Real-time, personalized prediction of sepsis onset using fusion of electronic medical records and in-sensor analog classifier
TimeWednesday, July 13th6pm - 7pm PDT
LocationLevel 2 Lobby
Event Type
Networking Reception
Work-in-Progress Poster
DescriptionWe present an artificial intelligence (AI) framework for real-time, personalized sepsis prediction 4 hours before onset through fusion of prediction scores from ECG and patient electronic medical record. An on-chip classifier combines analog reservoir-computer and artificial neural network to perform prediction without front-end data converter or feature extraction which reduces energy by 4x compared to state-of-the-art bio-medical AI circuits. The proposed AI framework predicts sepsis onset with 89.9% accuracy and is demonstrated on patient data collected from Emory University Hospital and MIMIC-III. The proposed framework is non-invasive and does not require laboratory tests which makes it suitable for at-home monitoring.