Multimodal brain monitoring and neuroinformatics
Hooman Kamel, J. Claude Hemphill · 2016
Abstract Multimodal brain monitoring involves the concurrent use of two or more methods of monitoring to assess aspects of brain physiology, metabolism, or function. While different methods of brain monitoring exist, no one method or monitor provides complete information. The purpose of multimodal monitoring is to utilize complementary monitoring tools in order to provide a more complete assessment with the goal of guiding individualized treatment and limiting secondary brain injury. Examples of brain monitoring methods used include measurement of intracranial pressure and subsequent derived values such as cerebral perfusion pressure or autoregulatory indices, brain tissue oxygen tension, microdialysis assessment of cerebral metabolites, cerebral blood flow, and electroencephalography. Given the complexity of acute brain injury and the high volume of data generated during multimodal monitoring, attention is now focused on methods of data acquisition, integration, and analysis that may provide improved insights into disease pathophysiology that are not apparent from current relatively crude methods of bedside presentation and assessment. Kiosk-based and distributed systems are emerging for the purpose of high-resolution data acquisition and archiving. Newer approaches to data analysis include data-driven methods such as artificial neural networks, or model-based methods such as dynamic Bayesian networks. Concepts of data visualization and workflow are also emerging as important considerations in moving multimodal brain monitoring from a research consideration into a routinely used paradigm in neurocritical care. This chapter provides a framework for the concept of multimodal brain monitoring in neurocritical care and outlines considerations for data acquisition, analysis, and real-time clinical use.