Learning Hierarchical Representations of Electronic Health Records for Clinical Outcome Prediction
This work addresses the challenge of improving healthcare quality through more accurate clinical outcome prediction, which is incremental as it builds on existing methods by introducing hierarchical modeling for temporal patterns.
The paper tackled the problem of predicting clinical outcomes from Electronic Health Records by addressing the limitations of conventional deep sequential models in capturing temporal patterns in long and irregular event sequences. It proposed a model that learns hierarchical representations to distinguish between short-range and long-range events, achieving AUC scores of 0.94 for death prediction and 0.90 for ICU admission prediction.
Clinical outcome prediction based on the Electronic Health Record (EHR) plays a crucial role in improving the quality of healthcare. Conventional deep sequential models fail to capture the rich temporal patterns encoded in the longand irregular clinical event sequences. We make the observation that clinical events at a long time scale exhibit strongtemporal patterns, while events within a short time period tend to be disordered co-occurrence. We thus propose differentiated mechanisms to model clinical events at different time scales. Our model learns hierarchical representationsof event sequences, to adaptively distinguish between short-range and long-range events, and accurately capture coretemporal dependencies. Experimental results on real clinical data show that our model greatly improves over previous state-of-the-art models, achieving AUC scores of 0.94 and 0.90 for predicting death and ICU admission respectively, Our model also successfully identifies important events for different clinical outcome prediction tasks