Technical Report XXVII
A Software Tool for Applying Bayes' Theorem in Medical Diagnostics
T Chatzimichail, AT Hatjimihail
Abstract
Background: In medical diagnostics, determining post-test or posterior probabilities for disease and understanding associated uncertainty and confidence intervals are essential for patient care.
Objective: This study introduces a software tool developed in the Wolfram Language for the parametric estimation, visualization, and comparison of Bayesian diagnostic measures and their uncertainty.
Methods: The tool employs Bayes' theorem to compute posterior probability for disease and absence of the disease, and diagnostic thresholds derived positive and negative predictive value. It also quantifies their standard sampling, measurement, and combined uncertainty using normal, lognormal, and gamma distributions and applying uncertainty propagation methods.
Results: : The tool generates diagnostic measures, standard uncertainty, and confidence intervals estimates and provides their plots, supporting clinical decision-making. A case study using fasting plasma glucose data from the National Health and Nutrition Examination Survey in the USA showcases its application in diagnosing diabetes mellitus, highlighting the significant role of measurement uncertainty.
Conclusion: : The software enhances the estimation and facilitates the comparison of Bayesian diagnostic measures, which are critical for medical practice. It provides a framework for analyzing uncertainty and assists in understanding and applying Bayes' theorem in medical diagnostics.
Published
2024
Citation
Chatzimichail T, Hatjimihail AT. A Software Tool for Applying Bayes' Theorem in Medical Diagnostics. Technical Report XXVII. Hellenic Complex Systems Laboratory; 2024. Available at: https://www.hcsl.com/TR/hcsltr27/hcsltr27.pdf
The material made freely available by Hellenic Complex Systems Laboratory is subject to its Terms of Use.