A probabilistic network for the diagnosis of acute cardiopulmonary diseases

TL;DR

This paper presents a probabilistic network model for diagnosing 63 cardiopulmonary diseases, combining expert knowledge and data-driven updates to improve diagnostic accuracy and interpretability.

Contribution

It introduces a novel probabilistic network with a low-parameter, physician-understandable parameterization, integrating expert elicitation and Bayesian updating for medical diagnosis.

Findings

Achieved satisfactory Concordance Index values for acute diseases

Enabled diagnosis of 63 diseases using up to 167 patient findings

Demonstrated effective inference on fictitious patient cases

Abstract

In this paper, the development of a probabilistic network for the diagnosis of acute cardiopulmonary diseases is presented. This paper is a draft version of the article published after peer review in 2018 (https://doi.org/10.1002/bimj.201600206). A panel of expert physicians collaborated to specify the qualitative part, that is a directed acyclic graph defining a factorization of the joint probability distribution of domain variables. The quantitative part, that is the set of all conditional probability distributions defined by each factor, was estimated in the Bayesian paradigm: we applied a special formal representation, characterized by a low number of parameters and a parameterization intelligible for physicians, elicited the joint prior distribution of parameters from medical experts, and updated it by conditioning on a dataset of hospital patient records using Markov Chain Monte Carlo simulation. Refinement was cyclically performed until the probabilistic network provided satisfactory Concordance Index values for a selection of acute diseases and reasonable inference on six fictitious patient cases. The probabilistic network can be employed to perform medical diagnosis on a total of 63 diseases (38 acute and 25 chronic) on the basis of up to 167 patient findings.