Fig. 6.

Variation of haemodynamic responses with changes in ${\mathrm{P}}_{\mathrm{a}}{\mathrm{CO}}_2$ (from baseline 40 mmHg) and blood pressure (baseline 100 mmHg). In the upper 3D graphs, responses are shown for simultaneous changes in both inputs. (Note the axis for ${\mathrm{P}}_{\mathrm{a}}$ in these plots increases from right to left.) The 2D graphs in the middle row show effects of changing ${\mathrm{P}}_{\mathrm{a}}$ and ${\mathrm{P}}_{\mathrm{a}}{\mathrm{CO}}_2$ alone. Shaded regions indicate where a false identification results when using a simple distance-based classifier. At the bottom, results from the same classifier are shown for both factors varying jointly over their ranges, with colour indicating the strength of the classification. A Systemic factors were varied with no change in metabolic demand. False positive responses (magenta in the bottom plot) occurred in the region of reduced pressure and increased ${\mathrm{CO}}_2$. They could be produced by ${\mathrm{CO}}_2$ changes alone, but not by pressure. B Metabolic demand was increased at the same time as the systemic changes, shifting the ${\mathrm{HbO}}_2$ response upwards and the HHb downwards. False negatives (cyan in the bottom plot) occurred across a wide range of combinations, but especially with reduced ${\mathrm{CO}}_2$ and increased pressure. They could also arise from changes in either factor alone.