Preparing For The Next Pandemic: Transfer Learning From Existing Diseases Via Hierarchical Multi-Modal BERT Models to Predict COVID-19 Outcomes
Developing prediction models for emerging infectious diseases from relatively small numbers of cases is a critical need for improving pandemic preparedness. Using COVID-19 as an exemplar, we propose a transfer learning methodology for developing predictive models from multi-modal electronic healthcare records by leveraging information from more prevalent diseases with shared clinical characteristics. Our novel hierarchical, multi-modal model (TransMED) integrates baseline risk factors from the natural language processing of clinical notes at admission using Spark NLP, time-series measurements of biomarkers obtained from laboratory tests, and discrete diagnostic, procedure and drug codes. We demonstrate the alignment of TransMED’s predictions with well-established clinical knowledge about COVID-19 through univariate and multivariate risk factor driven sub-cohort analysis. TransMED’s superior performance over state-of-the-art methods shows that leveraging patient data across modalities and transferring prior knowledge from similar disorders is critical for accurate prediction of patient outcomes, and this approach may serve as an important tool in the early response to future pandemics.
Senior Research Scientist at Pacific Northwest National Laboratory
Khushbu Agarwal is a senior computer scientist at PNNL with 15+ years of experience in data analytics. Her primary interests are in deriving human actionable knowledge from large corpus of data using a combination of symbolic and neural learning.
She has extensive experience working on scalable graph algorithms, time series data, knowledge bases and machine learning, applying them to solving real-world application needs.
Her recent work focuses on building intelligent medical systems that combine different modalities of healthcare data graphs into a single unified deep learning framework for prediction and reasoning about diseases.