An operating room intelligent instrument distribution and safety interlocking management system based on multi-source information fusion and full-process closed loop

Abstract

The application discloses a kind of based on multi-source information fusion and whole process closed loop operating room intelligent equipment and safety interlock management system and method.The system includes dynamic information carrier module, central server and global data center, preoperative intelligent scheduling and positioning module, operating room intelligent verification terminal, multi-level interlock safety verification unit, differentiated physical interlock and intelligent guidance execution unit, supply room handover and traceability module, and self-learning optimization module based on frequent pattern mining.Through five interlock verification mechanism and physical interlock execution synergistic effect, completely eliminate the instrument package opened by memory blindly, through step level preference matching and pre-validation function, realize the accurate verification of the instrument required for current operation step, through FP-Growth frequent pattern mining algorithm and artificial confirmation mechanism, make preference library continue to learn clinical real habits.The application can effectively improve the intelligent level and safety control ability of operating room instrument management.

Description

Technical Field

[0001] This invention relates to the field of medical information technology, and more specifically, to an intelligent operating room equipment dispensing and safety interlock management system based on multi-source information fusion and a closed-loop process. Background Technology

[0002] As modern surgery becomes increasingly sophisticated and complex, the variety and quantity of instruments required for a single operation have increased significantly. For example, a single laparoscopic liver resection may involve dozens or even hundreds of basic, specialized, and auxiliary instruments. Furthermore, instruments of different brands and models exhibit a high degree of homogeneity in appearance, making it difficult to accurately determine the actual contents of the instrument pack based solely on the labels attached to the pack. This information asymmetry presents a serious challenge to operating room instrument management.

[0003] Currently, operating room instrument management primarily relies on a combination of manual operation and static barcode labels. Specifically, instrument managers prepare instrument packs in advance based on the surgical schedule, and nurses, when opening and checking the packs pre-operatively, can only verify the instruments against the list displayed on the outer label. However, this existing management model has the following prominent problems: First, opening instrument packs is often unpredictable. Because nurses cannot foresee the specific configuration of the instruments inside before surgery, they often open the packs randomly based on memory or habit. If the actual configuration does not match the expectations, the pack must be replaced or temporarily reconfigured. This process not only compromises the sterile barrier but may also delay surgery and increase the risk of intraoperative infection. Second, there is a lag in information regarding changes to instruments. During surgery, due to the surgeon's habits or unforeseen circumstances, instrument managers may need to temporarily add or delete items from the instrument packs. However, existing static labels cannot be updated in real time, and relevant information cannot be promptly transmitted to the user, leading to discrepancies between the required instruments and the actual configuration, affecting the surgical process. Third, existing management methods are insufficient to meet personalized instrument requirements. Different specialty groups, different surgeons, and even different steps in the same surgery have significantly different requirements for instrument configuration. Existing static QR codes can only provide a fixed list of instruments and cannot dynamically verify or intelligently match them based on factors such as doctor preferences and surgical steps, resulting in insufficient accuracy in instrument allocation. Fourth, preoperative instrument scheduling is inefficient. When the surgical team needs a specific instrument (such as a retractor plate or a special-sized clamp), they often cannot quickly find out which instrument pack it is in. When the lead surgeon allocates instrument resources for multiple surgeries the following day, it is difficult to achieve optimal global configuration, leading to resource waste or conflicts. Fifth, the instrument circulation status lacks full lifecycle tracking. The entire circulation process of instruments, from use to cleaning, disinfection, and return to their place, lacks real-time monitoring. Information is fragmented between the operating room and the supply room, resulting in low instrument turnover efficiency and failing to provide effective status references for surgical scheduling. Sixth, the supply room handover process relies on paper documents. Postoperative instrument counting and tracing of problematic instruments still use traditional manual registration methods, which are inefficient and unclear in defining responsibility, easily leading to instrument loss or disputes.

[0004] In summary, the existing operating room instrument management model has significant limitations in terms of real-time information, accuracy of instrument allocation, and closed-loop process. There is an urgent need to introduce intelligent technologies to build a smart management system covering the entire process of preoperative planning, intraoperative control, and postoperative traceability. Summary of the Invention

[0005] This invention provides an intelligent operating room instrument packing and safety interlock management system and method based on multi-source information fusion and a closed-loop process, in order to solve the technical problems existing in the management of operating room instruments, such as blindly opening instrument packs, delayed information on instrument changes, insufficient accuracy of personalized instrument packing, low efficiency of preoperative scheduling, lack of full life cycle tracking of instrument circulation status, and reliance on paper documents for handover in the supply room.

[0006] This application establishes a wireless communication connection with the central server through a dynamic information carrier module, enabling remote dynamic updates of the instrument pack contents; it solves the pain points of difficulty in locating instruments and avoiding surgical conflicts by using a preoperative intelligent scheduling and positioning module for global searching and intelligent scheduling of the instrument pack; it ensures precise matching of intraoperative instrument configuration with surgical steps by performing multi-level interlocking safety verification through an intelligent verification terminal in the operating room; it automatically locks the shadowless lamp and operating table when verification fails through differentiated physical interlocking and intelligent guidance execution unit, and intelligently guides the acquisition of missing instruments; it realizes electronic postoperative inventory and direct traceability of abnormal instruments through a supply room handover and traceability module; and it continuously updates the four-level preference knowledge base through a self-learning optimization module based on frequent pattern mining, achieving adaptive evolution of the instrument allocation strategy.

[0007] As a preferred solution, an intelligent operating room instrument packing and safety interlock management system based on multi-source information fusion and a closed-loop process is characterized by comprising the following components: a dynamic information carrier module attached to the surface of each individual instrument pack; a central server and a global data platform deployed in the hospital data center, communicating with each subsystem; a mobile management APP for use by instrument management staff and supply room personnel; a preoperative intelligent scheduling and positioning module deployed on the central server, providing instrument search, scheduling allocation, and status tracking functions; an operating room intelligent verification terminal with a touch screen installed in each operating room; a multi-level interlock safety verification unit integrated into the verification terminal; a differentiated physical interlock and intelligent guidance execution unit, including a PLC controller, a shadowless lamp power control circuit, an operating table hydraulic lock circuit, and an intelligent instrument cabinet; a supply room handover and traceability module deployed on the mobile terminal and the central server; and a self-learning optimization module based on frequent pattern mining deployed on the central server.

[0008] This preferred solution configures the dynamic information carrier module as a remotely updatable e-ink display, enabling real-time synchronization of information on the instrument pack surface with the central server, avoiding the inherent drawbacks of traditional static labels that cannot be updated in a timely manner. By deploying a global data platform on the central server, unified management of multi-source data such as surgical scheduling, instrument dictionaries, preference knowledge bases, and status tables is achieved, providing a data foundation for intelligent decision-making. The mobile management app extends instrument management and supply room handover functions to mobile devices, improving operational efficiency and flexibility. The global search and conflict detection functions of the preoperative intelligent scheduling and positioning module address the core pain points of "difficulty in finding instruments and resolving conflicts" for lead surgeons when allocating instruments preoperatively. The multi-level verification and physical interlocking execution of the operating room intelligent verification terminal ensures that the instrument configuration for each surgery meets safety requirements. The electronic signature and direct anomaly traceability functions of the supply room handover and traceability module establish a complete closed loop for postoperative instrument circulation. The continuous evolution mechanism of the self-learning optimization module allows the system to continuously adapt to real clinical needs, reducing manual maintenance costs.

[0009] As a preferred embodiment, the dynamic information carrier module includes: a human-readable display area displaying a summary of key instruments in the package, applicable surgical types, and sterilization expiration dates in large font; a machine-readable encoding area storing a dynamic QR code, the encoded content of which includes at least a unique list of instruments in the package, version number, sterilization batch number, and current status indicator; and a wireless update unit receiving update commands pushed by the central server via Wi-Fi or 4G / 5G network, and triggering screen refresh and synchronous update of the QR code content. The specific hardware specifications of the dynamic information carrier module are: the main control chip is an ESP32-WROOM-32E, equipped with 4MB of RAM. Flash storage; 4.2-inch three-color e-ink screen; internal PCB is coated with conformal coating (temperature resistance ≥150℃) and filled with aerogel insulation layer; the module is covered with a medical-grade IP68 protective shell (material is polytetrafluoroethylene / PTFE), which can withstand 134℃ / 0.22MPa high-temperature and high-pressure steam sterilization (45min cycle) or hydrogen peroxide plasma sterilization; the module automatically enters deep sleep mode before sterilization (power consumption <1μA), and resumes working status through low-frequency 125kHz RFID wake-up signal after sterilization; the power supply uses CR2032 button battery (capacity 220mAh, battery life ≥3 years or ≥2000 refreshes), and the battery compartment and shell are integrated and sealed.

[0010] In this preferred solution, the human-readable display area uses a large font design, which facilitates nurses to quickly identify key information in the instrument pack before surgery, and can determine whether the pack meets the basic requirements of the current operation without scanning. The dynamic QR code in the machine-readable encoding area uses a version number mechanism. The version number increments with each content update. After receiving the update command, the dynamic information carrier module first compares the new version number with the local version number: if the new version number is greater than the local version number, the update is executed; if the new version number is less than or equal to the local version number, the expired command is discarded and the latest data is requested from the server. The central server pushes the update command with a timestamp. The module's cache validity period is 30 minutes. If it expires, it is automatically marked as expired. In conjunction with the timestamp verification of the central server, it can effectively prevent information inconsistency caused by the expiration of the local cache. The wireless update unit utilizes the existing hospital wireless network infrastructure, without the need to deploy additional communication cables, resulting in low transformation costs and a short implementation cycle.

[0011] As a preferred embodiment, the central server and global data platform include: an interface unit that connects with the HIS (Hospital Information System) via the HL7 / FHIR protocol to obtain patient information, connects with the surgical scheduling system via WebService to obtain surgical plans, and communicates with devices such as intelligent instrument cabinets and verification terminals via the MQTT protocol; the interface unit connects with the HIS system via the HL7 FHIR R4 protocol, uses FHIR Patient resources to obtain patient information, and maps key fields as follows: Patient ID ← FHIRPatient.identifier, Name ← FHIR Patient.name.text, Hospitalization Number ← FHIRPatient.identifier.system='hospital-mrn'; it connects with the surgical scheduling system via a RESTful API, uses the POST / schedule / operation interface to obtain surgical plans, and returns data in JSON format; and a database unit that maintains a surgical scheduling table, a four-level preference knowledge base, an instrument dictionary, an instrument package template library, an instrument lifecycle status table, an instrument-package ownership table, and a handover record table.

[0012] In this preferred solution, the interface unit adopts a standardized medical information exchange protocol, ensuring seamless integration with the hospital's existing information systems. The patient ID read from the wristband is compared with the patient ID obtained from the FHIR; if they do not match, a warning pops up prompting the nurse to verify, allowing the nurse to choose to verify based on either the wristband information or the system information. Abnormal situations are logged. The database unit uses a hybrid architecture of relational and document-oriented databases, ensuring the reliability of transactional operations while supporting efficient storage and retrieval of unstructured data such as historical timelines. The four-level preference knowledge base uses a multi-dimensional index structure of specialist group ID → doctor ID → surgical name → step number, supporting device preference matching accurate to the surgical step level, providing data support for precise device allocation at the step level.

[0013] As a preferred embodiment, the specific data structure of the four-level preference knowledge base includes: a specialty group identifier field, which records the specialty group number to which the instrument belongs; a doctor identifier field, which records the unique employee number of the surgeon; a surgery name field, which records the standard name or alias mapping of the surgery; a step sequence number field, which records the sequence number of the surgical step; a mandatory instrument ID list field, which records the set of instrument IDs that must be used in this step; and an optional instrument ID list field, which records the set of instrument IDs that can be used in this step.

[0014] This preferred solution solves the problem of naming differences for the same surgery in different departments or by different doctors through an alias mapping mechanism, thereby improving the matching success rate. By distinguishing between mandatory and optional instruments, it ensures the mandatory configuration of key instruments while providing flexibility for personalized preferences.

[0015] As a preferred embodiment, the preoperative intelligent scheduling and positioning module includes: a global instrument search unit, a real-time instrument status tracking unit, and an intelligent scheduling and allocation unit; wherein, the global instrument search unit performs the following steps: First, it matches the input instrument name in the instrument dictionary to obtain the instrument ID; second, it queries the instrument-package affiliation table to find the package ID to which the instrument ID currently belongs; third, it retrieves the package name and current status from the instrument package template library based on the package ID, and returns a list of all instrument packages containing the instrument and the current location of each package.

[0016] In this preferred solution, the global instrument search employs a fuzzy matching algorithm, supporting multiple search methods such as initials of pinyin and aliases, thus improving search usability. By linking the instrument status table with location sensor data, the returned location information can distinguish whether the instrument pack is located in the sterilization supply room's waiting area or has already been distributed to the operating room's instrument cabinet, providing accurate reference for the lead instructor's allocation decisions. The intelligent scheduling and allocation unit detects allocation conflicts of instrument packs in real time. When the time interval between two surgeries is less than the minimum turnaround time for a particular instrument pack, the system automatically alerts to the conflict and provides suggestions for replacing or adjusting the surgical sequence, avoiding surgical delays caused by insufficient instrument turnaround time.

[0017] As a preferred embodiment, the conflict detection algorithm of the intelligent scheduling and allocation unit is as follows: extract the actual duration of the most recent n turnovers of the instrument pack from the instrument circulation record table, and use a moving average calculation when n≥50. =(t(n-49)+…+tn) / 50, where the default value is used when n<50. =120 minutes; Obtain the planned start time of assigned surgery A. The planned start time of surgery B to be assigned Calculate the time interval ;like If surgery A and surgery B use the same instrument pack, it is considered a conflict and an alarm message is output.

[0018] This preferred solution introduces the dynamic parameter of historical average turnaround time, enabling conflict detection to adapt to the actual processing time of different instruments. It provides more buffer time for precision instruments with slower turnaround times, thereby improving the accuracy and practicality of the algorithm.

[0019] As a preferred embodiment, the intelligent verification terminal in the operating room includes: a barcode scanning module that supports reading one-dimensional barcodes, two-dimensional barcodes, and RFID tags; a patient wristband reading module that supports reading patient identity information via NFC or barcode; a display module configured as a 10-inch or larger industrial-grade touchscreen; and a multi-level interlocking security verification unit that performs five sequential verifications in software.

[0020] In this preferred solution, the barcode scanning module adopts a multimodal recognition design, which can adapt to the identification methods of different packaging forms of instrument packs and individual instruments; the patient wristband reading module interfaces with the hospital HIS system, and the obtained patient information can be automatically filled into the verification process, avoiding the risk of errors from manual input; the large-size touch screen design of the display module ensures the convenience of operations such as list viewing and step switching.

[0021] As a preferred embodiment, the five verifications performed by the multi-level interlocking safety verification unit are as follows: Verification 1, mandatory check confirmation, verifying whether the nurse has clicked the "I have checked the list" button; if not, a failure code E01 is output. Verification 2, instrument pack applicable surgical type matching, verifying whether the list of applicable surgical types in the QR code contains the name of the currently scheduled surgery; if not, a failure code E02 is output. Verification 3, patient identity matching, verifying whether the patient ID read from the wristband matches the patient ID in the scheduling table; if not, a failure code E03 is output. Verification 4, sterilization expiration date verification, verifying whether the sterilization date plus the expiration date in the QR code exceeds the current system time; if expired, a failure code E04 is output. Verification 5, current surgical step instrument satisfaction verification, calling the preference matching calculation unit to compare the necessity of the current step. The system selects the instrument set and the list within the package. If any instrument is missing, it outputs a failure code E05 and returns a list of missing instruments. In verification five, the system first obtains the standard procedure template for the current surgery from the surgical schedule table, calculates the current surgical stage based on the surgery start timestamp, and maps the surgical stage to the corresponding step number. Then, it calls the preference matching unit to locate the required instrument set for the current step in the four-level preference knowledge base based on the current surgery's specialty group, surgeon, and surgery name. This set is compared with the list of instruments within the package, and a list of missing instruments is output. If the current step does not exist in the preference base, it is downgraded to using the general required instrument set for matching. The general required instrument set is pre-stored in the instrument package template library, associated with the surgery name, and manually configured by the instrument administrator according to the surgery type.

[0022] This preferred solution ensures the integrity and rigor of the verification process through a five-step sequential execution mechanism. Failure of any verification interrupts the subsequent process and locks the equipment, effectively preventing unqualified instrument packs from entering the surgical procedure. By differentiating different failure codes and corresponding handling strategies (e.g., E02 only prompts for instrument pack replacement, while E05 triggers intelligent guidance), the system's differentiated response capability is improved.

[0023] As a preferred embodiment, the workflow of the preference matching operation unit is as follows: Input the current surgery ID and the current step number; obtain the specialist group ID, surgeon ID, and surgery name based on the surgery ID; locate the records of the surgeon, surgery, and step in the four-level preference knowledge base; extract the set of required instrument IDs and the set of optional instrument IDs; perform a difference operation between the set of required instrument IDs and the package list; and output a list of required instrument IDs that do not appear in the package list as the missing set. If the current step does not exist in the preference base, the universal set of required instruments for the entire surgical process is used for matching. The downgraded matching results and their triggering events are recorded in the log for the administrator to improve the preference base later.

[0024] This preferred solution ensures that new techniques or steps can be verified using a common configuration through a degradation mechanism, avoiding the embarrassing situation where the system cannot work due to a missing preference library. At the same time, it records the circumstances that trigger degradation through logging, providing data support for the subsequent improvement of the preference library.

[0025] As a preferred embodiment, the differentiated physical interlocking and intelligent guidance execution unit includes: a PLC controller that communicates with the verification terminal via the Modbus TCP protocol; a shadowless lamp power control circuit, with the PLC's digital output port connected in series to the shadowless lamp's power contactor coil; an operating table hydraulic lock circuit, with the PLC's digital output port controlling the power supply to the operating table's moving hydraulic valve; and an intelligent instrument cabinet, each compartment equipped with an electronic lock and a blue LED indicator, connected to the PLC via an RS485 bus. The electrical connection circuit between the PLC and the shadowless lamp and operating table includes a safety isolation relay and a solid-state optocoupler conforming to the IEC 60601-1 medical safety standard, connected in series. The PLC only outputs low-level control signals and is not directly connected to the equipment's main power supply circuit. The control circuit is configured with fail-safe logic; when PLC communication is interrupted or the power supply is abnormal, the relay switches to a normally closed safe state by default and restores power to the equipment. The intelligent medical device cabinet has an RS485 communication baud rate of 9600bps, 8 data bits, 1 stop bit, and no parity. The cabinet address encoding rule is: cabinet number (1-4) << 8 + cabinet number (1-16). The PLC maintenance device ID-cabinet address mapping table has an unlock command frame format of: device address + function code + target cabinet address + unlock command + CRC16 checksum.

[0026] This preferred solution uses a PLC controller as the core hub for interlocking execution, achieving a reliable conversion from software verification results to physical equipment control; by connecting the digital output port in series with the equipment power circuit, it ensures that the equipment immediately loses power when verification fails, preventing bypass startup through software; and through independent cabinet-level control and LED indicators of the intelligent medical equipment cabinet, it provides intuitive visual guidance for obtaining missing medical equipment.

[0027] As a preferred embodiment, the execution logic of the differentiated physical interlock and intelligent guidance execution unit is as follows: When the verification terminal generates any failure code, the terminal sends a lock command to the PLC via MQTT, carrying the failure code; the PLC immediately sets the digital output ports corresponding to the shadowless lamp and operating table to a low level, causing the equipment to lose power and lock; if the failure code is E05, the terminal simultaneously sends a list of missing instrument IDs to the PLC, the PLC queries the cabinet address mapping table based on the instrument IDs pre-issued by the server to find the cabinet where the missing instrument is located, sends an unlock command to the cabinet and causes the blue LED indicator to flash at a frequency of 2Hz, and the terminal screen displays guidance information; after the nurse takes the missing instrument and scans the instrument label, the terminal re-executes the fifth verification step, if the missing set is empty, it sends a verification pass command to the PLC, the PLC restores the power supply to the equipment and sends a global unlock command to the intelligent instrument cabinet.

[0028] This preferred solution provides a striking visual cue through 2Hz LED flashing, avoiding the hassle of nurses searching for target instruments among numerous cabinets; by requiring nurses to scan for missing instruments for secondary confirmation, a closed-loop operation of "retrieval-verification" is ensured, preventing secondary verification failures due to mistakenly retrieving other instruments; and through the PLC's global unlock command, all cabinets are restored to their locked state at once after successful verification, improving operational efficiency.

[0029] As a preferred embodiment, the differentiated physical interlocking and intelligent guidance execution unit also includes an anti-bypass design: the PLC is configured with an independent input interface (DI port, input voltage range 24VDC) to detect the status of the emergency bypass switch on the operating light panel; the bypass switch is a red emergency stop button with a protective cover, model GPE-16A, installed in the upper left corner of the operating light panel, using normally closed contacts. Under normal conditions, the contacts are closed, and the input port detects 0V; when the button is pressed, the contacts open, the input port detects 24V, and the PLC detects the falling edge signal with a 10ms scan cycle and determines it as a bypass signal trigger; when the bypass signal is detected, the PLC immediately performs the following actions: ① Drives the intermediate relay through the DO0 port to start a 90dB audible and visual alarm (model BZ-1205, sound pressure level ≥90dB@1m), the alarm continues until the bypass signal is released and the administrator is reset; ② Maintains a low-level output through the DO1 port to maintain the locked state of the operating table hydraulic valve; ③ Sends a JSON format message {packet_id, The system reset requires the administrator to enter their employee ID and password on the verification terminal to confirm the reset until the verification is successful.

[0030] This preferred solution, through its anti-bypass design, plugs the loopholes that allow bypassing the interlocking mechanism via hardware means, ensuring the integrity and unavoidability of the interlocking security strategy; it also uses a 90dB audible and visual alarm to deter violations on-site and provides evidence for post-event tracing through event recording.

[0031] As a preferred embodiment, the supply room handover and traceability module includes: a postoperative inventory submodule, which allows supply room staff to scan the instrument pack's QR code to display a standard list, check off each item for inventory, and then call the electronic signature for handover confirmation; an anomaly traceability submodule, which automatically pops up the surgical information of the last use of the instrument pack and provides a contact button when an anomaly is found during inventory; and a status push submodule, which scans the label and updates the status after each step of cleaning, disinfection, sterilization, etc., and pushes status notifications to relevant users based on subscription relationships.

[0032] In this preferred embodiment, the post-operative inventory submodule replaces paper signatures with electronic signatures, which not only improves handover efficiency, but also utilizes digital signature technology. The signature content includes the inventory personnel's employee number, inventory timestamp, and instrument list hash value. The signature generation process is as follows: the mobile terminal obtains the inventory personnel's digital certificate and uses the certificate's private key to sign the "inventory personnel's employee number + timestamp + SHA-256 digest of the instrument list JSON". The signature result, along with the original signature text, is stored in the central server. During signature verification, the public key is used to recalculate the hash for comparison. The server additionally calculates the SHA-256 digest for each signature record and stores it in the blockchain evidence storage service. Any modification to the database record will result in an inconsistent digest, thus ensuring the authenticity and immutability of the electronic record. The anomaly tracking submodule enables rapid location and communication of anomalies by directly connecting to the previous user, reducing the anomaly handling time from an average of 2 hours to less than 10 minutes. The status push submodule, through a message subscription mechanism, allows operating room nurses to know the processing progress of their requested instruments in real time, facilitating advance scheduling of surgical use.

[0033] As a preferred embodiment, the self-learning optimization module based on frequent pattern mining employs the FP-Growth algorithm for data mining. Specifically, it includes: a data acquisition unit that records the actual instrument data used in each surgery, including surgery ID, specialty group, doctor ID, surgery name, step number, and a list of actual instrument IDs; a learning algorithm unit that sets a minimum support threshold of 0.6 and a minimum confidence threshold of 0.7, mining frequent itemsets by specialty group, doctor, surgery name, and step; and an update strategy unit that automatically executes a learning task at 2:00 AM daily. For each specialty group-doctor-surgery-step combination, it calculates the matching degree between the required instruments in the current preference library and the actual usage data. If the usage frequency of an optional instrument in the last 30 cases exceeds 0.6, it is upgraded to a required instrument; if the usage frequency of a required instrument in the last 30 cases is less than 0.4, it is downgraded to an optional instrument. If a new strongly correlated instrument combination is found with a confidence level exceeding 0.7, it is recommended to add it to the optional set. All update suggestions are manually confirmed and then written into the production environment preference library.

[0034] The specific implementation steps of the FP-Growth algorithm are as follows: ① Data preparation: Export the surgical usage records of the most recent 90 days from the data acquisition unit to construct a transaction database. Each transaction consists of {Surgery ID, Specialty Group ID, Doctor ID, Surgery Name, Step Number, Instrument ID List}; ② Frequent 1-Itemset Filtering: Scan the transaction database and calculate the support of each instrument ID: support(A) = Count(A) / N, where N is the total number of the most recent 30 surgeries under this dimension. Filter frequent 1-itemsets with support(A) ≥ 0.6 and construct an item header table; ③ FP-Tree Construction: Insert the instrument ID list from each transaction into the FP-Tree according to the item header table order. Each node of the tree contains {Instrument ID, Count, Parent Node Pointer, ...} ④ Conditional pattern base mining: For each frequent item, recursively backtrack its conditional pattern base to generate frequent itemsets; ⑤ Association rule generation: For the frequent itemset {A,B}, calculate the confidence conf(A→B)=support(A∪B) / support(A), and filter association rules with conf(A→B)≥0.7; ⑥ Recommendation output: Output the association rule A→B that meets the threshold condition as a suggestion to add device B to the optional set of device A in the step.

[0035] The learning algorithm unit is also equipped with a preference conflict resolution mechanism: when there is a preference conflict among multiple surgeons for the same surgical procedure, the surgeons' job titles (chief physician: deputy chief physician: attending physician = 3:2:1) are weighted and fused with the actual usage frequency in the past 90 days, and the instruments with a weighted score ≥0.7 are included in the recommendation set.

[0036] The update strategy unit is also configured with cold start degradation logic: when a specialist group or doctor has less than 30 historical data, the system automatically calls the set of common required instruments for the same specialty for matching, and switches to personalized mining after the data accumulation reaches the standard.

[0037] The specific process for "writing to the production environment preference library after manual confirmation" is as follows: After the learning task is completed, an update suggestion list is generated and stored in the pending review table, and an email notification is sent to the preset administrator account. After logging into the system, the administrator confirms each suggestion one by one on the 'Preference Library Review' page, and can choose 'Agree,' 'Reject,' or 'Agree after modification.' After the review is completed, the system writes the approved suggestions to the production environment preference library in batches, effective at 0:00 the next day. If the administrator does not take any action within 72 hours, the system automatically sends a reminder message. Suggestions that are still not reviewed after 7 days are automatically invalidated. All update suggestions generated by the algorithm are put into a sandbox environment for at least 72 hours of clinical logic simulation verification. Only after the verification is passed can they be pushed to the manual confirmation queue.

[0038] This preferred solution leverages the efficient mining capabilities of the FP-Growth algorithm to automatically discover frequent patterns and association rules in device usage from massive amounts of historical usage data, replacing the inefficient traditional method of manually maintaining the preference database. By setting reasonable support and confidence thresholds, it ensures the reliability of update suggestions while maintaining the algorithm's sensitivity through a statistical window of 30 cases. A manual verification mechanism ensures that the system will not generate erroneous preference updates due to data noise, respecting the professional judgment of clinical experts.

[0039] As a preferred embodiment, the calculation formulas for the minimum support threshold and the minimum confidence threshold are as follows: Support , representing the frequency of simultaneous occurrence of device A and device B; confidence level Let represent the conditional probability of using device A simultaneously with device B; the system only operates if device A is used. and It is recommended to add instrument B to the optional set for this step only when necessary.

[0040] This preferred approach avoids incorrect recommendations due to a single indicator by using both support and confidence scores for screening. High support ensures that the association has sufficient generality in the statistical sample, while high confidence ensures that the association has sufficient reliability. Together, they ensure the high quality of the recommendations.

[0041] As a preferred embodiment, after generating the instrument usage record, the system further includes: a safety inspection unit, which checks whether the quantity of surgical instruments and medications meets preset values ​​according to the type of surgery before the surgery begins, and whether the surgical instruments meet preset safety standards; a compliance inspection unit, which assesses whether medical staff follow preset surgical operating procedures during the surgery; and a record integration unit, which adds the safety inspection results and compliance inspection results to the instrument usage record.

[0042] This preferred solution ensures that the surgery will not pose a safety hazard due to missing or substandard instruments by conducting quantity and safety standard checks before the operation; and by conducting compliance assessments during the operation, it forms a quality control system covering the entire surgical cycle, providing data support for medical quality traceability.

[0043] This application also provides an intelligent instrument packing and safety interlock management system for operating rooms, including a data communication module, a verification execution module, an interlock control module, and a self-learning optimization module. The data communication module establishes a wireless communication connection with a central server via a dynamic information carrier module to remotely update the contents and synchronize the status of the instrument pack. The verification execution module performs multi-level interlock safety verification at the intelligent verification terminal in the operating room, outputting verification success or failure codes. The interlock control module controls the PLC to execute physical interlock actions based on the verification failure code, including locking the power supply of the operating light, locking the movement of the operating table, and triggering the unlocking of the intelligent instrument cabinet compartments and LED indicators. The self-learning optimization module uses the FP-Growth algorithm to mine actual usage data and update the four-level preference knowledge base.

[0044] As a preferred embodiment, the verification execution module includes: a forced review confirmation unit, used to detect whether the nurse has clicked the "I have reviewed the list" button; an applicable surgery type matching unit, used to verify whether the list of applicable surgery types in the QR code contains the name of the currently scheduled surgery; a patient identity matching unit, used to verify whether the patient ID read by the wristband matches the patient ID in the scheduling table; a sterilization expiration date verification unit, used to verify whether the sterilization date plus the expiration date of the instrument pack exceeds the current system time; and a step instrument satisfaction verification unit, used to call the preference matching calculation unit to compare the required instrument set of the current step with the list in the pack, and output a list of missing instruments.

[0045] As a preferred embodiment, the interlocking control module includes: a PLC communication unit, used to receive locking commands sent by the verification terminal via the MQTT protocol and parse failure codes; an equipment control unit, used to control the digital output port of the PLC according to the failure codes to realize the power supply on / off of the shadowless lamp power supply and the hydraulic valve of the operating table; an intelligent cabinet control unit, used to query the cabinet address according to the missing instrument ID and send an unlocking command to control the electronic lock and LED indicator of the corresponding cabinet; and a bypass detection unit, used to detect the status signal of the emergency bypass switch, trigger an audible and visual alarm and push event records when a bypass signal is detected.

[0046] As a preferred embodiment, the self-learning optimization module includes: a data acquisition unit, used to obtain the actual set of instruments used in each surgery by having nurses select the instruments used on a terminal or by automatically reading them through an RFID smart inventory pad; a mining calculation unit, used to mine frequent itemsets in different dimensions using the FP-Growth algorithm based on a minimum support threshold of 0.6 and a minimum confidence threshold of 0.7; and an update management unit, used to convert the mining results into preference library update suggestions, which are then written into the production environment database after manual confirmation.

[0047] As a preferred embodiment, the system also includes a mobile management app, comprising: an instrument pack editing interface for instrument management staff to add or delete instruments within a pack after scanning a QR code; a global instrument search interface for users to enter an instrument name and return a list of all instrument packs containing that instrument and its current location; a surgical scheduling and allocation interface for lead staff to drag and drop instrument packs to allocate to each surgery and detect conflicts; a status reporting interface for supply room staff to scan tags and select the current status; and an anomaly handling interface for reporting anomalies and contacting the previous user.

[0048] This application also provides a computer-readable storage medium storing a computer program, which is called and executed by a processor to implement the above-described intelligent operating room instrument fitting and safety interlock management method.

[0049] Beneficial effects

[0050] Compared with the prior art, the present invention has the following beneficial effects:

[0051] 1. This invention enables contactless updating of medical device package information through real-time communication between a dynamic electronic ink screen and a central server, avoiding the tediousness and errors of manually changing labels.

[0052] 2. This invention, through the synergistic effect of a five-level interlocking verification mechanism and physical interlocking execution, completely eliminates the illegal behavior of blindly opening the instrument pack based on memory from three aspects: process, information, and execution.

[0053] 3. This invention achieves accurate verification of instruments required for the current surgical step through step-level preference matching and pre-verification functions, significantly reducing the frequency of temporary instrument retrieval during surgery.

[0054] 4. This invention achieves global optimization of medical device resources through full lifecycle status tracking and intelligent conflict detection.

[0055] 5. This invention organically combines the FP-Growth frequent pattern mining algorithm with a manual confirmation mechanism, enabling the preference database to continuously learn real clinical habits and keep pace with the evolution of surgical techniques. Attached Figure Description

[0056] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0057] Figure 1 A flowchart illustrating an intelligent instrument dispensing and safety interlock management method for an operating room, provided as an embodiment of this application;

[0058] Figure 2 This is a schematic diagram of the structure of an intelligent operating room equipment dispensing and safety interlock management system provided in an embodiment of this application. Detailed Implementation

[0059] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0060] In the description of this application, unless otherwise stated, "a number" means two or more.

[0061] The intelligent operating room instrument fitting and safety interlock management system and method based on multi-source information fusion and full-process closed loop provided in this application embodiment are mainly applied to the field of intelligent management of operating room instruments. Through dynamic information carriers, multi-level safety verification, physical interlock control and self-learning optimization and other technical means, it realizes full-process intelligent management of accurate preoperative instrument fitting, intraoperative safety interlock and postoperative traceability closed loop, thereby overcoming the technical problems existing in the management of operating room instruments, such as blind opening, lagging configuration, lack of personalization, inefficient scheduling, no tracking of circulation and reliance on paper documents for handover.

[0062] Example 1:

[0063] Please see Figure 1 The embodiments of this application provide a method for intelligent instrument distribution and safety interlock management in the operating room based on multi-source information fusion and a closed-loop process, including M1 to M6, and the specific implementation steps are as follows:

[0064] M1 establishes a wireless communication connection with the central server through the dynamic information carrier module, enabling remote dynamic updates of the medical device package contents.

[0065] In this embodiment of the application, step M1 includes M1.1 to M1.3, wherein M1.1 is the process of configuring the dynamic information carrier, M1.2 is the process of establishing wireless communication, and M1.3 is the process of realizing remote updates, specifically as follows:

[0066] M1.1 A dynamic information carrier module is attached to the surface of each individual instrument pack. This module includes: a human-readable display area displaying a summary of key instruments, applicable surgical types, and sterilization expiration dates in large font; a machine-readable encoding area storing a dynamic QR code, the encoded content of which includes at least a unique list of instruments within the pack, version number, sterilization batch number, and current status indicator; and a wireless update unit receiving update commands pushed by a central server via Wi-Fi or 4G / 5G networks, triggering screen refresh and synchronous updates of the QR code content.

[0067] The specific hardware specifications of the dynamic information carrier module are as follows: the main control chip is ESP32-WROOM-32E (operating temperature range -40℃~+85℃), with 4MB Flash storage; the display screen is a 4.2-inch three-color e-ink screen (resolution 400×300); the internal PCB is coated with conformal coating (temperature resistance ≥150℃) and filled with an aerogel heat insulation layer; the module is covered with a medical-grade IP68 protective shell (material is polytetrafluoroethylene / PTFE), which can withstand 134℃ / 0.22MPa high-temperature and high-pressure steam sterilization (45min cycle) or hydrogen peroxide plasma sterilization; before sterilization, the module automatically enters a deep sleep mode (power consumption <1μA), and after sterilization, it resumes working status through a low-frequency 125kHz RFID wake-up signal; the power supply is a CR2032 button battery (capacity 220mAh, battery life ≥3 years or ≥2000 refreshes), and the battery compartment and shell are integrated and sealed.

[0068] Before the patient's surgery begins, the dynamic information carrier module displays key instrument summary information of the instrument pack through a human eye-readable display area, including instrument name, quantity, and specifications. The key instrument summary is automatically extracted from the instrument pack template library according to the type of surgery. The extraction rule is to select the top 5 instruments with the largest quantity or the highest frequency of use from the standard instrument list corresponding to the type of surgery as the key instrument summary. The font size of the key instrument summary is no less than 24pt, ensuring that nurses can clearly see it from 3 meters away.

[0069] The update mechanism for the content displayed in the human eye-readable display area is as follows:

[0070] ① The central server generates update instructions based on changes in surgical schedules or changes in instrument pack contents and pushes them to the dynamic information carrier module via the wireless update unit;

[0071] ② After receiving the update command, the dynamic information carrier module first verifies the increment relationship of the version number. If the new version number is greater than the current version number, a refresh operation is performed.

[0072] ③ The refresh operation includes two parallel actions: updating the content displayed on the e-ink screen, and rewriting the content of the dynamic QR code and updating the version number simultaneously;

[0073] ④ After the update is completed, the dynamic information carrier module sends a confirmation receipt to the central server to complete the update process.

[0074] The version number mechanism for dynamic QR codes is as follows:

[0075] ① The version number uses an 8-digit numeric encoding in the format YYYYMMDD-NNN, where YYYYMMDD represents the update date and NNN represents the update sequence number for that day;

[0076] ② Each time content is updated, the version number automatically increments. The central server generates a new version number (the date is taken as the current day, and the sequence number is taken as the previous version number + 1). After receiving the update command, the dynamic information carrier module first compares the new version number with the local version number: if the new version number is greater than the local version number, the update is executed and the new version number is written to local storage; if the new version number is less than or equal to the local version number, it means that the command has expired, the module discards the command and requests the latest data from the server; when the central server pushes the update command, it includes a command generation timestamp. The module sets the cache validity period to 30 minutes. Cache data that exceeds the validity period is automatically marked as expired, and the latest data is requested first during the next update. The above timestamp verification mechanism prevents information inconsistency caused by local cache expiration.

[0077] ③ The QR code is encoded using Base64 and includes a list of medical devices in JSON format, version number, sterilization batch number, and status pointer, where the status pointer points to the current status record in the medical device lifecycle status table.

[0078] M1.2. A data interaction channel is established between the central server and the global data platform. The central server and the global data platform include: an interface unit that connects to the HIS (Hospital Information System) via the HL7 / FHIR protocol to obtain patient information, connects to the surgical scheduling system via WebService to obtain surgical plans, and communicates with devices such as intelligent instrument cabinets and verification terminals via the MQTT protocol. The interface unit connects to the HIS system via the HL7 FHIR R4 protocol, using FHIR Patient resources to obtain patient information. Key fields are mapped as follows: Patient ID ← FHIR Patient.identifier, Name ← FHIRPatient.name.text, Hospitalization Number ← FHIR Patient.identifier.system='hospital-mrn'; and it connects to the surgical scheduling system via a RESTful API using POST... The ` / schedule / operation` interface retrieves surgical plans. Interface parameters include the surgical date range. It returns a list of surgical plans in JSON format, containing fields such as surgical ID, patient ID, surgeon ID, estimated start time, surgical type, and procedure template ID. The database unit maintains a surgical scheduling table, a four-level preference knowledge base, an instrument dictionary, an instrument package template library, an instrument lifecycle status table, an instrument-package ownership table, and a handover record table.

[0079] The patient ID read by the wristband is compared with the patient ID obtained by FHIR. If they do not match, a warning pops up to prompt the nurse to verify. The nurse is allowed to choose to verify based on the wristband information or the system information. Abnormal situations are recorded in the log.

[0080] The specific data structure of the four-level preference knowledge base includes: a specialty group identifier field, which records the specialty group number to which the instrument belongs; a doctor identifier field, which records the unique employee number of the surgeon; a surgery name field, which records the standard name or alias mapping of the surgery; a step number field, which records the sequence number of the surgical step; a list of required instrument IDs field, which records the set of instrument IDs that must be used in this step; and a list of optional instrument IDs field, which records the set of instrument IDs that can be used in this step.

[0081] The alias mapping mechanism is implemented as follows:

[0082] ① In addition to storing the standard surgical name, the surgical name field also stores the common aliases for the surgical procedure in different specialty groups or by different doctors;

[0083] ② When performing surgical name matching, the system first performs an exact match. If no match is found, the system starts a fuzzy matching engine to search the alias list.

[0084] ③ Fuzzy matching uses the Levenshtein Distance algorithm, and a match is considered successful when the similarity exceeds 0.8;

[0085] ④ Each matching result is recorded in the log to continuously optimize the integrity of the alias mapping table.

[0086] M1.3 receives update instructions pushed by the central server through the wireless update unit, triggering screen refresh and synchronous update of QR code content;

[0087] The wireless update unit uses the MQTT protocol for communication, and the specific subscription and publish mechanism is as follows:

[0088] ① The dynamic information carrier module subscribes to the topic / device / package / {package_id} / update to receive update instructions from the central server;

[0089] ②When the surgical schedule changes, the contents of the instrument pack are adjusted, or the status changes, the central server issues update instructions to the corresponding topic;

[0090] ③ The update command contains a complete content data packet, which is parsed by the receiving end and written directly to local storage without any additional requests;

[0091] ④ The update frequency is adaptively adjusted according to the business scenario, with a 5-minute interval during peak surgical periods and a 30-minute interval during off-peak periods.

[0092] In this embodiment, M1 achieves contactless updates of instrument pack information through real-time communication between the dynamic e-ink screen and the central server, avoiding the tediousness and errors of manually changing labels. The human eye-readable display area adopts a large font design, which makes it easy for nurses to quickly identify key information of the instrument pack before surgery and determine whether the pack meets the basic requirements of the current surgery without scanning. The dynamic QR code adopts a version number mechanism, which, together with the timestamp verification of the central server, can effectively prevent information inconsistency caused by local cache expiration. The wireless update unit utilizes the existing hospital wireless network infrastructure, without the need for additional communication cables, resulting in low modification costs and a short implementation cycle.

[0093] M2 uses a preoperative intelligent scheduling and positioning module to perform global search and intelligent scheduling of instrument packs, solving the pain points of difficulty in finding instruments and avoiding surgical conflicts before surgery.

[0094] In this embodiment of the application, step M2 includes M2.1 to M2.3, where M2.1 is the global search process, M2.2 is the state tracking process, and M2.3 is the intelligent scheduling and allocation process, specifically as follows:

[0095] M2.1 The global instrument search unit performs the following steps: First, it matches the input instrument name in the instrument dictionary to obtain the instrument ID; Second, it queries the instrument-package affiliation table to find the package ID where the instrument ID is currently located; Third, it retrieves the package name and current status from the instrument package template library based on the package ID, and returns a list of all instrument packages containing the instrument and the current location of each package.

[0096] The global instrument search employs a fuzzy matching algorithm and supports multiple search methods, including initials of pinyin and aliases.

[0097] ① Pinyin initial letter matching: Convert the Chinese name entered by the user into a sequence of initial letters of pinyin and perform fuzzy search in the medical device dictionary;

[0098] ② Alias ​​Search: Perform full-text search in the alias field of the medical device dictionary, supporting partial matching;

[0099] ③Synonym expansion: Establish a thesaurus, and when a user enters a device name, automatically expand the search to include related devices with synonyms;

[0100] ④ Search results sorting: sorted in descending order of relevance score, which is calculated by considering the number of matching fields, matching position, and matching completeness.

[0101] M2.2 The real-time instrument status tracking unit, by linking the instrument status table with the location sensor data, can return location information that can distinguish whether the instrument pack is located in the waiting area of ​​the sterilization supply room or has been distributed to the operating room instrument cabinet, providing an accurate reference for the main teacher's allocation decision.

[0102] The positioning accuracy for real-time tracking of device status is divided into three levels:

[0103] ① Warehouse-level positioning: Positioning accuracy down to the warehouse location number of the disinfection supply room (e.g., CSSD-01-A area), achieved through RFID tags and fixed readers at the warehouse entrance;

[0104] ② Operating room level positioning: The positioning accuracy is down to the instrument cabinet number in the operating room (e.g., OR-01-cabinet A), which is achieved by the RFID reader at the entrance of the operating room and the embedded tag inside the cabinet;

[0105] ③ Cabinet-level positioning: The positioning accuracy is down to the specific cabinet compartment of the intelligent medical device cabinet, which is achieved through the built-in RFID module of the intelligent medical device cabinet, with a positioning accuracy of ±5 cm.

[0106] M2.3 The intelligent scheduling and allocation unit detects allocation conflicts of instrument packs in real time. When the time interval between two surgeries is less than the minimum turnaround time of the instrument pack, the system automatically prompts the conflict and provides suggestions for replacement or adjustment of the surgical order, thus avoiding surgical delays caused by insufficient instrument turnaround time.

[0107] The conflict detection algorithm of the intelligent scheduling and allocation unit is as follows:

[0108] Extract the actual duration of the most recent n turnovers of the medical device package from the medical device circulation record table to form a time series T=[t1,t2,…,t n ] Calculate the historical average turnaround time Obtain the planned start times of assigned surgery A and unassigned surgery B. and (Units are all minutes), calculate the time interval. ;

[0109] like If both surgeries use the same instrument pack, it is considered a conflict and an alarm message is output.

[0110] Among them, historical average turnaround time The calculation method is as follows:

[0111]

[0112] in, For this instrument pack recently Actual turnover time (unit: minutes). The value is 50; if the instrument pack is newly received or historical data is insufficient. Next, the default turnaround time (T) for this type of instrument will be used. default The default value is 120 minutes.

[0113] In this embodiment, M2 addresses the core pain points of "difficulty in finding instruments and resolving conflicts" for lead surgeons when allocating instruments before surgery through the global search and conflict detection functions of the preoperative intelligent scheduling and positioning module. The fuzzy matching algorithm supports multiple search methods such as initial letters of pinyin and aliases, improving the ease of use of the search. By associating the instrument status table with position sensor data, the returned location information can distinguish different storage locations of instrument packs, providing accurate reference for allocation decisions. The conflict detection algorithm introduces the dynamic parameter of historical average turnaround time, enabling conflict detection to adapt to the actual processing time of different instruments, providing more buffer time for precision instruments with slower turnaround times, and improving the accuracy and practicality of the algorithm.

[0114] M3. Multi-level interlocking safety verification is performed through the intelligent verification terminal in the operating room to ensure that the configuration of instruments and surgical steps are accurately matched during the operation.

[0115] In this embodiment of the application, step M3 includes M3.1 to M3.2, where M3.1 is the process of configuring the verification terminal, and M3.2 is the process of performing five verifications, specifically as follows:

[0116] The M3.1 operating room intelligent verification terminal includes: a barcode scanning module that supports reading one-dimensional barcodes, two-dimensional barcodes, and RFID tags; a patient wristband reading module that supports reading patient identity information via NFC or barcode; a display module configured as a 10-inch or larger industrial-grade touchscreen; and a multi-level interlocking security verification unit that performs five sequential verifications in software.

[0117] The barcode scanning module employs a multimodal recognition design, enabling it to adapt to different packaging forms of medical device packages and individual device identification methods. Specifically, it supports:

[0118] ① One-dimensional barcodes: Commonly used encoding formats include Code 128, Code 39, and EAN-13;

[0119] ② QR codes: Common encoding formats such as QR Code, Data Matrix, and PDF417;

[0120] ③ RFID tags: High-frequency RFID tags conforming to ISO 15693 or ISO 14443 standards, with a reading distance of not less than 10 cm.

[0121] The patient wristband reading module interfaces with the hospital's HIS system, and the acquired patient information can be automatically entered into the verification process, avoiding the risk of errors from manual input. The specific interface method is as follows:

[0122] ① Obtain basic patient information in real time via the HL7 FHIR protocol, including patient ID, name, hospital number, and surgical schedule information;

[0123] ② The wristband reader supports two modes: NFC mode to read the NFC wristband worn by the patient, and barcode mode to read the QR code printed on the wristband;

[0124] ③ After successful reading, the patient information field in the verification process will be automatically filled in, and the patient identity verification process will be started.

[0125] M3.2 The five verifications performed by the multi-level interlocking safety verification unit are as follows:

[0126] Verification 1: Forced review and confirmation. Check whether the nurse has clicked the "I have reviewed the list" button. If not, output failure code E01.

[0127] Verification 2: Instrument package applicable surgical type matching. Verify whether the list of applicable surgical types in the QR code contains the name of the currently scheduled surgery. If not, output failure code E02.

[0128] Verification 3: Patient identity matching. Verify whether the patient ID read by the wristband matches the patient ID in the scheduling table. If they do not match, output failure code E03.

[0129] Verification 4: Sterilization validity period verification. Verify whether the sterilization date plus the validity period in the QR code exceeds the current system time. If it has expired, output failure code E04.

[0130] Verification 5: Verification of instrument satisfaction for the current surgical step. Call the preference matching operation unit to compare the set of required instruments for the current step with the list in the package. If there are missing instruments, output failure code E05 and return the list of missing instruments.

[0131] In the fifth verification, the system first obtains the step number of the current surgery. Specifically, it obtains the standard step flow template of the current surgery from the surgery schedule table, calculates the current surgical stage based on the surgery start timestamp, and maps the surgical stage to the corresponding step number; then it calls the preference matching operation unit to perform matching.

[0132] The workflow of the preference matching operation unit is as follows:

[0133] Input the current surgery ID and the current step number. Based on the surgery ID, obtain the specialist group ID, the surgeon ID, and the surgery name. Locate the record of the doctor, the surgery, and the step in the four-level preference knowledge base. Extract the set of required instrument IDs and the set of optional instrument IDs. Compare the set of required instrument IDs with the list in the package. Output the list of required instrument IDs that do not appear in the list in the package as the missing set.

[0134] The comparison between the set of mandatory device IDs and the package list is implemented using Java HashSet difference operation: missingSet = mandatoryIDs - packageContentList; if missingSet is not empty, return failure code E05 and the missingSet list; if it is empty, return verification passed.

[0135] If the current step does not exist in the preference library, it will be downgraded to use the general mandatory instrument set for the entire surgical procedure for matching. The general mandatory instrument set is pre-stored in the instrument package template library and associated with the surgical name. It is manually configured by the instrument administrator according to the surgical type. When configuring, the administrator selects mandatory instruments from the standard instrument list for the surgical type, and the system automatically stores it as a general configuration.

[0136] The triggering conditions and processing flow of the downgrade mechanism are as follows:

[0137] ① When the preference matching operation unit fails to find the configuration record for the current step in the fourth-level preference knowledge base, the degradation process is automatically triggered;

[0138] ② The downgrade process extracts the set of required instruments from the general configuration at the surgical level and uses it to replace the configuration at the step level for matching;

[0139] ③ After downgrading and matching, if there are still missing instruments, the system outputs failure code E05 and enters the intelligent guidance process. At the same time, the guidance interface will display the message "The current step preference library is not configured, and the surgical general configuration has been used."

[0140] ④ The downgrade matching results and their triggering events are recorded in the log, including the surgery ID, step number, and timestamp that triggered the downgrade, providing data support for the improvement of the preference library in the future;

[0141] ⑤ The system sends a notification to the administrator that the preference library is missing, suggesting that the configuration data for this step be supplemented.

[0142] This embodiment of M3 ensures the integrity and rigor of the verification process through a five-step sequential execution mechanism. Failure of any verification interrupts the subsequent process and locks the equipment, effectively preventing unqualified instrument packs from entering the surgical procedure. By differentiating different failure codes and corresponding handling strategies (e.g., E02 only prompts for instrument pack replacement, while E05 triggers intelligent guidance), the system's differentiated response capability is improved. The barcode scanning module adopts a multimodal recognition design, adapting to different packaging forms of instrument packs and the identification methods of individual instruments. The patient wristband reading module interfaces with the hospital's HIS system, and the acquired patient information can be automatically filled into the verification process, avoiding the risk of errors from manual input. The degradation mechanism ensures that new surgical procedures or steps can be verified using a common configuration, avoiding the embarrassing situation where the system cannot function due to a missing preference library.

[0143] M4 automatically locks the shadowless lamp and operating table when verification fails through differentiated physical interlocking and intelligent guidance execution unit, and intelligently guides the acquisition of missing instruments.

[0144] In this embodiment of the application, step M4 includes M4.1 to M4.3, wherein M4.1 is the process of configuring the physical interlocking hardware, M4.2 is the process of executing the interlocking action, and M4.3 is the process of executing the intelligent guidance, specifically as follows:

[0145] The M4.1 differentiated physical interlocking and intelligent guidance execution unit includes: a PLC controller that communicates with the verification terminal via Modbus TCP protocol; a shadowless lamp power control circuit, with the PLC's digital output port connected in series to the shadowless lamp's power contactor coil; an operating table hydraulic lock circuit, with the PLC's digital output port controlling the power supply to the operating table's moving hydraulic valve; and an intelligent instrument cabinet, each compartment equipped with an electronic lock and a blue LED indicator, connected to the PLC via an RS485 bus.

[0146] The electrical connection circuit between the PLC and the shadowless lamp and operating table is connected in series with a safety isolation relay and a solid-state optocoupler that conform to the IEC 60601-1 medical safety standard. The PLC only outputs low-level control signals and is not directly connected to the main power supply circuit of the equipment. The control circuit is configured with fail-safe logic. When the PLC communication is interrupted, the power supply is abnormal, or a sudden change in the hydraulic pressure of the operating table is detected, the relay switches to the normally closed safety state by default and restores the power supply to the equipment. At the same time, it reports the interlocking abnormal event to the central server.

[0147] The intelligent medical device cabinet is implemented as follows: each compartment is equipped with an electromagnetic lock of model GZ-24V, rated voltage 24V DC, driving current 0.5A, and the contacts are open when locked; each compartment is equipped with a blue LED indicator, model LED-5mm, with a 220Ω current-limiting resistor in series, and the 2Hz flashing is controlled by the duty cycle of the PLC's PWM port; the intelligent medical device cabinet controller communicates with the PLC via RS485 bus, with a baud rate of 9600bps, 8 data bits, 1 stop bit, and no parity; the compartment address encoding rule is: cabinet number (1-4) << 8 + compartment number (1-16), for example, the address of compartment 3 of cabinet 1 is 0x0103; the PLC maintains the medical device ID-compartment address mapping table, and the unlock command frame format is: 0x01 (device address) + 0x05 (function code) + target compartment address + 0xFF00 (unlock) + CRC16 check.

[0148] The selection and configuration requirements for the PLC controller are as follows:

[0149] ①CPU module: It has at least 16 digital inputs and 16 digital outputs, and supports Modbus TCP protocol communication;

[0150] ② Communication interface: Equipped with an Ethernet port, it can be connected to the switch in the operating room via a network cable to realize MQTT communication with the verification terminal;

[0151] ③ Programming method: Ladder diagram programming is adopted, and the core logic includes interlocking state machine, cabinet control logic and bypass detection logic;

[0152] ④ Redundancy design: Key control loops are configured with hard-wired connections, and automatically switch to bypass mode and trigger an alarm when the PLC fails.

[0153] M4.2 When the verification terminal generates any failure code, the terminal sends a lock command to the PLC via MQTT, carrying the failure code; the PLC immediately sets the digital output ports corresponding to the shadowless lamp and the operating table to a low level, causing the equipment to lose power and lock.

[0154] The specific communication protocol for the locking command is as follows:

[0155] ① The verification terminal acts as an MQTT client, sending a lock message to the topic / interlock / lock. The message payload is in JSON format and includes the failure code, operating room number, and timestamp.

[0156] ②The PLC subscribes to the topic and receives the lock message, then executes the corresponding lock action after parsing the failure code;

[0157] ③ After the locking action is executed, the PLC sends status feedback to the topic / interlock / status, including the locking status and the device status;

[0158] ④ The latency of the entire communication process is controlled within 100 milliseconds to ensure the real-time nature of the locking action.

[0159] M4.3 If the failure code is E05, the terminal simultaneously sends the list of missing device IDs to the PLC. The PLC queries the cabinet address mapping table based on the device IDs pre-issued by the server to find the cabinet where the missing device is located, sends an unlock command to the cabinet, and causes the blue LED indicator to flash at a frequency of 2Hz. The terminal screen displays guidance information. After the nurse takes the missing device and scans the device label, the terminal re-executes the fifth verification. If the missing set is empty, it sends a verification pass command to the PLC. The PLC restores the power supply to the device and sends a global unlock command to the smart device cabinet.

[0160] The specific implementation process of intelligent guidance is as follows:

[0161] ①After the verification terminal generates a list of missing device IDs, it queries the cabinet address corresponding to the device ID.

[0162] ② The cabinet address is sent to the PLC via MQTT message, and the PLC sends an unlock command to the corresponding intelligent medical cabinet via RS485 bus;

[0163] ③ After receiving the unlock command, the intelligent medical device cabinet will drive the blue LED indicator of the cabinet compartment to flash at a frequency of 2Hz, and at the same time release the electronic lock;

[0164] ④ The verification terminal screen displays guidance information, including the name, quantity, and location of the missing device in the cabinet;

[0165] ⑤ After the nurse removes the missing instrument, she uses the verification terminal to scan the instrument label for secondary confirmation;

[0166] ⑥ The verification terminal calls the preference matching operation unit again to perform verification five. If the missing set is empty, the verification is deemed successful.

[0167] In this embodiment, M4 uses a PLC controller as the core hub for interlocking execution, achieving a reliable conversion from software verification results to physical device control. By connecting the digital output port in series with the device's power circuit, it ensures that the device immediately loses power in the event of verification failure, preventing bypass startup through software. The independent control of the smart instrument cabinet at the cabinet level and LED indicators provide intuitive visual guidance for retrieving missing instruments. The 2Hz LED flashing provides a striking visual cue, avoiding the hassle of nurses searching for the target instrument among numerous cabinets. By requiring nurses to scan the missing instrument for secondary confirmation, a closed-loop operation of "retrieval-verification" is ensured, preventing secondary verification failures caused by mistakenly retrieving other instruments.

[0168] M5 enables electronic postoperative inventory counting and direct traceability of abnormal instruments through the supply room handover and traceability module.

[0169] In this embodiment of the application, step M5 includes M5.1 to M5.3, wherein M5.1 is the postoperative inventory process, M5.2 is the anomaly tracking process, and M5.3 is the status push process, specifically as follows:

[0170] M5.1, Postoperative Inventory Submodule, allows supply room staff to scan the QR code on the instrument pack to display a standard list, check off each item for inventory, and then call the electronic signature and conduct handover confirmation after confirmation;

[0171] The specific procedure for post-operative inventory checks is as follows:

[0172] ① Supply room staff use a mobile management app to scan the dynamic QR code on the instrument pack;

[0173] ② The system displays a standard list of the medical device package, including the device name, serial number, and quantity;

[0174] ③ Staff members check off each item to confirm, and the system automatically records the counted equipment;

[0175] ④ After the inventory is completed, staff members confirm the electronic signature on their mobile devices;

[0176] The electronic signature is implemented using digital signature technology. The signature content includes the employee number of the inventory personnel, the inventory timestamp, and the hash value of the inventory equipment list. The signature generation process is as follows: the mobile terminal obtains the digital certificate of the inventory personnel, uses the certificate private key to sign the "employee number of the inventory personnel + timestamp + SHA-256 digest of the JSON of the equipment list", and the signature result, together with the original signature text, is stored in the central server database.

[0177] ⑤ After the electronic signature is confirmed, the system automatically updates the status of the equipment package to "handed over" and generates a handover record; during signature verification, the hash value of the original signature is recalculated using the public key of the person counting the items in the signature, and compared with the stored signature result. If they match, the verification is successful; to prevent the signature data from being tampered with, the server additionally calculates the SHA-256 digest for each signature record and stores it in the blockchain evidence storage service. Any modification to the database record will result in an inconsistent digest.

[0178] M5.2, Anomaly Tracing Submodule: When an anomaly is found during inventory, the last surgical information of the instrument pack used will automatically pop up and a contact button will be provided.

[0179] The implementation mechanism for anomaly tracing is as follows:

[0180] ① When staff members check and find discrepancies in the number of instruments or damage to instruments, they should mark the abnormality on the mobile device and fill in the abnormality description;

[0181] ② The system queries the surgical information of the most recent use of the instrument pack based on the ownership of the instrument pack, including the operating room number, the surgeon, the operation time, and the circulating nurse;

[0182] ③ The system pop-up window displays the last usage information and provides two quick dial buttons: "Contact the attending surgeon" and "Contact the circulating nurse";

[0183] ④ The abnormal information is simultaneously pushed to the central server, generating an abnormal handling work order and assigning it to the corresponding responsible person.

[0184] ⑤ After the exception is handled, the person in charge fills in the handling result on the mobile device and confirms the archiving.

[0185] M5.3 Status Push Submodule: After each step of cleaning, disinfection, sterilization, etc. is completed, the system scans the label and updates the status. The system pushes status notifications to relevant users based on the subscription relationship.

[0186] The subscription mechanism for status push notifications is as follows:

[0187] ① When submitting an instrument usage request, operating room nurses may choose to subscribe to status notifications for that batch of instruments;

[0188] ②After completing each step of cleaning, disinfection, and sterilization, the supply room staff use a mobile device to scan the instrument label and select the current status;

[0189] ③ The system pushes notifications to subscribers based on status changes. Notification methods include APP push, SMS and WeChat Enterprise Account;

[0190] ④ The push notification includes the medical device package number, current status, estimated completion time, and information of the responsible personnel.

[0191] In this embodiment, the M5 establishes a complete closed loop for postoperative instrument circulation through the electronic signature and direct anomaly traceability functions of the supply room handover and traceability module. The postoperative inventory submodule replaces paper signatures with electronic signatures, which not only improves handover efficiency but also ensures the authenticity and immutability of electronic records through signature hashes. The anomaly traceability submodule enables rapid location and communication of anomalies by directly connecting to the previous user, reducing anomaly handling time from an average of 2 hours to less than 10 minutes. The status push submodule, through a message subscription mechanism, allows operating room nurses to know the processing progress of their requested instruments in real time, facilitating advance scheduling of surgical use.

[0192] M6. The four-level preference knowledge base is continuously updated through a self-learning optimization module based on frequent pattern mining, so as to achieve adaptive evolution of the matching strategy.

[0193] In this embodiment of the application, step M6 includes M6.1 to M6.3, where M6.1 is the data acquisition process, M6.2 is the pattern mining process, and M6.3 is the update strategy execution process, specifically as follows:

[0194] M6.1, Data Acquisition Unit, records the actual instrument data used in each surgery, including surgery ID, specialty group, doctor ID, surgery name, step number, and a list of actual instrument IDs used;

[0195] The actual method for collecting data on medical devices in use is as follows:

[0196] ① After the surgery, the nurse checks the instruments actually used in the surgery through the verification terminal, and the system automatically records them as the actual instrument set used;

[0197] ② For surgeries that do not use verification terminals, the usage status of instruments can be automatically read through RFID smart inventory pads, and the data read by the inventory pads can serve as a supplementary source of actual usage data;

[0198] ③ If the nurse does not complete the check mark within 24 hours after the operation, the system will automatically send a reminder message; if the check mark is still not completed after 48 hours, the data will be marked as "unconfirmed" and will not be included in subsequent data mining calculations;

[0199] ④ After being cleaned, the collected data is stored in the raw data table of the data acquisition unit, awaiting subsequent mining and processing.

[0200] M6.2, the learning algorithm unit, sets the minimum support threshold to 0.6 and the minimum confidence threshold to 0.7, and mines frequent itemsets by specialty group, by doctor, by surgery name, and by step.

[0201] The specific implementation steps of the FP-Growth algorithm are as follows:

[0202] ① Data preparation: Export the surgical usage records of the most recent 90 days from the data acquisition unit, build a transaction database, and each transaction is {Surgery ID, Specialty Group ID, Doctor ID, Surgery Name, Step Number, Instrument ID List};

[0203] ② Frequent 1-itemset filtering: Scan the transaction database, calculate the support for each device ID (support(A) = Count(A) / N), where N is the total number of the 30 most recent surgeries under this dimension, filter frequent 1-itemsets with support(A) ≥ 0.6, and construct an item header table;

[0204] ③ FP-Tree Construction: Insert the list of device IDs in each transaction into the FP-Tree according to the order of the item header table. Each node of the tree contains {device ID, count, parent node pointer, child node linked list}.

[0205] ④ Conditional pattern base mining: For each frequent item, recursively backtrack its conditional pattern base to generate a frequent itemset;

[0206] ⑤ Association rule generation: For frequent itemsets {A,B}, calculate the confidence score conf(A→B)=support(A∪B) / support(A), and filter association rules with conf(A→B)≥0.7;

[0207] ⑥ Recommended output: For association rules A→B that meet the threshold conditions, the output is a suggestion to add device B to the optional set of steps in which device A is located.

[0208] The specific formulas for calculating the minimum support threshold and the minimum confidence threshold are as follows:

[0209] Support S(A→B) = count(A∪B) / total samples , which represents the frequency with which instrument A and instrument B appear simultaneously in the same surgical procedure;

[0210] The confidence level C(A→B) = count(A∪B) / count(A) represents the conditional probability of using device B at the same time as using device A.

[0211] The system recommends adding device B to the optional set of this step only if S(A→B)≥0.6 and C(A→B)≥0.7.

[0212] The learning algorithm unit is also equipped with a preference conflict resolution mechanism: when there is a preference conflict among multiple surgeons for the same surgical procedure, the surgeons' job titles (chief physician: deputy chief physician: attending physician = 3:2:1) are weighted and fused with the actual usage frequency in the past 90 days, and the instruments with a weighted score ≥0.7 are included in the recommendation set.

[0213] The learning algorithm unit is also configured with cold start degradation logic: when a specialist group or doctor has less than 30 historical data, the system automatically calls the set of common required instruments for the same specialty for matching, and switches to personalized mining after the data accumulation reaches the standard.

[0214] M6.3 Update Strategy Unit: Automatically executes learning tasks at 2:00 AM daily. For each specialty group-doctor-surgery-step combination, calculates the matching degree between the required devices in the current preference library and the actual usage data. If the usage frequency of an optional device in the last 30 cases exceeds 0.6, it is upgraded to a required device. If the usage frequency of a required device in the last 30 cases is less than 0.4, it is downgraded to an optional device. If a new strongly correlated device combination is found and the confidence level exceeds 0.7, it is recommended to add it to the optional set. All update suggestions are written into the production environment preference library after manual confirmation.

[0215] The execution flow of the update strategy is as follows:

[0216] ①At 2:00 AM every day, the system automatically checks if there is any new data that meets the conditions (at least 30 unprocessed data). If so, it triggers a learning task and starts the FP-Growth mining task to extract historical data from the data collection unit for the most recent 90 days.

[0217] ② The mining task is executed according to four dimensions: specialty group, doctor, surgical name, and step number, generating frequent itemsets for each dimension;

[0218] ③ Calculate the matching degree and upgrade / downgrade suggestions for each combination, and generate a list of preference library update suggestions;

[0219] ④ After the learning task is completed, an updated suggestion list is generated and stored in the pending review table. At the same time, an email notification is sent to the preset administrator account. After logging into the system, the administrator confirms each suggestion one by one on the "Preference Library Review" page and can choose "Agree", "Reject" or "Agree after modification". After the review is completed, the system writes the approved suggestions into the production environment preference library in batches, and the effective time is 0:00 am the next day.

[0220] ⑤ If the administrator does not take any action within 72 hours, the system will automatically send a reminder message; suggestions that are still not reviewed after 7 days will be automatically invalidated.

[0221] ⑥ All update suggestions generated by the algorithm are put into a sandbox environment for at least 72 hours of clinical logic simulation verification. Only after the verification is passed can they be pushed to the manual confirmation queue.

[0222] This embodiment, M6, leverages the efficient mining capabilities of the FP-Growth algorithm to automatically discover frequent patterns and association rules in device usage from massive amounts of historical usage data, replacing the inefficient traditional method of manually maintaining preference databases. By setting reasonable support and confidence thresholds, it ensures the reliability of updated recommendations while maintaining the algorithm's sensitivity through a statistical window of 30 cases. A manual verification mechanism ensures that the system will not generate erroneous preference updates due to data noise, respecting the professional judgment of clinical experts. By using both support and confidence indicators for screening, it avoids incorrect recommendations caused by a single indicator. High support ensures that the association has sufficient generality in the statistical sample, while high confidence ensures that the association has sufficient reliability; together, they ensure the high quality of the recommendations.

[0223] Overall, this embodiment has the following beneficial effects:

[0224] This application achieves contactless updates of instrument pack information through real-time communication between a dynamic e-ink screen and a central server, avoiding the tediousness and errors of manually changing labels. Through the synergistic effect of a five-level interlocking verification mechanism and physical interlocking execution, it completely eliminates the unauthorized practice of blindly opening instrument packs based on memory, addressing issues in process, information, and execution. Through step-level preference matching and pre-verification functions, it achieves accurate verification of instruments required for the current surgical step, significantly reducing the frequency of searching for instruments during surgery. Through full lifecycle status tracking and intelligent conflict detection, it achieves global optimization of instrument resource allocation. Through the organic combination of the FP-Growth frequent pattern mining algorithm and the manual confirmation mechanism, the preference database can continuously learn from real clinical habits, maintaining synchronization with surgical techniques.

[0225] Example 2:

[0226] Please see Figure 2 The embodiments of this application provide an intelligent operating room equipment and safety interlock management system based on multi-source information fusion and full-process closed loop, including a dynamic information carrier module 100, a central server and full-domain data platform 200, a mobile management terminal APP 300, a preoperative intelligent scheduling and positioning module 400, an operating room intelligent verification terminal 500, a multi-level interlock safety verification unit 600, a differentiated physical interlock and intelligent guidance execution unit 700, a supply room handover and traceability module 800, and a self-learning optimization module 900 based on frequent pattern mining;

[0227] Among them, the dynamic information carrier module 100 is used to attach to the surface of each individual instrument pack to realize remote dynamic updates of the contents of the instrument pack.

[0228] The central server and global data platform 200 are deployed in the hospital data center to communicate with various subsystems and maintain surgical scheduling tables, four-level preference knowledge bases, medical device dictionaries, medical device package template libraries, medical device life cycle status tables, medical device-package ownership tables, and handover record tables.

[0229] The mobile management app 300 is for use by instrument management teachers and supply room staff. It provides functions such as instrument pack editing, global instrument search, surgical scheduling allocation, status reporting and anomaly handling.

[0230] The preoperative intelligent scheduling and positioning module 400 is deployed on the central server to provide instrument search, scheduling and status tracking functions;

[0231] The 500 intelligent verification terminal for operating rooms, equipped with a touch screen, is installed in each operating room to perform multi-level interlocking security verification.

[0232] The multi-level interlocking safety verification unit 600 is integrated into the verification terminal and is used to perform five sequential verifications in software form.

[0233] The differentiated physical interlocking and intelligent guidance execution unit 700 includes a PLC controller, a shadowless lamp power control circuit, an operating table hydraulic lock circuit, and an intelligent instrument cabinet, which is used to automatically lock the shadowless lamp and operating table when verification fails, and intelligently guide the acquisition of missing instruments.

[0234] The supply room handover and traceability module 800 is deployed on mobile devices and a central server to enable electronic postoperative inventory and direct traceability of abnormal instruments.

[0235] The self-learning optimization module 900, based on frequent pattern mining, is deployed on a central server to continuously update the four-level preference knowledge base, enabling adaptive evolution of the allocation strategy.

[0236] In one embodiment, the dynamic information carrier module 100 includes a display submodule 101, an encoding submodule 102, and a wireless update submodule 103;

[0237] Among them, the display submodule 101 is used to display the summary of key instruments in the package, the applicable surgical type and the sterilization expiration date in large font;

[0238] The coding submodule 102 is used to store dynamic QR codes, whose coding content includes at least a unique list of medical device codes in the package, version number, sterilization batch number and current status pointer;

[0239] The wireless update submodule 103 is used to receive update instructions pushed by the central server via Wi-Fi or 4G / 5G network, and trigger screen refresh and synchronize the update of QR code content.

[0240] This embodiment displays the submodule 101, encoding submodule 102, and wireless update submodule 103 by configuring the dynamic information carrier module as a remotely updatable e-ink screen display, enabling the information on the instrument pack surface to be synchronized with the central server in real time, avoiding the inherent defect of traditional static labels that cannot be updated in a timely manner. The display submodule 101 adopts a large font design, which makes it easy for nurses to quickly identify key information in the instrument pack before surgery. The dynamic QR code in the encoding submodule 102 adopts a version number mechanism, with the version number incrementing each time the content is updated. Combined with the timestamp verification of the central server, it can effectively prevent information inconsistency problems caused by local cache expiration. The wireless update submodule 103 utilizes the existing hospital wireless network infrastructure, without the need to deploy additional communication cables, resulting in low transformation costs and a short implementation cycle.

[0241] In one embodiment, the central server and global data platform 200 include an interface submodule 201 and a database submodule 202.

[0242] Among them, the interface submodule 201 is used to obtain patient information by connecting with HIS through the HL7 / FHIR protocol, obtain surgical plans by connecting with the surgical scheduling system through WebService, and communicate with devices such as smart instrument cabinets and verification terminals through the MQTT protocol.

[0243] Database submodule 202 is used to maintain the surgical scheduling table, the four-level preference knowledge base, the instrument dictionary, the instrument package template library, the instrument life cycle status table, the instrument-package ownership relationship table, and the handover record table.

[0244] In this embodiment, the interface submodule 201 adopts a standardized medical information exchange protocol, ensuring seamless integration between the system and the hospital's existing information system. The database submodule 202 uses a hybrid architecture of relational and document databases, which ensures the reliability of transactional operations and supports efficient storage and querying of unstructured data such as historical status timelines. The four-level preference knowledge base adopts a multi-dimensional index structure of specialist group ID → doctor ID → surgery name → step number, supporting device preference matching accurate to the surgical step level, providing data support for precise device allocation at the step level.

[0245] In one embodiment, the preoperative intelligent scheduling and positioning module 400 includes a search submodule 401, a tracking submodule 402, and a scheduling submodule 403;

[0246] The search submodule 401 is used to match the input device name in the device dictionary to obtain the device ID, query the device-package affiliation table to find the package ID where the device ID is currently located, obtain the package name and current status from the device package template library based on the package ID, and return a list of all device packages containing the device and the current location of each package.

[0247] The tracking submodule 402 is used to return the current location of the instrument pack in real time by associating the instrument status table with the location sensor data, and to distinguish whether the instrument pack is located in the waiting area of ​​the sterilization supply room or has been distributed to the operating room instrument cabinet.

[0248] The scheduling submodule 403 is used to detect instrument pack allocation conflicts in real time. When the time interval between two surgeries is less than the minimum turnaround time of the instrument pack, the system will automatically prompt the conflict and provide suggestions for replacing or adjusting the surgical order.

[0249] In this embodiment, the search submodule 401, tracking submodule 402, and scheduling submodule 403 address the core pain points of "difficulty in finding instruments and difficulty in resolving conflicts" for the lead instructor when allocating instruments before surgery through global search and conflict detection functions. The search submodule 401 uses a fuzzy matching algorithm and supports multiple search methods such as initial letters of pinyin and aliases, improving the ease of use of the search. The scheduling submodule 403 introduces the dynamic parameter of historical average turnaround time, enabling conflict detection to adapt to the actual processing time of different instruments, providing more buffer time for precision instruments with slower turnaround times, and improving the accuracy and practicality of the algorithm.

[0250] In one embodiment, the operating room intelligent verification terminal 500 includes a barcode scanning submodule 501, a patient reading submodule 502, a display submodule 503, and a verification submodule 504.

[0251] Among them, the barcode scanning submodule 501 is used to support the reading of one-dimensional barcodes, two-dimensional barcodes and RFID tags;

[0252] The patient reading submodule 502 is used to support reading patient identity information via NFC or barcode.

[0253] Display submodule 503 is configured as an industrial-grade touchscreen of 10 inches or larger;

[0254] Verification submodule 504 is used to perform five sequential verifications in software.

[0255] In this embodiment, the barcode scanning submodule 501 adopts a multimodal recognition design, which can adapt to the identification methods of instrument packs with different packaging forms and individual instruments; the patient reading submodule 502 interfaces with the hospital's HIS system, and the obtained patient information can be automatically filled into the verification process, avoiding the risk of errors from manual input; the large-size touch screen design of the display submodule 503 ensures the convenience of operations such as list viewing and step switching; the verification submodule 504 ensures the integrity and rigor of the verification process through a five-step verification sequential execution mechanism. If any verification fails, the subsequent process is interrupted and the device is locked, effectively preventing unqualified instrument packs from entering the surgical process.

[0256] In one embodiment, the five verifications performed by the multi-level interlocking safety verification unit 600 are as follows:

[0257] The mandatory check confirmation unit 601 is used to check whether the nurse has clicked the "I have checked the list" button. If not, it outputs failure code E01.

[0258] The applicable surgery type matching unit 602 is used to verify whether the list of applicable surgery types in the QR code contains the name of the currently scheduled surgery. If it does not contain it, the failure code E02 is output.

[0259] The patient identity matching unit 603 is used to verify whether the patient ID read by the wristband matches the patient ID in the scheduling table. If they do not match, it outputs a failure code E03.

[0260] The sterilization validity period verification unit 604 is used to verify whether the sterilization date plus the validity period days of the instrument pack exceeds the current system time. If it has expired, it outputs a failure code E04.

[0261] The step instrument satisfaction verification unit 605 is used to call the preference matching operation unit to compare the required instrument set of the current step with the package list and output the missing instrument list.

[0262] In this embodiment, the mandatory verification unit 601, the applicable surgical type matching unit 602, the patient identity matching unit 603, the sterilization validity period verification unit 604, and the procedure instrument satisfaction verification unit 605, through the sequential execution mechanism of five verifications, ensure the integrity and rigor of the verification process. If any verification fails, the subsequent process is interrupted and the equipment is locked, effectively preventing unqualified instrument packs from entering the surgical process. By distinguishing different failure codes and corresponding different handling strategies (e.g., E02 only prompts to replace the instrument pack, while E05 triggers intelligent guidance), the system's differentiated response capability is improved.

[0263] In one embodiment, the differentiated physical interlocking and intelligent guidance execution unit 700 includes a control submodule 701, a lighting control submodule 702, a bed control submodule 703, and a cabinet control submodule 704;

[0264] Among them, the control submodule 701 is used to communicate with the verification terminal via Modbus TCP protocol and configure an independent input interface for detecting the status of the emergency bypass switch on the shadowless lamp panel.

[0265] The lighting control submodule 702 is used to connect the digital output port of the PLC in series with the power contactor coil of the shadowless lamp to realize the on-off control of the power supply of the shadowless lamp.

[0266] The bed control submodule 703 is used to control the power supply of the hydraulic valve for moving the operating table via the digital output port of the PLC, thereby enabling the locking and unlocking of the operating table movement.

[0267] The cabinet control submodule 704 is used to equip each cabinet with an electronic lock and a blue LED indicator. It is connected to the PLC via an RS485 bus to realize independent control of the cabinet.

[0268] In this embodiment, the control submodule 701 serves as the core hub for interlocking execution, enabling a reliable conversion from software verification results to physical device control. The lighting control submodule 702 ensures that the device immediately loses power upon verification failure by connecting the digital output port in series with the device's power circuit, preventing bypass startup through software. The cabinet control submodule 704 provides intuitive visual guidance for obtaining missing medical devices through independent cabinet-level control and LED indicators in the intelligent medical device cabinet. The anti-bypass design plugs the loopholes that allow bypassing the interlocking mechanism through hardware means, ensuring the integrity and unavoidability of the interlocking safety strategy.

[0269] In one embodiment, the execution logic of the differentiated physical interlocking and intelligent guidance execution unit 700 is as follows:

[0270] When the verification terminal generates any failure code, the terminal sends a lock command to the PLC via MQTT, carrying the failure code.

[0271] The PLC immediately sets the digital output ports corresponding to the shadowless lamp and operating table to a low level, causing the equipment to lose power and lock.

[0272] If the failure code is E05, the terminal will simultaneously send the list of missing device IDs to the PLC. The PLC will look up the cabinet where the missing device is located in the cabinet address mapping table based on the device IDs pre-sent by the server, send an unlock command to the cabinet, and make the blue LED indicator flash at a frequency of 2Hz. The terminal screen will display the guidance information.

[0273] After the nurse retrieves the missing instrument and scans the instrument label, the terminal re-executes the fifth verification step. If the missing set is empty, it sends a verification pass command to the PLC. The PLC then restores power to the device and sends a global unlock command to the smart instrument cabinet.

[0274] This embodiment provides a striking visual cue through 2Hz LED flashing, avoiding the hassle of nurses searching for target instruments among numerous cabinets. By requiring nurses to scan for missing instruments for secondary confirmation, a closed-loop operation of "retrieval-verification" is ensured, preventing secondary verification failures due to mistakenly retrieving other instruments. Through the PLC's global unlock command, all cabinets are restored to their locked state at once after successful verification, improving operational efficiency. With the bypass prevention design, when a bypass signal is detected, the PLC immediately issues a 90dB audible and visual alarm and pushes the interlocking bypass event record to the central server. At this time, the operating table remains locked until successful verification before it can be reset.

[0275] In one embodiment, the supply room handover and traceability module 800 includes an inventory submodule 801, a traceability submodule 802, and a push submodule 803;

[0276] Among them, the inventory submodule 801 is used for supply room staff to scan the QR code of the instrument pack to display the standard list, check each item for inventory, and call the electronic signature and handover confirmation after confirmation.

[0277] The traceability submodule 802 is used to automatically pop up the surgical information of the last use of the instrument pack and provide a contact button when an abnormality is found during the inventory.

[0278] The push submodule 803 is used to scan the label and update the status after each step of cleaning, disinfection, sterilization, etc. is completed. The system pushes status notifications to relevant users according to the subscription relationship.

[0279] In this embodiment, the inventory submodule 801 replaces paper signatures with electronic signatures, which not only improves handover efficiency but also ensures the authenticity and immutability of electronic records through signature hashes. The traceability submodule 802 enables rapid location and communication of abnormal situations by directly connecting to the previous user, reducing the abnormal handling time from an average of 2 hours to less than 10 minutes. The push submodule 803, through a message subscription mechanism, allows operating room nurses to know the processing progress of their requested instruments in real time, facilitating advance scheduling of surgical use.

[0280] In one embodiment, the self-learning optimization module 900 based on frequent pattern mining includes a collection submodule 901, a mining submodule 902, and an update submodule 903.

[0281] Among them, the data acquisition submodule 901 is used to obtain the actual set of instruments used in each surgery by having the nurse select the instruments used on the terminal or by automatically reading them through the RFID smart inventory pad.

[0282] Mining submodule 902 is used to mine frequent itemsets by dimension based on the minimum support threshold of 0.6 and the minimum confidence threshold of 0.7 using the FP-Growth algorithm;

[0283] The update submodule 903 is used to convert the mining results into preference library update suggestions, which are then written to the production environment database after manual confirmation.

[0284] In this embodiment, the data acquisition submodule 901, mining submodule 902, and update submodule 903 utilize the efficient mining capabilities of the FP-Growth algorithm to automatically discover frequent patterns and association rules of device use from massive amounts of historical usage data, replacing the inefficient traditional method of manually maintaining the preference database. By setting reasonable support and confidence thresholds, the reliability of update suggestions is ensured, while the sensitivity of the algorithm is maintained through a statistical window of 30 cases. The manual confirmation mechanism ensures that the system will not produce erroneous preference updates due to data noise, respecting the professional judgment of clinical experts.

[0285] In one embodiment, the mobile management terminal APP 300 includes an editing interface 301, a search interface 302, an allocation interface 303, a reporting interface 304, and an exception handling interface 305.

[0286] Among them, the editing interface 301 allows the equipment management teacher to add or delete equipment in the package after scanning the QR code;

[0287] The search interface 302 allows users to enter the name of a device and returns a list of all device packs containing that device, along with the current location.

[0288] The allocation interface 303 allows the lead instructor to drag and drop instrument packs to allocate for each surgery and detect conflicts.

[0289] The reporting interface shows a 304 error, which allows supply room personnel to scan the label and select the current status.

[0290] The exception handling interface 305 is used to report exceptions and contact the previous user when an exception is found.

[0291] In this embodiment, the mobile management APP 300 extends functions such as equipment management and supply room handover to the mobile terminal, improving operational efficiency and flexibility. Through the editing interface 301, the equipment management teacher can adjust the contents of the equipment pack at any time and synchronize it to the dynamic information carrier module in real time through the wireless update unit. Through the search interface 302, the head teacher can quickly locate the current location of the required equipment. Through the allocation interface 303, the drag-and-drop operation simplifies the scheduling and allocation process, and the conflict detection function avoids resource waste.

[0292] Overall, this embodiment has the following beneficial effects:

[0293] This application configures the dynamic information carrier module as a remotely updatable e-ink display, enabling real-time synchronization of information on the instrument pack surface with the central server, avoiding the inherent shortcomings of traditional static labels that cannot be updated in a timely manner. By deploying a global data platform on the central server, unified management of multi-source data such as surgical scheduling, instrument dictionaries, preference knowledge bases, and status tables is achieved, providing a data foundation for intelligent decision-making. The mobile management APP extends functions such as instrument management and supply room handover to mobile devices, improving operational efficiency and flexibility. The global search and conflict detection functions of the preoperative intelligent scheduling and positioning module solve the core pain points of "difficulty in finding instruments and resolving conflicts" for the lead surgeon when allocating instruments before surgery. The multi-level verification and physical interlocking execution of the intelligent verification terminal in the operating room ensure that the instrument configuration for each surgery meets safety requirements. The electronic signature and direct anomaly traceability functions of the supply room handover and traceability module establish a complete closed loop for postoperative instrument circulation. The continuous evolution mechanism of the self-learning optimization module enables the system to continuously adapt to real clinical needs and reduce manual maintenance costs.

[0294] Example 3:

[0295] This application provides a computer-readable storage medium, which includes a stored computer program. When the computer program is executed, it controls the device where the computer-readable storage medium is located to perform the aforementioned intelligent operating room equipment fitting and safety interlock management method based on multi-source information fusion and full-process closed loop.

[0296] The aforementioned intelligent operating room instrument distribution and safety interlock management method based on multi-source information fusion and a closed-loop process, when implemented as a software functional unit and used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0297] The above are preferred embodiments of the present invention. It should be noted that, for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. An intelligent operating room instrument dispensing and safety interlock management system based on multi-source information fusion and a closed-loop process, characterized in that, include: The dynamic information carrier module is attached to the surface of each individual instrument pack to display and store the instrument information inside the pack, and communicates wirelessly with the central server to receive update instructions, so as to realize the synchronous update of the displayed content and the stored content. The central server and the global data platform are used to interact with the hospital information system, the surgical scheduling system and various functional modules of this system, and to maintain surgical scheduling data, preference knowledge base, device master data and device life cycle status data. The preoperative intelligent scheduling and positioning module is used to perform a global search based on the input instrument name to obtain all instrument packs containing that instrument and their current location, and to perform intelligent scheduling and conflict detection based on surgical scheduling information and instrument pack turnaround time. The intelligent verification terminal in the operating room is used to read patient identity information and instrument pack identification information, and trigger the execution of multi-level interlocking security verification. A multi-level interlocking security verification unit is integrated into the verification terminal or communicates with the verification terminal. It is used to perform multiple sequential verifications and output the corresponding failure code when any verification fails. A differentiated physical interlocking and intelligent guidance execution unit is used to receive the verification results from the verification unit and execute corresponding physical interlocking actions and intelligent guidance operations according to different verification results; The supply room handover and traceability module is used to realize the electronic recording of postoperative instrument count and the traceability of abnormal instruments; The self-learning optimization module based on frequent pattern mining is used to mine frequent itemsets and association rules based on the actual instrument data used in historical surgeries, so as to update the preference knowledge base.

2. The intelligent operating room instrument distribution and safety interlock management system based on multi-source information fusion and a closed-loop process as described in claim 1, characterized in that, The dynamic information carrier module includes a display submodule, an encoding submodule, and a wireless update submodule; the display submodule uses an electronic ink screen to display the summary of key instruments in the package, the applicable surgical type, and the sterilization expiration date in large font. The encoding submodule stores dynamic QR codes, the encoded content of which includes at least a unique list of medical devices in the package, version number, sterilization batch number, and current status pointer; the wireless update submodule receives update instructions pushed by the central server through a wireless network and triggers the screen refresh of the display submodule to update the QR code content of the encoding submodule synchronously.

3. The intelligent operating room instrument distribution and safety interlock management system based on multi-source information fusion and a closed-loop process as described in claim 1, characterized in that, The central server and global data platform include an interface submodule and a database submodule. The interface submodule connects to the HIS system to obtain patient information via the HL7 FHIR R4 protocol, connects to the surgical scheduling system to obtain surgical plans via the RESTful API, and communicates with various functional modules of the system via the MQTT protocol. The database submodule maintains the surgical scheduling table, the four-level preference knowledge base, the instrument dictionary, the instrument package template library, the instrument life cycle status table, the instrument-package ownership table, and the handover record table.

4. The intelligent operating room instrument distribution and safety interlock management system based on multi-source information fusion and a closed-loop process as described in claim 1, characterized in that, The preoperative intelligent scheduling and positioning module includes a search submodule, a tracking submodule, and a scheduling submodule. The search submodule matches the input instrument name in the instrument dictionary to obtain a list of all instrument packs containing that instrument and the current location of each pack. The tracking submodule correlates the instrument status table with position sensor data and returns the current location of the instrument pack in real time. The scheduling submodule performs conflict detection by obtaining the historical average turnaround time of the currently unassigned instrument pack. Obtain the planned start times of assigned surgery A and unassigned surgery B. and Calculate the time interval ,like If surgery A and surgery B use the same instrument pack, it is considered a conflict and an alarm message is output.

5. The intelligent operating room instrument distribution and safety interlock management system based on multi-source information fusion and a closed-loop process as described in claim 1, characterized in that, The multi-level interlocking safety verification unit performs five verifications in sequence: Verification 1, mandatory check and confirmation; if not checked, output failure code E01; Verification 2, instrument pack applicable surgical type matching; if mismatch, output failure code E02; Verification 3, patient identity matching; if mismatch, output failure code E03; Verification 4, sterilization expiration date verification; if expired, output failure code E04; Verification 5, current surgical step instrument satisfaction verification; if missing instruments exist, output failure code E05 and return the missing instrument list.

6. The intelligent operating room instrument distribution and safety interlock management system based on multi-source information fusion and a closed-loop process as described in claim 5, characterized in that, In the fifth verification, the system first obtains the step number of the current surgery, then calls the preference matching operation unit to locate the set of required instruments for the current step in the four-level preference knowledge base according to the specialty group, surgeon and surgery name of the current surgery. The system then compares the set with the instrument list in the package and outputs the list of missing instruments. If the current step does not exist in the preference library, it will be downgraded to use the general required instrument set for the procedure for matching.

7. The intelligent operating room instrument distribution and safety interlock management system based on multi-source information fusion and a closed-loop process as described in claim 1, characterized in that, The differentiated physical interlocking and intelligent guidance execution unit includes a PLC controller, a shadowless lamp power control circuit, an operating table hydraulic lock circuit, and an intelligent instrument cabinet. The PLC controller communicates with the verification terminal via the Modbus TCP protocol. The shadowless lamp power control circuit controls the on / off state of the shadowless lamp power supply via the digital output port of the PLC. The operating table hydraulic lock circuit controls the locking and unlocking of the operating table movement via the digital output port of the PLC. Each compartment of the intelligent instrument cabinet is equipped with an electronic lock and indicator light, connected to the PLC via an RS485 bus, and the PLC controls the unlocking and indication of the corresponding compartment according to the missing instrument list.

8. The intelligent operating room instrument distribution and safety interlock management system based on multi-source information fusion and a closed-loop process as described in claim 7, characterized in that, The PLC controller is also equipped with an independent input interface for detecting the status of the emergency bypass switch; when a bypass signal is detected, the PLC triggers an audible and visual alarm and pushes the interlocking bypass event record to the central server, while maintaining the locked state of the operating table.

9. A method for intelligent instrument distribution and safety interlock management in the operating room based on multi-source information fusion and a closed-loop process, characterized in that, Includes the following steps: Information update and synchronization steps: Wireless communication is established between the instrument pack and the central server through a dynamic information carrier attached to the surface of the instrument pack, receiving and executing update commands to achieve remote dynamic updates of the instrument pack information; Preoperative scheduling and positioning steps: A global search is performed based on the input instrument name to obtain all instrument packs containing that instrument and their current locations; intelligent scheduling and conflict detection are performed based on surgical scheduling information and instrument pack turnover time; Multi-level safety verification steps: Patient and instrument pack information is read through the intelligent verification terminal in the operating room, and multiple sequential verifications are performed. If any verification fails, the subsequent process is interrupted and the corresponding failure code is output; Differentiated interlocking and guidance steps: Based on the failure code output by the verification steps, the corresponding physical interlocking action and intelligent guidance operation are executed; Postoperative traceability steps: Electronic recording of postoperative instrument count and traceability of abnormal instruments are achieved; Self-learning optimization steps: Based on the actual instrument data used in historical surgeries, a frequent pattern mining algorithm is used to mine frequent itemsets and association rules to update the preference knowledge base.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which is called and executed by a processor to implement the intelligent operating room instrument fitting and safety interlock management method based on multi-source information fusion and full-process closed loop as described in any one of claims 1-9.

Images