Artificial neural networks for the classification of potentially intoxicated patients treated with digoxin

This paper proposes a global solution for classifying patients with risk of being intoxicated when treated with oral digoxin. This is a widely used drug for treating cardiac diseases although it may produce intoxication in the patient. The goal of this work is to classify patients accurately into two categories: those with high risk of being intoxicated and those with low risk. Several artificial neural networks have been used to deal with it. As a direct consequence of the excellent outcomes obtained, a new software tool for clinical-decision has been developed which is currently used with success and confidence in several hospitals and pharmacy services