Recent advances in deep learning (DL) have prompted the development of high-performing early warning score (EWS) systems, predicting clinical deteriorations such as acute kidney injury, acute myocardial infarction, or circulatory failure. DL models have proven to be powerful tools for various tasks but come with the cost of lacking interpretability and limited generalizability, hindering their clinical applications. To develop a practical EWS system applicable to various outcomes, we propose causally-informed explainable early prediction model, which leverages causal discovery to identify the underlying causal relationships of prediction and thus owns two unique advantages: demonstrating the explicit interpretation of the prediction while exhibiting decent performance when applied to unfamiliar environments. Benefiting from these features, our approach achieves superior accuracy for 6 different critical deteriorations and achieves better generalizability across different patient groups, compared to various baseline algorithms. Besides, we provide explicit causal pathways to serve as references for assistant clinical diagnosis and potential interventions. The proposed approach enhances the practical application of deep learning in various medical scenarios.
Models that can predict the occurrence of events ahead of time with low false-alarm rates are critical to the acceptance of decision support systems in the medical community. This challenging task is typically treated as a simple binary classification, ignoring temporal dependencies between samples, whereas we propose to exploit this structure. We first introduce a common theoretical framework unifying dynamic survival analysis and early event prediction. Following an analysis of objectives from both fields, we propose Temporal Label Smoothing (TLS), a simpler, yet best-performing method that preserves prediction monotonicity over time. By focusing the objective on areas with a stronger predictive signal, TLS improves performance over all baselines on two large-scale benchmark tasks. Gains are particularly notable along clinically relevant measures, such as event recall at low false-alarm rates. TLS reduces the number of missed events by up to a factor of two over previously used approaches in early event prediction.
The recent success of machine learning methods applied to time series collected from Intensive Care Units (ICU) exposes the lack of standardized machine learning benchmarks for developing and comparing such methods. While raw datasets, such as MIMIC-IV or eICU, can be freely accessed on Physionet, the choice of tasks and pre-processing is often chosen ad-hoc for each publication, limiting comparability across publications. In this work, we aim to improve this situation by providing a benchmark covering a large spectrum of ICU-related tasks. Using the HiRID dataset, we define multiple clinically relevant tasks in collaboration with clinicians. In addition, we provide a reproducible end-to-end pipeline to construct both data and labels. Finally, we provide an in-depth analysis of current state-of-the-art sequence modeling methods, highlighting some limitations of deep learning approaches for this type of data. With this benchmark, we hope to give the research community the possibility of a fair comparison of their work.
Fontan patients often develop circulatory failure and are in desperate need of a therapeutic solution. A blood pump surgically placed in the cavopulmonary pathway can substitute the function of the absent sub-pulmonary ventricle by generating a mild pressure boost. However, there is currently no commercially available device designed for the cavopulmonary application; and the risks and benefits of implanting a ventricular assist device (VAD) originally designed for the left ventricular application on the right circulation of failing Fontan patients is not yet clear. Moreover, further research is needed to compare the hemodynamics between the two clinically-considered surgical configurations (Full Support and IVC Support) for cavopulmonary assist, with Full and IVC Support corresponding to the entire venous return or only the inferior venous return, respectively, being routed through the VAD. In this study, we used a numerical model of the failing Fontan physiology to evaluate the Fontan hemodynamic response to a left VAD during the IVC and Full supports. We observed that during the Full support the VAD improved the cardiac output while maintaining blood pressures within safe ranges, and lowered the IVC pressure to <15mmHg; however, we found a potential risk of lung damage at higher pump speeds due to the excessive pulmonary pressure elevation. IVC support the other hand, did not benefit the hemodynamics of the example failing Fontan patients, resulting in the superior vena cava pressure increasing to an unsafe level of >20 mmHg. The findings in this study may be helpful to surgeons for recognizing the risks of a cavopulmonary VAD and developing coherent clinical strategies for the implementation of cavopulmonary support.
Intensive care clinicians are presented with large quantities of patient information and measurements from a multitude of monitoring systems. The limited ability of humans to process such complex information hinders physicians to readily recognize and act on early signs of patient deterioration. We used machine learning to develop an early warning system for circulatory failure based on a high-resolution ICU database with 240 patient years of data. This automatic system predicts 90.0% of circulatory failure events (prevalence 3.1%), with 81.8% identified more than two hours in advance, resulting in an area under the receiver operating characteristic curve of 94.0% and area under the precision-recall curve of 63.0%. The model was externally validated in a large independent patient cohort.