Communications of the ACM
Early Warning System Model Predicts Cancer Patients' Deterioration
Washington University in St. Louis' Chenyang Lu observed, There are early signs hidden in the data that suggest that a person will develop clinical deterioration in a few hours or a few days."
Credit: iStock
Chenyang Lu and colleagues at Washington University in St. Louis have developed a machine learning predictive model for hospitalized cancer patients that integrates heterogeneous data in electronic health records.
The end-to-end CrossNet recurrent neural network model combines static data collected at time of admission with time-series data gathered repeatedly during the hospital stay.
CrossNet learns to forecast deterioration while accurately entering any missing static or time-series data.
Said Lu, “Humans cannot see these hidden patterns or trends in the data, so this is where machine learning is very good at picking up these patterns.”
The research is part of an effort to design an early warning system for predicting cancer patients' deterioration while hospitalized, as well as enhancing patient outcomes.
From The Source (Washington University in St. Louis)
View Full Article
No entries found