Data-Driven Modelling and Control for Robot Needle Insertion in Deep Anterior Lamellar Keratoplasty.

Area of Science:

• Ophthalmology
• Robotics
• Biomedical Engineering

Background:

• Deep anterior lamellar keratoplasty (DALK) is a cornea transplantation technique with reduced patient morbidity.
• DALK presents challenges for human surgeons due to small scales, fine control needs, and visualization difficulties.
• Accurate needle insertion in DALK is critical for reliable clinical outcomes.

Purpose of the Study:

• To develop a data-driven model for tool-tissue interaction in DALK.
• To create a model predictive controller for robot-assisted needle insertion in DALK.
• To improve the accuracy of needle placement during robot microsurgery for DALK.

Main Methods:

• Developed a data-driven autoregressive dynamic model of surgical tool-cornea tissue interaction.
• Implemented a model predictive controller to guide robot needle insertion.
• Evaluated the controller's performance in an ex vivo model.

Main Results:

• The developed controller significantly improved needle positioning accuracy by over 40% compared to previous methods.
• The data-driven model effectively captured small-scale tool-tissue interactions.
• Enhanced accuracy in needle placement was achieved in the ex vivo DALK model.

Conclusions:

• Robot microsurgery, guided by advanced modeling and control, can significantly enhance DALK precision.
• The developed approach offers a promising solution for improving surgical outcomes in cornea transplantation.
• This work paves the way for more accurate and reliable robot-assisted ophthalmic microsurgery.