Medical device management system
The medical device management system uses local features to accurately identify and manage medical devices, addressing symbol wear issues and ensuring reliable data storage through a distributed ledger.
Patent Information
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- CONTEC CO LTD
- Filing Date
- 2024-12-18
- Publication Date
- 2026-06-30
Smart Images
Figure 2026106490000001_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a medical device management system for managing medical devices.
Background Art
[0002] Patent Document 1 describes a management system for surgical instruments that manages data of surgical instruments using a centralized management device. In the management system described in Patent Document 1, a two-dimensional symbol as a data carrier is engraved on a surgical instrument, and information read from the two-dimensional symbol is stored in the centralized management device.
Prior Art Documents
Patent Documents
[0003]
Patent Document 1
Summary of the Invention
Problems to be Solved by the Invention
[0004] However, since surgical instruments are frequently cleaned, the engraved two-dimensional symbol gradually wears out. Since information may not be accurately read from the worn two-dimensional symbol, there is a risk that the data of surgical instruments may not be correctly managed.
[0005] In view of the above problems, an object of the present invention is to provide a medical device management system capable of correctly managing data of medical devices.
Means for Solving the Problems
[0006] To solve the above problems, a medical device management system as an example of an embodiment of the present invention is a medical device management system for managing medical devices marked with symbols, and includes a storage unit that stores a work database in which work data describing events that have occurred with respect to the medical devices is registered, and at least one data registration device that registers the work data to the work database. The storage unit stores comparison information that shows the correspondence between local features unique to the medical device and identification information that individually identifies the medical device. The data registration device includes an imaging unit that captures an image of the medical device and an extraction unit that extracts the local features of the medical device from the image of the medical device. The extraction unit identifies the position of the symbol in the image of the medical device and extracts the local features from an extraction range determined based on the position of the symbol. The data registration device identifies the identification information corresponding to the medical device by comparing the local features extracted by the extraction unit with the comparison information, and registers the work data relating to the medical device corresponding to the identified identification information to the work database. [Effects of the Invention]
[0007] According to the present invention, since identification information corresponding to a medical device is identified using local characteristics unique to the medical device, the data of the medical device is correctly managed even if the symbol is worn out. [Brief explanation of the drawing]
[0008] [Figure 1] A schematic diagram illustrating the overview of a medical device management system. [Figure 2] A diagram showing an example of data stored in the memory unit. [Figure 3] A diagram showing an example of the extraction range of symbols and local features in medical devices. [Figure 4] A diagram showing an example of matching information. [Figure 5] A schematic diagram illustrating the processing performed by the data registration device. [Figure 6]A schematic diagram illustrating the overview of code reading from symbols and matching using local features. [Figure 7] A schematic diagram showing the correspondence between identification information and work data. [Figure 8] A schematic diagram illustrating the data structure in a blockchain. [Figure 9] A diagram showing an example of equipment information. [Figure 10] A diagram showing an example of work data. [Figure 11] A diagram showing an example of historical data. [Figure 12] A flowchart illustrating the processing flow when matching information is registered. [Figure 13] A flowchart illustrating the processing flow using a medical device management system. [Modes for carrying out the invention]
[0009] Referring to the drawings, a medical device management system 10 according to an example of an embodiment of the present invention will be described. Figure 1 is a schematic diagram showing an overview of the medical device management system 10. Multiple locations (nodes) participate in this medical device management system 10, and Figure 1 shows one of these participating locations 20. These participating locations 20 can communicate with each other via a network 50 (a computer network such as the Internet).
[0010] The medical device management system 10 manages multiple medical devices 25 handled at each of the participating locations 20. In Figure 1, for illustrative purposes, the shapes of the multiple medical devices 25 are all shown as forceps, but the medical devices 25 are not limited to forceps.
[0011] Medical device management system 10 can handle all so-called steel-made instruments, such as forceps (scissors), scalpels, scalpel holders, tweezers, hooks, needle holders, rib spreaders, etc., as medical devices 25. For the sake of convenience, although the above is expressed as steel-made instruments, the material of medical device 25 is not limited to steel, and medical device 25 may include various materials such as silicon, ceramic, synthetic resin, natural rubber, etc.
[0012] Also, medical device 25 may be used as a single instrument, or may be an instrument composed of multiple parts. For example, in the case of an instrument composed of multiple parts like a trocar or a cleaning and suction device used in laparoscopic surgery, not only the whole instrument formed by the combination of parts, but also each individual part may be treated as an individual medical device 25. Furthermore, the whole instrument formed by the combination of parts may also be treated as another individual medical device 25 different from the individual parts.
[0013] Each of the plurality of participating sites 20 is a site that handles medical device 25. For example, facilities of medical institutions such as hospitals, facilities of manufacturers that manufacture medical device 25, facilities of maintenance providers that perform maintenance such as sterilization, cleaning, repair, and part replacement of medical device 25, etc. can be the participating sites 20. Here, the maintenance of medical device 25 refers to all actions performed to maintain the function of medical device 25 (to make it in a state where it can be used in medical procedures such as surgery).
[0014] Note that there may be multiple participating sites 20 (nodes) within one facility. For example, in one medical institution (such as a hospital), each area where medical device 25 is handled, such as the surgical area where medical device 25 is used in surgery, the cleaning area where medical device 25 is cleaned, the storage area where medical device 25 is stored, etc., may be a participating site 20.
[0015] Each of the participating sites 20 of the medical device management system 10 includes a data registration device 30 and a storage unit 45. The participating site 20 illustrated in FIG. 1 includes a management device 40, a display unit 41, and an input unit 42 in addition to the storage unit 45 and the data registration device 30.
[0016] In FIG. 1, the management device 40 of the participating site 20 is connected to other participating sites 20 via a network 50. Each of the other participating sites 20 also includes a storage unit 45 and a data registration device 30. Separately from the participating sites 20, a reference server 29 for each participating site 20 to refer to various information may be connected to the network 50. Alternatively, the reference server 29 may function as one of the participating sites 20.
[0017] In FIG. 1, the data registration device 30, the display unit 41, the input unit 42, and the storage unit 45 are connected to the management device 40. The management device 40 is a device (e.g., a computer) for managing the operations performed at the participating site 20. The display unit 41 is a display device (e.g., a liquid crystal display device) for displaying various information to the user. The input unit 42 is a device (e.g., a mouse and a keyboard) for receiving the user's input operations for the management device 40.
[0018] The storage unit 45 is a storage device (e.g., a storage device such as a hard disk drive or a solid state drive) for storing various electronic information (data). The data stored in the storage unit 45 includes a work database in which work data describing events that occurred with respect to the medical device 25 is registered.
[0019] Each of the participating locations 20 includes a storage unit 45, but it is sufficient that a storage unit 45 accessible from the participating locations 20 is provided corresponding to each participating location 20; it is not necessary for each participating location 20 to have its own separate storage unit 45. In other words, one storage unit 45 may be shared by multiple participating locations 20. For example, the surgical area, cleaning area, and storage area within a single medical institution may each be considered a participating location 20, while these participating locations 20 within the same medical institution may share the same storage unit 45. Note that in Figure 1, for illustrative purposes, the storage unit 45 is shown as being included in the participating location 20, but the storage unit 45 does not necessarily have to be located within the participating location 20. For example, a storage unit 45 accessible from the participating locations 20 via a network 50 (such as the Internet) may be located in a facility separate from the participating locations 20 (such as a data center). As a specific example, a portion of the storage capacity of a storage device owned by a reference server 29 connected to the network 50 may be provided as a storage unit 45 corresponding to the participating location 20. Furthermore, a storage unit 45 (such as a shared server) accessible via a network 50 may be shared among multiple participating sites 20, which are separate facilities.
[0020] Information registered in the work database stored in the memory unit 45 can be accessed from each of the multiple participating locations 20 (nodes). In other words, the work database is shared among the multiple participating locations 20. Specifically, it is preferable that the work database be a distributed ledger in which data is held at each of the multiple participating locations 20. The validity of the data in the distributed ledger is guaranteed by agreement among the participating locations 20. Each participating location 20 records events that occur with each individual medical device 25 by registering work data describing events that occurred with that medical device 25 in a work database such as a distributed ledger. For example, when the state of the medical device 25 changes, such as when it is used in surgery or undergoes maintenance at a medical site, the events that occurred with the medical device 25 are recorded in the work database as work data. However, the work database is not limited to a distributed ledger and may be a centrally managed database. For example, a work database in the form of a centrally managed database on a reference server 29 that can be accessed from each of the multiple participating locations 20 may be provided.
[0021] The storage unit 45 can also store data other than the work database. Figure 2 shows an example of data stored in the storage unit 45. For example, in addition to the work database 60, the storage unit 45 can store management information 46 for the medical devices 25.
[0022] The management information 46 for the medical device 25 is stored in the storage unit 45 as data separate from the work database 60. The management information 46 for the medical device 25 includes information used to manage the medical device 25 in the medical device management system 10 that is not registered in the work database 60. Since information regarding the usage history of the medical device 25, such as how the medical device 25 has been used in medical settings, is registered in the work database 60 in the form of work data, it is preferable that the management information 46 includes information separate from the usage history of the medical device 25. This management information 46, which is stored as data separate from the work database 60, may be stored in a reference server 29 that can be accessed from the participating sites 20, instead of being stored in the storage unit 45 of each participating site 20.
[0023] For example, the management information 46 in Figure 2 includes verification information 47 and instrument information 48. Verification information 47 is data that is verified when the data registration device 30 receives a medical instrument 25 in order to identify the identification information that individually identifies the medical instrument 25. Instrument information 48 is information that is determined at the time of manufacture of the medical instrument 25, such as the instrument name and manufacturer name, and is preferably data that does not change regardless of how the medical instrument 25 is used (it does not depend on the usage history).
[0024] Thus, at least a portion of the information regarding the medical device 25 (in this case, management information 46) may be stored as data separate from the work database 60. In this case, it is preferable that the work data registered in the work database 60 includes link information to data containing the management information 46, so that each participating site 20 can refer to the management information 46 regarding the medical device 25.
[0025] Each participating base 20 includes a data registration device 30 (Figure 1), which is a device (medical device data registration device) for registering work data describing events that occurred with respect to medical devices 25 into the work database 60 of the storage unit 45. The data registration device 30 can identify identification information that individually identifies the received medical device 25 by receiving the actual medical device 25. The data registration device 30 may include, for example, an imaging unit that captures images and a communication unit that communicates with the management device 40. The data registration device 30 generates work data corresponding to the identified identification information and registers the generated work data into the work database 60 of the storage unit 45 (via the management device 40 in Figure 1).
[0026] Here, when we say that the data registration device 30 accepts the physical form of the medical device 25, we mean that the data registration device 30 reads some information from the physical form of the medical device 25. The data registration device 30 is equipped with the function of reading information from the physical form of the medical device 25.
[0027] In this embodiment, as shown in Figure 3, each medical device 25 is affixed with a symbol 26C that indicates encoded information (symbol information 26). Figure 3 is a diagram showing an example of the symbol 26C and the extraction range 31 of local features 28 on the medical device 25. The symbol 26C may be a graphic that indicates specific information (symbol information) by a predetermined combination of patterns, such as a barcode or a two-dimensional code. The symbol 26C may be attached, for example, by being affixed to a specific position on the medical device 25, but it is preferable that the symbol 26C be irreversibly attached, such as by being stamped onto the medical device 25, so that it does not come off due to washing or the like. When the data registration device 30 receives the actual medical device 25, it is preferable that an attempt is made to read the symbol information 26 indicated by the symbol 26C.
[0028] In Figure 3, a two-dimensional code is attached to the medical device 25 as symbol 26C. Here, the two-dimensional code is stamped onto the medical device 25. Preferably, the position where symbol 26C is attached is fixed for each type of medical device 25. For example, in the case of forceps shown as medical device 25 in Figure 3, it is preferable that symbol 26C be attached at a certain distance from the pivot point of the forceps.
[0029] If the symbol information 26 indicated by symbol 26C is included in the matching information 47 and is associated with identification information that individually identifies the medical device 25 (such as an instrument ID that is individually determined for each medical device 25), the data registration device 30 can identify the identification information of the medical device 25 by reading the symbol information 26.
[0030] However, if the symbol 26C wears down due to cleaning or other reasons, the symbol information 26 may no longer be accurately read from the symbol 26C. In that case, the identification information of the medical device 25 may not be accurately identified, and there is a risk that the data of the medical device 25 will not be properly managed. Therefore, the medical device management system 10 of this embodiment uses local features 28 unique to the medical device 25 to ensure that the data of the medical device 25 is properly managed even if the symbol 26C is worn down.
[0031] Local features 28 are unique to each individual medical device 25, consisting of surface patterns (such as fine surface irregularities) of the medical device 25, and are sometimes called object fingerprints. Even if medical devices 25 are mass-produced using molds, individual differences in surface patterns naturally occur during the manufacturing process. By utilizing these individual differences in surface patterns as local features 28, the medical device management system 10 can individually identify each medical device 25.
[0032] For example, the data registration device 30 may have a function to capture an image of the medical device 25 and a function to extract local features 28 from the image of the medical device 25. When the data registration device 30 extracts local features 28 from the image of the medical device 25, it is preferable that the method of acquiring the image and the method of extracting local features 28 from the image be predetermined, such as what imaging conditions (lighting conditions, camera angle, etc.) to use to capture the image, what image processing to perform on the captured image to extract local features 28, and which part of the surface pattern of the medical device 25 to use as the basis for extracting local features 28. In particular, the extraction range 31 from which local features 28 are extracted should be determined based on the position of the symbol 26C in the image of the medical device 25.
[0033] Furthermore, the captured image does not necessarily have to be used directly as the local feature 28; data calculated based on the image of the medical device 25 may be used as the local feature 28. For example, the data registration device 30 may use an AI model (AI: Artificial Intelligence) that performs image recognition processing to extract image features from the image of the medical device 25, and use the extracted features as the local feature 28.
[0034] The process for extracting local features 28 does not necessarily have to be performed within the data registration device 30. The data registration device 30 may transmit the captured images to another device (for example, the reference server 29 in Figure 1), and the local features 28 (and their matching with the matching information 47) may be extracted by the other device that receives the images.
[0035] In this embodiment, first, the imaging unit of the data registration device 30 captures an image of the medical device 25, and local features 28 are extracted from an extraction range 31 determined based on the position of the symbol 26C in the image. In this embodiment, matching information 47 indicating the correspondence between these local features 28 and the identification information of the medical device 25 is stored in the storage unit 45 of the medical device management system 10.
[0036] The medical device management system 10 of this embodiment can identify identification information corresponding to a medical device 25 by comparing local features 28 extracted from an image of the medical device 25 with matching information 47. Here, since the local features 28 are not lost due to wear caused by normal cleaning processes, the medical device management system 10 can identify the identification information without any problems even if the medical device 25 is repeatedly cleaned and the symbol 26C is worn down.
[0037] Figure 4 shows an example of the matching information 47 stored in the memory unit 45. The matching information 47 in Figure 4 shows the correspondence between local features 28 unique to the medical device 25 and the device ID assigned to each individual medical device 25. The device ID is an example of identification information 27 that individually identifies the medical device 25. In this embodiment, the identification information 27 (device ID) that individually identifies the medical device 25 is stored in the memory unit 45 in association with the data (image data or feature quantities, etc.) of the local features 28.
[0038] Identification information 27 (device ID) is assigned to individual medical devices 25 at one of the participating locations 20 included in the medical device management system 10. Hereinafter, a participating location 20 that assigns identification information 27 to a medical device 25 will be referred to as an identification information assignment location. For example, the facility of the manufacturer of the medical device 25 may participate in the medical device management system 10 as an identification information assignment location (one of the participating locations 20). The identification information assignment location may, for example, assign identification information 27 (device ID) to the medical device 25 when the medical device 25 is shipped from the identification information assignment location.
[0039] Then, when identification information 27 is assigned to the medical device 25 at the identification information assignment point, local features 28 are extracted from an extraction range 31 determined based on the position of symbol 26C. The extracted local features 28 are associated with the identification information 27 assigned to the medical device 25 and registered as matching information 47. Here, "registered as matching information 47" means that matching information 47 showing the correspondence between local features 28 and identification information 27 is stored in the storage unit 45. The matching information 47 registered at the identification information assignment point should be accessible from other participating points 20.
[0040] Furthermore, it is preferable that the identification information 27 (device ID) is centrally managed within the medical device management system 10 so that it is unique data without duplication throughout the entire medical device management system 10. For example, a reference server 29 (Figure 1) is provided within the medical device management system 10, which can access information from each participating site 20, and the identification information 27 is centrally managed as unique data on this reference server 29. For example, when a medical device 25 is shipped from the manufacturer (identification information assignment site), the manufacturer's data registration device 30 accesses the reference server 29 and receives the issuance of new identification information 27. The manufacturer's data registration device 30 then extracts local features 28 from an image of the medical device 25 scheduled for shipment and stores matching information 47, which shows the correspondence between the identification information 27 issued by the reference server 29 and the extracted local features 28, in the storage unit 45. Here, the matching information 47 may be stored in the reference server 29 instead of being stored in the storage unit 45 of each participating site 20. Furthermore, while it is preferable that the matching information 47 is stored as data separate from the work database 60, it may also be data included in the work database 60.
[0041] When a medical device 25 is shipped from an identification information assignment point, the matching information 47 is stored in the storage unit 45 or the reference server 29. This allows the data registration devices 30 at participating points 20 other than the identification information assignment point to use this matching information 47 to identify the identification information 27 that individually identifies the medical device 25. In other words, when a data registration device 30 at each participating point 20 of the medical device management system 10 receives a physical medical device 25, it can identify the identification information 27 corresponding to the medical device 25 by comparing the local features 28 of the medical device 25 with the matching information 47.
[0042] Furthermore, the matching information 47 in Figure 4 includes the correspondence between the identification information 27 (device ID) of the medical device 25 and the code indicated by the symbol 26C attached to the medical device 25. The code in Figure 4 is an example of the symbol information 26 indicated by the symbol 26C. For example, when the identification information 27 is attached to the medical device 25 at the identification information assignment point, it is desirable that the correspondence between the identification information 27 and the symbol information 26 (code) be included and stored in the matching information 47.
[0043] If the data registration device 30 successfully reads the symbol information 26 (code) from the symbol 26C attached to the medical device 25, it can identify the identification information 27 (device ID) corresponding to the medical device 25 by comparing the read symbol information 26 with the matching information 47. Reading the symbol information 26 is often faster and less computationally intensive than extracting local features 28. Therefore, by using the reading of symbol information 26 in conjunction with the extraction of local features 28, it is possible to quickly identify the identification information 27 of medical devices 25 whose symbol 26C is not worn, and to accurately identify the identification information 27 of medical devices 25 whose symbol 26C is worn, thereby ensuring proper data management.
[0044] The processing performed by the data registration device 30 will be explained using Figure 5. Figure 5 is a schematic diagram showing an overview of the processing performed by the data registration device 30. The data registration device 30 includes an identification unit 35, a generation unit 36, and a registration unit 38. The identification unit 35 in Figure 5 includes an imaging unit 32 that captures an image of the medical device 25, and an extraction unit 34 that extracts local features 28 of the medical device 25 from the image of the medical device 25. The data registration device 30 includes a processing unit (processor) such as a CPU (Central Processing Unit), GPU (Graphics Processing Unit), and NPU (Neural Processing Unit), and a storage device such as ROM (Read Only Memory) and RAM (Random Access Memory). The functions of the identification unit 35, generation unit 36, registration unit 38, etc. are realized by the processor executing a program stored in the storage device. If high processing power is required to perform calculations in the data registration device 30, the calculations may be performed on a device other than the data registration device 30. For example, the data registration device 30 may communicate via the network 50 with a high-level computing device (such as a supercomputer) with high computing power located at another participating site 20, and the functions of the identification unit 35 and generation unit 36 may be realized by the high-level computing device executing a program. Alternatively, the functions of the identification unit 35 and generation unit 36 may be realized by distributed processing via the network 50 performed by computing devices located at multiple participating sites 20.
[0045] In the data registration device 30, the identification unit 35 first receives the physical medical device 25 and identifies the identification information 27 (device ID) corresponding to the received medical device 25. In this embodiment, a symbol 26C (such as a two-dimensional code) indicating symbol information 26 is engraved on the medical device 25, so the data registration device 30 reads the symbol information 26. The data registration device 30 then identifies the identification information 27 by comparing the read symbol information 26 with the matching information 47. In this case, the identification unit 35 functions as a code reader (such as a barcode reader or two-dimensional code reader). If a memory element (such as an RFID tag) that stores the identification information 27 is attached to the medical device 25, the data registration device 30 identifies the identification information 27 by reading the memory element. In this case, the identification unit 35 functions as a memory element reader (such as an RFID reader).
[0046] The identification unit 35 can also identify identification information 27 by extracting local features 28 from the medical device 25. The identification unit 35 captures an image of the medical device 25 using the imaging unit 32 (camera unit, etc.), and extracts local features 28 from the captured image using the extraction unit 34 (program, etc., that performs image analysis processing). The identification unit 35 then compares the local features 28 extracted by the extraction unit 34 with the matching information 47 stored in the medical device management system 10. By comparing the local features 28 with the matching information 47, the identification information 27 (device ID) corresponding to the medical device 25 is identified. The identification unit 35 then notifies the registration unit 38 of the identified identification information 27.
[0047] If the data registration device 30 is installed at an identification information assignment base (such as a manufacturer), the identification information 27 corresponding to the medical device 25 may not yet be included in the matching information 47. At the identification information assignment base facility, instead of matching the local features 28 with the matching information 47, the data registration device 30 may generate new identification information 27 corresponding to the local features 28 (via a reference server 29, etc.) and include it in the matching information 47. The data registration device 30 then notifies the registration unit 38 of the generated identification information 27 as the identified identification information 27.
[0048] Meanwhile, the generation unit 36 of the data registration device 30 generates work data 62 that is registered in the work database 60 in association with the identification information 27 (instrument ID). This work data 62 includes data related to the work performed on the medical instrument 25 at the participating base 20 where the data registration device 30 is installed (for example, information on the date and time when cleaning, sterilization, and maintenance were performed).
[0049] The registration unit 38 associates the work data 62 generated by the generation unit 36 with the identification information 27 (instrument ID) identified by the identification unit 35 and registers it in the work database 60. Preferably, the registration unit 38 can register the work data 62 in the work database 60 only when the identification unit 35 has identified the identification information 27 corresponding to the received medical device 25. In other words, it is preferable that the data registration device 30 cannot register work data 62 related to the medical device 25 in the work database 60 when the actual medical device 25 has not been received. For example, even if the user of the data registration device 30 knows the identification information 27 of the medical device 25, they cannot register work data 62 related to the medical device 25 in the work database 60 when the medical device 25 is not in their possession. Therefore, it is certain that the work data 62 registered in the work database 60 is related to the work performed when the medical device 25 was actually handled at the site (participating base 20), thus increasing the reliability of the data.
[0050] Furthermore, in order to restrict the devices that can perform the registration of work data 62 to only the data registration device 30, it is preferable that the work database 60 be designed such that, for example, each data registration device 30 is pre-assigned a unique device number (ID), and that work data 62 cannot be registered from devices other than those with the pre-assigned device numbers. Also, as a specific example of a configuration that enables the registration of work data 62 when the identification unit 35 identifies the identification information 27 of the medical device 25, it is preferable that the registration unit 38 be configured to start accessing the work database 60 on the condition that the identification unit 35 identifies the identification information 27, or that the generation unit 36 be configured to start generating work data 62 on the condition that the identification unit 35 identifies the identification information 27.
[0051] Figure 6 shows a schematic diagram illustrating the overview of reading the code (symbol information 26) from symbol 26C and the overview of matching using local features 28. In the example shown at the top of Figure 6, the entire symbol 26C is included in the image captured by the imaging unit 32. In this case, the identification unit 35 can read the code (symbol information 26) from symbol 26C.
[0052] On the other hand, in the example shown at the bottom of Figure 6, the surface of the medical instrument 25 is worn, and a portion (nearly half) of the symbol 26C is covered by the worn portion 31B. In this state, the identification unit 35 cannot read the symbol information 26 of the symbol 26C. However, symbol recognition programs or symbol recognition AI models used for code reading, such as barcode readers and 2D code readers, can often recognize that the symbol 26C exists in the input image even when the symbol information 26 cannot be accurately read because the symbol 26C is worn or hidden. In other words, the extraction unit 34 of the identification unit 35 can identify the location of the symbol 26C even when the symbol information 26 cannot be read. Therefore, when the symbol information 26 cannot be read, the extraction unit 34 of the identification unit 35 determines an extraction range 31 based on the identified location of the symbol 26C and extracts local features 28 from the extraction range 31. The data registration device 30 then compares the extracted local features 28 with the matching information 47.
[0053] The identification unit 35 determines that the medical device 25 to be identified is a medical device 25 that has the local features 28 recorded as matching information 47, if the degree of match between the local features 28 extracted from the extraction range 31 and the local features 28 recorded as matching information 47 is sufficiently high. In the lower example of Figure 6, the area of the exposed part 31A that is not covered by the worn part 31B within the extraction range 31 matches a wide range of the local features 28 in the matching information 47. In the illustrated example, more than half of the extraction range 31 matches the local features 28 in the matching information 47. There are several types of algorithms for determining the match of local features 28, some of which determine a match to be valid with a degree of match of about 50%. Therefore, by using such an algorithm, the identification information 27 of the medical device 25 can be identified even when nearly half of the symbol 26C is covered by the worn part 31B, as in the illustrated example.
[0054] The identification unit 35 may also use the symbol 26C solely to determine the extraction range 31 of the local feature 28 without reading the symbol information 26, and identify the identification information 27 using only the local feature 28. In this case, the identification unit 35 can always accurately identify the identification information 27 of the medical device 25, regardless of whether the symbol 26C is worn or not.
[0055] Incidentally, the details of the work data 62 registered in the work database 60 when the data registration device 30 identifies the identification information 27 of the medical device 25 vary depending on the participating base 20 where the data registration takes place. For example, when the medical device 25 is shipped from the identification information assignment base, it is preferable that the work data 62, including at least the identification information 27 of the medical device 25, be registered using the data registration device 30.
[0056] Figure 7 is a schematic diagram showing the correspondence between identification information 27 and work data 62. As an example, Figure 7 explains the work database 60 as a distributed ledger. In the work database 60 of Figure 7, for each piece of identification information 27 (equipment ID), a series of transaction data corresponding to that identification information 27 is registered as work data 62. In a distributed ledger, the data corresponding to identification information 27 (equipment ID) cannot be directly modified or deleted once it has been registered; data updates are only performed by adding new transaction data (work data 62).
[0057] However, with respect to newly manufactured medical devices 25, since data regarding the identification information 27 (device ID) of the medical device 25 is not registered in the distributed ledger, it is desirable that transaction data (work data 62) including the identification information 27 be registered when the device is shipped from the manufacturer.
[0058] The data in the work database 60, which functions as a distributed ledger, is stored in the respective storage units 45 of the multiple participating locations 20, and its contents are shared among all participating locations 20. Because the work database 60 is stored in multiple participating locations 20, even if a failure occurs in one of the participating locations 20, the other participating locations 20 can use the medical device management system 10 without any problems, resulting in a highly fault-tolerant system.
[0059] Furthermore, if the work database 60 is a distributed ledger, in order to assert that the data in the work database 60 stored in the storage unit 45 of any participating site 20 is legitimate, it is necessary to obtain the agreement of the other participating sites 20 (confirming that it is consistent with the data stored at the other participating sites 20). Therefore, it is difficult to tamper with data already registered in the work database 60, and the system has high robustness.
[0060] Furthermore, if the work database 60 is a distributed ledger, it is preferable that the distributed ledger be based on blockchain technology. The schematic diagram in Figure 8 shows the data structure in blockchain 64. In blockchain 64, a series of blocks 65 are registered for each identification information 27 (device ID). Each block 65 contains transaction data (work data 62) related to the medical device 25. More precisely, transaction data (work data 62) that occur at the same time are grouped into one block 65 (i.e., transaction data related to multiple medical devices 25 are included in one block 65), but in Figure 6, for illustrative purposes, it is illustrated that one block 65 contains one transaction data (work data 62).
[0061] Furthermore, within block 65, which is newly added to blockchain 64, in addition to transaction data (working data 62), the hash value 66 of the previous block is included. The hash value 66 of the previous block is data corresponding to the data of the existing block 65, and it is data whose legitimacy can be verified by a third party. Note that the data corresponding to the data of the existing block 65 included in block 65 does not necessarily have to be a hash value, as long as it is data whose legitimacy can be verified according to a certain rule, but here it will be explained as a hash value.
[0062] The hash value 66 of the previous block contained in each block 65 must correctly correspond to the data in the existing block 65. Therefore, if the data in any block 65 is tampered with, it will no longer match the hash value 66 of the previous block contained in the next block 65, making the tampering easily detectable. For this reason, it is difficult to tamper with the data in blockchain 64, and the system has high robustness.
[0063] Furthermore, when the medical device 25 is shipped from the identification information assignment point (such as the manufacturer), it is preferable that device information 48 related to the medical device 25 is also recorded. Figure 9 shows an example of device information 48. The device information 48 in Figure 9 corresponds to the identification information 27 (device ID) and includes information such as the device name (type of device or product name), manufacturer name (manufacturer name), and special notes (such as that it is a custom-made product and information of the physician who ordered it). This information is determined at the time of manufacturing the medical device 25 and does not change even if multiple transaction data (work data 62) are registered thereafter. Therefore, it is preferable that such device information 48 be stored as data separate from the work database 60 (in this case, a distributed ledger) rather than being registered in the work database 60. The data separate from the work database 60 may be stored in the storage unit 45 of each participating point 20, or it may be stored in a reference server 29 separate from the participating points 20.
[0064] However, if at least a portion of the information regarding the medical device 25, such as the device information 48, is stored as data separate from the work database 60, it is preferable that the transaction data (work data 62) includes link information to the data separate from the work database 60, so that the user can reach that data by following the transaction data (work data 62).
[0065] Next, we will explain the transaction data (work data 62) at participating locations 20 other than the identification information assignment location. For example, when a medical device 25 is delivered to a medical institution (such as a hospital), the medical institution, as a participating location 20, should use the data registration device 30 to register transaction data (work data 62) that includes at least the date and time of delivery of the medical device 25. By including the date and time of delivery information in the transaction data (work data 62), users can obtain information from the work database 60 about when the medical device 25 began to be used at the medical institution.
[0066] Furthermore, when the medical device 25 is used in a medical institution, it is preferable that transaction data (work data 62), including at least the date and time of use of the medical device 25, be registered using the data registration device 30.
[0067] Figure 10 shows an example of transaction data (work data 62). The transaction data (work data 62) in Figure 10 is data that is registered in the work database 60 (distributed ledger) when a medical device 25 is used in surgery at a medical institution (or after it has been used).
[0068] The work data 62 in Figure 8 includes information such as the date and time of the surgery, the location where the surgery was performed (information of the medical institution), the surgeon (information of the doctor), the ID of the data registration device 30 that registers the work data 62, the type of work performed (in this case, surgery), the ID of the instrument for which the work data 62 is registered (instrument ID, i.e., identification information 27), and remarks (such as the details of the surgery). By including this information in the work data 62 and registering it, the user can obtain information from the work database 60 about when, how many times, and in what type of surgery the medical instrument 25 was used (events that occurred regarding the medical instrument 25). In addition, although not shown in Figure 10, the work data 62 may also include information on items requiring special management, such as whether or not it was used in surgery on a patient with a specific infectious disease, which are historical items of use that require special attention in the management of the medical instrument 25. Such information on items requiring special management may be recorded as information on the type of work performed or as remarks.
[0069] If there is a limit to the amount of information that can be stored in the work data 62 at once, the data registration device 30 may include only the minimum necessary information (such as the date and time of execution) and link information to another database in the work data 62, and store the other information in the other database.
[0070] The reason why the ID of the data registration device 30 is included in the work data 62 in Figure 8 is to indicate that the data registration device 30 attempting to register this work data 62 is a device predetermined to be capable of registering work data 62 in the work database 60. However, if the communication protocol used within the medical device management system 10 is configured to automatically reject the registration of work data 62 from devices other than those with a specific device ID, the ID of the data registration device 30 does not need to be included in the work data 62.
[0071] Furthermore, when the medical device 25 is delivered to a participating facility 20 that performs maintenance on the medical device 25, it is desirable that the work data 62, including at least the date and time of delivery of the medical device 25, be registered using the data registration device 30. Also, when the medical device 25 is delivered from the participating facility 20 that performs maintenance on the medical device 25, it is desirable that the work data 62, including at least the maintenance status of the medical device 25, be registered using the data registration device 30. By including this information in the work data 62, users can obtain information from the work database 60 about where the medical device 25 is currently located and when, how many times, and what kind of maintenance (sterilization, cleaning, repair, parts replacement, etc.) it has received (information about events that have occurred regarding the medical device 25).
[0072] Furthermore, when the medical device 25 is disposed of by a waste disposal company, it is preferable that the work data 62, including at least the disposal certificate information for the medical device 25, is registered using the data registration device 30. If the work database 60 is a distributed ledger, the data for the medical device 25 is not erased when the medical device 25 is disposed of, and the information that it has been "disposed of" (disposal certificate information) is registered in the distributed ledger. By including the disposal certificate information in the work data 62, users can obtain information from the work database 60 regarding whether the medical device 25 brought to the waste disposal company has been properly disposed of.
[0073] At each participating site 20, users can obtain historical data 90 related to medical devices 25 from the work database 60. For example, a user can access the storage unit 45 and obtain data stored in the work database 60 by operating the management device 40 (Figure 1) at the participating site 20 via the input unit 42. The data stored in the work database 60 is processed for display by the management device 40 and displayed on the display unit 41.
[0074] Figure 11 shows an example of history data 90. History data 90 is information obtained when the management device 40 integrates work data 62 corresponding to the identification information 27 of the medical device 25, and, if necessary (depending on link information), also includes information such as management information 46 that is stored as data different from the work database 60.
[0075] The history data 90 in Figure 9 includes information such as the instrument name, manufacturer name, number of uses, last use date and time, number of sterilizations, and last sterilization date and time, corresponding to the instrument ID (identification information 27). As mentioned above, the instrument name, manufacturer name, etc., may be stored as instrument information 48, which is different data from the work database 60. However, if link information to that data is included in the work data 62 of the work database 60, the management device 40 can also retrieve the instrument information 48.
[0076] Although the number of uses and sterilizations may not be directly included in the work data 62, the control device 40 can include the number of such operations in the history data 90 by counting how many data points in the work data 62 corresponding to the identification information 27 indicate that the medical instrument 25 was used or sterilized. Similarly, for the last use date and time and last sterilization date and time, the control device 40 can include the date and time of the last operation in the history data 90 by checking the date and time of the most recent data in the work data 62 indicating that the same type of operation was performed.
[0077] This historical data 90 can be viewed even at participating locations 20 that do not currently handle the medical device 25. In other words, while the registration of new work data 62 related to the medical device 25 cannot be performed unless the actual medical device 25 is in the user's possession, users can view information about the medical device 25 even if the actual medical device 25 is not in their possession.
[0078] In the above embodiment, only one reference server 29 in Figure 1 is shown as a server different from each participating base 20, but multiple servers different from each participating base 20 may be provided. For example, the server that issues identification information 27, the server that stores matching information 47, and the server that stores at least a part of the information related to the medical device 25 (such as device information 48) separately from the work database 60 may each be separate servers.
[0079] Furthermore, the work database 60 does not necessarily have to be based on blockchain technology, nor does it have to be a distributed ledger. The work database 60 can be any database in which information about the medical device 25 is updated by repeatedly registering work data 62 related to the medical device 25 that corresponds to the identification information 27.
[0080] Next, the processing flow in the medical device management system 10 will be explained with reference to Figures 1, 5, 12, and 13. Figure 12 is a flowchart showing the processing flow when the matching information 47 is registered. Figure 13 is a flowchart showing the processing flow in the medical device management system 10.
[0081] At identification information assignment points, such as the facilities of manufacturers of medical devices 25, the identification information 27 is assigned to the medical device 25 by performing a process to register the matching information 47, as shown in Figure 12.
[0082] The identification information assignment point is one of the participating points 20 that participate in the medical device management system 10, and includes a data registration device 30. When the data registration device 30 receives a medical device 25 at the identification information assignment point, the process of registering matching information 47 is started (START).
[0083] First, in step S11, the imaging unit 32 of the data registration device 30 captures an image of the medical device 25. In the medical device management system 10, it is preferable that the imaging conditions (lighting conditions, camera angle, etc.) of the medical device 25 remain constant. For example, the imaging unit 32 may be a unit in which the camera and lighting device are fixed so that the imaging conditions remain constant.
[0084] After the imaging unit 32 captures an image of the medical device 25, in step S12, the extraction unit 34 searches for the symbol 26C in the image of the medical device 25 and identifies the position of the symbol 26C.
[0085] Here, since the symbol 26C is not yet worn at the time the identification information 27 is assigned (especially during the manufacture of the medical device 25), the extraction unit 34 can reliably identify the position of the symbol 26C at the identification information assignment point (such as the manufacturer's facility).
[0086] In this embodiment, the extraction unit 34 identifies the position of symbol 26C by performing image processing on the captured image. For example, the extraction unit 34 identifies the position of symbol 26C as the position where a specific pixel pattern appears in the image according to a predetermined procedure. To give one specific example, the extraction unit 34 analyzes the arrangement pattern of points with large brightness changes in the image and identifies the position where a specific arrangement pattern appears as symbol 26C. For example, if symbol 26C is a two-dimensional code, the density of points with large brightness changes will be high at the position of symbol 26C, so the position where points with large brightness changes are densely clustered in a specific pattern is identified as the position of symbol 26C.
[0087] Alternatively, the extraction unit 34 may identify the position of symbol 26C using an AI model that describes an input-output relationship where an image is taken as input and the position of symbol 26C within the image is output. In this case, it is preferable to have an image recognition AI model prepared in advance that has been trained to output the position of symbol 26C using various images of symbol 26C as training data.
[0088] After the extraction unit 34 identifies the position of symbol 26C, in step S13 it extracts local features 28 from an extraction range 31 determined based on the position of symbol 26C. In step S13, in addition to extracting local features 28, symbol information 26 from symbol 26C may also be read.
[0089] The extraction range 31, which is determined based on the position of symbol 26C, preferably includes the entire symbol 26C and is wider than symbol 26C. For example, if symbol 26C is a square two-dimensional code, the extraction unit 34 defines the extraction range 31 as the area of the square expanded outward from each side of that square (for example, an 800-pixel × 800-pixel area).
[0090] Furthermore, if the dimensions of the symbol 26C in the medical device 25 are specified to be constant (for example, 5 mm x 5 mm), the extraction unit 34 may define the extraction range 31 as an area corresponding to a certain dimensional range that is larger than the dimensions of the symbol 26C. For example, the extraction unit 34 may identify an area around the symbol 26C in the image that corresponds to a specific dimensional range (for example, 10 mm x 10 mm) by comparing the size of the area occupied by the symbol 26C in the image with the specified dimensions of the symbol 26C, and define that area as the extraction range 31.
[0091] Furthermore, as a method for determining the extraction range 31 when the shape of symbol 26C is not constant, the extraction unit 34 may, for example, calculate the centroid of the symbol 26C and define the extraction range 31 as a square area (for example, an 800-pixel x 800-pixel area) centered on that centroid.
[0092] The shape of the extraction range 31 is not limited to a square; it can be any range that can be determined by a certain rule based on the position of symbol 26C. For example, the extraction range 31 may be a rectangle, any polygon, a circle, an ellipse, etc.
[0093] The extraction unit 34 determines the extraction range 31 based on the position of the symbol 26C, and then extracts local features 28 from the extraction range 31. The extracted local features 28 may be image data within the extraction range 31, or they may be feature quantities calculated based on the image data within the extraction range 31.
[0094] Once local features 28 within the extraction range 31 are extracted, the data registration device 30 proceeds to step S14 and receives the issuance of identification information 27 (device ID) to be assigned to the medical device 25. For example, the data registration device 30 accesses a device that issues identification information 27 (such as a reference server 29) and receives the issuance of new identification information 27. The timing of the issuance of identification information 27 does not necessarily have to be after the extraction of local features 28; it is sufficient if the identification information 27 is issued before the registration of the matching information 47. For example, the issuance of identification information 27 may occur simultaneously with imaging by the imaging unit 32 (step S11).
[0095] The data registration device 30 then proceeds to step S15, where it stores the matching information 47, which indicates the correspondence between the local features 28 extracted in step S13 and the identification information 27 issued in step S14, in the storage unit 45, thereby registering the matching information 47. If symbol information 26 is read from symbol 26C, the correspondence between the identification information 27 and symbol information 26 should also be included in the matching information 47 and stored (registered) in the storage unit 45. Once the matching information 47 is registered in this manner, the process of assigning the identification information 27 to the medical device 25 at the identification information assignment base is completed (END).
[0096] Next, referring to Figure 13, the procedure for identifying the medical device 25 at the participating site 20 and registering the work data 62 will be explained. Processing begins (START) when the medical device 25 is brought into the participating site 20 (node) participating in the medical device management system 10. First, the data registration device 30 receives the actual medical device 25 (step S21). Specifically, the data registration device 30 attempts to read the symbol information 26 of the symbol 26C attached to the medical device 25.
[0097] In step S22, the data registration device 30 determines whether it is possible to read symbol information 26 from symbol 26C. If the reading of symbol information 26 is successful (YES in step S22), the data registration device 30 proceeds to step S23, where it compares the symbol information 26 with the matching information 47 to identify the identification information 27 (device ID) corresponding to the medical device 25.
[0098] On the other hand, if the symbol 26C is worn or obscured by an obstacle, it may be impossible to read the symbol information 26 from the symbol 26C. If reading the symbol information 26 fails (NO in step S22), the data registration device 30 proceeds to step S24 and performs a comparison with the matching information 47 using the local features 28 of the medical instrument 25.
[0099] In step S24, the extraction unit 34 extracts local features 28 of the medical device 25 in the same manner as in steps S12 and S13 of Figure 12. In step S24, the extraction unit 34 first identifies the position of symbol 26C in the captured image by identifying the location where a specific pixel pattern appears through image processing, or by using an image recognition AI model that has been trained to output the position of symbol 26C with an image as input. Even if it is not possible to read the symbol information 26, the extraction unit 34 can still identify the position of symbol 26C in the image.
[0100] When the extraction unit 34 identifies the position of symbol 26C, it determines the extraction range 31 based on that position. The extraction range 31 includes, for example, the entirety of symbol 26C and is wider than symbol 26C. Once the extraction range 31 is determined based on the position of symbol 26C, the extraction unit 34 extracts local features 28 from the extraction range 31. The extracted local features 28 may be image data within the extraction range 31, or they may be feature quantities calculated based on the image data within the extraction range 31.
[0101] Here, it is preferable that the imaging conditions of the imaging unit 32 at each participating site 20 are the same as those at the identification information assignment site, so that the local features 28 are extracted under conditions as close as possible to those at the identification information assignment site. For example, it is preferable that a common data registration device 30 is used at each participating site 20 of the medical device management system 10, configured so that the imaging conditions of the imaging unit 32 remain constant.
[0102] Once the local features 28 are extracted, the data registration device 30 proceeds to step S23, where the extraction unit 34 compares the extracted local features 28 with the matching information 47 to identify the identification information 27 (device ID) corresponding to the medical device 25. Even if the symbol 26C is worn or obscured by an obstacle, the local features 28 can still be extracted. Therefore, even if it is impossible to read the symbol information 26 of the symbol 26C, the data registration device 30 can identify the identification information 27 using the local features 28.
[0103] Once the identification information 27 corresponding to the medical device 25 is identified, the data registration device 30 acquires history data 90 related to that identification information 27 via the management device 40, storage unit 45, etc. (step S25). The acquired history data 90 is displayed on the display unit 41. The user checks from the contents of the history data 90 whether the status of the medical device 25 is consistent with that of the current participating site 20 (for example, whether sterilization, cleaning, etc., have been carried out in the prescribed order and whether it has been brought into the surgical area). Step S25 may be skipped if it is not necessary to check the history data 90, such as when the medical device 25 is shipped from the manufacturer.
[0104] Once the identification information 27 corresponding to the medical device 25 is identified, the data registration device 30 becomes capable of recording work data 62 corresponding to the identification information 27 into the work database 60. The generation unit 36 of the data registration device 30 then generates work data 62 to be registered in the work database 60 in association with the identification information 27 (step S26). For example, in the surgical area, the generation unit 36 generates work data 62 indicating that a surgery using the medical device 25 has been performed.
[0105] Then, the registration unit 38 of the data registration device 30 associates the work data 62 generated by the generation unit 36 with the identification information 27 identified by the identification unit 35 and registers it in the work database 60 (step S27).
[0106] Once the work data 62 is registered in the work database 60, the processing for the medical device 25 received this time is completed (END), and the medical device management system 10 enters a standby state until the next medical device 25 is brought in.
[0107] According to the above embodiment, even if the symbol 26C is worn, the data registration device 30 can correctly identify the identification information 27 corresponding to the medical device 25 using local features 28 unique to the medical device 25. Therefore, according to the medical device management system 10 of this embodiment, the data of the medical device 25 is correctly managed.
[0108] The present invention is not limited to the embodiments described above, and can be implemented in various forms. Furthermore, the attached drawings are merely schematic representations of the components for ease of understanding, and the dimensions, number, etc., of each component shown may differ from the actual dimensions. In addition, the shapes, properties, etc., of each component shown in the embodiments described above are examples only, and are not limited to the above-described forms; they can be modified in various ways within the scope of the spirit of the present invention. [Explanation of Symbols]
[0109] 10. Medical device management system 20 Participating locations 25 Medical devices 26 Symbol Information 26C Symbol 27 Identification Information 28 Local Features 29 Reference Server 30 Data Registration Device 31 Extraction range 32 Imaging Unit 34 Extraction part 35 Identification unit 36 Generation part 38 Registration Department 40 Management device 41 Display section 42 Input section 45 Storage section 46 Management information 47. Verification Information 48. Equipment Information 50 Networks 60 Work Databases 62 Work Data 64 Blockchains 65 blocks 66 Hash value of the previous block 90 Historical Data
Claims
1. In a medical device management system that manages medical devices marked with symbols, The system includes a storage unit that stores a work database in which work data describing events that occurred with respect to the medical device is registered, and at least one data registration device that registers the work data to the work database, The memory unit stores matching information that shows the correspondence between local characteristics unique to the medical device and identification information that individually identifies the medical device. The data registration device comprises an imaging unit for capturing an image of the medical device and an extraction unit for extracting the local features of the medical device from the image of the medical device. The extraction unit identifies the position of the symbol in the image of the medical device and extracts the local features from an extraction range determined based on the position of the symbol. The data registration device identifies the identification information corresponding to the medical device by comparing the local features extracted by the extraction unit with the matching information, and registers the work data relating to the medical device corresponding to the identified identification information into the work database, the medical device management system.
2. The medical device management system includes an identification information assignment point that assigns the identification information to the medical device, The medical device management system according to claim 1, wherein when the identification information is assigned to the medical device at the identification information assignment point, the local features extracted from the extraction range determined based on the position of the symbol attached to the medical device are registered as matching information in association with the identification information.
3. The matching information includes the correspondence between the identification information of the medical device and the symbol information indicated by the symbol attached to the medical device. The aforementioned data registration device is An attempt is made to read the symbol information indicated by the symbol in the image of the medical device. If the reading of the symbol information is successful, the identification information corresponding to the medical device is identified by comparing the symbol information with the matching information. If reading the symbol information fails, the medical device management system according to claim 1, wherein the local features extracted by the extraction unit are compared with the matching information to identify the identification information corresponding to the medical device.
4. The medical device management system according to claim 1, wherein the work database is a distributed ledger based on blockchain technology, and the data registration device registers the identification information of the medical device and the corresponding transaction data in the distributed ledger.