What are the accuracy and limitations of published algorithm using surface ECG to localize overt AV accessory pathways?