mlsanityData to 7 Oct 2026

AttDMM: An Attentive Deep Markov Model for Risk Scoring in Intensive Care Units

arXiv 2102.04702 ↗First posted 9 Feb 2021cs.LG cs.AIcited by 14

Not on the map. Why?

Abstract (arXiv)

Clinical practice in intensive care units (ICUs) requires early warnings when a patient's condition is about to deteriorate so that preventive measures can be undertaken. To this end, prediction algorithms have been developed that estimate the risk of mortality in ICUs. In this work, we propose a novel generative deep probabilistic model for real-time risk scoring in ICUs. Specifically, we develop an attentive deep Markov model called AttDMM. To the best of our knowledge, AttDMM is the first ICU prediction model that jointly learns both long-term disease dynamics (via attention) and different disease states in health trajectory (via a latent variable model). Our evaluations were based on an established baseline dataset (MIMIC-III) with 53,423 ICU stays. The results confirm that compared to state-of-the-art baselines, our AttDMM was superior: AttDMM achieved an area under the receiver operating characteristic curve (AUROC) of 0.876, which yielded an improvement over the state-of-the-art method by 2.2%. In addition, the risk score from the AttDMM provided warnings several hours earlier. Thereby, our model shows a path towards identifying patients at risk so that health practitioners can intervene early and save patient lives.

Key facts. Cited by 14 papers in the map (data through 7 Oct 2026). Builds on 16 of its 60 references that are in the map. First posted on arXiv 9 Feb 2021.

Builds on

16 of 60 references.

References · 16

  1. 2016Recurrent Neural Networks for Multivariate Time Series with Missing Valuescited by 548
  2. 2016RETAIN: An Interpretable Predictive Model for Healthcare using Reverse Time Attention Mechanismcited by 344
  3. 2013Black Box Variational Inferencecited by 317
  4. 2017Multitask learning and benchmarking with clinical time series datacited by 287
  5. 2016Structured Inference Networks for Nonlinear State Space Modelscited by 182
  6. 2017Dipole: Diagnosis Prediction in Healthcare via Attention-based Bidirectional Recurrent Neural Networkscited by 114
  7. 2021Estimating Average Treatment Effects via Orthogonal Regularizationcited by 25
  8. 2016Personalized Risk Scoring for Critical Care Prognosis using Mixtures of Gaussian Processescited by 19
  9. 2016Medical Concept Representation Learning from Electronic Health Records and its Application on Heart Failure Predictioncited by 9
  10. 2017Benchmark of Deep Learning Models on Large Healthcare MIMIC Datasetscited by 8
  11. 2018Modeling Irregularly Sampled Clinical Time Seriescited by 5
  12. 2018Deep Ensemble Tensor Factorization for Longitudinal Patient Trajectories Classificationcited by 3
4 more
  1. 2020Patient Similarity Analysis with Longitudinal Health Datacited by 3
  2. 2017Individualized Risk Prognosis for Critical Care Patients: A Multi-task Gaussian Process Modelcited by 1
  3. 2019Hierarchical Hidden Markov Jump Processes for Cancer Screening Modelingcited by 1
  4. 2020Estimating Risk-Adjusted Hospital Performancecited by 1

Built on by

Cited by year

Citing papers per year: 2021: 3; 2022: 5; 2023: 2; 2024: 3; 2025: 1.

14 papers.

Sort:

  1. 2022Contrastive Learning for Unsupervised Domain Adaptation of Time Seriescited by 36
  2. 2022Combining Observational and Randomized Data for Estimating Heterogeneous Treatment Effectscited by 32
  3. 2023Bayesian Neural Controlled Differential Equations for Treatment Effect Estimationcited by 26
  4. 2021Sequential Deconfounding for Causal Inference with Unobserved Confounderscited by 24
  5. 2021Generalizing Off-Policy Learning under Sample Selection Biascited by 16
  6. 2021Deconfounding Temporal Autoencoder: Estimating Treatment Effects over Time Using Noisy Proxiescited by 16
  7. 2024IGC-Net for conditional average potential outcome estimation over timecited by 10
  8. 2024Stabilized Neural Prediction of Potential Outcomes in Continuous Timecited by 7
  9. 2022Learning Optimal Dynamic Treatment Regimes Using Causal Tree Methods in Medicinecited by 4
  10. 2022Modeling Long-term Dependencies and Short-term Correlations in Patient Journey Data with Temporal Attention Networks for Health Predictioncited by 3
  11. 2023Estimating Treatment Effects in Continuous Time with Hidden Confounderscited by 3
  12. 2025Deep Belief Markov Models for POMDP Inferencecited by 2
2 more
  1. 2022Integrated Convolutional and Recurrent Neural Networks for Health Risk Prediction using Patient Journey Data with Many Missing Valuescited by 1
  2. 2024Data-driven subgrouping of patient trajectories with chronic diseases: Evidence from low back paincited by 0