Bayesian approach to uncertainty quantification for cerebral autoregulation index

TL;DR

This paper introduces a Bayesian method using Markov-chain Monte Carlo to quantify uncertainty in the cerebral autoregulation index, providing error estimates for a commonly used clinical measure.

Contribution

It presents a novel Bayesian framework for calculating the ARI with associated uncertainty, enhancing the interpretability of autoregulation assessments.

Findings

Bayesian approach yields probability distributions for ARI.

Provides error estimates for the ARI.

Improves reliability of cerebral autoregulation assessment.

Abstract

Cerebral autoregulation refers to the brain's ability to maintain cerebral blood flow at an approximately constant level, despite changes in arterial blood pressure. The performance of this mechanism is often assessed using a ten-scale index called the ARI (autoregulation index). Here, $0$ denotes the absence of, while $9$ denotes the strongest, autoregulation. Current methods to calculate the ARI do not typically provide error estimates. Here, we show how this can be done using a bayesian approach. We use Markov-chain Monte Carlo methods to produce a probability distribution for the ARI, which gives a natural way to estimate error.