Using entropies to monitoring intracranial pressure, evidence from an animal model

FRANCISCO REDELICO

BIOMEDICAL SIGNAL PROCESSING AND CONTROL

ELSEVIER SCI LTD

Lugar: Amsterdam; Año: 2023

1746-8094

Intracranial hypertension (ICH) is associated with worse neurological outcomes and increasedmortality. Therefore, its correct monitoring is very important in neurological intensive care andthe operating room. There have been successful attempts to use entropic quantifiers to monitorintracranial pressure (ICP);however, they have not been compared against each other to analyzetheir properties. In this study, we conducted an animal experiment on intracranial hypertensionand analyzed the data to determine the efficacy of the most commonly used entropies in literature,namely, Approximate Entropy, Sample Entropy, Permutation Entropy, and Wavelet Entropy. Ouranalysis revealed that Wavelet Entropy exhibited the best early warning properties, detecting amedian insult value of 10.34 ml, 147 seconds before the ICP reached 20 mmHg, when the ICPmedian value was 8.37 mmHg. Although all the entropies showed a decomplexing effect on theICP signal, Wavelet Entropy was the most sensitive, possibly due to the frequency-dependentnature of brain compliance.