A Novel Switching State-Space Model for Post-ICU Mortality Prediction and Survival Analysis

Deep Survival Analysis With Latent Clustering and Contrastive Learning

Survival analysis is employed to analyze the time before the event of interest occurs, which is broadly applied in many fields. The existence of censored data with incomplete supervision information about survival outcomes is one key challenge in survival analysis tasks. Although some progress has been made on this issue recently, the present methods generally treat the instances as separate ones while ignoring their potential correlations, thus rendering unsatisfactory performance. In this study, we propose a novel Deep Survival Analysis model with latent Clustering and Contrastive learning (DSACC). Specifically, we jointly optimize representation learning, latent clustering and survival prediction in a unified framework. In this way, the clusters distribution structure in latent representation space is revealed, and meanwhile the structure of the clusters is well incorporated to improve the ability of survival prediction. Besides, by virtue of the learned clusters, we further propose a contrastive loss function, where the uncensored data in each cluster are set as anchors, and the censored data are treated as positive/negative sample pairs according to whether they belong to the same cluster or not. This design enables the censored data to make full use of the supervision information of the uncensored samples. Through extensive experiments on four popular clinical datasets, we demonstrate that our proposed DSACC achieves advanced performance in terms of both C-index (0.6722, 0.6793, 0.6350, and 0.7943) and Integrated Brier Score (IBS) (0.1616, 0.1826, 0.2028, and 0.1120).

LGTRL-DE: Local and Global Temporal Representation Learning with Demographic Embedding for in-hospital mortality prediction

Predicting the patient's in-hospital mortality from the historical Electronic Medical Records (EMRs) can assist physicians to make clinical decisions and assign medical resources. In recent years, researchers proposed many deep learning methods to predict in-hospital mortality by learning patient representations. However, most of these methods fail to comprehensively learn the temporal representations and do not sufficiently mine the contextual knowledge of demographic information. We propose a novel end-to-end approach based on Local and Global Temporal Representation Learning with Demographic Embedding (LGTRL-DE) to address the current issues for in-hospital mortality prediction. LGTRL-DE is enabled by (1) a local temporal representation learning module that captures the temporal information and analyzes the health status from a local perspective through a recurrent neural network with the demographic initialization and the local attention mechanism; (2) a Transformer-based global temporal representation learning module that extracts the interaction dependencies among clinical events; (3) a multi-view representation fusion module that fuses temporal and static information and generates the final patient's health representations. We evaluate our proposed LGTRL-DE on two public real-world clinical datasets (MIMIC-III and e-ICU). Experimental results show that LGTRL-DE achieves an area under receiver operating characteristic curve of 0.8685 and 0.8733 on the MIMIC-III and e-ICU datasets, respectively, outperforming state-of-the-art approaches.

Forecasting medical state transition using machine learning methods

Early circulatory failure detection is an effective way to reduce medical fatigue and improve state pre-warning ability. Instead of using 0-1 original state, a transformed state is proposed in this research, which reflects how the state is transformed. The performance of the proposed method is compared with the original method under three models, including logistic regression, AdaBoost and XGBoost. The results show that the model XGBoost generally has the best performance measured by AUC, F1 and Sensitivity with values around 0.93, 0.91 and 0.90, at the prediction gaps 5, 10 and 20 separately. Under the model XGBoost, the method with transformed response variable has significantly better performance than that with the original response variable, with the performance metrics being around 1% to 4% higher, and the t values are all significant under the level 0.01. In order to explore the model performance under different baseline information, a subgroup analysis is conducted under sex, age, weight and height. The results demonstrate that sex and age have more significant influence on the model performance especially at the higher gaps than weight and height.

Early ICU Mortality Prediction for Respiratory Failure by A Regression-Based Hazard Markov Model

Respiratory failure is one of the major causes of death in critical care units. While respiratory failure could come with the acute symptoms progressively, an early warning model is urgently required to assess mortality risks in advance. To this end, an early mortality risk prediction in patients who suffer respiratory failures can provide support for timely decision making of clinical treatment and medical resource management. In the study, we propose a dynamic modeling approach for early mortality risk prediction of at-risk patients with respiratory failure based on the first 24 hours of ICU physiological data. Our proposed model is validated on the eICU Collaborate Research Database. We achieve high AUROC performance of around 80% and significantly improved AUCPR by 4% from Day 4 to Day 6 since ICU admission, compared to the state-of-art prediction models. Furthermore, we show the survival probability curve that contains the time-varying information for early ICU admission patients.