We present a novel semi-autonomous clinician-in-the-loop strategy to perform the laparoscopic cryoablation of small kidney tumors. To this end, we introduce a model-independent bimanual tissue manipulation technique. In this method, instead of controlling the robot, which inserts and steers the needle in the deformable tissue (DT), the cryoprobe is introduced to the tissue after accurate manipulation of a target point on the DT to the desired predefined insertion location of the probe. This technique can potentially reduce the risk of kidney fracture, which occurs due to the incorrect insertion of the probe within the kidney. The main challenge of this technique, however, is the unknown deformation behavior of the tissue during its manipulation. To tackle this issue, we proposed a novel real-time deformation estimation method and a vision-based optimization framework, which do not require prior knowledge about the tissue deformation and the intrinsic/extrinsic parameters of the vision system. To evaluate the performance of the proposed method using the da Vinci Research Kit, we performed experiments on a deformable phantom and an ex vivo lamb kidney and evaluated our method using novel manipulability measures. Experiments demonstrated successful real-time estimation of the deformation behavior of these DTs while manipulating them to the desired insertion location(s).