Updates include an enlarged and refined training dataset, the use of CSF data (in addition to grey matter and white matter), PCA to reduce dimensionality and individual 95% confidence intervals derived from the Gaussian Processes regression model. Minor bug fixes since v2.0.