Predictive Agorithms and Learning for Medics (PAL-MEDIC)

This proposal is for Phase II continuation of our work on Predictive Algorithms and Learning for Medics (PAL-MEDIC).PAL-MEDIC is an integrated mobile application that will provide critical support to Army combat medics and civilian paramedics during pre-hospital patient care, and will also enhance the subsequent care process.Our Phase I program successfully met all of its objectives.Our PAL-MEDIC system design incorporates vitals input from standard field measures and an analysis engine that uses predictive algorithms and an existing protocol library to output alarms, predictions and care recommendations.Most important, our successful Proof-of-Concept Demonstration incorporated all the PAL-MEDIC elements, and showed that:Medical crisis alerting and prediction is feasible and accurate using Causal State Modeling (CSM) algorithms; The PeerAtlas EMS procedures control system can provide a framework for patient-oriented care recommendations; The PAL-MEDIC app can be logically integrated with the NETT Warrior End User Device (EUD) and other mobile devices. Our Phase II work will make two major contributions in applying intelligent computation to military and civilian emergency medical care; these are: One, Predictive Algorithms and Machine Learning for alarms and patient state prediction; and Two, Adaptive Protocols for care recommendations directly related topatient state.