Unsupervised phenotyping of sepsis using nonnegative matrix factorization of temporal trends from a multivariate panel of physiological measurements.

Menghan DingYuan Luo

Published in: BMC medical informatics and decision making (2021)

Our results suggest that these phenotypes can be used to develop targeted therapies based on phenotypic heterogeneity and algorithms designed for monitoring, validating and intervening clinical decisions for sepsis patients.

Keyphrases

machine learning
end stage renal disease
acute kidney injury
intensive care unit
newly diagnosed
ejection fraction
chronic kidney disease
septic shock
prognostic factors
single cell
high throughput
peritoneal dialysis
deep learning
patient reported outcomes
patient reported