Fig. 5.

Simulated false positives and false negatives arising from systemic and extracerebral contamination. A Haemodynamic response characteristic of functional activation. B Example responses that would not be identified as functional activation. C Simulated responses optimised to resemble the corresponding plots above in panel A (with no activation) and B (with the same level of activation as in panel A). In the first row of panel C, only arterial pressure and ${\mathrm{CO}}_2$ were allowed to vary. In the middle row, a drop in arterial ${\mathrm{O}}_2$ saturation was also permitted. In the bottom row, a contribution from the scalp compartment was also included and the superficial blood flux allowed to vary. Markers below each plot indicate the direction of change of the systemic variables giving rise to the observed signals. (More detailed plots are provided in Supplementary Data 2. Individual values for all the inputs in each case are provided in Supplementary Data 1.).