An Intelligent Control of Flexible Needle Insertion Based on Model Predictive Control for Soft Tissue Intervention

The insertion of flexible needle into soft tissue is subject to uncertainties arising from tissue deformation and organ displacement, which impose technically demanding constraints on achieving precise control. An intelligent control algorithm based on Model Predictive Control (MPC) for flexible needle insertion into soft tissue is proposed in this paper. To characterize the deformation of needles, a kinematic model describing the deflection behavior of the needle during insertion into soft tissue is formulated. The Control Strategy Optimizer of MPC determines the optimal control command for the next step through the predicted future insertion trajectory by integrating the reconstructed trajectory with the needle deformation model, ensuring the needle tip trajectory aligns closely with the predefined reference path. A hydrogel phantom based on polyvinyl alcohol (PVA) is developed to mimic the mechanical behavior of real pulmonary soft tissue for experimental validation. The experimental results show that the root mean square error of targeting error is 0.89 mm. The proposed algorithm effectively compensates for positional deviations during flexible needle insertion, thereby enhancing both the precision and safety of percutaneous interventions.