Detection of autoregulation in the brain of premature infants using a novel subspace-based technique

A study by Tsuji et al. (2000) suggests that, under certain circumstances, concordant changes in cerebral intravascular oxygenation and mean arterial blood pressure reflect impaired cerebrovascular autoregulation. In this paper, we propose a new measure to quantify this concordance, derived from the common subspace of these signals. The method is compared to correlation and coherence analysis with respect to our application, but it is also suited to other biomedical signals. Furthermore, this model-based approach is not restricted to applications involving only two signals.