Does the 64-pole basket catheter provide adequate spatial coverage and electrode contact for mapping atrial fibrillation in humans?