The project has two main goals. We would like to implement a public data base of the existing multi-channel ECG measurements, together with tools for visualization and internet access. In the next phase we would like to implement the first application on the MECG database, dealing with the optimal selection of minimal number of electrodes needed for the reproduction of a standard 12-leads ECG. The investigation methodology will be divided into two mainstreams. First, optimization methods will be used including genetic algorithms and multivariate statistics. Second, advanced interactive visualization methods will be used, which are able to support human ideas and hypothesis to find unknown data relations. All results will be verified on the reference 32-channel measurements performed on different patients and healthy volunteers in terms of comparison of measured and reproduced electrocardiograms.