Intraoperative instrument full-process closed-loop management system and method based on intelligent scanning

Abstract

The invention discloses an intraoperative instrument full-process closed-loop management system and method based on intelligent scanning, and belongs to the field of medical information technology and machine vision. The method comprises the following steps: constructing an expected state hash table; cancel-after-verification is carried out on the recycled surgical instrument in real time, cancel-after-verification is carried out on the basis of the hash value of the current unique identity feature obtained through calculation of the recycled instrument, searching and removing operation is carried out in the expected state hash table, and the integrity state is synchronously verified; and executing a final counting decision, checking the final state of the expected state hash table after receiving a counting end instruction, and triggering a corresponding counting result according to whether the final state is empty or has an abnormal mark. According to the method and the device, the delay of final counting and judgment is reduced to a microsecond level through feedforward construction of an expected state data structure and a mechanism of post-operation real-time verification, zero-delay verification is realized, the technical problem of easy omission or misjudgment of manual counting is fundamentally eliminated, and the operation safety is greatly improved.

Description

Technical Field

[0001] This invention relates to the field of medical information technology, and in particular to a closed-loop management system and method for the entire process of intraoperative instruments based on intelligent scanning. Background Technology

[0002] Accurate instrument counting during surgical procedures is crucial for ensuring patient safety and preventing medical accidents. Traditional instrument counting relies primarily on nurses' manual visual identification, counting, and recording. This process is not only time-consuming and labor-intensive but also highly susceptible to omissions or misjudgments due to staff fatigue, inattention, or similar instrument shapes, potentially leading to serious medical accidents such as instruments being left inside the patient. To address this issue, existing technologies have proposed several auxiliary counting methods, such as attaching barcodes or RFID tags to instruments. However, these methods still have limitations. For example, barcodes are easily worn away during cleaning and sterilization, and RFID tags may be affected by the shielding effect of metal instruments. More importantly, most of these methods still follow the logic of "postoperative batch comparison," meaning that all recovered instruments are counted after surgery and then compared with the preoperative list. The inherent flaw of this model is that its verification process has a significant delay, and problems cannot be detected before the counting is completed. Therefore, how to provide a system and method that can automate, real-time, and zero-delay verify intraoperative instrument counting to completely eliminate the risk of errors from manual counting is a pressing technical problem to be solved in this field. Summary of the Invention

[0003] The purpose of this application is to provide a closed-loop management system and method for the entire intraoperative instrument process based on intelligent scanning, so as to solve the problems of low efficiency, easy error and high verification delay of traditional instrument counting methods mentioned in the background art.

[0004] In a first aspect, this application provides a closed-loop management method for intraoperative instruments based on intelligent scanning, comprising: constructing a predictive state hash table, wherein the predictive state hash table is generated based on the unique identity feature and integrity status feature of each surgical instrument obtained by preoperative scanning, and storing a hash value of the unique identity feature as the key and information associated with the surgical instrument as the value; real-time verification of the recovered surgical instruments, wherein the verification is based on scanning the recovered surgical instruments to obtain a current unique identity feature, calculating a current hash value of the current unique identity feature, and using the current hash value to perform lookup and removal operations in the predictive state hash table; and executing a final inventory decision, wherein after receiving an inventory completion instruction, the final state of the predictive state hash table is checked, and a corresponding inventory result is triggered according to the final state.

[0005] Optionally, the step of obtaining the unique identity feature and integrity status feature of each surgical instrument includes: acquiring a high-resolution image of the surgical instrument through an image acquisition device; and performing image processing on the high-resolution image to extract the unique identity feature and the integrity status feature.

[0006] Optionally, the unique identity feature includes at least one of a set of key point descriptors extracted based on a scale-invariant feature transformation algorithm or an accelerated robust feature algorithm, and a set of geometrically invariant moments calculated based on the contour of the surgical instrument.

[0007] Optionally, the step of using a hash value of the unique identity feature as a key includes: using a deterministic hash function to calculate the unique identity feature to generate a unique hash key as the key.

[0008] Optionally, the step of real-time verification and recycling of the surgical instruments, after finding a matching key-value pair in the expected state hash table using the current hash value, further includes: obtaining a current integrity status feature of the recycled surgical instruments; comparing the current integrity status feature with the integrity status feature stored in the matching key-value pair; and triggering an integrity anomaly alarm if the comparison result is inconsistent.

[0009] Optionally, if the comparison results are inconsistent, after triggering the integrity anomaly alarm, the matching key-value pair will be marked as an abnormal entry in the expected state hash table, instead of being removed.

[0010] Optionally, the step of executing the final inventory decision includes: if the final state of the expected state hash table is empty, triggering an inventory success result; if the final state of the expected state hash table is not empty, triggering an inventory result for missing instruments, and displaying the surgical instrument information corresponding to all remaining unremoved key-value pairs in the expected state hash table.

[0011] Secondly, this application provides a closed-loop management system for the entire surgical instrument process based on intelligent scanning, comprising: a preoperative scanning and expected state construction module, used to construct an expected state hash table based on the unique identity feature and integrity status feature of each surgical instrument obtained by preoperative scanning, wherein the expected state hash table stores a hash value of the unique identity feature as the key and information associated with the surgical instrument as the value; a postoperative real-time verification and anomaly adjudication module, used to perform real-time verification of the recovered surgical instruments, the module being configured to scan the recovered surgical instruments to obtain a current unique identity feature, calculate a current hash value of the current unique identity feature, and perform lookup and removal operations in the expected state hash table using the current hash value; the postoperative real-time verification and anomaly adjudication module is also used to execute a final inventory adjudication after receiving an inventory completion instruction, which checks a final state of the expected state hash table and triggers a corresponding inventory result based on the final state; and a full-process data archiving and traceability module, used to store the construction record of the expected state hash table, the verification process record, and the result of the final inventory adjudication.

[0012] Optionally, the postoperative real-time verification and anomaly adjudication module is further configured to: after finding a matching key-value pair in the expected state hash table using the current hash value, obtain a current integrity status feature of the recycled surgical instrument; compare the current integrity status feature with the integrity status feature stored in the matching key-value pair; and, if the current integrity status feature is inconsistent with the integrity status feature, trigger an integrity anomaly alarm.

[0013] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in any of the preceding claims.

[0014] This application introduces a feedforward hash table of expected state and a real-time verification mechanism to pre-position and amortize the verification logic of traditional postoperative batch comparison. In the preoperative stage, the system constructs an efficient data structure for real-time verification, ensuring that each instrument scanned postoperatively requires only one hash lookup and removal operation with a time complexity of O(1). When the count is complete, the final decision requires only a single instantaneous check of whether the hash table is empty, reducing latency to the microsecond level and achieving "zero-latency" verification. This architectural innovation not only fundamentally eliminates the verification latency caused by batch comparison but also smoothly amortizes the computational load from the postoperative peak to each independent scanning action, ensuring high responsiveness and stability of the system. This solves the technical problems of high count latency and inability to detect anomalies in real time in existing technologies, achieving the beneficial effects of reducing the risk of errors in manual counts and improving surgical safety. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating a closed-loop management method for intraoperative instruments based on intelligent scanning, provided in an embodiment of this application. Figure 2 This is a schematic diagram of a closed-loop management system for intraoperative instruments based on intelligent scanning. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0017] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0019] This application provides a closed-loop management system and method for the entire intraoperative instrument process based on intelligent scanning. In a specific implementation, this method employs a feedforward mechanism using a hash table of expected states and real-time verification. This moves the verification logic of traditional postoperative batch data comparison to the preoperative stage for data structure construction, and performs individual, real-time verification during postoperative retrieval. This enables near-zero-latency identity and status verification of retrieved instruments, distributing the computational load from postoperative peak values ​​across the entire retrieval process. This method solves the technical problems of existing technologies where surgical instrument counting relies on manual labor, is prone to omissions or misjudgments, and suffers from unacceptable delays in the verification process. It achieves the beneficial effects of reducing the risk of errors in manual counting and improving surgical safety.

[0020] The following will refer to Figure 1 The methods of the embodiments of this application will be described in detail. Figure 1 This is a flowchart illustrating a closed-loop management method for intraoperative instruments based on intelligent scanning, provided in one embodiment of this application.

[0021] S100: Construct a expected state hash table, which is generated based on the unique identity feature and integrity status feature of each surgical instrument obtained from the preoperative scan, and stores the hash value of the unique identity feature as the key and the information associated with the surgical instrument as the value.

[0022] In one embodiment, the core task of S100 is to establish a digital archive for all surgical instruments to be used before the surgery begins, allowing for real-time verification. This digital archive is organized into a specific data structure, namely the Expected State Hash Table (ESHT). A hash table, as a data structure that maps inputs (keys) of arbitrary length to fixed-length outputs (hash values) using a hash function, offers the significant engineering advantage of providing lookup, insertion, and deletion operations with an average time complexity of O(1). This application leverages this characteristic to build a high-performance verification engine.

[0023] Specifically, the execution process of S100 can be broken down into a series of sequential sub-steps.

[0024] First, the system initializes an empty ESHT. In implementation, this can be a dictionary, hash map, or similar data structure allocated in system memory.

[0025] The operator (such as an operating room nurse) places the first surgical instrument to be used on a dedicated, high-precision scanning platform. This platform typically integrates a high-resolution image acquisition device and a shadowless light source to eliminate shadows and reflections. For example, the image acquisition device could be an industrial-grade CMOS or CCD camera with a resolution of 20 megapixels or higher, equipped with a telecentric lens to eliminate perspective errors where objects appear larger than they are, ensuring accurate imaging of the instrument's size and shape. The shadowless light source could be a ring-shaped LED array with adjustable brightness and color temperature to accommodate the imaging needs of instruments made of different materials (such as highly reflective stainless steel or matte polymer).

[0026] After the instrument is securely positioned, the system triggers the image acquisition device to obtain at least one high-resolution image that fully and clearly shows the instrument's overall appearance and key details. The acquired raw image data, for example, a... The RGB three-channel image of a pixel first undergoes a series of image preprocessing operations. These operations aim to enhance image quality and prepare for subsequent feature extraction, and may include: Gaussian filtering to reduce noise and eliminate random noise introduced by sensor thermal noise or ambient light fluctuations; adaptive threshold binarization to convert the grayscale image into a black and white binary image to highlight the contours of the instrument; and Canny edge detection or Hough transform, etc., for accurately extracting the edge contours of the instrument.

[0027] After preprocessing, the core feature extraction stage begins. This step aims to extract two types of key information from the image: unique identity features and integrity status features.

[0028] Unique identification features aim to assign a unique digital fingerprint to each physically independent instrument. Even if two instruments have the same model number, their unique identification features should differ distinguishably due to manufacturing tolerances, wear and tear, or even minor surface scratches. In a preferred embodiment of this application, this feature is composite. One part can be extracted based on Scale-Invariant Feature Transform (SIFT) or Speeded-Up Robust Features (SURF) algorithms. These algorithms obtain features that are invariant to image scaling, rotation, and brightness changes by detecting extreme points in different scale spaces of the image and generating a descriptor vector describing the direction and magnitude of the gradient in its neighborhood for each extreme point (i.e., keypoint). Another part can be based on the global shape information of the instrument, for example, by calculating its Hu invariant moments from the extracted instrument contour. Hu invariant moments are a set of seven values ​​composed of second and third central moments that are invariant to translation, rotation, and scaling. By concatenating or combining the local feature descriptor set extracted by SIFT / SURF with the global shape feature vector calculated by Hu invariant moments, a unique identity feature vector with higher dimension and stronger robustness can be formed.

[0029] Integrity status features aim to record the ideal integrity of instruments before surgery. This is crucial for instruments composed of multiple separable parts (such as screwdrivers with removable screws, needle holders connected by multiple joints, etc.). This feature extraction can be achieved through methods such as template matching or deep learning object detection. For example, for a screwdriver with a screw, the system can match a predefined screw head template within a specific region of interest (ROI) of its image. If the matching score is higher than a preset threshold, the screw is considered to be in place, and the integrity status is "true". For more complex multi-part instruments, an object detection model such as YOLO (You Only Look Once) or Faster R-CNN can be trained to simultaneously identify and locate all key components on the instrument. As long as all expected components are detected, the integrity status is considered "true".

[0030] For example, suppose the object being scanned is a "Synthes" brand Phillips screwdriver with a removable screw. The system extracts a total of 256 keypoints from its metal shaft and handle using the SIFT algorithm, each keypoint described by a 128-dimensional floating-point vector. Simultaneously, its contour is calculated to obtain 7 Hu invariant moment values. These 256 128-dimensional vectors and one 7-dimensional vector are serialized and concatenated to form a unique identity feature vector of 32,775 dimensions. At the same time, the system successfully located the screw on the screwdriver head through template matching, thus confirming its integrity status characteristics. It is recorded as the boolean value True.

[0031] A unique identity feature vector was obtained. Then, the system uses a deterministic hash function to calculate a unique hash key. Deterministic hash functions, such as SHA-256 (Secure Hash Algorithm 256-bit), can process inputs of any length (here, ) into hash keys. The serialized byte string is converted into a fixed-length (256 bits, typically represented as 64 hexadecimal characters) output. Its key characteristic is that the same input will always produce the same output, while even a tiny change in the input will cause the output hash value to change dramatically, and it is virtually impossible to deduce the input from the output. This ensures that the identity characteristics of each device can be mapped to a unique, compact, and unforgeable hash key. .

[0032] For example, the aforementioned 32775-dimensional feature vector Inputting the SHA-256 function may yield a hash key. It is '1a8f...ce92'.

[0033] Finally, the system creates an instrument information object as the "value" to be stored in ESHT. This object is a structured collection of data that includes at least: the instrument's generic name (e.g., "screwdriver"), model number (e.g., "Synthes Phillips 2.5mm"), a unique asset number retrieved from the instrument database (if any), and previously extracted integrity status characteristics. (True), and the storage path or the image itself of the original high-resolution image from this scan. Then, the system uses the hash key generated in the previous step. This device information object is stored as a key-value pair in ESHT.

[0034] For example, a new record is added to ESHT: {'1a8f...ce92': {'name': 'screwdriver', 'model': 'Synthes Phillips 2.5mm', 'integrity': True, 'image_path': ' / data / 20231027 / op123 / pre_op_001.png'}}.

[0035] This process (from instrument placement to storage in the ESHT) will be repeated for each instrument used in the procedure until all instruments have been scanned and recorded. Finally, after step S100 is completed, a complete ESHT representing the expected state of all instruments in this surgery will be built in the system memory. This data structure is the benchmark and sole basis for all subsequent real-time operations.

[0036] S200: (Optional) Intraoperative dynamic tracking

[0037] In some embodiments of this application, to achieve more comprehensive end-to-end management, the system may also include an intraoperative dynamic tracking step. This step is not the core of the inventory logic, but it provides more data dimensions for postoperative traceability and analysis. This step can be implemented through various technical means. For example, multiple wide-angle cameras can be installed on the operating room ceiling to identify surgical instruments within the field of view in real time using computer vision technology, and their three-dimensional coordinates on the operating table can be roughly located through multi-view geometric reconstruction. Alternatively, key instruments can be equipped with miniaturized, high-temperature and high-pressure sterilization-resistant RFID tags or Bluetooth Low Energy (BLE) beacons. Through a network of readers or receivers deployed in the operating room, the signal strength indication (RSSI) of the instruments can be acquired periodically (e.g., once per second), and their positions can be estimated using triangulation algorithms. The system can associate and record this dynamic location information and status information (such as "in use," "contaminated," "to be cleaned," etc.) manually entered by nurses or perceived by other sensors with the corresponding instrument entries in the ESHT. Importantly, the data generated in this step is stored in an attached log database without modifying the core content of ESHT itself (i.e., identity and preoperative integrity status), to ensure the purity and stability of the S300 and S400 inventory logic.

[0038] S300: Real-time verification of the recovered surgical instruments, the verification being based on scanning the recovered surgical instruments to obtain a current unique identity feature, calculating a current hash value of the current unique identity feature, and using the current hash value to perform lookup and removal operations in the expected state hash table.

[0039] This step is activated when the surgery ends and the instruments begin to be retrieved. The process is highly symmetrical to the scanning and feature extraction sections in S100, but its purpose is no longer "construction" but "verification".

[0040] The operator places a recovered instrument on the scanning platform. The system triggers a scan to acquire a current high-resolution image of the recovered instrument. Subsequently, the system processes this new image using the exact same image preprocessing algorithm and unique identity feature extraction algorithm as in S100 (e.g., the same SIFT parameters and Hu invariant moment calculation method) to extract the current unique identity feature vector of the recovered instrument. .

[0041] The key point is that the system then uses the exact same deterministic hash function (e.g., SHA-256) as in S100 to compute this current feature vector. The hash value is used to obtain a current hash value. Due to the similarity between the algorithm and the function, if the recovered instrument is indeed one that was scanned preoperatively, and its identity characteristics did not fundamentally change during the operation, then theoretically... It should be exactly the same as the hash key calculated for the device before the procedure.

[0042] The system immediately uses this calculated current hash value. As the lookup key, a search is performed in the ESHT in memory. Since the ESHT is a hash table structure, the average time complexity of this lookup operation is O(1), meaning that its time consumption is almost independent of the number of remaining instruments in the ESHT, and it can usually be completed in microseconds or nanoseconds.

[0043] After finding the result, the system enters a logical judgment branch.

[0044] In a basic embodiment, if the system successfully finds a match in the ESHT... A matching key indicates that the recovered device is one of the expected devices. The system then performs a removal operation, which permanently deletes the matching key-value pair from the ESHT. This atomic "lookup-delete" operation constitutes the "cancellation" of a device.

[0045] For example, suppose the first instrument recovered is an Aesculap No. 1 scalpel scanned preoperatively. The system calculates its current hash value after scanning. Found in ESHT Then, the corresponding entry is deleted from the ESHT. At this point, the size of the ESHT is reduced by one.

[0046] If the system does not find a match in ESHT The matching key constitutes an anomaly. This could mean two things: first, the instrument is not on the pre-operative list for this surgery (e.g., an instrument mistakenly taken from another surgical tray); second, although the instrument is on the list, its surface has been severely altered (e.g., broken, heavily soiled), causing a significant change in its identity feature vector, thus calculating a completely new hash value. In either case, the system will immediately trigger an "unplanned instrument" or "identity unrecognized" alert, prompting the operator to intervene manually.

[0047] In a more preferred embodiment, the reversal process also integrates synchronous verification of integrity status. Specifically, when using the current hash value... After finding a matching key-value pair in ESHT, the system does not immediately delete it. Instead, it first reads the stored preoperative integrity status characteristics of the device from the "value" portion of the key-value pair. Simultaneously, the system performs the same integrity status feature extraction algorithm as in S100 on the currently scanned image of the retrieved instrument to obtain the current integrity status features. .

[0048] Then the system will and Compare them.

[0049] If both are consistent (e.g., the screws were in place before the operation and are still in place after the operation), it means that the instrument is not only correctly identified but also in a complete state. Only then will the system perform the removal operation and delete the key-value pair from ESHT.

[0050] If the two images are inconsistent (e.g., screws were in place before surgery, but missing after surgery), this is a serious anomaly. The system will immediately trigger a high-priority "Integrity Anomaly" audible and visual alarm and clearly display the preoperative and current images of the instrument on the screen, highlighting the inconsistencies. In this case, to preserve evidence and traceability information, the system will not delete this entry in the ESHT, but will instead add an anomaly marker to the entry, for example, 'status': 'ANOMALY_INTEGRITY'.

[0051] For example, suppose the recovered item is a "Synthes" brand screwdriver, but the screw at its tip is missing after the procedure. The current hash value calculated after the system scan. It remains '1a8f...ce92' because the identity characteristics of its main part have not changed. The system successfully found a match in ESHT. However, the system then performs an integrity analysis on the current image, detecting a missing screw and obtaining the current integrity status. The value is False. This does not match the preoperative integrity status True stored in the ESHT. The system immediately triggers an alert and updates the corresponding entry in the ESHT to: {'1a8f...ce92': {'name': 'Screwdriver', ..., 'integrity': True, ..., 'status': 'ANOMALY_INTEGRITY'}}.

[0052] This cycle of "scan-calculate-find-verify-cancel / mark" will be performed once for each recycled device. The entire process is real-time and performed one by one, smoothly distributing the heavy comparison task throughout the entire recycling period.

[0053] S400: Execute the final count decision. After receiving a count completion instruction, check the final state of the expected state hash table and trigger a corresponding count result based on the final state.

[0054] Once the operators confirm that all visible instruments have been retrieved from the surgical site and the S300 scanning and verification process has been completed, they will click a "Count Complete" button on the system interface, or issue a count complete command through other means (such as voice commands).

[0055] Upon receiving the instruction, the system immediately executes the final inventory decision logic. This logic is extremely simple and efficient: it checks the final state of the ESHT in current memory. This check consumes almost no computation time, and the response is instantaneous.

[0056] The logical branches of the ruling are as follows:

[0057] If the check finds that the ESHT is empty (i.e., its size is 0), this means that all instruments recorded preoperatively were successfully scanned and identified postoperatively, and (in the preferred embodiment) their integrity was verified, and they were ultimately successfully cancelled (removed). This is an ideal, anomaly-free closed loop. The system will immediately trigger a "count successful" result, for example, by displaying a green success icon and text on the screen, accompanied by a clear and calm voice announcement: "Count complete, instrument quantity and integrity verified."

[0058] If the ESHT is found to be non-empty (i.e., its size is greater than 0), this indicates that an instrument is missing. Each key-value pair still existing in the ESHT represents an instrument that was scanned and recorded preoperatively but was never scanned and checked off during postoperative retrieval. The system will immediately trigger a high-priority "Instrument Missing" alarm, typically accompanied by a rapid, high-decibel audible signal and a flashing red warning icon on the screen. More importantly, the system will immediately display on the screen, in a list or graphical format, the surgical instrument information corresponding to all remaining entries in the ESHT, including its name, model, specifications, and especially its high-resolution preoperative image. This provides visual cues for on-site personnel to conduct a final, targeted search in the operating room, avoiding aimless, aimless searching.

[0059] In the preferred embodiment that includes integrity anomaly markers, the adjudication logic is more nuanced. Even if the ESHT is not empty, the system distinguishes the status of entries. Entries that have not been written off and do not have anomaly markers are classified as "missing devices." Entries marked with 'ANOMALY_INTEGRITY' are classified as "instruments with integrity anomalies." The final report lists these two types of anomalies separately, providing a precise basis for subsequent incident handling and liability determination.

[0060] For example, let's continue with the aforementioned scenario. Suppose a total of 3 instruments were brought in. The scalpel is missing, the screwdriver's screw is missing, and only the tissue shears were successfully checked out. When the nurse clicks "Count Complete," two records remain in the ESHT: one is the original record for the scalpel, and the other is the record for the screwdriver that was marked as abnormal. The system immediately triggers a red alert and displays the following on the screen:

[0061] Warning: Inventory failed!

[0062] "Missing instrument (1):"

[0063] "- Name: Aesculap No. 1 Surgical Knife (with preoperative images)"

[0064] "Integrity-abnormal device (1):"

[0065] "- Name: Synthes Phillips Screwdriver (with before and after comparison images, highlighting the area where the screw was missing)"

[0066] Finally, regardless of the inventory results, the entire process record of this surgery, including the initial ESHT setup, the logs of each write-off or anomaly marking operation (including timestamps and operator information), and the final ruling and report, will be packaged into a complete electronic archive. This archive will be sent to a full-process data archiving and traceability module for long-term, secure storage. This module can be a database system deployed on a hospital server, ensuring the immutability and traceability of the data, providing a solid and reliable electronic chain of evidence for medical quality management, teaching and research, and even the handling of potential medical disputes.

[0067] System Implementation Examples

[0068] To achieve the above method, this application also provides a closed-loop management system for the entire intraoperative instrument process based on intelligent scanning. This system can be a dedicated device integrating hardware and software, or a software system deployed on existing computing equipment in the operating room, working collaboratively with external scanning hardware. In a typical embodiment, the system includes the following core modules:

[0069] A preoperative scanning and expected state construction module serves as the entry point for data acquisition and baseline establishment. Physically, it includes the aforementioned high-precision industrial camera, shadowless light source, placement stage, and processor for controlling the hardware and performing calculations. Software-wise, it embeds an image acquisition driver, an image preprocessing algorithm library (such as OpenCV-based filtering and binarization algorithms), a feature extraction engine (implementing algorithms such as SIFT, SURF, and Hu invariant moments), and a deterministic hash calculation unit (calling standard cryptographic libraries such as OpenSSL). This module's responsibility is to execute the entire S100 process, and its output is a completed expected state hash table (ESHT) stored in the system's high-speed memory (RAM).

[0070] A postoperative real-time verification and anomaly adjudication module shares scanning hardware with the preoperative module, but its core is purely software-based. This module maintains the ESHT instance received from the preoperative module and implements an event listener to respond to each postoperative scan completion event. Upon receiving an event, it immediately schedules the processor to execute the feature extraction, hash calculation, lookup, verification, and verification / marking logic described in S300. All these operations are performed directly on the ESHT in memory, ensuring extremely low latency. The module also includes a user interface or API for receiving "count complete" commands and, upon receiving the command, executes the final ESHT status check and adjudication logic described in S400. The module's output is a real-time alarm signal (which can drive an audible and visual alarm) and the final count result (displayed on the screen or output as a digital report).

[0071] A full-process data archiving and traceability module is typically a database service or log management system, which can be deployed on a local server or in the cloud. It defines standardized data interfaces to receive all process and result data from the first two modules. It is responsible for structured storage of this data, for example, creating surgical record tables, instrument lists, and operation log tables in a relational database, and ensuring the correlation and consistency between the data. It also provides a query interface, allowing authorized users to retrieve and analyze historical records based on criteria such as surgical number, date, and instrument name, thereby achieving full-process traceability.

[0072] Those skilled in the art will understand that the system modules of the above embodiments can be implemented by a computer program, which can be stored in a computer-readable storage medium, such as a hard disk, optical disk, or flash memory. When the program is executed by a processor, it completes the steps in the above method embodiments. Accordingly, this application also protects the computer-readable storage medium and the computer program product containing the program.

[0073] In summary, this application reconstructs the underlying logic of surgical instrument counting through an innovative technical solution based on feedforward control and real-time verification, transforming an error-prone and high-latency batch processing task into an automated, highly reliable, and zero-latency real-time verification process, providing solid technical support for ensuring surgical safety.

[0074] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit described above can be implemented in hardware.

[0075] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A closed-loop management method for intraoperative instruments based on intelligent scanning, characterized in that, include: A predictive state hash table is constructed. The predictive state hash table is generated based on the unique identity feature and integrity status feature of each surgical instrument obtained from the preoperative scan. The hash value of the unique identity feature is used as the key, and the information associated with the surgical instrument is used as the value for storage. The surgical instruments that have been recycled are verified in real time. The verification is based on scanning the recycled surgical instruments to obtain a current unique identity feature, calculating a current hash value of the current unique identity feature, and using the current hash value to perform lookup and removal operations in the expected state hash table. as well as, The final count decision is executed by checking the final state of the expected state hash table after receiving a count completion instruction, and triggering a corresponding count result based on the final state.

2. The method according to claim 1, characterized in that, The steps of obtaining the unique identification characteristics and integrity status characteristics of each surgical instrument include: A high-resolution image of the surgical instrument is acquired using an image acquisition device; The high-resolution image is processed to extract the unique identity feature and the integrity status feature.

3. The method according to claim 2, characterized in that, The unique identity features include at least one of a set of key point descriptors extracted based on a scale-invariant feature transformation algorithm or an accelerated robust feature algorithm, and a set of geometrically invariant moments calculated based on the contour of the surgical instrument.

4. The method according to claim 1, characterized in that, The step of using a hash value of the unique identity feature as a key includes: A deterministic hash function is used to calculate the unique identity feature to generate a unique hash key as the key.

5. The method according to claim 1, characterized in that, The step of real-time verification and recycling of the surgical instruments, after finding a matching key-value pair in the expected state hash table using the current hash value, further includes: Obtain a current integrity status feature of the recovered surgical instrument; The current integrity status feature is compared with the integrity status feature stored in the matched key-value pair; If the comparison results are inconsistent, an integrity anomaly alarm will be triggered.

6. The method according to claim 5, characterized in that, If the comparison results are inconsistent, after triggering the integrity anomaly alarm, the matching key-value pair will be marked as an abnormal entry in the expected state hash table, instead of being removed.

7. The method according to claim 1, characterized in that, The steps for executing the final inventory decision include: If the final state of the expected state hash table is empty, a successful count result is triggered. If the final state of the expected state hash table is not empty, an inventory of missing instruments is triggered, and the surgical instrument information corresponding to all remaining key-value pairs that have not been removed from the expected state hash table is displayed.

8. A closed-loop management system for the entire intraoperative instrument process based on intelligent scanning, characterized in that, include: A preoperative scanning and expected state construction module is used to construct an expected state hash table based on the unique identity feature and integrity status feature of each surgical instrument obtained by preoperative scanning. The expected state hash table stores a hash value of the unique identity feature as the key and information associated with the surgical instrument as the value. A postoperative real-time verification and anomaly resolution module is used to perform real-time verification of the recovered surgical instruments. The module is configured to scan the recovered surgical instruments to obtain a current unique identification feature, calculate a current hash value of the current unique identification feature, and use the current hash value to perform lookup and removal operations in the expected state hash table. The postoperative real-time verification and anomaly resolution module is also used to perform a final inventory resolution after receiving an inventory completion instruction. It checks a final state of the expected state hash table and triggers a corresponding inventory result based on the final state. as well as, A full-process data archiving and traceability module is used to store the construction records of the expected state hash table, the write-off process records, and the results of the final inventory decision.

9. The system according to claim 8, characterized in that, The postoperative real-time verification and anomaly assessment module is also configured as follows: After finding a matching key-value pair in the expected state hash table using the current hash value, a current integrity status feature of the recycled surgical instrument is obtained; The current integrity status feature is compared with the integrity status feature stored in the matched key-value pair; Furthermore, if the current integrity status characteristics are inconsistent with the integrity status characteristics, an integrity anomaly alarm is triggered.

10. A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method as described in any one of claims 1 to 7.

Images