A causal probabilistic model for the estimation of risk of sudden death in athletes

Author

Irler, W.J. ; Bertoldi, A. ; Bettini, R. ; Antolini, R. ; Furlanello, F.

Author_Institution

IRST Povo, Italy

fYear

1991

fDate

23-26 Sep 1991

Firstpage

465

Lastpage

468

Abstract

A causal probabilistic model is described which unifies the related arrhythmological and cardiopathological concepts in a homogeneous diagnostic network and gives a mechanism to estimate the risk when competitive sport activity is exercised. The records from 766 amateur and top-level athletes from 1974 to 1991 provided the medical expertise for the weighting of the nodes in the network. Some cognitive issues of Pearl´s belief propagation for singly connected graphs explain the design. The model holds 60 finding nodes and 50 internal concepts. For a number of chosen athletes, the fitness estimation is compared with the clinical decision

Keywords

cardiology; expert systems; medical diagnostic computing; physiological models; Pearl´s belief propagation; amateur athletes; arrhythmology; cardiopathology; causal probabilistic model; clinical decision; cognitive issues; competitive sport activity; fitness estimation; homogeneous diagnostic network; singly connected graphs; top-level athletes; Bayesian methods; Belief propagation; Cardiac disease; Cardiology; Cardiovascular diseases; Certification; Medical diagnostic imaging; Pathology; Probability; Testing;

fLanguage

English

Publisher

ieee

Conference_Titel

Computers in Cardiology 1991, Proceedings.

Conference_Location

Venice

Print_ISBN

0-8186-2485-X

Type

conf

DOI

10.1109/CIC.1991.169145

Filename

169145