A multi-task Gaussian process self-attention neural network for real-time prediction of the need for mechanical ventilators in COVID-19 patients

Abstract

Objective: The Coronavirus Disease 2019 (COVID-19) pandemic has overwhelmed the capacity of healthcare resources and posed a challenge for worldwide hospitals. The ability to distinguish potentially deteriorating patients from the rest helps facilitate reasonable allocation of medical resources, such as ventilators, hospital beds, and human resources. The real-time accurate prediction of a patient's risk scores could also help physicians to provide earlier respiratory support for the patient and reduce the risk of mortality.

Methods: We propose a robust real-time prediction model for the in-hospital COVID-19 patients' probability of requiring mechanical ventilation (MV). The end-to-end neural network model incorporates the Multi-task Gaussian Process to handle the irregular sampling rate in observational data together with a self-attention neural network for the prediction task.

Results: We evaluate our model on a large database with 9,532 nationwide in-hospital patients with COVID-19. The model demonstrates significant robustness and consistency improvements compared to conventional machine learning models. The proposed prediction model also shows performance improvements in terms of area under the receiver operating characteristic curve (AUROC) and area under the precision-recall curve (AUPRC) compared to various deep learning models, especially at early times after a patient's hospital admission.

Conclusion: The availability of large and real-time clinical data calls for new methods to make the best use of them for real-time patient risk prediction. It is not ideal for simplifying the data for traditional methods or for making unrealistic assumptions that deviate from observation's true dynamics. We demonstrate a pilot effort to harmonize cross-sectional and longitudinal information for mechanical ventilation needing prediction.

Keywords: Deep neural network; Gaussian process; Mechanical ventilation prediction.

Graphical abstract

Fig. 1

MGP-MS model overview. The model combines a Multi-task Gaussian Process module with a self-attention neural network for trajectory prediction.

Fig. 2

Data completeness of lab tests and vital signs (100% means a feature's data is fully complete).

Fig. 3

The average risk score trajectories of the two classes of patients with the shaded area denote the +/− 1 standard deviation. The right panel shows the two risk score distributions at the 64th hour, and the Wilcoxon rank-sum test yields a p-value of $6.00\times {10}^{-38}$ when assuming the null hypothesis to be two distributions are the same.

Fig. 4

Two sample patients' risk score trajectory prediction using different models. (a) The risk score pathway of a randomly selected patient with COVID-19 who would need MV after 3 days since admission. (b) The risk score pathway of a randomly selected patient with COVID-19 who would not need MV after 3 days since admission.

Fig. 5

Performance evaluation of different models, (a) Consistency (b) Robustness.

Fig. 6

Scatter plots of 200 sample patients' trajectory robustness and slopes. (a) Logistic Regression, (b) XGBoost, (c) Cox Proportional-Hazard Model, (d) the proposed MGP-MS model.