Current atrial fibrillation (AF) management guidelines suggest that initially a decision must be made to apply either a rate control or rhythm control strategy in the individual patient. However, patients' selection remains substantially empirical and the strategy initially chosen often proves unsuccessful and alternative therapies must be adopted. Thus, it seems desirable to develop and apply tests that quantify AF disease state and guide AF management. The overall hypothesis of this paper is that time-frequency analysis of AF from modified 12-lead ECG configurations will improve AF management beyond current diagnostic and therapeutic standards. In particular, we present a novel concept in which 12-lead ECG configurations are modified for time-frequency analysis of AF (electroatriography). While five electrodes (VR, VL, VF, V1, V2) are placed in the conventional position, the other four electrodes (V3, V4, V5, V6) are empirically repositioned anterior or posterior over the atria. By applying spatiotemporal QRST cancellation and time-frequency analysis to these recordings in 19 patients with persistent AF, fibrillatory rate dispersion among individual anterior (25+/-14 fibrillations per minute, fpm) and posterior leads (16+/-11fpm) as well as individual anterior/posterior rate gradients ranging between -24 and +116fpm could be identified. Consequently, the portrayed techniques may form the conceptual basis for individualized noninvasive characterization of AF. Initiation of further studies using the described techniques in different AF subsets, for comparisons with intracardiac recordings and outcome of different therapies, e.g. cardioversion, antiarrhythmic drug and ablation therapy may be stimulated.