Sample size and power determination when limited preliminary information is available

Table 3 Estimated sample size needed per group in various scenarios

DOSI Measure	Power	Significance level	Estimated correlation coefficient^a
			0.50		0.80		0.90
			Common STD^b (σ)	2 × Common Variance	Common STD^b (σ)	2 × Common Variance	Common STD^b (σ)	2 × Common Variance
Water			σ = 2.121	σ = 2.999	σ = 1.420	σ = 2.008	σ = 1.088	σ = 1.539
	80	0.01	69	136	32	62	20	38
		0.05	47	92	22	42	13	25
	90	0.01	88	173	41	79	25	47
		0.05	62	122	29	56	17	33
Deoxyhemoglobin (ctHHb)			σ = 0.287	σ = 0.406	σ = 0.198	σ = 0.279	σ = 0.158	σ = 0.224
	80	0.01	72	142	35	68	23	45
		0.05	48	95	24	46	16	30
	90	0.01	90	180	45	86	29	56
		0.05	64	127	31	61	21	40
Lipid			σ = 2.633	σ = 3.724	σ = 1.816	σ = 2.568	σ = 1.443	σ = 2.041
	80	0.01	96	191	47	92	31	59
		0.05	65	128	32	62	21	40
	90	0.01	122	242	59	117	38	75
		0.05	86	171	42	80	27	53

^aThe correlation coefficient was estimated between pre- and post-treatment in simulated MRI data and in simulated DOSI measure
^bThe standard deviation is denoted as STD

ISSN: 1471-2288