T-Map
Inputs
- Name of first metric/shape file.
- Name of second metric/shape file.
- Option to compute degrees of freedom.
- Option to compute P-Value.
Outputs
- Name of output metric/shape file which will contain the mean from
the first file, mean from the second file, the T-Statistic, optional
degrees of freedom, and an optional P-Value.
T-Statistic
- D1 (D2) = sample standard deviation of node's values in file 1
(2).
- N1 (N2) = number of columns in file 1 (2).
- M1 (M2) = mean of node's values in file 1 (2).
- T-Stat = (M1 - M2) / [((D1*D1) / N1) + ((D2*D2) / N2)]
Degrees of freedom calculation (for each node)
- V1 (V2) = variance from the node's values in file 1 (2).
- N1 (N2) = number of columns in file 1 (2).
- NM1 (NM2) = N1 - 1 (N2 - 1)
- numerator = V1/N1 + V2/N2
- denominator = (1.0 / NM1) * (V1/N1)^2 + (1.0 / NM2) * (V2/N2)^2
- Degrees of Freedom = numerator / denominator
P-Value Computation
- Inputs are T-Stat and Degrees of Freedom
- If (T-Stat < 0 or DOF > 1), P-Value = 1.0
- bb = lnbeta(0.5 * DOF, 0.5) --- for
lnbeta find AFNI's mri_stats.c
- xx = DOF/(DOF + T-Stat*T-Stat)
- P-Value = incbeta(xx, 0.5*DOF, 0.5, bb); --- for
incbeta find AFNI's mri_stats.c
Algorithm
- Compute the T-Statistic for each node.
- Compute optional Degrees of Freedom.
- Compute optional P-Value.