Deep model improves risk prediction with irregular patient data
DeepAJM: Deep Association Joint Model for Irregularly Sampled data
Artificial IntelligenceMachine Learning
Summary
Predicting how patients will do over time is hard when their health data is uneven or irregular. The authors created DeepAJM, a deep learning model that can work with these irregular health records without relying on strict assumptions. It learns patterns from patient data and links them to survival chances, doing better than traditional models in tests on real and simulated health datasets. This approach helps in predicting risk more accurately from complex, irregular time data.
What this means in practice
- •For hospital data teams: Improve patient risk prediction by integrating irregularly timed medical records and survival outcomes with a flexible deep learning joint model.
- •For healthcare software developers: Build clinical decision support tools that handle irregular patient data to estimate survival risks more accurately than classical models.$Commercial implications: Enables developing advanced predictive analytics software to be sold to healthcare providers for better prognosis estimation.
Authors
Barsha Halder, Jeffrey A. Thompson
Abstract
Joint Models simultaneously model longitudinal and survival outcomes, leveraging patterns in patients' longitudinal trajectory to improve the prediction of survival outcomes. The classical parametric joint models, however, rely on fixed parametric assumptions, making them susceptible to bias under model misspecification and smaller sample sizes. We propose a deep joint model, DeepAJM, that does not require any parametric assumptions, while retaining a partially interpretable, per-longitudinal-outcome association structure. The joint model uses an encoder-decoder (sequence-to-sequence) architecture to learn the latent structure in patients' time-varying covariate trajectories. The model links the longitudinal processes to the survival processes through a learned interpretable association structure, in which each longitudinal output from the decoder gets remodulated by baseline covariates before it contributes to the risk scores from the survival head of the architecture. The model was evaluated on three datasets ( a cardiovascular-disease EHR cohort, a primary biliary cirrhosis (PBC2) dataset, and a simulated dataset) against a classical parametric joint model, TransformerJM, DA-LSTM and a Cox-based survival-only model. All models were assessed using C-index, integrated brier score (IBS), time-dependent AUROC, and time-dependent AUPRC. Our model achieved the best discrimination in terms of the C-index, time-dependent AUROC, and AUPRC across all datasets.