Introduction: Atrial ablation has been recently utilized to treat atrial fibrillation (AF) by isolation or destruction of arrhythmia drivers. In chronic or persistent AF patients these drivers often consist of one or few rotors at unknown exact locations, and several ablation attempts are commonly conducted before arrhythmia activity is terminated. However, the irreversible damage done to the atrial tissue may lead to recurrence of AF within months or few years after the procedure. Here we propose an alternative strategy to terminate rotor activity by its attraction to a low-energy depolarizing probe and its traction into a peripheral non 1:1 conducting region. Methods: The feasibility of the proposed method was numerically tested in 2D models of chronic AF human atrial tissue. Left-to-right gradients of either acetylcholine (ACh) or potassium conductance were employed to generate distinct regions of 1:1 and non 1:1 conduction, characterized by their dominant frequency (DF) ratios. Spiral waves were established in the 1:1 conducting region, and raster scanning pattern was employed using a stimulating probe to attract the spiral wave tip. The probe was then linearly moved towards the boundary between the two regions. Results and conclusions: Successful attraction and anchoring of spiral waves to the probe was demonstrated for all scanning configurations when the probe was <8mm from the spiral wave tip. Maximal traction velocity without loss of anchoring increased in a non-linear, monotonic way with increasing values of ACh concentration. Success rate of spiral wave termination was over 90% for regional DF ratios of as low as 1:1.2 and 100% for ratios over 1:1.4. Given that normally much higher ratios are measured in physiological atrial tissues, we envision this technique to provide a feasible, safer alternative to ablation procedures performed in persistent AF patients.