Finding the needle by modeling the haystack: Pulmonary embolism in an emergency patient with cardiorespiratory manifestations

作者：

Highlights：

• Medical diagnosis is performed from a patient’s findings without requiring any conjecture.

• Diagnostic predictions are empirically calibrated and consistent with medical reasoning.

• Causal knowledge allows the system to learn from any type of clinical data.

• Causal assumptions can be revised in the light of model performance assessments.

摘要

•Medical diagnosis is performed from a patient’s findings without requiring any conjecture.•Diagnostic predictions are empirically calibrated and consistent with medical reasoning.•Causal knowledge allows the system to learn from any type of clinical data.•Causal assumptions can be revised in the light of model performance assessments.

论文关键词：Bayesian networks,Markov chain Monte Carlo,Causal modeling,Latent variables,Emergency medicine,Calibration analysis

论文评审过程：Received 23 February 2021, Revised 16 August 2021, Accepted 9 October 2021, Available online 15 October 2021, Version of Record 23 October 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.116066