Statistical discrimination methods are suitable not only for classification but also for characterisati on of differences between a reference group of patterns a nd the population under investigation. In the last ye ars, statistical methods have been proposed to classify and analyse morphological and anatomical structures of medical images. Most of these techniques work in high-dimensional spaces of particular features such as shapes or statistical parametric maps and have overcome the difficulty of dealing with the inheren t high dimensionality of medical images by analysing segmented structures individually or performing hypothesis tests on each feature separately. In th is paper, we present a general multivariate linear framework to identify and analyse the most discriminating hyper-plane separating two populations. The goal is to analyse all the intens ity features simultaneously rather than segmented versions of the data separately or feature-by-featu re. The conceptual and mathematical simplicity of the approach, which pivotal step is spatial normalisati on, involves the same operations irrespective of the complexity of the experiment or nature of the data, giving multivariate results that are easy to interp ret. To demonstrate its performance we present experimental results on artificially generated data set and real medical data.