Photo Credit: Chinnapong
The following is a summary of “Dynamic assessment of signal entropy for prognostication and secondary brain insult detection after traumatic brain injury,” published in the December 2024 issue of Critical Care by Bögli et al.
Researchers conducted a retrospective study to explore the origin and value of entropy time trends in traumatic brain injury (TBI) and its potential to predict unfavorable outcomes.
They accessed 232 continuous recordings of arterial blood pressure and intracranial pressure (ICP) from patients with TBI with available clinical data and 6-month outcomes (Glasgow Outcome Scale) from the Brain Physics database. Multiscale entropy (MSE), calculated at 20 coarse graining steps starting from 0.1 Hz, was used to estimate biosignal entropy. The MSE was computed for consecutive, overlapping 6-hour segments and evaluated the percentage of monitoring time (ptime) or dosage (duration*level/hour) below various cutoffs against outcomes through univariable and multivariable analyses, along with propensity score matching. Correlations with clinical and monitoring metrics were explored using correlation coefficients. Finally, they assessed secondary brain insults, including deviations in ICP, cerebral perfusion pressure (CPP), and pressure reactivity in relation to MSE changes.
The results showed that increased MSE of arterial blood pressure (ABP) and cerebral perfusion pressure (CPP) time (OR 1.28 [1.07–1.58] for ABP, OR 1.50 [1.16–2.03] for CPP) and dose (OR 1.12 [1.02–1.27] for ABP, OR 1.21 [1.06–1.46] for CPP) were linked to poor outcomes, even after correcting for intracranial pressure (ICP), CPP, and the pressure reactivity index in multivariable models with propensity score matching. The MSE trajectories differed significantly based on outcome. The entropy metrics showed weak correlations with clinical parameters. Episodes of deranged physiology were associated with decreased MSE from both cerebral and systemic biosignals.
Investigators concluded the dynamic changes in biosignal entropy after TBI offer potential for individualized outcome prediction, identification of secondary insults, and enhanced utilization of existing physiological data in critical care.
Source: ccforum.biomedcentral.com/articles/10.1186/s13054-024-05228-z
