A common component in many protocols for the evaluation of cerebral autoregulation is the comparison of transcranial Doppler ultrasound (TCD) velocities with blood pressure recordings, in which correlations between these two signals correspond to impaired autoregulation. With long data sets and complicated paradigms, however, visual inspection alone cannot adequately distinguish random coincidence from consistent correlation in a statistically valid fashion. We suggest and illustrate the use of the coherence index for this purpose. To illustrate this technique, long-term recordings of TCD velocity and blood pressure were obtained from 6 normal subjects and using 23 data segments from 8 patients following subarachnoid hemorrhage. Each signal was first normalized to its mean, and coherence calculated by dividing the data into overlapping subintervals and computing an average. Coherence was specifically examined over time periods of 30 sec. Coherence calculations identified correlations between signals for which interpretation by visual inspection was unclear, and obvious correlations could be quantified. In 4 of the 6 normal subjects, the coherence was less than 0.60 but slightly greater than 0. Five of the 8 patients showed segments with coherence of greater than 0.60. The coherence index provides a quantitative tool for the evaluation of comparisons between two complex signals. As this task becomes more common in the evaluation of cerebral autoregulation, algorithms of this sort will become increasingly necessary
Giller, C.A., Iacopino, D.G. (1997). Use of middle cerebral velocity and blood pressure for the analysis of cerebral autoregulation at various frequencies: the coherence index. NEUROLOGICAL RESEARCH, 19(19), 634-640.