Advancing oncology with federated learning: transcending boundaries in breast, lung, and prostate cancer. A systematic review

📅 2024-08-08
🏛️ medRxiv
📈 Citations: 2
Influential: 0
📄 PDF
🤖 AI Summary
Multi-center tumor data sharing is severely constrained by privacy regulations, limiting the generalizability of AI models in oncology. Method: This study systematically reviews and— for the first time—comprehensively evaluates federated learning (FL) in real-world clinical settings for breast, lung, and prostate cancers. It employs a distributed training framework that integrates heterogeneous imaging and histopathological data while incorporating privacy-enhancing technologies, including differential privacy. Contribution/Results: Across 25 empirical studies, FL outperformed centralized learning in 15 cases, significantly improving cross-site model generalizability and robustness. It enables effective multimodal data fusion without raw data exchange, advancing privacy-preserving, multi-center precision oncology. The findings establish FL as a trustworthy technical pathway for clinically deployable AI decision support systems.

Technology Category

Application Category

📝 Abstract
Federated Learning (FL) has emerged as a promising solution to address the limitations of centralised machine learning (ML) in oncology, particularly in overcoming privacy concerns and harnessing the power of diverse, multi-center data. This systematic review synthesises current knowledge on the state-of-the-art FL in oncology, focusing on breast, lung, and prostate cancer. Distinct from previous surveys, our comprehensive review critically evaluates the real-world implementation and impact of FL on cancer care, demonstrating its effectiveness in enhancing ML generalisability, performance and data privacy in clinical settings and data. We evaluated state-of-the-art advances in FL, demonstrating its growing adoption amid tightening data privacy regulations. FL outperformed centralised ML in 15 out of the 25 studies reviewed, spanning diverse ML models and clinical applications, and facilitating integration of multi-modal information for precision medicine. Despite the current challenges identified in reproducibility, standardisation and methodology across studies, the demonstrable benefits of FL in harnessing real-world data and addressing clinical needs highlight its significant potential for advancing cancer research. We propose that future research should focus on addressing these limitations and investigating further advanced FL methods, to fully harness data diversity and realise the transformative power of cutting-edge FL in cancer care.
Problem

Research questions and friction points this paper is trying to address.

Overcoming privacy concerns in oncology using federated learning
Enhancing ML generalisability and performance in cancer care
Addressing reproducibility and standardisation challenges in FL studies
Innovation

Methods, ideas, or system contributions that make the work stand out.

Federated Learning enhances ML generalisability and privacy
FL integrates multi-modal data for precision medicine
FL outperforms centralized ML in diverse clinical applications
A
A. Ankolekar
Department of Precision Medicine, GROW Research Institute for Oncology and Reproduction, Maastricht University, Maastricht, the Netherlands
S
S. Boie
Pfizer Pharma GmbH, Berlin, Germany
M
M. Abdollahyan
Centre for Cancer Biomarkers and Biotherapeutics, Barts Cancer Institute, Queen Mary University of London, United Kingdom
E
E. Gadaleta
Centre for Cancer Biomarkers and Biotherapeutics, Barts Cancer Institute, Queen Mary University of London, United Kingdom
S
S. A. Hasheminasab
Centre for Statistics in Medicine, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, United Kingdom
G
G. Yang
Bioengineering Department and Imperial -X, Imperial College London, London, UK
C
C. Beauville
Flower Labs, Hamburg, Germany
N
N. Dikaios
Mathematics Research Center, Academy of Athens, Athens, Greece
G
G. Kastis
Mathematics Research Center, Academy of Athens, Athens, Greece
M
M. Bussmann
Helmholtz -Zentrum Dresden -Rossendorf, Dresden, Germany
S
S. Khalid
Centre for Statistics in Medicine, Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, United Kingdom
H
H. Kruger
Pfizer Pharma GmbH, Berlin, Germany
P
P. Lambin
Department of Precision Medicine, GROW Research Institute for Oncology and Reproduction, Maastricht University, Maastricht, the Netherlands
G
G. Papanastasiou
Pfizer Inc, New York, New York, USA