Introduction: Quantitative time of flight in echo mode ultrasound computed tomography (TFEM USCT) is an ill-posed and ill-conditioned inverse problem, where the goal is to reconstruct the speed-of-sound (SoS) distribution from time-of-flight measurements. Compared to the more established transmission-mode USCT, TFEM USCT poses a greater challenge due to increased indeterminacy, as the travel paths of the ultrasound signals are undefined. As such, regularization is essential for stable and physically meaningful reconstructions. Aim: This study introduces an anisotropic Tikhonov regularization using a second-derivative operator to better capture directional SoS variations. The goal is to enable cardiac imaging using a single sparse linear transducer. Such a configuration could allow rapid quantitative imaging of the entire heart without requiring repositioning of the transducer, reducing operator dependency. Methods: Computational simulations were conducted using the k-Wave toolbox for MATLAB and a simplified numerical phantom representing an axial cross-section of the human thorax, centered on the heart region. The transducer array was modeled with a 120mm length, consisting of 128 punctual elements with a 0.9091mm pitch. Reconstructions were performed for anisotropic regularization coefficients K ranging from 0.00 to 1.00 in 0.05 increments, and quantitatively evaluated using root mean square error (RMSE). Results: The lowest RMSE values were observed for K values between 0.35 and 0.45, depending on the structure. Across all structures, reconstructions using anisotropic regularization outperformed the isotropic case (K = 0.5), enabling clearer delineation of features. Conclusions: Results support the feasibility of quantitative cardiac imaging using a sparse transducer array and show that anisotropic regularization provides superior reconstructions.