Interpretable Graph Learning on Irregular Clinical Time Series

Real-world medical data often includes measurements from multiple signals that are collected at irregular and asynchronous time intervals. For example, different types of blood tests can be measured at different times and frequencies, resulting in fragmented and unevenly scattered temporal data. Effectively learning from such data requires models that can handle sets of temporally sparse and heterogeneous signals. In this paper, we propose Graph Mixing Additive Networks (GMAN), a novel and interpretable-by-design model for learning over irregular sets of temporal signals. Our method achieves state-of-the-art performance in a real-world medical tasks, predicting the onset of Crohn's Disease (CD) from routine biomarkers. We further demonstrate how its interpretable design allows to gain real medical insights such as dysregulation patters in the pre-diagnostic phase of complex diseases.