Dascena Presents Poster Highlighting Machine Learning Approach to Predict Acute Heart Failure Risk at AHA Scientific Sessions
AHF syndromes are associated with significant morbidity and mortality. Patients at high risk of AHF are likely to experience re-hospitalization or death post-discharge. Dascena’s machine learning algorithm demonstrated an area under the receiver operating curve (AUROC) of 86% for the prediction of AHF diagnosis with prescription of a vasodilator using vital signs and laboratory results, and an AUROC of 76.7% using a vitals-only model. Predictions from the vitals-only model were made an average of 38.5 hours prior to AHF diagnosis, and predictions incorporating laboratory results were made an average of 34 hours prior to AHF diagnosis.
“AHF risk stratification remains a significant unmet need in the patient community, and we are encouraged that our machine learning model is a useful method for predicting AHF,” said Jana Hoffman, Ph.D., vice president of science at Dascena. “The ability to identify patients at high or low risk for AHF may lead to earlier diagnoses and treatment. This will not only accelerate effective treatment and reduce the duration of hospital stays, but will ultimately improve patient outcomes in the long term.”
Dascena is developing machine learning diagnostic algorithms to enable early disease intervention and improve care outcomes for patients. For more information, visit dascena.com.
Dan Budwick, 1AB
Source: Dascena, Inc.