information traceability, comprising a visual acquisition module, an intelligent recognition algorithm module, a dynamic risk assessment module, a process control module, an information traceability module, a central management platform, and a database module.

[0023] The visual acquisition module includes a top visual camera, a side visual camera, a ring-shaped fill light unit, and an image synchronization acquisition unit. The top visual camera is used to acquire planar contour images of the instrument, the side visual camera is used to acquire three-dimensional structural images of the instrument, the ring-shaped fill light unit is used to reduce the interference caused by the reflective metal areas of the instrument on image acquisition, and the image synchronization acquisition unit is used to synchronously acquire images from different angles.

[0024] The intelligent recognition algorithm module includes an image preprocessing unit, a structural feature extraction unit, a contamination recognition unit, and a classification recognition unit. The image preprocessing unit performs grayscale normalization, edge enhancement, and background separation on the instrument image. The structural feature extraction unit extracts the instrument's contour node features, movable joint features, connection structure features, and cavity region features. The contamination recognition unit identifies blood residue areas and highly contaminated areas on the instrument surface. The classification recognition unit establishes instrument classification results based on the instrument's structural features and performs region filling, contour completion, and gap closure processing on non-highly reflective hollow structures in the instrument to eliminate the influence of hollow holes and gaps on structural segmentation, thereby achieving accurate recognition of the overall shape and key parts of the instrument.

[0025] Among them, the intelligent recognition algorithm module adopts a deep convolutional neural network model based on the attention mechanism. It extracts local structural features of the instrument through convolutional layers, reduces the feature dimension through pooling layers, and assigns high weights to the active connection area, slit area and cavity area of ​​the instrument through the attention mechanism to improve the recognition accuracy of complex structure instruments. The intelligent recognition algorithm module is also equipped with a hollow structure processing unit. The hollow structure processing unit uses a contour interpolation and completion algorithm to fill holes, complete contours and close gaps in non-high reflective hollow structures in instrument images, thereby eliminating the interference of hollow structures on feature extraction and classification recognition.

[0026] The contour interpolation completion algorithm uses a linear interpolation formula to complete the broken points of the instrument contour. The formula is as follows:

[0027] in, The completed contour points generated by interpolation. This is the current interpolation point number. , These are the two ends of the break in the instrument's profile. This represents the total number of points that need to be inserted between the two points. These are the coordinates of the contour points; This formula generates a continuous and smooth completed contour, realizing the contour repair and region closure of non-highly reflective hollow structures. Combined with morphological closing operation and region connectivity filling algorithm, the hollow area is filled.

[0028] The non-reflective hollow structure described in this invention includes, but is not limited to, hollow and slotted structures that do not produce mirror reflection, such as weight reduction holes for surgical instrument handles, hollow slots for forceps heads, positioning hollow holes, anti-slip hollow slots, and auxiliary ventilation holes.

[0029] The dynamic risk assessment module includes a complexity calculation unit, a contamination level calculation unit, a material sensitivity calculation unit, an anomaly feedback analysis unit, and a risk value generation unit. The complexity calculation unit calculates the structural complexity of the device based on the number of device outline nodes, the number of moving joints, the number of connection parts, and the number of slit areas. The contamination level calculation unit calculates the contamination level based on the area and density of the contaminated area. The material sensitivity calculation unit generates material sensitivity based on the high-temperature resistance and corrosion resistance of the device material. The anomaly feedback analysis unit records historical data on device processing anomalies. The risk value generation unit generates the device risk value based on the calculation results.

[0030] In this embodiment, the device risk level algorithm satisfies the following relationship:

[0031] in, This refers to the risk level value of the medical device. This represents the structural complexity coefficient of the instrument. This is the pollution level coefficient. The sensitivity coefficient of the instrument material. This is the coefficient for the number of historical abnormalities of the instrument. This is the abnormal time decay coefficient. The time interval between anomalies. , , as well as These are the corresponding weight coefficients. The abnormality rate per unit period, This represents the abnormal impact coefficient.

[0032] Furthermore, the dynamic weight update algorithm periodically adjusts the weight coefficients α, β, γ, and δ based on the historical abnormality rate of the device. Its basic formula is as follows:

[0033] in, The current anomaly rate, The average abnormality rate, To learn the correction factor; When the abnormality rate of the device exceeds the preset threshold in the corresponding processing stage, the system automatically increases the weight value of the corresponding influencing factor; when the abnormality rate of the device does not exceed the preset threshold for multiple consecutive processing cycles, the system automatically decreases the weight value of the corresponding influencing factor.

[0034] The process control module includes a parameter matching unit, an equipment control unit, and an anomaly adjustment unit. The parameter matching unit is used to match the corresponding cleaning time, ultrasonic power, sterilization temperature, and sterilization pressure parameters according to the instrument risk value. The equipment control unit is used to send the corresponding control parameters to the cleaning equipment and sterilization equipment. The anomaly adjustment unit is used to automatically extend the sterilization time and increase the sterilization level when the instrument risk value is higher than the preset threshold.

[0035] The information traceability module includes a timestamp generation unit, a node recording unit, a status association unit, and an anomaly location unit. The timestamp generation unit is used to record the time information of the device entering each processing node; the node recording unit is used to record the status data of the device during the stages of retrieval, cleaning, sterilization, assembly, and distribution; the status association unit is used to establish a dynamic association data chain of each node data in chronological order; and the anomaly location unit is used to automatically locate the abnormal node and generate early warning information when the node data is abnormal.

[0036] The working principle of this embodiment is as follows: After imported surgical instruments enter the recycling area, the vision acquisition module acquires multi-angle images of the instruments, and the intelligent recognition algorithm module identifies the structural features and contamination status of the instruments. Subsequently, the dynamic risk assessment module generates instrument risk values ​​based on the complexity of the instruments, the level of contamination, and historical anomalies, and adaptively corrects the risk model through a dynamic weight update algorithm. The process control module automatically adjusts the operating parameters of the cleaning and sterilization equipment based on the instrument risk values. The information traceability module dynamically records and correlates data throughout the entire instrument processing process to achieve intelligent identification, dynamic risk control, and information traceability of imported surgical instruments throughout the entire process.

[0037] Example 2: This example is basically the same as Example 1, except that the algorithm parameters in the dynamic risk assessment module are specifically limited in this example.

[0038] In this embodiment: Instrument structural complexity coefficient The value is 0.42; Pollution level coefficient The value is 0.38; Instrument material sensitivity coefficient The value is 0.12; Instrument historical abnormality frequency coefficient The value is 0.08; Abnormal influence coefficient The value is 0.35; Abnormal time decay coefficient The value is 0.20; Learning correction factor The value is 0.15.

[0039] When the system detects that the number of moving joints of the instrument is greater than 8, the number of slit areas is greater than 12, and the contaminated area exceeds 35% of the total area of ​​the instrument, the dynamic risk assessment module automatically increases the contamination level coefficient. and structural complexity coefficient The dynamic weight value is used to improve the sterilization level of the corresponding instrument.

[0040] The process control module automatically adjusts equipment operating parameters based on the instrument risk value. when When <0.35, use the standard sterilization method; When 0.35≤ When the value is less than 0.65, the ultrasonic cleaning power will be automatically increased and the cleaning time will be extended. when When the temperature is ≥0.65, the sterilization temperature, sterilization pressure, and sterilization duration will be automatically increased.

[0041] In this embodiment, the information traceability module establishes a dynamic data chain by using the status association unit to link the data of the instruments during the stages of collection, cleaning, sterilization, assembly and distribution. When an abnormality occurs in the handling of the instrument, the abnormality location unit automatically generates an abnormality warning information and synchronizes the abnormal data to the central management platform and database module so that the risk model can perform dynamic learning and weight correction in the future.

[0042] The embodiments disclosed in this invention are preferred embodiments, but are not limited thereto. Those skilled in the art can easily understand the spirit of this invention based on the above embodiments and make different extensions and variations, but as long as they do not depart from the spirit of this invention, they are all within the protection scope of this invention.

Claims

1. A management system for imported surgical instruments based on intelligent visual recognition and information traceability, characterized in that, It includes a visual acquisition module, an intelligent recognition algorithm module, a dynamic risk assessment module, and an information traceability module; The visual acquisition module is used to acquire image data of external surgical instruments; The intelligent recognition algorithm module is used to identify the structural features, contaminated area features and instrument status features in instrument images based on a deep learning recognition model. It also performs region filling, contour completion and gap closing processing on non-high reflective hollow structures in the instrument to eliminate the influence of hollow holes and gaps on structural segmentation, realize the accurate recognition of the overall shape and key parts of the instrument, and output the corresponding instrument type data and contamination status data. The dynamic risk assessment module is used to establish a device risk level model based on the device's structural complexity, contamination level, material sensitivity, and historical abnormal feedback data, and to generate device risk values ​​through a dynamic weight update algorithm. The dynamic weight update algorithm adaptively corrects the weight parameters in the risk level model based on the pollution anomaly, sterilization anomaly, and structural anomaly data generated during the historical processing of the medical devices, so that different types of medical devices correspond to different risk levels and processing processes in subsequent processing. The information traceability module is used to record the node association of instrument identification results, risk level results, and process control results to establish a dynamic traceability data chain in the instrument processing process.

2. The management system for imported surgical instruments based on intelligent visual recognition and information traceability according to claim 1, characterized in that, The visual acquisition module includes a top visual camera, a side visual camera, a ring-shaped fill light unit, and an image synchronous acquisition unit; The top vision camera is used to acquire a planar contour image of the instrument; The lateral vision camera is used to acquire three-dimensional structural images of the instrument; The ring-shaped supplementary lighting unit is used to reduce the interference of the reflective metal area of ​​the instrument on image acquisition; The image synchronization acquisition unit is used to acquire images from different angles in a time-synchronized manner.

3. The management system for imported surgical instruments based on intelligent visual recognition and information traceability according to claim 1, characterized in that, The intelligent recognition algorithm module includes an image preprocessing unit, a structural feature extraction unit, a pollution recognition unit, and a classification recognition unit; The image preprocessing unit is used to perform grayscale normalization, edge enhancement, and background separation on the instrument image. The structural feature extraction unit is used to extract the device contour node features, movable joint features, connection structure features, and cavity region features; The contamination identification unit is used to identify areas of residual blood and highly contaminated areas on the surface of the instrument. The classification and recognition unit is used to establish instrument classification results based on the instrument's structural features.

4. The management system for imported surgical instruments based on intelligent visual recognition and information traceability according to claim 1, characterized in that, The intelligent recognition algorithm module adopts a deep convolutional neural network model based on the attention mechanism; The deep convolutional neural network model extracts local structural features of the device through convolutional layers, reduces the feature dimension through pooling layers, and assigns high weights to the device's active connection region, slit region, and cavity region through an attention mechanism to improve the accuracy of complex device recognition. The intelligent recognition algorithm module is also equipped with a hollow structure processing unit. The hollow structure processing unit uses a contour interpolation and completion algorithm to fill holes, complete contours and close gaps in non-high reflective hollow structures in instrument images, thereby eliminating the interference of hollow structures on feature extraction and classification recognition.

5. The management system for imported surgical instruments based on intelligent visual recognition and information traceability according to claim 1, characterized in that, The dynamic risk assessment module includes a complexity calculation unit, a pollution level calculation unit, a material sensitivity calculation unit, an anomaly feedback analysis unit, and a risk value generation unit. The complexity calculation unit is used to calculate the structural complexity of the device based on the number of device outline nodes, the number of movable joints, the number of connection parts, and the number of slit areas. The pollution level calculation unit is used to calculate the pollution level based on the area of ​​the polluted area and the pollution density. The material sensitivity calculation unit is used to generate material sensitivity based on the high temperature resistance and corrosion resistance of the instrument material. The anomaly feedback analysis unit is used to record historical anomaly data of the device. The risk value generation unit is used to generate device risk values ​​based on the calculation results.

6. The management system for imported surgical instruments based on intelligent visual recognition and information traceability according to claim 1, characterized in that, The risk value generation unit generates a device risk value based on a device risk level algorithm, which satisfies the following relationship: ； in, This represents the risk level value of the medical device. This represents the structural complexity coefficient of the instrument. This is the pollution level coefficient. The sensitivity coefficient of the instrument material. This is the coefficient for the number of historical abnormalities of the instrument. This is the abnormal time decay coefficient. The time interval between anomalies. , , as well as These are the corresponding weight coefficients. The abnormality rate per unit period, This represents the abnormal impact coefficient.

7. A management system for imported surgical instruments based on intelligent visual recognition and information traceability according to claim 6, characterized in that, The dynamic weight update algorithm periodically adjusts the weight coefficients α, β, γ, and δ based on the historical abnormality rate of the medical device. Its basic formula is as follows: ； in, The current anomaly rate, The average abnormality rate, To learn the correction factor; When the abnormality rate of the instrument exceeds the preset threshold in the corresponding processing stage, the system automatically increases the weight value of the corresponding influencing factor. When the abnormality rate of the instrument does not exceed the preset threshold for multiple consecutive processing cycles, the system automatically reduces the weight value of the corresponding influencing factor.

8. A management system for imported surgical instruments based on intelligent visual recognition and information traceability according to claim 7, characterized in that, The system also includes a process control module, which is used to adaptively adjust the cleaning and sterilization parameters of the corresponding instruments according to the instrument risk value. The process control module includes a parameter matching unit, an equipment control unit, and an anomaly adjustment unit. The parameter matching unit is used to match the corresponding cleaning time, ultrasonic power, sterilization temperature and sterilization pressure parameters according to the instrument risk value. The equipment control unit is used to send corresponding control parameters to the cleaning equipment and sterilization equipment; The abnormality adjustment unit is used to automatically extend the sterilization time and increase the sterilization level when the instrument risk value is higher than a preset threshold.

9. A management system for imported surgical instruments based on intelligent visual recognition and information traceability according to claim 1, characterized in that, The information traceability module includes a timestamp generation unit, a node recording unit, a status association unit, and an anomaly location unit; The timestamp generation unit is used to record the time information of the instrument entering each processing node; The node recording unit is used to record the status data of the instrument during the stages of retrieval, cleaning, sterilization, assembly, and distribution; The state association unit is used to establish a dynamic association data chain for the data of each node in chronological order. The anomaly location unit is used to automatically locate abnormal nodes and generate early warning information when node data is abnormal.

10. A management system for imported surgical instruments based on intelligent visual recognition and information traceability according to claim 9, characterized in that, The system also includes a central management platform and a database module, wherein the central management platform is used for centralized management of the device processing status; The database module is used to store medical device standard atlas data, medical device processing parameter data, historical anomaly data, and medical device traceability data. The central management platform establishes data communication connections with the hospital's disinfection supply system, operating room management system, and medical device supplier database.

Images