Towards safer robot-assisted skull bone drilling: A real-time force model under an elastic clinical environment

The safety of cranial surgical drilling is often challenged by the risk of complications, especially mechanical damage to soft tissues from accidental penetration, highlighting the need for a deeper understanding of this procedure. While robot-assisted drilling contributes to the stability and precision, it is short of clinical experience to understand the force-depth relationship during this process. Traditional bone drilling models working under experimental conditions are challenged by several complexities in clinical cranial utilization, fluctuating feedrate, manual interruption, and low system stiffness. To address these questions, a real-time force model is introduced in this study to predict the thrust force and the actual drilling depth, aiming to enhance the monitoring capacity in craniotomy for better safety. An offline model is included to elucidate the force generation mechanism under varied feedrate using specialized drilling tool, as an expansion of traditional research. Then a novel online prediction method complements this foundation by considering the current robot position and system stiffness, providing real-time depth and force estimation based on partial differentiation. This approach is well-suited for adjusting to manual interruptions and density variations dynamically, which reveals the pattern of force variation in the clinical environment. Experimental validation demonstrated a reasonable prediction accuracy and a submillimeter depth error, indicating the feasibility to monitor the skull drilling procedure. This capability broadens the traditional drilling force model’s application and significantly contributes to the development of multi-modal safety protocols in clinical robot-assisted skull drilling.