WO2024202488A1 - 学習モデル作成装置、学習モデル作成方法 - Google Patents
学習モデル作成装置、学習モデル作成方法 Download PDFInfo
- Publication number
- WO2024202488A1 WO2024202488A1 PCT/JP2024/002574 JP2024002574W WO2024202488A1 WO 2024202488 A1 WO2024202488 A1 WO 2024202488A1 JP 2024002574 W JP2024002574 W JP 2024002574W WO 2024202488 A1 WO2024202488 A1 WO 2024202488A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- learning
- learning model
- data
- request
- model creation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
Definitions
- the present invention relates to a learning model creation device and a learning model creation method.
- learning models In various fields, including the medical field, it is common to create learning models using machine learning. As is well known, when creating learning models using machine learning, a more accurate learning model can be created by performing machine learning (hereinafter sometimes simply referred to as learning) using a larger amount of data.
- Patent Document 1 describes a configuration in which medical data is provided in exchange for payment, and by using the data provided in this way and increasing the amount of data used for learning, a more accurate learning model can be created.
- one data (medical data) provided in the above Patent Document 1 can be used in multiple machine learning (to create multiple learning models). That is, for example, data of an 80-year-old man can be used to create multiple learning models, such as a learning model for men only, a learning model for adult men, a learning model for elderly men, a learning model for adults regardless of gender, and a learning model for elderly people regardless of gender.
- the data also includes various information (for example, height, weight, blood pressure, chronic diseases, and past medical conditions, etc., if the data is medical data).
- the more types of information included in the data the more learning models can be created using this data. For this reason, once medical data is provided, there is a problem in that it is not possible to grasp the manner in which the provided data is used (such as which learning model it was used to create and how many times it was used).
- the present invention has been made in consideration of the above background, and aims to provide a learning model creation device and a learning model creation method that can grasp the usage patterns of data.
- the learning model creation device of the present invention creates a learning model by machine learning using data stored in a data management server, and includes a learning request receiving unit that receives a learning request sent from a learning request source, the learning request specifying a learning algorithm to be used in creating the learning model and a range of data to be used in creating the learning model, a data reading unit that reads data within the range specified in the learning request from the data management server, and a learning model creation unit that creates a learning model using the data read from the data management server and the learning algorithm specified in the learning request.
- the data may be medical data.
- the system may also include a learning model transmission unit that transmits the created learning model to the learning request source.
- the system may also include a learning process maintenance unit that stores the learning process, including identification information for identifying the data used to create the learning model, on the blockchain.
- the system may also include a fee calculation unit that calculates the fee for using the data used to create the learning model.
- the learning model creation unit may create multiple learning models by performing multiple machine learning operations using multiple groups of data, some of which overlap.
- the compensation calculation unit may calculate the compensation for use of data that has been used multiple times in machine learning according to the number of times the data has been used multiple times.
- the learning model creation unit may treat each of the multiple learning models as a local learning model, and create one global learning model by integrating these local learning models.
- the learning request source may treat each of the multiple learning models as a local learning model, and integrate these local learning models to create a single global learning model.
- the learning model creation method of the present invention creates a learning model by machine learning using data stored in a data management server, and includes a learning request receiving step for receiving a learning request transmitted from a learning request source, the learning request specifying a learning algorithm to be used in creating the learning model and a range of data to be used in creating the learning model, a data reading step for reading from the data management server the data within the range specified in the learning request, and a learning model creation step for creating a learning model using the data read from the data management server and the learning algorithm specified in the learning request.
- the present invention provides a learning model creation device and a learning model creation method that can grasp how data is used.
- FIG. 1 is a schematic diagram showing a configuration of a learning system.
- FIG. 1 is an explanatory diagram of medical data.
- FIG. 2 is an explanatory diagram illustrating functions of a processor. 13 is a flowchart showing a process flow of a processor.
- FIG. 2 is an explanatory diagram illustrating functions of a processor.
- FIG. 2 is an explanatory diagram illustrating functions of a processor.
- the learning system 10 is composed of a medical data provider environment 12 , a learning requester environment 14 (learning requester), and a learning environment 16 .
- the medical data source environment 12 is, for example, a medical institution such as a hospital, or an environment consisting of a part of such a medical institution, and is equipped with a medical data management server 22 (data management server) that stores medical data 20 (data).
- the medical data 20 is medical information obtained by the medical institution during examinations, etc., such as the results of a health check.
- medical data 20 includes n pieces of medical data 20(1) to 20(n) for n patients with patient numbers 1 to n (n is an arbitrary integer) stored in a medical data management server 22.
- Each piece of medical data 20(1) to 20(n) includes information such as the patient number mentioned above, as well as name, age, sex, height, and weight. Note that this embodiment will be described using an example in which the data is medical data, but the present invention is not limited to this. The data may be other than medical data.
- the learning requesting environment 14 is, for example, an environment configured by a medical manufacturer that develops and manufactures pharmaceuticals and medical devices, or a part of such a manufacturer, and makes a learning request (learning request) to the learning environment 16 for the purpose of product development and improvement.
- the learning requesting environment 14 is equipped with a requesting terminal 30, and the learning request is made via the requesting terminal 30.
- the type of learning algorithm to be used for learning e.g., deep learning using a neural network
- the range of medical data to be used for learning e.g., men over 80 years old
- the learning request specifies the (type of) learning algorithm to be used for learning, and the range of medical data to be used for learning.
- the learning environment 16 is an environment in which a learning model is generated by machine learning using medical data, and includes a processor 40 (learning model creation device).
- the processor 40 includes a memory 42 and a central control unit 44.
- the memory 42 stores a plurality of learning programs corresponding to various learning algorithms used in machine learning, and a control program for causing the processor 40 (central control unit 44) to function as a learning model creation device.
- the central control unit 44 reads and executes the control program from the memory 42 when the processor 40 is started.
- the processor 40 central control unit 44 functions as a learning request receiving unit 50, a medical data reading unit 52 (data reading unit), a learning model creation unit 54, and a learning model transmission unit 56.
- the learning request receiving unit 50 receives a learning request sent from the learning request source environment 14 (see FIG. 1). As described above, the learning request specifies the (type of) learning algorithm to be used for learning, and the range of medical data to be used for learning. The learning request is transmitted to the processor 40 via the request source terminal 30. The learning request receiving unit 50 receives (accepts) the learning request transmitted in this manner. In this manner, the learning request receiving unit 50 executes the learning request receiving step of the present invention.
- the medical data reading unit 52 accesses the medical data management server 22 and reads out medical data that falls within the range specified in the learning request (see FIG. 1). For example, if the range of medical data specified in the learning request is for men aged 80 or older, in this embodiment, as shown in FIG. 2, medical data 20(1) falls within the specified range of medical data. Therefore, the medical data reading unit 52 reads out this medical data 20(1) from the medical data management server 22. In this way, the medical data reading unit 52 executes the data reading step of the present invention.
- the learning model creation unit 54 performs machine learning using the medical data read by the medical data reading unit 52 to create a learning model.
- the learning model creation unit 54 also performs machine learning using a learning algorithm specified in the learning request (a learning program that learns using the learning algorithm specified in the learning request) to create a learning model. For example, if the learning algorithm specified in the learning request is deep learning using a neural network, the learning model is created using a learning program that learns using this learning algorithm. In this way, the learning model creation unit 54 creates a learning model using the medical data read by the medical data reading unit 52 and the learning algorithm specified in the learning request. In other words, the learning model creation unit 54 executes the learning model creation step of the present invention.
- the learning model transmission unit 56 transmits the learning model created by the learning model creation unit 54 to the learning request source environment 14 (in this embodiment, the request source terminal 30) (see Figure 1).
- the process flow (learning model creation method) performed by the processor 40 will be described below with reference to FIG. 4.
- the processor 40 accepts it (learning request acceptance step) and starts the process of creating a learning model.
- the processor 40 reads medical data that falls within the range specified in the learning request from the medical data management server 22 (data reading step).
- the processor 40 then creates a learning model using the read medical data and the learning algorithm specified in the learning request (learning model creation step), and transmits the created learning model to the request source terminal 30.
- the learning requesting environment 14 creates a desired learning model without providing medical data to the learning requesting environment 14 (requesting terminal 30), so the learning requesting environment 14 can understand how the medical data is used. Furthermore, according to the present invention, it can also contribute to appropriate management of medical data. That is, when medical data is provided to the learning requesting environment 14, the medical data may be used in the learning requesting environment 14 in a manner not intended by the provider, but according to the present invention, such problems can be prevented. Furthermore, when medical data is provided to the learning requesting environment 14, there is a problem that even information that is not related to the creation of the learning model is disclosed to the learning requesting environment 14. That is, for example, even if only information on age and medical history is required to create a learning model, information such as name, height, and weight are also disclosed to the learning requesting environment 14. According to the present invention, such disclosure of unnecessary information can also be prevented.
- the processor 40 functions as a learning process maintenance unit 60 in addition to the above-mentioned units.
- the same members as those in the first embodiment are denoted by the same reference numerals and the explanation thereof will be omitted.
- the learning process preservation unit 60 preserves the creation process of this learning model (learning process). Specifically, it preserves information (hereinafter, process information) necessary for understanding and confirming the learning process (i.e., the process of creating the learning model).
- process information includes at least identification information for identifying the medical data used to create the learning model.
- the learning process maintenance unit 60 maintains the process information by storing it on the blockchain 62.
- the blockchain 62 is constructed by a blockchain network consisting of the medical data provider environment 12 (medical data management server 22), the learning requester environment 14 (requester terminal 30), the learning environment 16 (processor 40), etc., and is a chronological linking of blocks 64 containing process information.
- the learning process maintenance unit 60 creates a new block 64 each time a learning model is created and links this to the blockchain 62. In this way, by storing the process information on the blockchain 62, tampering with the learning process can be prevented.
- the processor 40 functions as a fee calculation unit 70 in addition to the above-mentioned units.
- the fee calculation unit 70 calculates the fee for using the medical data 20 used to create the learning model.
- the fee for use is calculated individually for each of the medical data 20(1) to 20(n).
- the fee for use is calculated as an amount (value) according to the number of times the medical data 20(1) to 20(n) is used (the number of times the medical data is used repeatedly), such as 10 yen per use.
- medical data 20(1) when creating a learning model for a male over 80 years old, medical data 20(1) (see FIG. 2) is used once. Therefore, a fee for one use of medical data 20(1) is calculated as a fee for use. Furthermore, for example, when creating a learning model for a person over 80 years old regardless of gender, medical data 20(1) and medical data 20(3) (see FIG. 2) are both used once. Therefore, a fee for one use of medical data 20(1) and medical data 20(3) is calculated as a fee for use. Furthermore, when creating the two learning models described above, that is, when creating a learning model for a male over 80 years old and a learning model for a person over 80 years old regardless of gender, medical data 20(1) is used twice and medical data 20(3) is used once. Therefore, the fee for use is calculated as two uses of medical data 20(1) and one use of medical data 20(3).
- an appropriate compensation can be calculated for each of the medical data 20(1) to 20(n) according to the number of times it is used. Then, it becomes possible to pay such an appropriate compensation to the provider (medical data provider environment 12) or to bill the user (learning requester environment 14).
- a learning model for adult men and a learning model for adult women can be integrated to create a learning model for adults regardless of gender, or a learning model for Kanto (created from medical data of patients living in Kanto) and a learning model for Kansai can be integrated to create a learning model for the whole of Japan.
- a learning model for Kanto created from medical data of patients living in Kanto
- a learning model for Kansai can be integrated to create a learning model for the whole of Japan.
- each of the multiple learning models can be a local learning model, and these can be integrated to create a single global learning model.
- the learning environment 16 When creating a global learning model by integrating local learning models as described above, the learning environment 16, specifically the learning model creation unit 54 described above, may be configured to create the global learning model.
- the learning requester environment 14 e.g., the requester terminal 30
- the global learning model may also be configured to create the global learning model.
- the processor 40 is described as having multiple learning programs for learning using various learning algorithms, but the present invention is not limited to this.
- the learning requesting environment 14 e.g., the requesting terminal 30
- the requesting terminal 30 may be configured to transmit the learning program to be used in learning to the processor 40 together with the learning request.
- the medical data management server 22 was provided separately from the processor 40, but the present invention is not limited to this.
- the medical data management server 22 may be provided integrally with the processor 40. In this case, for example, it is conceivable that the medical data 20 may be stored in the memory 42.
- the hardware structure of the processing units that execute various processes are various processors as shown below.
- the various processors include a CPU (Central Processing Unit), which is a general-purpose processor that executes software (programs) and functions as various processing units, a Programmable Logic Device (PLD), which is a processor whose circuit configuration can be changed after manufacture, such as an FPGA (Field Programmable Gate Array), and a dedicated electrical circuit, which is a processor with a circuit configuration designed specifically to execute various processes.
- a CPU Central Processing Unit
- PLD Programmable Logic Device
- FPGA Field Programmable Gate Array
- dedicated electrical circuit which is a processor with a circuit configuration designed specifically to execute various processes.
- a single processing unit may be configured with one of these various processors, or may be configured with a combination of two or more processors of the same or different types (for example, multiple FPGAs, or a combination of a CPU and an FPGA). Multiple processing units may also be configured with one processor.
- multiple processing units may also be configured with one processor.
- first there is a form in which one processor is configured with a combination of one or more CPUs and software, as represented by computers such as clients and servers, and this processor functions as multiple processing units.
- a processor is used that realizes the functions of the entire system, including multiple processing units, with a single IC (Integrated Circuit) chip, as represented by System On Chip (SoC).
- SoC System On Chip
- the hardware structure of these various processors is an electrical circuit that combines circuit elements such as semiconductor elements.
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Medical Treatment And Welfare Office Work (AREA)
Abstract
Description
図1において、学習システム10は、医療データ提供元環境12、学習依頼元環境14(学習依頼元)、学習環境16、とから構成される。
図5に示すように、第2実施形態では、プロセッサ40が、前述した各部に加えて、学習プロセス保全部60としても機能する。なお、図5以降の図面を用いた説明では、上述した第1実施形態と同様の部材については同様の符号を付して説明を省略する。
図6に示すように、第3実施形態では、プロセッサ40が、前述した各部に加えて、対価算出部70としても機能する。対価算出部70は、学習モデルが作成されると、この学習モデルの作成に用いた医療データ20の利用の対価を算出する。利用の対価は、医療データ20(1)~20(n)の各々について個別に算出される、また、利用の対価は、例えば、1回の利用につき10円といったように、医療データ20(1)~20(n)の利用回数(利用の重複した回数)に応じた額(値)が算出される。
12 医療データ提供元環境
14 学習依頼元環境
16 学習環境
20 医療データ(データ)
22 医療データ管理サーバ(データ管理サーバ)
30 依頼元端末
40 プロセッサ(学習モデル作成装置)
42 メモリ
44 中央制御部
50 学習依頼受付部
52 医療データ読み出し部(データ読み出し部)
54 学習モデル作成部
56 学習モデル送信部
60 学習プロセス保全部
62 ブロックチェーン
64 ブロック
70 対価算出部
Claims (10)
- データ管理サーバに格納されたデータを用いた機械学習により学習モデルを作成する、学習モデル作成装置、において、
学習依頼元から送信される学習依頼であり、学習モデルの作成に用いる学習アルゴリズムと、学習モデルの作成に用いるデータの範囲と、が指定された学習依頼を受け付ける、学習依頼受付部と、
前記学習依頼で指定された範囲のデータを、前記データ管理サーバから読み出す、データ読出部と、
前記データ管理サーバから読み出したデータと、前記学習依頼において指定された学習アルゴリズムと、を用いて学習モデルを作成する、学習モデル作成部と、を備える、
学習モデル作成装置。 - 前記データが医療データである、
請求項1に記載の学習モデル作成装置。 - 作成した学習モデルを、学習依頼元に送信する、学習モデル送信部、を備える、
請求項2に記載の学習モデル作成装置。 - 学習モデルの作成に用いたデータを識別するための識別情報を含む学習プロセスを、ブロックチェーン上に格納する、学習プロセス保全部、を備える、
請求項3に記載の学習モデル作成装置。 - 学習モデルの作成に用いたデータの利用の対価を算出する、対価算出部、を備える、
請求項4に記載の学習モデル作成装置。 - 前記学習モデル作成部は、
一部が重複した複数グループのデータを用いた複数回の機械学習により、複数の学習モデルを作成する、
請求項5に記載の学習モデル作成装置。 - 前記対価算出部は、
複数回の機械学習に重複して用いられたデータについては、利用の重複した回数に応じて利用の対価を算出する、
請求項6に記載の学習モデル作成装置。 - 前記学習モデル作成部は、
前記複数の学習モデルの各々をローカルな学習モデルとし、これらローカルな学習モデルを統合することにより、1つのグローバルな学習モデルを作成する、
請求項6または7に記載の学習モデル作成装置。 - 前記学習依頼元において、
前記複数の学習モデルの各々をローカルな学習モデルとし、これらローカルな学習モデルを統合することにより、1つのグローバルな学習モデルを作成する、
請求項6または7に記載の学習モデル作成装置。 - データ管理サーバに格納されたデータを用いた機械学習により学習モデルを作成する、学習モデル作成方法、において、
学習依頼元から送信される学習依頼であり、学習モデルの作成に用いる学習アルゴリズムと、学習モデルの作成に用いるデータの範囲と、が指定された学習依頼を受け付ける、学習依頼受付ステップと、
前記学習依頼で指定された範囲のデータを、前記データ管理サーバから読み出す、データ読出ステップと、
前記データ管理サーバから読み出したデータと、前記学習依頼において指定された学習アルゴリズムと、を用いて学習モデルを作成する、学習モデル作成ステップと、を備える、
学習モデル作成方法。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2025509816A JPWO2024202488A1 (ja) | 2023-03-27 | 2024-01-29 |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2023049847 | 2023-03-27 | ||
| JP2023-049847 | 2023-03-27 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2024202488A1 true WO2024202488A1 (ja) | 2024-10-03 |
Family
ID=92904066
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/JP2024/002574 Pending WO2024202488A1 (ja) | 2023-03-27 | 2024-01-29 | 学習モデル作成装置、学習モデル作成方法 |
Country Status (2)
| Country | Link |
|---|---|
| JP (1) | JPWO2024202488A1 (ja) |
| WO (1) | WO2024202488A1 (ja) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007279887A (ja) * | 2006-04-04 | 2007-10-25 | Nippon Telegr & Teleph Corp <Ntt> | 特異パターン検出システム、モデル学習装置、特異パターン検出装置、特異パターン検出方法、及び、コンピュータプログラム |
| JP2019526851A (ja) * | 2016-07-18 | 2019-09-19 | ナント ホールディングス アイピー エルエルシーNant Holdings IP, LLC | 分散型機械学習システム、装置、および方法 |
| WO2021075091A1 (ja) * | 2019-10-15 | 2021-04-22 | 日本電気株式会社 | 対価算出装置、制御方法、及びプログラム |
-
2024
- 2024-01-29 JP JP2025509816A patent/JPWO2024202488A1/ja active Pending
- 2024-01-29 WO PCT/JP2024/002574 patent/WO2024202488A1/ja active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2007279887A (ja) * | 2006-04-04 | 2007-10-25 | Nippon Telegr & Teleph Corp <Ntt> | 特異パターン検出システム、モデル学習装置、特異パターン検出装置、特異パターン検出方法、及び、コンピュータプログラム |
| JP2019526851A (ja) * | 2016-07-18 | 2019-09-19 | ナント ホールディングス アイピー エルエルシーNant Holdings IP, LLC | 分散型機械学習システム、装置、および方法 |
| WO2021075091A1 (ja) * | 2019-10-15 | 2021-04-22 | 日本電気株式会社 | 対価算出装置、制御方法、及びプログラム |
Also Published As
| Publication number | Publication date |
|---|---|
| JPWO2024202488A1 (ja) | 2024-10-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10957451B2 (en) | Patient healthcare interaction device and methods for implementing the same | |
| CN109523067B (zh) | 基于预测模型的费用预测方法、装置、服务器及存储介质 | |
| Cadarette et al. | Relative effectiveness of osteoporosis drugs for preventing nonvertebral fracture | |
| US9443002B1 (en) | Dynamic data analysis and selection for determining outcomes associated with domain specific probabilistic data sets | |
| CN109637615B (zh) | 异常医疗处方的判断方法、装置、设备及可读存储介质 | |
| Cho et al. | Effect of improved medication adherence on health care costs in osteoporosis patients | |
| Halldorsson et al. | A clinical decision support system for the diagnosis, fracture risks and treatment of osteoporosis | |
| US20150339778A1 (en) | System and method for assessing healthcare risks | |
| Lizzio et al. | Electronic patient-reported outcome collection systems in orthopaedic clinical practice | |
| Yeung et al. | Interventions for treating ankle fractures in children | |
| Guerard et al. | The influence of respondent characteristics on the validity of self‐reported survey responses | |
| Heringa et al. | Better specification of triggers to reduce the number of drug interaction alerts in primary care | |
| Shever et al. | The effect of high nursing surveillance on hospital cost | |
| Rekawek et al. | Artificial intelligence: the future of maxillofacial prognosis and diagnosis? | |
| Jiang et al. | Expected lifetime numbers, risks, and burden of osteoporotic fractures for 50-year old Chinese women: a discrete event simulation incorporating FRAX | |
| KR20170022007A (ko) | 건강정보관리 시스템 및 그 방법이 구현된 컴퓨터로 판독 가능한 기록매체 | |
| CN112907045A (zh) | 一种基于5g物联网的智慧医疗数据采集方法 | |
| WO2024202488A1 (ja) | 学習モデル作成装置、学習モデル作成方法 | |
| CN114519380A (zh) | 一种基于隐私保护及图像识别的给药方法与装置 | |
| CN117809841B (zh) | 一种基于大模型技术的皮肤专病患者管理方法及系统 | |
| Gotmaker et al. | Bayesian adaptive design: the future for regional anesthesia trials? | |
| JP5731616B2 (ja) | レセプトチェック装置、レセプトチェック方法およびレセプトチェックプログラム | |
| Chalmers et al. | Human experts’ and a fuzzy model's predictions of outcomes of scoliosis treatment: A comparative analysis | |
| CN117633340A (zh) | 一种基于大数据的专病和症状的常用药品推荐方法和系统 | |
| CN114694815A (zh) | 基于人工智能的诊疗评估方法及相关设备 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 24778634 Country of ref document: EP Kind code of ref document: A1 |
|
| ENP | Entry into the national phase |
Ref document number: 2025509816 Country of ref document: JP Kind code of ref document: A |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 2025509816 Country of ref document: JP |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 2024778634 Country of ref document: EP |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| ENP | Entry into the national phase |
Ref document number: 2024778634 Country of ref document: EP Effective date: 20251027 |
|
| ENP | Entry into the national phase |
Ref document number: 2024778634 Country of ref document: EP Effective date: 20251027 |
|
| ENP | Entry into the national phase |
Ref document number: 2024778634 Country of ref document: EP Effective date: 20251027 |
|
| ENP | Entry into the national phase |
Ref document number: 2024778634 Country of ref document: EP Effective date: 20251027 |
|
| ENP | Entry into the national phase |
Ref document number: 2024778634 Country of ref document: EP Effective date: 20251027 |