Leveraging Federated Learning for Automatic Detection of Clopidogrel Treatment Failures

TL;DR

This study tackles the variability of clopidogrel response by developing a privacy-preserving predictive framework using federated learning across UK Biobank data. It compares localized, centralized, and federated training, implementing FedAvg via NVIDIA NVFlare with two neural architectures (FCN and GRU) to detect treatment failure within one year of initial clopidogrel prescription. Key findings show that centralized training yields the best performance, but federated learning can substantially narrow the performance gap, with GRU achieving very high AUC (0.940) using data from a subset of centers. The work highlights the potential of privacy-preserving predictive modeling in healthcare and points to future improvements in aggregation strategies to further enhance federated learning performance across heterogeneous centers.

Abstract

The effectiveness of clopidogrel, a widely used antiplatelet medication, varies significantly among individuals, necessitating the development of precise predictive models to optimize patient care. In this study, we leverage federated learning strategies to address clopidogrel treatment failure detection. Our research harnesses the collaborative power of multiple healthcare institutions, allowing them to jointly train machine learning models while safeguarding sensitive patient data. Utilizing the UK Biobank dataset, which encompasses a vast and diverse population, we partitioned the data based on geographic centers and evaluated the performance of federated learning. Our results show that while centralized training achieves higher Area Under the Curve (AUC) values and faster convergence, federated learning approaches can substantially narrow this performance gap. Our findings underscore the potential of federated learning in addressing clopidogrel treatment failure detection, offering a promising avenue for enhancing patient care through personalized treatment strategies while respecting data privacy. This study contributes to the growing body of research on federated learning in healthcare and lays the groundwork for secure and privacy-preserving predictive models for various medical conditions.

Figures (6)

• Figure 1: Experimental scenarios to illustrate how data is handled; 1) localized training, 2) centralized training, and 3) federated learning.
• Figure 2: Number of patients with clopidogrel prescription in individual centers
• Figure 3: Number of patients that are selected as test set in idividual centers.
• Figure 4: AUCs of locally trained models with their own data.
• Figure 5: Scatter plot of the AUCs of locally trained models with respect to their cohort sizes.