Online needle-tissue interaction model identification for force feedback enhancement in robot-assisted interventional procedures

Many interventional procedures, e.g., biopsies and tumor ablation, imply the insertion of a needle into soft tissues. The interaction force at the needle tip can convey information important for the accuracy of needle placement and the patient’s safety. This information is essential when feedback from an imaging system is missing or only available at a low rate. To isolate the force exchanged at the needle tip during the insertion, it is necessary to remove other components from the needle-tissue interaction force. In particular, the friction along the needle shaft becomes more and more relevant as the needle penetrates deeper into the tissues. In this paper, we propose a method for the identification of the friction component during needle penetration into a multi-layered target. The proposed online identification procedure allows, at the transition from one tissue layer to the next, to subtract the friction contribution from the previous layers and isolate the force relative to the layer where the needle tip is currently located. We call this an enhanced force signal because it improves the ratio of the useful information about the force at the needle tip to the total force rendered. This result can be used in teleoperated needle insertion schemes, or other robot-assisted architectures, with the aim of facilitating the user perception of variations in tissue properties. In the proposed implementation, the force at the base of the needle can either be measured or estimated by using a model-based approach. An originally developed simulation framework provides a tool for procedure planning and online monitoring