Multi Model Rule Generation for Automatic Drug Administration Systems
Filipe Quinaz, Paulo Fazendeiro, Pedro Nunes Bernotavicius Araujo, Miguel Castelo-Branco, N. Tavares, R. A. da Costa · Procedia Technology · 2013
Many solutions have been proven to effectively and safely control drug dosage in several medical problems. They encounter however barriers in practical application due to the following characteristics: medical models are very challenging to acquire, the need for deep control knowledge by the user, lack of safe measurement and actuation devices, and finally, the abstraction from real world problems. We developed a model that aims to reduce the impact of these difficulties in the deployment of real world automatic administration systems. Our system is based on multi model architecture that allows for adaptative rule creation and algorithm effectiveness evaluation, focused on the automatic control of hypertension in the intensive care unit.