Real-time Deformation-aware Control for Autonomous Robotic Subretinal Injection under iOCT Guidance

2024-11-10

arXiv.org

AI Summary

In autonomous robotic subretinal injection, tissue deformation induces needle-tip localization errors and misalignment with the target retinal layer. Method: This paper proposes an iOCT-guided, deformation-aware closed-loop control framework. It achieves millisecond-scale tracking of instrument-tissue relative pose via dense B5-scan sampling, real-time segmentation, dynamic virtual targeting-layer modeling between the ILM and RPE, and 3D scene reconstruction. It introduces, for the first time, deformation-compensated motion control based on virtual anatomical layers during subretinal injection, enabling adaptive adjustment of insertion depth. Results: In ex vivo porcine eye experiments, subretinal bleb formation success increased from 35% to 100%, with significant improvements in needle-tip localization accuracy and system robustness. This work represents the first integration of real-time deformation sensing and virtual-layer-based closed-loop targeting within an autonomous ophthalmic surgical system, establishing a new paradigm for high-precision, minimally invasive intraocular interventions.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Computer Vision: Vision for Robotics & Autonomous Driving
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation

Abstract

Robotic platforms provide repeatable and precise tool positioning that significantly enhances retinal microsurgery. Integration of such systems with intraoperative optical coherence tomography (iOCT) enables image-guided robotic interventions, allowing to autonomously perform advanced treatment possibilities, such as injecting therapeutic agents into the subretinal space. Yet, tissue deformations due to tool-tissue interactions are a major challenge in autonomous iOCT-guided robotic subretinal injection, impacting correct needle positioning and, thus, the outcome of the procedure. This paper presents a novel method for autonomous subretinal injection under iOCT guidance that considers tissue deformations during the insertion procedure. This is achieved through real-time segmentation and 3D reconstruction of the surgical scene from densely sampled iOCT B-scans, which we refer to as B5-scans, to monitor the positioning of the instrument regarding a virtual target layer defined at a relative position between the ILM and RPE. Our experiments on ex-vivo porcine eyes demonstrate dynamic adjustment of the insertion depth and overall improved accuracy in needle positioning compared to previous autonomous insertion approaches. Compared to a 35% success rate in subretinal bleb generation with previous approaches, our proposed method reliably and robustly created subretinal blebs in all our experiments.

Problem

• Autonomous robotic subretinal injection under iOCT guidance.
• Addressing tissue deformations during needle insertion.
• Improving accuracy and success rate of subretinal bleb generation.

Innovation

• Real-time tissue deformation-aware robotic control
• iOCT-guided 3D surgical scene reconstruction
• Dynamic needle depth adjustment for precision