Zobrazeno 1 - 10
of 2 033
pro vyhledávání: '"Van Ooijen, P."'
Autor:
Mol, Frank N., van der Hoek, Luuk, Ma, Baoqiang, Nagam, Bharath Chowdhary, Sijtsema, Nanna M., van Dijk, Lisanne V., Bunte, Kerstin, Vlijm, Rifka, van Ooijen, Peter M. A.
The superior soft tissue differentiation provided by MRI may enable more accurate tumor segmentation compared to CT and PET, potentially enhancing adaptive radiotherapy treatment planning. The Head and Neck Tumor Segmentation for MR-Guided Applicatio
Externí odkaz:
http://arxiv.org/abs/2412.06610
In this paper, we delve into the susceptibility of federated medical image analysis systems to adversarial attacks. Our analysis uncovers a novel exploitation avenue: using gradient information from prior global model updates, adversaries can enhance
Externí odkaz:
http://arxiv.org/abs/2310.13893
Optimization-based regularization methods have been effective in addressing the challenges posed by data heterogeneity in medical federated learning, particularly in improving the performance of underrepresented clients. However, these methods often
Externí odkaz:
http://arxiv.org/abs/2310.09444
This paper explores the security aspects of federated learning applications in medical image analysis. Current robustness-oriented methods like adversarial training, secure aggregation, and homomorphic encryption often risk privacy compromises. The c
Externí odkaz:
http://arxiv.org/abs/2310.08681
Federated learning offers a privacy-preserving framework for medical image analysis but exposes the system to adversarial attacks. This paper aims to evaluate the vulnerabilities of federated learning networks in medical image analysis against such a
Externí odkaz:
http://arxiv.org/abs/2310.06227
Autor:
Nikos Sourlos, Rozemarijn Vliegenthart, Joao Santinha, Michail E. Klontzas, Renato Cuocolo, Merel Huisman, Peter van Ooijen
Publikováno v:
Insights into Imaging, Vol 15, Iss 1, Pp 1-12 (2024)
Abstract Various healthcare domains have witnessed successful preliminary implementation of artificial intelligence (AI) solutions, including radiology, though limited generalizability hinders their widespread adoption. Currently, most research group
Externí odkaz:
https://doaj.org/article/7bd1a19456934e9b9fa9aa090ad2aa7d
Autor:
Li, Jingxiong, Zheng, Sunyi, Shui, Zhongyi, Zhang, Shichuan, Yang, Linyi, Sun, Yuxuan, Zhang, Yunlong, Li, Honglin, Ye, Yuanxin, van Ooijen, Peter M. A., Li, Kang, Yang, Lin
Karyotyping is of importance for detecting chromosomal aberrations in human disease. However, chromosomes easily appear curved in microscopic images, which prevents cytogeneticists from analyzing chromosome types. To address this issue, we propose a
Externí odkaz:
http://arxiv.org/abs/2306.14129
Publikováno v:
Scandinavian Journal of Work, Environment & Health, Vol 50, Iss 6, Pp 437-446 (2024)
OBJECTIVES: This study examined the associations between implemented disability-related policies and practices (DPP) and sustained employment among partially disabled employees in The Netherlands. METHODS: Employer survey data on implemented DPP were
Externí odkaz:
https://doaj.org/article/321b632d631c4da7a909efe7ece14ada
Deep learning is effective in diagnosing COVID-19 and requires a large amount of data to be effectively trained. Due to data and privacy regulations, hospitals generally have no access to data from other hospitals. Federated learning (FL) has been us
Externí odkaz:
http://arxiv.org/abs/2303.16141