Table 1:

Regression-derived effects of scan duration on various hemodynamic measurements

Hemodynamic Measurement	Deconvolution Algorithm	Function Used for CBV Integration	t-statistic	Slope^{^a}	Potential Truncation Effect^{^b}	Lesion Reversal Frequency
CBVc	None	Concentration	41.53	1.08/log₁₀(s)	47.6%	37% (21/57)
sCBVr	sSVD	Response	17.99	1.46/log₁₀(s)	64.2%	35% (20/57)
oCBVr	oSVD	Response	16.72	1.17/log₁₀(s)	51.4%	46% (26/57)
sCBF	sSVD	NA	9.62	0.09/log₁₀(s)	4.10%	4% (2/57)
oCBF	oSVD	NA	3.58	0.04/log₁₀(s)	1.96%	2% (1/57)
sMTTc	sSVD	Concentration	34.49	3.03/log₁₀(s)	133%	47% (27/57)
sMTTr	sSVD	Response	35.57	3.73/log₁₀(s)	164%	28% (16/57)
oMTTc	oSVD	Concentration	42.50	3.79/log₁₀(s)	166%	54% (31/57)
oMTTr	oSVD	Response	51.15	4.67/log₁₀(s)	205%	39% (22/57)
sTmax	sSVD	NA	25.91	14.10/log₁₀(s)	6.19 seconds	44% (25/57)
oTmax	oSVD	NA	28.62	18.2/log₁₀(s)	8.00 seconds	42% (24/57)

Note:—NA indicates not applicable.
↵a The slopes of all 11 hemodynamic parameters with respect to the logarithm of scan duration were significantly greater than zero (P < .001). Therefore, t-statistics rather than P values are reported.
↵b “Potential truncation effect” refers to the expected reduction in the calculated parameter value that would result from decreasing the scan duration from 110 seconds to 40 seconds postinjection. For example, if the CBVc value derived from a 110-second scan were 107.6% of normal, the expected CBVc using a 40-second scan would be 60.0% of normal.