Modelling C-arm fluoroscopy and Operating table Kinematics via Machine Learning

Pascal Fallavollita^1*Faria Jaheen²Vinod Gutta²

• ¹University of Ottawa, Ottawa, Canada
• ²University of Ottawa School of Electrical Engineering and Computer Science, Ottawa, Canada

This work presents a machine learning driven framework for data-efficient kinematic modeling and workspace optimization in modular C-arm fluoroscopy systems integrated with operating tables. Addressing the limitations of traditional analytical solvers in handling joint redundancy, nonlinearity, and real-time adaptability, the proposed approach formulates both forward and inverse kinematics as supervised regression tasks. A comprehensive dataset of joint configurations and end-effector poses annotated with voxelized collision status enables the training of predictive models across multiple system configurations ranging from 5 to 9 degrees of freedom. Extensive benchmarking across five machine learning algorithms including tree ensembles and deep neural networks reveals the effectiveness of ensemble-based models and dimensionality reduction strategies in achieving high spatial precision while maintaining collision avoidance. The proposed framework demonstrates the viability of data-driven trajectory planning in multi-degree of freedom C-arm systems, providing a clinically relevant solution for improving imaging access and reducing intraoperative collision risks. This work contributes to the advancement of intelligent surgical navigation by enabling responsive, AI-assisted decision-making in constrained operating environments.