Key result
A Convolutional Neural Network model predicted the risk of disseminated intravascular coagulation in ICU patients with an AUC of 0.986, significantly outperforming the XGBoost model.
Why the study?
Disseminated intravascular coagulation is a complex, life-threatening syndrome, and early prediction of DIC risk is urgently needed to reduce adverse outcomes.
Does a CNN-based time series model accurately predict the risk of DIC in adult ICU patients?
Cohort (n=6,631)
No
Does a CNN-based time series model accurately predict the risk of DIC in adult ICU patients?
Effect estimate: AUC 0.986
Absolute Event Rate: 0.986% vs 0.851%
p-value: p=<0.001
An interpretable CNN model using time-series EHR data can accurately predict the risk of DIC in ICU patients, potentially enabling earlier clinical intervention.
Supports CNN time-series models for DIC prediction in ICU cohorts; leaves open prospective validation and outcome impact.
Disseminated intravascular coagulation (DIC) is a complex, life-threatening syndrome associated with the end-stage of different coagulation disorders. Early prediction of the risk of DIC development is an urgent clinical need to reduce adverse outcomes. In this study, we developed a novel interpretable deep learning based time series for predicting the risk of DIC. The study cohort included ICU patients from a 4,300-bed academic hospital between January 1, 2019, and January 1, 2022. Experimental results show that our model achieves excellent performance (AUC: 0.986, accuracy: 95.7%, and F1:0.935). We used Gradient-weighted Class Activation Mapping (Grad-CAM) to explain how predictive models identify patients with DIC. The decision basis of the model was displayed in the form of a heat map. The model can be used for the early identification of high-risk patients with DIC, which will help in the early intervention of DIC patients and improve the treatment effect.
No takes yet. Share an insight, caveat, or question.
Yang et al. (2022) conducted a cohort in Disseminated intravascular coagulation (DIC) (n=6,631). Convolutional Neural Network (CNN) prediction model vs. XGBoost model was evaluated on Prediction of DIC (Area Under the Curve) (AUC 0.986, p=<0.001). A Convolutional Neural Network model predicted the risk of disseminated intravascular coagulation in ICU patients with an AUC of 0.986, significantly outperforming the XGBoost model.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: