专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

11. 发明公开

EP4270402A1 GENOGRAM CREATION AND DIAGNOSIS 审中-实审
公开(公告)号：EP4270402A1
公开(公告)日：2023-11-01
申请号：EP22382392.3
申请日：2022-04-25
申请人： Fujitsu Limited
发明人： GARCIA SANTA, Nuria , CETINA, Kendrick
IPC分类号： G16H10/20 , G16H10/60 , G16H15/00 , G16H50/70
摘要： A computer-implemented method comprising: using a trained family medical history, FMH, model trained to use Named Entity Recognition, NER, and Relation Extraction, RE, to extract entities and relations between the entities from unstructured data, extracting information from unstructured medical data for at least one family member of a target patient, the information including: a family member role indicating a relationship of the family member with the target patient; a status of the family member; and at least one observation of the family member; and transforming the extracted information into a genogram comprising a plurality of nodes respectively representing a plurality of people including the target patient and the at least one family member, wherein an edge between two nodes indicates a relationship between the people represented by the two nodes.

12. 发明公开

EP4009220A1 METHOD AND APPARATUS FOR DECENTRALIZED SUPERVISED LEARNING IN NLP APPLICATIONS 审中-公开
公开(公告)号：EP4009220A1
公开(公告)日：2022-06-08
申请号：EP20383052.6
申请日：2020-12-03
申请人： FUJITSU LIMITED
发明人： GARCIA SANTA, Nuria , CETINA, Kendrick
IPC分类号： G06F40/30 , G06N3/02 , G06N20/00 , G06F21/60
摘要： A method of training a neural network as a natural language processing, NLP, model, comprises: inputting annotated training data to first architecture portions of the neural network, the first architecture portions being executed respectively in a plurality of distributed client computing devices in communication with a server computing device, the training data being derived from text data private to the client computing device in which the first architecture portion is executed, the server computing device having no access to any of the private text data; deriving from the training data, using the first architecture portions, weight matrices of numeric weights which are decoupled from the private text data; concatenating the weight matrices, in a second architecture portion of the neural network executed in the server computing device, to obtain a single concatenated weight matrix; and training, on the second architecture portion, the NLP model using the concatenated weight matrix.

13. 发明公开

EP3382584A1 A SYSTEM AND A METHOD TO PREDICT PATIENT BEHAVIOUR 无效
公开(公告)号：EP3382584A1
公开(公告)日：2018-10-03
申请号：EP17164007.1
申请日：2017-03-30
申请人： FUJITSU LIMITED
发明人： VILLAZON-TERRAZAS, Boris , GARCIA SANTA, Nuria , DE LA TORRE, Victor
IPC分类号： G06F19/00
CPC分类号： G16H50/20 , G16H50/50 , G16H50/70
摘要： A system arranged to predict patient clinical behaviour, the processor being arranged to provide: a patient clinical object, PCO, engine to accept historical clinical behaviour of each patient in a population of patients and to create a PCO for each patient linking a patient identification with a clinical node for each historical piece of clinical information for that patient; a patient activity object, PAO, engine to input historical activity behaviour of each patient in the population of patients in the form of pieces of activity information identifying one or more visits to healthcare facilities; and to create a PAO for each patient linking the patient identification with an activity node for each historical piece of activity information for that patient; a clinical pattern miner, to use machine learning on the patient population PCOs to cluster the patients into two or more clinical patterns according to the PCO of each of the population of patients; a patient behaviour estimator, to: input the PAO and PCO for each of the population of patients clustered within each clinical pattern and to use machine learning to create a model of the clinical patterns; to input activity behaviour of a new patient in the form of pieces of activity information as to any visits of the new patient to healthcare facilities; to create a PAO for the new patient linking the new patient identification with an activity node for each piece of activity information for the new patient; to use machine learning to classify the new patient into one of the clinical patterns ; and to predict new patient clinical behaviour from the clinical nodes of the clinical pattern in which the new patient has been classified.

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式