Dascena presents at ATS conference in San Diego, CA
Two novel machine learning algorithms were presented at the American Thoracic Society International Conference, May 18-23, 2018, in San Diego, CA.
A pediatric sepsis prediction and detection algorithm was validated for four hours before pediatric severe sepsis presentation. This algorithm analyzed pediatric vital sign data and achieved a high sensitivity for pediatric patient data at UCSF.
A machine learning algorithm for mortality prediction was developed and validated across multicenter patient datasets, maintaining high sensitivity and specificity up to 48 hours in advance of in-hospital death while using only six commonly measured patient vital signs as input.