Abstract
BackgroundPredictive modeling with longitudinal electronic health record (EHR) data offers great promise for accelerating personalized medicine and better informs clinical decision-making. Recently, deep learning models have achieved state-of-the-art performance for many healthcare prediction tasks. However, deep models lack interpretability, which is integral to successful decision-making and can lead to better patient care. In this paper, we build upon the contextual decomposition (CD) method, an algorithm for producing importance scores from long short-term memory networks (LSTMs). We extend the method to bidirectional LSTMs (BiLSTMs) and use it in the context of predicting future clinical outcomes using patients’ EHR historical visits.MethodsWe use a real EHR dataset comprising 11071 patients, to evaluate and compare CD interpretations from LSTM and BiLSTM models. First, we train LSTM and BiLSTM models for the task of predicting which pre-school children with respiratory system-related complications will have asthma at school-age. After that, we conduct quantitative and qualitative analysis to evaluate the CD interpretations produced by the contextual decomposition of the trained models. In addition, we develop an interactive visualization to demonstrate the utility of CD scores in explaining predicted outcomes.ResultsOur experimental evaluation demonstrate that whenever a clear visit-level pattern exists, the models learn that pattern and the contextual decomposition can appropriately attribute the prediction to the correct pattern. In addition, the results confirm that the CD scores agree to a large extent with the importance scores generated using logistic regression coefficients. Our main insight was that rather than interpreting the attribution of individual visits to the predicted outcome, we could instead attribute a model’s prediction to a group of visits.ConclusionWe presented a quantitative and qualitative evidence that CD interpretations can explain patient-specific predictions using CD attributions of individual visits or a group of visits.
Highlights
Predictive modeling with longitudinal electronic health record (EHR) data offers great promise for accelerating personalized medicine and better informs clinical decision-making
We extend the work of Murdoch et al.[23] to Bidirectional long short-term memory network (BiLSTM), in the context of patient visit-level decomposition for analyzing patient-specific predictions made by standard BiLSTMs
Models training To validate the performance of the proposed interpretability approach, we train Long short-term memory network (LSTM) and BiLSTM models on the asthma dataset, which has two classes: c=1 for cases, and c=0 for controls
Summary
Predictive modeling with longitudinal electronic health record (EHR) data offers great promise for accelerating personalized medicine and better informs clinical decision-making. The exponential surge in the amount of digital data captured in electronic health record (EHR) offers promising opportunities for predicting the risk of potential diseases and better informs decision-making. The high-dimensionality and sparsity of medical features captured in the EHR makes it more complex for clinicians to interpret the relative impact of features and patterns which are potentially important in decisions. The majority of previous work has focused on assigning importance scores to individual features As a result, these techniques only provide limited local interpretations and do not model fine-grained interactions of groups of input features. In the case of LSTMs, a recent work by Murdoch et al [23] introduced contextual decomposition (CD), an algorithm for producing phrase-level importance scores from LSTMs without any modifications to the underlying model, and demonstrated it on the task of sentiment analysis
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
