In children with congenital heart disease, strain imaging might help in choosing the time-point for intervention. We evaluated three time delay estimation methods that use a-priori information on the cross-correlation function. The autocorrelation function and a parametric function using the Gaussian shape of the autocorrelation function were tested using phantom and patient acquisitions. These experiments revealed a performance similar to conventional peak detection of the cross-correlation function in phantoms and a decreased performance for in vivo data. Furthermore, it is possible to estimate the strain in vivo in the myocardium of patients and normals using an rf-based strain imaging approach.