Single ventricle disease is a congenital condition where the heart ventricles do not form correctly during in utero development but instead form a single ventricle (SV). The abnormal shape of the SV directly influences the underlying biomechanics of the patient’s heart. Despite its clinical importance, we understand very little about SV mechanics, which impacts our ability to develop predictive models of post-treatment. As a first step in developing high-fidelity models, we determined the in vivo SV deformations over a single cardiac cycle to obtain 3D strains fields. 3D in vivo images of 3 patients with SV were segmented over the cardiac cycle, with 15-20 frames per heart. This resulted in endo- and epi-cardial point clouds that were simultaneously fit to determine the 3D wall strains. Fits were created using truncated hierarchical B-splines (THB-spline). We exploited the local support of THB-splines to compute local deformations and had the ability to utilize mesh refinement in more complex regions. The resulting personalized THB-Spline mesh convects with the SV over the cardiac cycle, allowing 3D shape and strains to be directly computed. Use of THB-splines resulted in high-fidelity fits with far fewer control points than our previous NURBS based method, with minimal loss of accuracy. Results from all three patients demonstrated a very complex, asymmetric, and generally abnormal deformations patterns. These included focal regions of high deformation (>30% strain) and pronounced irregular end-diastole to end-systolic deformations. We also observed significant transmural kinematics not previously observed. A direct advantage of our approach is that it can be directly used to develop patient-specific cardiac models of SV diseased hearts. Use of THB-splines facilitates efficient and accurate convecting of 3D solid models for SV disease. Above all, this first study underscored the need for personalized patient-specific models for modeling SV disease.