Plateau waves are recurrent phenomena observed in traumatic brain injury (TBI) patients, characterised by an increase in intracranial pressure (ICP) above 40 mmHg combined with an almost zero arterial blood pressure (ABP) variation and, hence, a decrease in cerebral perfusion pressure (CPP). A raised ICP for a long period of time, namely plateau waves, can lead to a secondary brain injury. Due to the impaired cerebral autoregulation mechanism these TBI patients present, they are admitted to neurocritical care units (NCCUs) to be under continuous multimodal monitoring, which allows a correct diagnosis for each patient. Plateau waves can end naturally by activating a vasoconstriction mechanism which decreases the amount of blood available in the brain. Alternatively, the phenomenon can end with therapeutic treatment.In this sense, the present study consists in the development of an algorithm capable of automatically detecting plateau waves using offline data, i.e. data already collected from patients. This creates an extra tool which allows for faster detection of events to assist their identification and final diagnosis. Despite the additional steps that can be included to improve the algorithm, the results show good performance, and thus it may be applied in NCCUs.
Keywords: Arterial blood pressure; Event detection; Intracranial pressure; Neurocritical care unit; Plateau waves; Traumatic brain injury.