Federated Learning for Thyroid Ultrasound Image Analysis to Protect Personal Information: Validation Study in a Real Health Care Environment

BACKGROUND: Federated learning is a decentralized approach to machine learning; it is a training strategy that overcomes medical data privacy regulations and generalizes deep learning algorithms. Federated learning mitigates many systemic privacy risks by sharing only the model and parameters for tr...

Descripción completa

Detalles Bibliográficos
Autores principales: Lee, Haeyun, Chai, Young Jun, Joo, Hyunjin, Lee, Kyungsu, Hwang, Jae Youn, Kim, Seok-Mo, Kim, Kwangsoon, Nam, Inn-Chul, Choi, June Young, Yu, Hyeong Won, Lee, Myung-Chul, Masuoka, Hiroo, Miyauchi, Akira, Lee, Kyu Eun, Kim, Sungwan, Kong, Hyoun-Joong
Formato: Online Artículo Texto
Lenguaje:English
Publicado: JMIR Publications 2021
Materias:
Original Paper
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8170555/
https://www.ncbi.nlm.nih.gov/pubmed/33858817
http://dx.doi.org/10.2196/25869

Ejemplares similares

Ejemplares similares