DANIEL: A Distributed and Scalable Approach for Global Representation Learning with EHR Applications
This work addresses the problem of handling high-dimensional, heterogeneous data with privacy constraints in healthcare applications, offering an incremental improvement over existing methods for federated learning in EHR systems.
The authors tackled the challenge of scalable and privacy-preserving representation learning from large-scale binary data in multi-institutional electronic health record (EHR) settings by developing a distributed framework based on the Ising model, achieving superior performance in global representation learning and downstream clinical tasks on datasets from 58,248 patients.
Classical probabilistic graphical models face fundamental challenges in modern data environments, which are characterized by high dimensionality, source heterogeneity, and stringent data-sharing constraints. In this work, we revisit the Ising model, a well-established member of the Markov Random Field (MRF) family, and develop a distributed framework that enables scalable and privacy-preserving representation learning from large-scale binary data with inherent low-rank structure. Our approach optimizes a non-convex surrogate loss function via bi-factored gradient descent, offering substantial computational and communication advantages over conventional convex approaches. We evaluate our algorithm on multi-institutional electronic health record (EHR) datasets from 58,248 patients across the University of Pittsburgh Medical Center (UPMC) and Mass General Brigham (MGB), demonstrating superior performance in global representation learning and downstream clinical tasks, including relationship detection, patient phenotyping, and patient clustering. These results highlight a broader potential for statistical inference in federated, high-dimensional settings while addressing the practical challenges of data complexity and multi-institutional integration.