Method of identifying a medical device

Abstract

The invention relates to a method of identifying a medical device, referred to as a target medical device (50), comprising the steps of: a. collecting images of an external part of the target medical device (50), each image being labeled with an identifier of the target medical device (50), the collected images forming a training image database, b. training an identification model from the training image database to obtain a trained identification model, the identification model assigning an identifier to a target medical device (50) to be identified from an image of an external part (54) of the medical device (50), c. receiving an image of an external part (54) of a medical device (50) to be identified, d. assigning an identifier to the target medical device (50) imaged on the received image by the trained classification model, and e. characterizing the target medical device (50) according to the assigned identifier.

Description

[0001] This invention relates to a method for identifying medical devices. The invention also relates to a related computer program.

[0002] Different medical devices can be used to administer intravenous fluids to patients. Generally, these devices include internal parts for insertion into the patient's body, such as those inserted into veins, arteries, esophagus (or other parts of the digestive system), cerebrospinal fluid (epidural catheters), or subcutaneous devices (such as implantable chambers); and external parts visible from the outside for manipulation by healthcare personnel, such as for injecting medications or nutritional solutions.

[0003] However, once a medical device has been implanted in a patient, it is not always easy for healthcare professionals to identify the type of device. Incorrect identification, such as using the device inappropriately or injecting unsuitable substances, can potentially lead to serious side effects for the patient.

[0004] The risk of confusion is particularly high in the case of PICC (peripherally inserted central venous catheter) and midline catheters (peripherally inserted venous catheters) because both catheters are located in the patient's arm. The risk of confusion is specifically described in Letournel C. Gestion du risque de confusion entre PICC Lines et midline catheters [Management of the risk of mix-up between PICC Lines and midline catheters J PharmClin 2018;37(1):19-26 doi:10.1684 / jpc.2017.0374]. If the maximum usage period is not adhered to, confusion between PICC and midline catheters is likely to result in the risk of intravenous toxicity or infection through the midline catheter.

[0005] Furthermore, it seems that healthcare professionals are not always clear about the specific uses and optimal procedures for different medical devices (Recommendations SF2H: Prevention of intravascular catheter-related infections. [Guidelines for the prevention of intravascular catheter-related infections] May 2019, Hygiènes Vol XXVII–no.2; Infusion Therapy Standards of Practice 8th Edition, Jan / Feb 2021, Journal of Infusion Nursing, Vol 44, no.51), whether they are inserted into the body, in contact with the skin, or connected to devices that are inserted into the body or in contact with the skin. This confusion can also lead to complications in patients due to incorrect use of the devices.

[0006] Therefore, it is necessary to adopt methods to reduce the risk of healthcare workers misidentifying the medical devices carried by patients, thereby reducing the risk of misuse or non-compliance with best practices for the use of such devices.

[0007] Therefore, the subject of this specification is a method for identifying a medical device (referred to as a target medical device), which is a device for insertion into or in contact with a patient's body, or for connection to a device inserted into or in contact with a patient's body, each target medical device having an external portion outside the patient's body, the method comprising the following steps:

[0008] - Images of the external portion of the target medical device are collected, each image labeled with an identifier of the target medical device from a set of predefined identifiers. The collected images form a training image database. This collection step is performed by a computer.

[0009] - A recognition model is trained from the training image database to obtain a trained recognition model. This model assigns an identifier from a predefined set of identifiers to the target medical device based on images of the external portion of the medical device. The training step is performed by a computer.

[0010] - Receive an image of the external portion of the medical device (50) to be identified, the receiving step being performed by a computer.

[0011] - Identifiers are assigned to target medical devices imaged on the received images by a trained classification model; the classification step is performed by a computer.

[0012] - The target medical device is characterized according to the assigned identifier.

[0013] According to other specific implementations, the method includes one or more of the following features, which can be used individually or in any technically possible combination:

[0014] The method includes the step of processing the target medical device based on the characterization performed.

[0015] - The training image database consists of images of at least two different types of target medical devices, and the at least two types of target medical devices imaged on the images in the training image database are selected from the following types: medical tubing, intravenous catheters, arterial catheters, enteral nutrition catheters, parenteral nutrition catheters, drainage catheters, epidural catheters, systemic catheters, peripherally inserted central venous catheters (PICC), peripheral venous catheters (midline catheters), central venous catheters (CVC), umbilical catheters, Huber needles, percutaneous medical tubing fixation systems, covering dressings, valves, closure systems, and plugs connected to the medical tubing interface.

[0016] - Each identifier is associated with data related to the type of medical device, and the characterization step includes determining the type of the target medical device imaged on the received image based on the assigned identifier.

[0017] - Each identifier is associated with a usage recommendation for the target medical device, and the characterization step includes determining a usage recommendation for the target medical device imaged on the received image based on the assigned identifier.

[0018] -The usage recommendations include one or more of the following:

[0019] o is the first route of administration for inserting the target medical device.

[0020] o The maximum rate at which liquid is injected into the target medical device.

[0021] o The maximum failure pressure of the target medical device.

[0022] o If dressings are used, the frequency of dressing replacement on the target medical device.

[0023] o. Compatibility of the target medical device with other types of medical devices.

[0024] The compatibility of the target medical device with the drug solution, and

[0025] o The compatibility of the target medical device with predefined medical examinations and, where appropriate, the conditions for conducting the examinations.

[0026] - Each identifier is associated with an internet link and / or a remote database through which visualization support regarding the use of the target medical device is available. The characterization step includes displaying visualization media corresponding to the internet link or the database associated with the assigned identifier, such as videos, images, or articles.

[0027] - Each identifier is associated with at least one structural feature of the target medical device, the structural feature or at least one structural feature being selected from: the brand name of the target medical device, the contact information of the legitimate manufacturer of the target medical device, the manufacturing material of the target medical device, the potentially allergenic raw materials used in manufacturing the target medical device (if any), the size of the target medical device, and, when the target medical device is a catheter or medical tube, the diameter and initial length of the catheter or medical tube, the number of lumens of the catheter or medical tube, the dead volume of each lumen of the catheter or medical tube, and the flow rate of the contrast agent of the catheter or medical tube when injected under gravity or by a pump or syringe, the characterization step comprising determining the structural features of the target medical device (50) imaged on the received image according to the assigned identifier.

[0028] - Each target medical device has an internal portion connected to the external portion, the internal portion being a tube or needle inserted into the patient's body, and the external portion including at least a tube and / or a connector connected to the tube.

[0029] - The identifier assigned to the target medical device is associated with a remote database from which usage information of the medical device can be obtained.

[0030] - The usage information includes data defined by the manufacturer and / or data defined by medical professionals when using the medical device on a patient.

[0031] The method includes the steps of updating the usage information of the medical device for a given patient and / or a given purpose, and archiving the usage information to ensure patient follow-up and / or traceability of the medical device.

[0032] - The identifier is associated with information related to the environmental impact and / or eco-design of the target medical device.

[0033] This specification also relates to a computer program product, including program instructions stored on a computer-readable medium for performing the methods described above when a computer program is executed on a computer.

[0034] This description further relates to a readable information medium on which a computer program product as described above is stored. Attached Figure Description

[0035] Other features and benefits of the invention will be apparent from the following description of embodiments of the invention (by way of example only) and with reference to the accompanying drawings:

[0036] [ Figure 1 ], Figure 1 This is a schematic diagram of a computer example implementing a method for identifying medical devices.

[0037] [ Figure 2 ], Figure 2 This is a flowchart illustrating an example of a method for identifying medical devices.

[0038] [ Figure 3 ], Figure 3 This is a schematic diagram of a catheter example.

[0039] [ Figure 4 ], Figure 4 This is a schematic diagram of a medical tube example.

[0040] [ Figure 5 ], Figure 5 This is a diagram illustrating a fixed Huber pin instance.

[0041] [ Figure 6 ], Figure 6 This is a schematic diagram of an example of a medical device that comes into contact with a patient's skin (a transparent dressing covering the percutaneous insertion site of a catheter).

[0042] [ Figure 7 ], Figure 7 This is a diagram illustrating another example of a medical device that comes into contact with a patient's skin (catheter insertion dressing), and

[0043] [ Figure 8 ], Figure 8 This is a schematic diagram of an example of a closed system connected to the interface of a conduit. Detailed Implementation

[0044] Figure 1 The calculator 10 and computer program product 12 are displayed.

[0045] Calculator 10 is preferably a computer.

[0046] More generally, calculator 10 is an electronic calculator suitable for processing and / or converting data into other similar data corresponding to physical data in storage, registers or other display types, said data being presented as electronic or physical quantities in the registers and / or memory of calculator 10.

[0047] The calculator 10 interacts with the computer program product 12.

[0048] like Figure 1 As shown, the calculator 10 includes a processor 14, which includes a data processing unit 16, a memory 18, and a data storage medium 20. Figure 1 In the example shown, the calculator 10 includes a keyboard 22 and a display unit 24.

[0049] The computer program product 12 includes a storage medium 26.

[0050] Storage medium 26 is a readable medium of calculator 10, typically readable by data processing unit 16. The readable storage medium 26 is suitable for storing electronic instructions and is adapted for connection to the bus of a computer system.

[0051] For example, storage medium 26 is a floppy disk or floppy disk, optical disk, CD-ROM, magneto-optical disk, ROM, RAM, EPROM, EEPROM, magnetic card or optical card.

[0052] A computer program 12 containing program instructions is stored on storage medium 26.

[0053] The computer program 12 can be loaded into the data processing unit 16, and when the computer program 12 is executed on the processing unit 16 of the calculator 10, it is suitable for guiding the implementation of the medical device identification method.

[0054] Although Figure 1 The example shown is a calculator in the form of a physical computer, but this instruction manual is not limited to this type of computer. This manual extends to any device that integrates a calculator linked to a computer program (computer application).

[0055] Therefore, in one implementation, the calculator and computer program product 12 is integrated into a smartphone, touch-sensitive tablet, or any other connected object.

[0056] Now refer to Figure 2 and Figures 3 to 8 To describe the operation of Calculator 10, Figure 2 A flowchart illustrating a method for identifying medical devices is shown. Figures 3 to 8 An example of a medical device considered within the framework of this method is shown.

[0057] The medical device considered within the framework of this method is referred to as the target medical device 50.

[0058] The target medical device 50 includes, for example, an internal portion 52 and an external portion 54 connected to the internal portion 52.

[0059] The internal portion 52 is for insertion into the patient's body. When the medical device 50 is already inserted into the patient's body, the internal portion 52 is therefore not visible to medical personnel. The internal portion 52 is typically a tube or needle.

[0060] The external portion 54 is intended to remain outside the patient's body for manipulation by healthcare personnel, such as for injecting medication. The external portion 54 includes at least a tube and / or a tube connector.

[0061] The target medical device 50 is considered to have an internal part 52 connected to the external part 54, such as a catheter (e.g.) Figure 3 (as shown), a wider range of medical tubes (such as) Figure 4 (as shown) and / or Huber pins (such as Figure 5 (As shown).

[0062] like Figure 3 As shown, the outer portion 54 of the catheter includes a tube 59, typically with an interface 60, a clip 62, and a connector 64 (Luer lock). A valve (not shown) is typically inserted at the end of the connector 64. The interface 60 is typically covered with an adhesive or subcutaneous anchoring system and a transparent dressing.

[0063] Catheters include, for example, venous catheters, arterial catheters, enteral / parenteral nutrition catheters, drainage catheters, epidural catheters, systemic catheters, or intravenous catheters. Clearly, for all such catheters and medical tubes, there are problems related to confusion over different routes: particularly central versus peripheral, enteral versus parenteral, pulmonary versus other routes, and intrathecal versus intravenous (cf. Outils de Sécurisation et d'auto-évaluation de l'administration des médicaments [Tools for safety and self-assessment of drug administration], HAS, May 2013; ISO 80369-1:2018). Multiple routes of administration and failure to recognize catheters or medical tubes can sometimes lead to serious errors. For example, administering highly neurotoxic vinca alkaloids via the intrathecal route can result in patient death. Furthermore, injecting molecules classified as blistering agents into inappropriate peripheral catheters (S. Manrique-Rodriguez, Standardization and Chemical Characterization of intravenous Therapy in Adult Patients: A Step Further in Medication Safety, Drugs in R&D, December 2020) may lead to risks of phlebitis, necrosis, or even extravasation.

[0064] More specifically, the catheter is, for example, a peripherally inserted central catheter (referred to as PICC), a peripheral intravenous catheter (referred to as a midline catheter), a central venous catheter (referred to as CVC), a central umbilical catheter, or an enteral nutrition catheter.

[0065] A PICC is a venous catheter that is inserted above the elbow into the deep vein of the arm and reaches the cavoatrial junction. Therefore, its distal end is located at the entrance of the heart. PICC can particularly be used for the following injections: chemotherapy, antibiotic therapy, parenteral feeding and infusion. A PICC can be implanted in a patient's body for up to 6 months.

[0066] A midline catheter is a catheter that is inserted into the vein of the arm from the bend of the elbow and then reaches the axilla. Therefore, the distal end is located below the clavicle. Midline catheters are particularly used for injecting non-irritating products (<900mOs / l, 5<pH<9) such as antibiotics, analgesics or hydration substances during long-term treatment (more than 8 days). A midline catheter can be implanted in a patient's body for up to 30 days.

[0067] A CVC is a catheter that is inserted into the neck, chest or femur. The distal end of this catheter is located at the entrance of the heart.

[0068] Another type of catheter is such as Figure 4 shown in . In Figure 4 the example shown in , the medical device 50 is a tube 65, one end of which is configured to be inserted into a patient's body, thus forming an internal portion 52, and the other end forms an external portion 54. In Figure 4 the specific example of , the tube is an oxygen tube. However, the medical device 50 may be any other type of tube.

[0069] A Huber needle is a device used for an implantable catheter port (CCI). A port CCI is a subcutaneously implantable device composed of an injection container and a central catheter. The Huber needle performs injection and sampling by puncturing the subcutaneous port housing. The Huber needle particularly comprises a (tangential bevel) needle that is inserted into the patient's skin at the housing, and a connector, which is configured to remain externally to be connected to, for example, a tube.

[0070] As shown in Figure 5 , the internal portion 52 of the Huber needle is a (tangential bevel) needle configured to be inserted into the patient's skin at the housing. The external portion 54 comprises a connector 70 configured to remain externally to be connected to, for example, a tube.

[0071] Furthermore, or in a variant, the target medical device 50 only comprises the external portion 54 that is visible outside the patient (there is no internal portion inserted into the body).

[0072] In this case, the external portion 54 can be used to contact the patient's skin, or to connect to a medical device inserted into the patient's body or in contact with the patient's skin.

[0073] The target medical device 50, which is considered to have only one external part 54, is, for example, a percutaneous system, such as: a percutaneous fixation system for medical tubes (catheters, drainage tubes, etc.), a covering dressing, a valve, a sealing system, and a plug connected to the interface of the medical tube (e.g., the interface 60 of the catheter).

[0074] Figure 6 An example is given of a target medical device 50 that comes into contact with a patient's skin. In this example, the medical device 50 is a transparent dressing 71 that covers the percutaneous insertion site 72 of the catheter 73.

[0075] Figure 7 This is another example of the target medical device 50 coming into contact with the patient's skin. In this example, the medical device 50 is a dressing 74 used to secure the catheter 75.

[0076] Figure 8 This is another example of a medical device 50 being connected to a medical device for insertion into a patient's body or contact with the patient's skin. In this example, the medical device 50 is a closed system 76 connected to an interface 77 of a catheter.

[0077] The identification method includes step 100 of collecting images of the external portion of the target medical device 50. The collected images thus form a database of training images. Step 100 is performed by interaction between the calculator 10 and the computer program product 12, i.e., by a computer.

[0078] Images in the database are captured, for example, by a smartphone camera.

[0079] Preferably, the collected images are of the target medical device 50 in use, i.e., when the device 50 includes an internal part 52, the internal part is inserted into the patient's body; when the device 50 includes only an external part 54, the external part 54 is in contact with the patient's skin, or connected to other medical devices inserted into or in contact with the patient's skin. Therefore, only the external part of the medical device can be seen in the images.

[0080] In one variant, the collected images also include images of the target medical device 50 not being worn by the patient.

[0081] The training image database typically consists of images of at least two different types of target medical devices.

[0082] In one instance, at least two types of medical devices 50 imaged on images in the training image database are selected from the following types: medical tubes, venous catheters, arterial catheters, enteral nutrition catheters, parenteral nutrition catheters, drainage catheters, epidural catheters, systemic catheters, peripherally inserted central venous catheters (PICC), peripheral venous catheters (midline catheters), central venous catheters (CVC), umbilical catheters, Huber needles, percutaneous fixation systems for medical tubes (catheters, drainage tubes, etc.), covering dressings, valves, closure systems, and caps connected to the interfaces of the medical tubes.

[0083] In a specific instance, at least one imaging device is a peripherally inserted central venous catheter, referred to as a PICC, and at least one imaging device is a peripheral venous catheter, referred to as a midline catheter, in order to distinguish between the two types of catheters.

[0084] Each image is labeled with an identifier for the target medical device 50 from a set of predefined identifiers. The identifiers are determined by the operator.

[0085] Each predefined identifier corresponds to the same type of medical device, or even a specific model of medical device (same reference). Therefore, medical devices with the same identifier share common characteristics or correspond to the same device.

[0086] Preferably, each identifier is associated with relevant data for the type of medical device. The type of medical device may be one or more of the following: medical tubing, intravenous catheters, arterial catheters, enteral nutrition catheters, parenteral nutrition catheters, drainage catheters, epidural catheters, systemic catheters, peripherally inserted central venous catheters (PICC), peripherally inserted venous catheters (CVC), central venous catheters (CVC), umbilical catheters, Huber needles, percutaneous fixation systems for medical tubing (catheters, drainage tubes, etc.), covering dressings, valves, closure systems, and caps connected to the interfaces of the medical tubing.

[0087] In a specific instance, at least one device is a peripherally inserted central venous catheter, referred to as a PICC, and at least one device is a peripherally inserted venous catheter, referred to as a midline catheter, in order to distinguish between the two types of catheters.

[0088] Preferably, each identifier is associated with usage recommendations for the target medical device 50. These recommendations are designed to guide healthcare professionals in the proper use of the device.

[0089] In one instance, the recommendations include one or more of the following:

[0090] a. First route of administration (peripheral, central, intestinal, etc.)

[0091] b. The maximum rate at which liquid (by gravity, a contrast agent pump, or a syringe) is injected into the target medical device 50.

[0092] c. The maximum failure pressure of the target medical device 50.

[0093] d. Frequency of dressing (if any) replacement on target medical device 50,

[0094] e. Compatibility of the target medical device 50 with other types of medical devices.

[0095] f. Compatibility between the target medical device 50 and the medication, and

[0096] g. Compatibility of the target medical device 50 with predetermined medical examinations and the conditions under which the examinations are performed, where appropriate.

[0097] Preferably, each identifier is associated with an internet link and / or a remote database through which visual media related to the use of the target medical device 50 can be obtained. The visual media is provided for healthcare personnel to view or read to aid in the proper use of the target medical device 50. The visual media may be, for example, videos, images, or articles.

[0098] Preferably, each identifier is associated with a structural feature of the target medical device 50. Structural features include, for example, one or more of the following: the brand name of the target medical device 50, contact information for the legitimate manufacturer of the target medical device 50, the manufacturing materials of the target medical device 50, potential allergenic raw materials of the target medical device 50 (if any), the dimensions of the target medical device 50, and, when the target medical device 50 is a catheter or medical tube, the diameter and initial length of the catheter or medical tube, the number of lumens in the catheter or medical tube, the dead volume of each lumen (path) of the catheter or medical tube, and the flow rate of the contrast agent in the catheter or medical tube when injected by gravity or via a pump or syringe.

[0099] The identification method includes step 110, which involves training a recognition model from a training image database. Step 110 is used to obtain a recognition model trained to assign an identifier to the target medical device 50 from a set of predefined identifiers based on an image of the external portion 54 of the target medical device 50. Step 100 is performed interactively by the calculator 10 and the computer program product 12, i.e., by a computer.

[0100] In one implementation example, the recognition model is trained according to a learning method applied to the database.

[0101] The recognition model is, for example, a neural network. The learning method model is used to configure the neural network when learning from a database. In one instance, one part of the database is used to configure the neural network, and another part is used to validate the configuration.

[0102] Typically, the trained model tends to first identify the specific model of the medical device corresponding to the identifier based on the device model's specific physical characteristics (see reference). Physical characteristics are, for example, the color, shape, or accessories of the medical device. Then, other characteristics of the device (e.g., device type, usage recommendations, etc.) are looked up based on the identifier.

[0103] The identification method includes step 120 of receiving an image of the external portion 54 of the target medical device 50 to be identified. This step 120 is performed interactively by the calculator 10 and the computer program product 12, i.e., by the computer.

[0104] The imaging target medical device 50 is, for example, being used on a patient. In another case, the imaging medical device 50 is not being used by a patient.

[0105] For example, the received image is acquired by a smartphone camera. This acquisition may be initiated by healthcare personnel to identify a target medical device 50, such as one or more target medical devices 50 inserted into or in contact with the patient's skin.

[0106] The identification method includes step 130, which assigns an identifier to a target medical device 50 imaged on a received image by a trained classification model. The classification step is performed interactively by a calculator 10 and a computer program product 12, i.e., by a computer.

[0107] Preferably, the probability representing the confidence level is also determined by identifying the identifier assigned to the target medical device 50. Therefore, healthcare personnel are informed that there is a possibility of error in the identification performed. For example, if the probability is below a predetermined threshold (e.g., 85%), the identification is considered potentially erroneous.

[0108] The identification method includes step 140 of characterizing the target medical device 50 according to an assigned identifier. Therefore, at the end of step 140, at least one feature relating to the target medical device 50 imaged on the received image can be obtained.

[0109] For example, when the identifier is associated with data related to the type of medical device, the characterization step 140 includes determining the type of the target medical device 50 that is imaged on the received image based on the assigned identifier.

[0110] For example, when an identifier is associated with a usage recommendation for the target medical device 50, the characterization step 140 includes determining, based on the assigned identifier, a usage recommendation for the target medical device 50 that is imaged on the received image.

[0111] For example, when the identifier is associated with an internet link or a remote database, corresponding to visual support related to the use of the target medical device 50, the characterization step 140 includes displaying visual media corresponding to the internet link or the database associated with the assigned identifier.

[0112] For example, when an identifier is associated with one or more structural features, the characterization step 140 includes determining one or more structural features of the target medical device 50 imaged on the received image based on the assigned identifier.

[0113] Characterization step 140 is performed, for example, by the interaction between calculator 10 and computer program product 12, i.e., by a computer. For example, features are extracted from a database based on assigned identifiers.

[0114] In one variant, characterization step 140 is performed by an operator, such as a healthcare professional (changing a medical device, injecting a medication). For example, the operator would search a database for features corresponding to the assigned identifier.

[0115] In one implementation example, the identification method includes step 150 of processing the target medical device 50 based on the characterization performed. For example, the processing may be performed by an operator (such as a healthcare professional).

[0116] For example, based on the characterization performed, healthcare professionals will handle the target medical device 50 differently. For instance, the manipulation and injected substances for PICC and midline catheters may differ.

[0117] Therefore, this method can identify one or more medical devices from a set of reference medical devices. This approach can thus help healthcare professionals identify devices and reduce the risk of confusion.

[0118] By understanding the usage recommendations and / or characteristics of the identified devices, the identification of target medical devices can further achieve the following objectives:

[0119] - Limit the risk of injecting molecules that are not suitable for the drug delivery route of the device.

[0120] - Follow best practices for using and maintaining equipment, and

[0121] - Understand the incompatibility of the device with other medical devices.

[0122] Those skilled in the art will understand that the above-described embodiments and variations can be combined to form new embodiments, provided that they are technically compatible.

[0123] More specifically, as a variation or optional supplement to the aforementioned implementation scheme, an identifier assigned to the target medical device is associated with a remote database through which usage information of the medical device can be obtained. This makes it possible to obtain information that patients cannot readily access at home or even in a hospital. However, this information is crucial for the safe use of the medical device. More specifically, without access to usage information, the following two typical examples could have serious consequences for patients:

[0124] - First case: The maximum injection rate was not observed, which led to catheter rupture and potentially extravasation.

[0125] -Second example: Little is known about the presence of catheters in central or peripheral venous networks, which could lead to the injection of venous toxic molecules into peripheral veins, damaging the patient's veins.

[0126] Therefore, the method can obtain this usage information when the identifier assigned to the target medical device is associated with a remote database. Usage information is typically manufacturer-specific (to ensure correct use) and / or established / supplemented by medical professionals based on actual usage data. Thus, the usage information can be, for example:

[0127] - Information from the manufacturer (maximum flow rate), meaning that relevant information cannot be found through internet searches or bibliographical articles, especially if relevant references cannot be identified, and / or

[0128] -From professional healthcare workers themselves (center or periphery).

[0129] In one implementation example, usage information initially includes default information, such as information related to manufacturer-defined characteristics of the medical device, such as central or peripheral. Secondly, personnel handling the medical device (medical professionals) supplement the usage information by entering additional information into a database, such as the name of the vein used for the first time, the length of the implanted catheter, and the date of implantation. Therefore, this additional information complements the default information. Preferably, usage information is also supplemented by any user of the medical device (medical professionals) by incorporating additional information to ensure the traceability and safety of the medical device's use; this additional information includes, for example, the type of treatment administered and the quality of the injection (no pain, redness, etc.).

[0130] As an optional add-on or variation, usage information can be archived in a database (app or storage) to provide follow-up for patients. For example, once a medical device is identified, a healthcare professional can return to the app to find explanatory information from previous uses (manufacturer and their own follow-up). The healthcare professional will be able to find all scanned and used medical devices and their intended use, ensuring the traceability and security of their actions.

[0131] Preferably, the usage information is suitable for extraction from the follow-up (e.g., all catheters placed / used by the operator) to generate statistical reports, which are available to healthcare professionals, for example, at the medical law level, or for monitoring clinical activity, or as a post-market surveillance tool (post-CE) for manufacturers under new MDR regulations. In the latter framework, the goal is to establish a database of medical device, particularly catheter, usage.

[0132] As a variant or optional additional feature, the identifier is associated with information related to the environmental impact and / or eco-design of the target medical device. This information is published, for example, by the manufacturer. This information includes, for example, a rating of the target medical device. Specifically, this information includes one or more of the following: carbon impact, any additives contained in the components of the target medical device, and information regarding the recycling (packaging) of the target medical device.

Claims

1. A method for identifying a medical device, the medical device being referred to as a target medical device (50), the target medical device (50) being a device for insertion into or in contact with a patient's body, or a device for connection to a device for insertion into or in contact with a patient's body, each target medical device (50) having an external portion (54) outside the patient's body, the method comprising the steps of: a. Images of the external portion of the target medical device (50) are collected, each image labeled with an identifier of the target medical device (50) from a set of predefined identifiers. The collected images form a training image database. This collection step is performed by a computer. b. A recognition model is trained from the training image database to obtain a trained recognition model, which assigns an identifier from a set of predefined identifiers to the target medical device (50) to be identified based on an image of the external part (54) of the medical device (50). This training step is performed by a computer. c. Receive an image of the external portion (54) of the target medical device (50) to be identified; this receiving step is performed by a computer. d. Identifiers are assigned to target medical devices (50) imaged on the received images by a trained classification model; this classification step is performed by a computer. e. Characterize the target medical device (50) according to the assigned identifier; Each identifier is associated with an Internet link and / or a remote database through which visualization support regarding the use of the target medical device (50) is available, and the characterization step includes displaying visualization support media corresponding to the database associated with the Internet link or the assigned identifier.

2. The method of claim 1, wherein the method includes the step of processing the target medical device (50) according to the characterization performed.

3. The method according to claim 1 or 2, wherein the training image database consists of images of at least two different types of target medical devices (50), and the at least two different types of target medical devices (50) imaged on the images in the training image database are selected from the following types: medical tubes, venous catheters, arterial catheters, enteral nutrition catheters, parenteral nutrition catheters, drainage catheters, epidural catheters, systemic catheters, peripherally inserted central venous catheters (PICC), peripheral venous catheters (midline catheters), central venous catheters (CVC), umbilical catheters, Huber needles, percutaneous fixation systems for medical tubes, covering dressings, valves, closure systems, and plugs connected to the medical tube interface.

4. The method of claim 1 or 2, wherein each identifier is associated with data relating to the type of medical device, and the characterization step includes determining the type of the target medical device (50) imaged on the received image based on the assigned identifier.

5. The method of claim 1, wherein each identifier is associated with a usage recommendation for the target medical device (50), and the characterization step includes determining a usage recommendation for the target medical device (50) imaged on the received image based on the assigned identifier.

6. The method of claim 5, wherein the usage recommendations include one or more of the following recommendations: a. A first route of administration for inserting the target medical device (50), b. The maximum rate at which liquid is injected into the target medical device (50), c. The maximum failure pressure of the target medical device (50), d. If dressings are used, the frequency of dressing replacement on the target medical device (50), e. Compatibility of the target medical device (50) with other types of medical devices, f. The compatibility of the target medical device (50) with the drug solution, and g. The compatibility of the target medical device (50) with predefined medical examinations and, where appropriate, the conditions under which the examinations are performed.

7. The method of claim 1, wherein each identifier is associated with at least one structural feature of the target medical device (50), the at least one structural feature being selected from: the brand name of the target medical device (50), the contact information of the legitimate manufacturer of the target medical device (50), the manufacturing material of the target medical device (50), the potentially allergenic raw materials used to manufacture the target medical device (50), the size of the target medical device (50), and, when the target medical device (50) is a catheter or medical tube, the diameter and initial length of the catheter or medical tube, the number of lumens of the catheter or medical tube, the dead volume of each lumen of the catheter or medical tube, and the flow rate of the contrast agent of the catheter or medical tube when injected under gravity or by a pump or syringe, the characterization step comprising determining the structural feature of the target medical device (50) imaged on the received image according to the assigned identifier.

8. The method of claim 1, wherein each target medical device (50) has an internal portion (52) connected to the external portion (54), the internal portion (52) being a tube or needle inserted into a patient, and the external portion (54) comprising at least one tube (59) and / or a connector connected to the tube (64; 70).

9. The method of claim 1, wherein the identifier assigned to the target medical device is associated with a remote database from which usage information of the medical device is available.

10. The method of claim 9, wherein the usage information includes manufacturer-defined data and / or data defined by a medical professional when using the medical device on a patient.

11. The method of claim 9, wherein the method includes the step of updating usage information of the medical device for a given patient and / or a given use, and archiving the usage information to ensure follow-up of the patient and / or traceability of the medical device.

12. The method of claim 1, wherein the identifier is associated with information relating to the environmental impact and / or eco-design of the target medical device.

13. A computer program product comprising program instructions stored on a computer-readable storage medium, which, when executed on a computer, implement the steps of the identification method according to claim 1.

Images