Accuracy of on-site teleoperated milling with haptic assistance

Sergey Drobinsky; Matías de la Fuente; Behrus Puladi; Klaus Radermacher

doi:10.1007/s11548-023-02983-2

Accuracy of on-site teleoperated milling with haptic assistance

Int J Comput Assist Radiol Surg. 2023 Nov;18(11):1969-1976. doi: 10.1007/s11548-023-02983-2. Epub 2023 Jul 16.

Authors

Sergey Drobinsky¹, Matías de la Fuente², Behrus Puladi^{3

4}, Klaus Radermacher²

Affiliations

¹ Chair of Medical Engineering, RWTH Aachen University, Aachen, Germany. drobinsky@hia.rwth-aachen.de.
² Chair of Medical Engineering, RWTH Aachen University, Aachen, Germany.
³ Department of Oral and Maxillofacial Surgery, University Hospital RWTH Aachen, Aachen, Germany.
⁴ Institute of Medical Informatics, University Hospital RWTH Aachen, Aachen, Germany.

Abstract

Purpose: In bone surgery specialties, like orthopedics, neurosurgery, and oral and maxillofacial surgery patient safety and treatment success depends on the accurate implementation of computer-based surgical plans. Unintentional plan deviations can result in long-term functional damage to the patient. With on-site teleoperation, the surgeon operates a slave robot with a physically-decoupled master device, while being directly present at the operation site. This allows the surgeon to perform surgical tasks with robotic accuracy, while always remaining in the control loop.

Methods: In this study the master- and slave-side accuracy of an on-site teleoperated miniature cooperative robot (minaroHD) is evaluated. Master-side accuracy is investigated in a user study regarding scale factor, target feed rate, movement direction and haptic guidance stiffness. Scale factors are chosen to correspond to primarily finger, hand, and arm movements. Slave-side accuracy is investigated in autonomous milling trials regarding stepover, feed rate, movement direction, and material density.

Results: Master-side user input errors increase with increasing target feed rate and scale factor, and decrease with increasing haptic guidance stiffness. Resulting slave-side errors decrease with increasing scale factor and are < 0.07 mm for optimal guidance parameters. Slave-side robot position errors correlate with the feed rate but show little correlation with stepover distance. For optimal milling parameters, the 95th percentile of tracked slave-side position error is 0.086 mm with a maximal error of 0.16 mm.

Conclusion: For optimal guidance and milling parameters, the combined error of 0.23 mm is in the range of the dura mater thickness (< 0.27 mm) or mandibular canal wall (~ 0.85 mm). This corresponds to safety margins in high-demand surgical procedures like craniotomies, laminectomies, or decortication of the jaw. However, for further clinical translation, the performance and usability of on-site teleoperated milling must be further evaluated for real-life clinical application examples with consideration of all error sources in a computer-assisted surgery workflow.

Keywords: Cooperative robotic assistants; Haptic assistance; Virtual fixtures.