Next: Introduction
USING A PHYSIOLOGICAL MODEL FOR PREDICTION
OF THERAPY EFFECTS IN HEART DISEASE
William J. Long, Shapur Naimi, M. G. Criscitiello, Robert Jayes
M.I.T. Laboratory for Computer Science, Cambridge and
Tufts-New England Medical Center, Boston, Massachusetts
Abstract:
The multitude of interacting processes in the cardiovascular
system make reasoning about the possible effects of therapies very
difficult and these difficulties are compounded by the various
diseases that change the physiological relationships. As part of a
project whose overall goal is to develop a program to assist the
physician in reasoning about the diagnosis and management of patients
in heart failure, we have developed a methodology, based on signal flow
analysis, for predicting hemodynamic changes using a model of
physiological relationships. The methodology has the advantage that
it provides a justification for the predictions in terms of the
physiological pathways from the intervention to the effect, accounting
for the feedback paths that affect the pathway. Using this
methodology and cardiovascular relationships derived from the
physiological literature and clinical expertise, we are developing a
model for anticipating and understanding the effects of therapies and
other interventions on the patient with cardiovascular
disease. This paper presents the methodology and its application to
mitral stenosis, aortic stenosis, and aortic regurgitation.