NECA: Neighbor Event Curve Augmentation for Multi-Horizon ICU Deterioration Prediction

Chenyang Wang; Jiawei Theresa Tang; Xinran Liu; Xinyu Lu; Xulei Yang

cs.LGOct 29, 2026

NECA: Neighbor Event Curve Augmentation for Multi-Horizon ICU Deterioration Prediction

Authors: Chenyang Wang, Jiawei Theresa Tang, Xinran Liu, Xinyu Lu, Xulei Yang

Organizations: Department of Statistics and Data Science, National University of Singapore, Singapore · School of Computing, National University of Singapore, Singapore · Institute for Infocomm Research (I2R), A*STAR, Singapore

Abstract

Early prediction of ICU organ dysfunction events supports timely clinical monitoring and intervention, but estimating short-term organ-specific risk from early ICU records remains challenging due to irregular measurements, heterogeneous clinical variables, and the need to predict risk across multiple future horizons. Existing supervised models usually treat each patient as an isolated sample and do not explicitly use the observed event timing of similar historical patients. We propose Neighbor Event-Curve Augmentation (NECA), a model-agnostic method that retrieves training-set patients with similar early-window profiles and converts their subsequent organ dysfunction event onset times into cumulative event-curve features over 6h, 12h, 24h, and 42h. These features are concatenated with the original patient representation and used by standard downstream predictors. On PhysioNet 2012 and a 4,000-stay MIMIC-IV subset, NECA improves average PR-AUC and ROC-AUC across four organ dysfunction event tasks and seven supervised learners. The largest gain appears for neurological dysfunction events, where task-level average PR-AUC increases from 0.461 to 0.557 on PhysioNet 2012. Ablations show that clinically meaningful retrieval and horizon-specific neighbor risk drive most of the improvement, while the full cumulative curve adds task-dependent temporal context. These results support similar-patient event timing as a structured temporal risk signal for ICU organ dysfunction event prediction. The source code will be publicly released upon acceptance of this paper.

Figures & tables

Figure 1: Overview of the proposed NECA framework for early ICU organ dysfunction event prediction. The framework consists of data preprocessing, neighbor event-curve augmentation, model integration, and multi-horizon output prediction.

Endpoint	Recorded event rule	Early-window variables
Renal	Maximum Cr $>1.5$ mg/dL or BUN $>30$ mg/dL	Age, gender, minimum MAP, maximum Cr and BUN, last urine output, maximum lactate
Neurological	Last GCS $\leq 12$	Age, gender, minimum and last GCS, minimum MAP, maximum HR
Electrolyte / metabolic	Minimum pH $<7.35$ , minimum HCO 3 $<22$ , or maximum lactate $>2.0$	Age, gender, maximum lactate, minimum pH and HCO 3 , maximum Cr and BUN
Inflammatory	Maximum WBC $>12$ or temperature $>38^{\circ}$ C	Age, gender, maximum WBC, maximum temperature, minimum platelets, maximum lactate

Table 2: EHR-observed endpoint rules and selected first-six-hour predictor variables.

		PR-AUC		ROC-AUC
Dataset	Endpoint	Raw	Curve-NECA	Raw	Curve-NECA
PN12	Neurological	0.461	0.557	0.689	0.766
	Renal	0.353	0.387	0.711	0.740
	Electrolyte/metabolic	0.357	0.378	0.671	0.692
	Inflammatory	0.514	0.538	0.692	0.727
MIMIC	Neurological	0.419	0.520	0.749	0.809

Table 5: Mean test performance across four horizons and seven supervised learners. PN12 is PhysioNet 2012; MIMIC is the 4,000-stay MIMIC-IV subset.

Clinical Outcome Prediction

Clinical outcome prediction uses patient records, imaging, signals, and biomarkers to estimate risks, disease trajectories, treatment response, and survival.

See all 268 papers

Predicting Future Organ Dysfunction in ICU Patients Using Temporal Convolutional Networks on MIMIC-IV Data
Aug 29, 2026Razan Albouq, Asra AslamClinical Outcome Prediction Temporal Convolutional Networks
University of Sheffield, UK
INTERVenE: Temporal-Abstraction-Interval Based Transformers for Short-Horizon Medical Event Prediction
Aug 30, 2026Shahar Oded, Yuval ShaharEvent Time Prediction Interpretable ML
The Stein Faculty of Computer and Information Science, Ben-Gurion University of the Negev, Beer-Sheva, Israel
Multimodal Deep Learning for Early Prediction of Patient Deterioration in the ICU: Integrating Time-Series EHR Data with Clinical Notes
Mar 16, 2026Binesh SadanandanClinical Risk Prediction Longitudinal EHR Modeling

Aug 29, 2026J/K move · Enter open · S save

Predicting Future Organ Dysfunction in ICU Patients Using Temporal Convolutional Networks on MIMIC-IV Data

Razan Albouq, Asra Aslam

University of Sheffield, UK

Clinical Outcome Prediction Temporal Convolutional Networks

Open paper AskPDF arXiv

Predicting future organ dysfunction in Intensive Care Unit (ICU) patients is critical for early clinical intervention, yet existing machine learning approaches have largely treated the Sequential Organ Failure Assessment (SOFA) score as an input to binary mortality prediction rather than as a continuous clinical outcome in its own right. We investigate the extent to which a Temporal Convolutional Net work (TCN) can predict next-day SOFA scores from multivariate ICU time-series data extracted from MIMIC-IV, characterise the relative contribution of each organ system to total SOFA variance and deterioration, and identify distinct trajectory patterns across ICU stays. A residual TCN trained on three-day sliding windows achieved a five-fold cross-validation R2 of 0.740 +- 0.013 and MAE of 1.431 +- 0.022, outperforming a naive persistence baseline on RMSE and R2. SHAP interpretability analysis revealed that the model functions primarily as a severity-anchoring mechanism rather than a true sequence model, with predictions dominated almost entirely by the most recent observation day. Cardiovascular dysfunction emerged as the strongest discriminator of both cross-sectional severity and acute deterioration, and unsupervised trajectory clustering identified two clinically meaningful phenotypes, an improving group (58.9%) and a persistently severe group (41.1%), differentiated by cardiovascular, hepatic, coagulation, and renal involvement. We conclude that TCNs can extract meaningful predictive signal from ICU physiological data, but that short input windows and complete-case selection bias currently limit their clinical utility, motivating future work on longer input horizons, alternative missing-data strategies, and external validation.

Abstract

Clinical Outcome Prediction

Explore similar work

Predicting Future Organ Dysfunction in ICU Patients Using Temporal Convolutional Networks on MIMIC-IV Data

INTERVenE: Temporal-Abstraction-Interval Based Transformers for Short-Horizon Medical Event Prediction

Multimodal Deep Learning for Early Prediction of Patient Deterioration in the ICU: Integrating Time-Series EHR Data with Clinical Notes