
After estimating the mixingmatrix , the source estimates are calculated according to equation 6 by projecting each voxel's time course onto the time courses contained in the columns of the unmixing matrix .
[McKeown et al., 1998] suggest transforming the spatial maps to scores (transform the spatial maps to have zero mean and unit variance) and thresholding at some level (e.g, ). The spatial maps, however, are the result of an ICA decomposition where the estimation optimises for nonGaussianity of the distribution of spatial intensities. This is explicit in the case of the fixedpoint iteration algorithm employed here, but also true for the Infomax or similar algorithms where the optimisation for nonGaussian sources is implicit in the choice of nonlinearity. As a consequence, the spatial intensity histogram of an individual IC map is not Gaussian and a simple transformation to voxelwise scores and subsequent thresholding will necessarily result in an arbitrary and uncontrolled falsepositive rate: the estimated mean and variance will not relate to an underlying nulldistribution. Figure 1 shows an example where the estimated Gaussian (dashdotted line) neither represents the 'background noise' Gaussian nor the entire image histogram, and any threshold value based on the expected number of falsepositives becomes meaningless with respect to the spatial map.
Instead, consider the estimated residual noise at a single voxel location :
Under the nullhypothesis of no signal and after variancenormalisation, the estimated sources are just random regression coefficients which, after this transformation, will have a clearly defined and spatially stationary voxelwise falsepositive rate at any given threshold level.^{3}While, for reasons outlined above, the nullhypothesis test is generally not appropriate, the voxelwise normalisation also has important implication under the alternative hypothesis; it normalises what has been estimated as effect (the raw IC maps) relative to what has been estimated as noise and thus makes different voxel locations comparable in terms of their signalto noise characteristics for a now given basis (the estimated mixing matrix). This is important since the mixing matrix itself is data driven. As such, the estimated mixing matrix will give a better temporal representation at different voxel locations than at others and this change in 'specificity' is reflected in the relative value of residual noise.
In order to assess the maps for significantly activated voxels, we follow [Everitt and Bullmore, 1999] and [Hartvig and Jensen, 2000] and employ mixture modelling of the probability density for spatial map of scores.
Equation 6 implies that
In cases where the number of 'active' voxels is small, however, a single Gaussian mixture may actually have the highest model evidence, simply due to the fact that the model evidence is only approximated in the current approach. In this case, however, a transformation to spatial scores and subsequent thresholding is appropriate, i.e. reverting to null hypothesis testing instead of the otherwise preferable alternative hypothesis testing.
If the mixture model contains more than a single Gaussian, we can calculate the probability of any intensity value being background noise by evaluating the probability density function of the single Gaussian that models the density of background noise. Conversely, we can evaluate the set of additional Gaussians and calculate the probability under the alternative hypothesis of 'activation'^{4} with respect to the associated time course, i.e. we obtain the estimate of the posterior probability for activation of voxel in the score map as [Everitt and Bullmore, 1999]:
Figure 3 illustrates the process for a spatial map extracted from a data set with artificial activation introduced into FMRI resting data (see section 5 for details). Voxels with an estimated posterior probability of activation exceeding a certain threshold value are labeled active. The threshold level, though arbitrary, directly relates to the loss function we like to associate with the estimation process, e.g. a threshold level of 0.5 places an equal loss on false positives and false negatives [Hartvig and Jensen, 2000]. Alternatively, because we have explicitly modelled the probabilities under the null and alternative hypothesis, we can choose a threshold level based on the desired false positive rate over the entire brain or at the cluster level simply by evaluating the probabilities under the null and alternative hypotheses.