Table 2.

The categorised groups and variability of signals in each quality group.

Groups (n)	CV% of MABP	CV% of MFv
		RMCA	LMCA
Un-A (raw data; n=166)	4.72 (3.98 – 5.46)	7.53 (6.11 – 8.94)	7.78 (6.96 – 8.60)
Group A (n=167)	4.05 (3.79 – 4.31)x	6.33 (5.34 – 7.32)	5.62 (5.27 – 5.98)***
Group B (n=65)	4.14 (3.73 – 4.55)	6.80 (5.37 – 8.23)	11.7 (-0.07 – 23.4)
Group C (n=25)	7.05 (5.56 – 8.53)	5.12 (3.92 – 6.33)	5.02 (3.86 – 6.17)
Group D (n=42)	5.53 (4.51 – 6.55)	5.86 (5.14 – 6.57)	6.30 (5.35 – 7.25)
Group E (n=154)	4.95 (4.50 – 5.40)‡	No Data Side	Signal Side
		61.8 (48.5 – 75.2)‡	12.5 (7.9 – 17.1)†

Un-A: Median-filtered but no manual cleaning; Group A: All optimal in both ABP and CBFV; Group B: Either side or both sides of CBFV are adequate; Group C: ABP is adequate or unusable and CBFV is optimal; Group D: ABP is adequate or unusable and CBFV is optimal or adequate; and Group E: ABP is optimal, adequate, or unusable, one side of CBFV is optimal or adequate but the other side of CBFV is unusable (i.e. quality = 1: severe artefact) or no data (i.e. quality = 0; the measured signal is noise, with no usable information of CBFV being measured).

Data are presented as mean (95% Confidence Interval). MABP, mean arterial blood pressure; MFv, mean cerebral blood flow velocity; CV%, percentage of coefficient of variation;

^xp = 0.069;

^*** p <0.0001 (for Un-A vs. Group A, by paired sample t-test).

^†p-trend <0.05;

^‡ p-trend <0.0001 for Group A–E.

Two patients with only left-sided window are in Groups A and B respectively.