Cardiac ablation (CA) is increasingly used to treat atrial fibrillation. However, long-term success is relatively low, and the procedure carries serious risks. To this end, we are developing a β (coefficient of nonlinearity) imaging method that may be employed to perform both tissue characterization and real time temperature estimation to respectively plan, monitor, and execute optimal CA. Starting from a one-dimensional generalized form of the Westervelt equation, we derived an analytical procedure for extracting β which is then further adapted to echo-mode. To evaluate the method performances, in-silico and in-vitro experiments were performed. First, one- to three-dimensional simulations including linear array scanning of three-dimensional objects were obtained with the INCS method. Next, the ULA-OP scanner was used with an Esaote LA332 linear-array to image a phantom consisting of 2-layers obtained as a mixture of oil, gelatin and water. Varying the percentage of oil (β similar to fat), different β values were obtained for the two layers. Both in-silico and in-vitro results show the capability of the method to estimate β variations. Compared to existing methods, the proposed approach provides more stable estimations (spatially) and does not require a special transducer or set-up, being more easily applicable in a clinical setting.