The Use of Model and Artificial Intelligence Techniques in Patient Monitoring

Abstract

There exist a large number of techniques in patient monitoring, which have been well developed to measure physiological parameters. A skilled doctor can assess the vital level of the patient by integrating the information derived from multiple variables. The next stage will be to develop a technique which anyone can use for the aids of the diagnosis. In the previous studies Forrester et al reported to characterize patients by the haemodynamic deviation focused on the pulmonary capillary wedge pressure (PCWP) and the cardiac index (CI), proposed various kinds of drug therapy. Despite this, the natural course of haemodynamic change during the acute phase is not completely known, since most studies have not made serial measurements. Moreover, they studied only the selected groups of high risk patients, and the patients with initially less severe symptoms have not been fully represented. The aim of this study is to predict the haemodynamic change by a simple model. The advantage of the model approach is that it can manage the time-series of multiple physiological parameters systematically. The limit of it is that it cannot infer drastic change of the parameters, and the drugs are needed on that particular occasion. So the dosage of drugs still depends more or less on the doctor’s thought process. The AI-techniques will be the suitable aids of the drug therapy in the future.