Real-time Deformation-aware Control for Autonomous Robotic Subretinal Injection under iOCT Guidance (Ex-vivo Porcine Eye Experiments)

Core Concepts

This research paper introduces a novel method for autonomous robotic subretinal injection that dynamically adapts to tissue deformation during the procedure, improving needle positioning accuracy and injection success rate in ex-vivo porcine eyes.

Abstract

• Bibliographic Information: Arikan, D., Zhang, P., Sommersperger, M., Dehghani, S., Esfandiari, M., Taylor, R. H., Nasseri, M. A., Gehlbach, P., Navab, N., & Iordachita, I. (2024). Real-time Deformation-aware Control for Autonomous Robotic Subretinal Injection under iOCT Guidance. arXiv preprint arXiv:2411.06557.
• Research Objective: To develop a real-time, deformation-aware control system for autonomous robotic subretinal injections that enhances needle positioning accuracy and injection success rate.
• Methodology: The researchers developed a system that utilizes intraoperative optical coherence tomography (iOCT) to guide a robotic arm during subretinal injections. The system segments iOCT images in real-time to identify the needle, retinal layers (ILM and RPE), and calculates the needle tip's position relative to a virtual target layer. This virtual layer dynamically adjusts based on tissue deformation, ensuring accurate needle placement even with tissue movement. The system was tested on ex-vivo porcine eyes and compared to a fixed-point targeting method.
• Key Findings: The proposed method demonstrated superior performance compared to fixed-point targeting, achieving a 100% success rate in creating subretinal blebs (fluid accumulation indicating successful injection) compared to a 35% success rate with the conventional approach. The average error in needle tip positioning relative to the virtual target layer was 29µm.
• Main Conclusions: Real-time, deformation-aware control is crucial for accurate and successful robotic subretinal injections. The proposed method using a virtual target layer effectively compensates for tissue deformation, improving needle placement accuracy and injection outcomes.
• Significance: This research significantly contributes to advancing autonomous robotic surgery in ophthalmology. The developed method has the potential to improve the safety and efficacy of subretinal injections, which are crucial for treating various retinal diseases.
• Limitations and Future Research: The study was conducted on ex-vivo porcine eyes, which may not fully represent the complexities of in-vivo human eyes. Future research should focus on validating the system's performance in in-vivo settings and addressing limitations related to tissue bounce-back effects after needle penetration.

Stats

Subretinal injections are performed by inserting a microsurgical cannula through the retinal surface (ILM) and guiding it between the photoreceptor cells and the RPE.
For optimal therapeutic results, a bleb is formed by injecting fluid, creating a localized separation between the ILM and RPE.
Human retinal thickness is on average 250 µm.
Surgeon hand tremor is around 180 µm.
The study achieved an average axial error of 29µm between the target and the needle tip using the proposed method.
The fixed-point targeting method resulted in a 35% success rate in subretinal bleb generation.
The proposed method reliably and robustly created subretinal blebs in 100% of the experiments.

Quotes

"Due to the fragile and non-regenerative nature of the photoreceptor and RPE cells, any tissue damage can cause irreversible harm to the patient’s eyesight."
"Robotic systems have been developed for ophthalmic applications, which are not affected by limitations of human dexterity and enable precise and repeatable needle control inside the eye."
"To address this challenge, we propose a real-time, OCT-based feedback system."
"Our method utilizes OCT B5-scans, which we define as five densely sampled 2D OCT B-scans generating small 3D OCT volumes."
"This dynamic targeting method adapts to tissue deformation during the insertion procedure by continuously updating the target distance based on the current retinal conformation."

Key Insights Distilled From

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

by Demir Arikan... at arxiv.org 11-12-2024

https://arxiv.org/pdf/2411.06557.pdf

Deeper Inquiries

How can this robotic system be adapted for other delicate surgical procedures beyond ophthalmology?

This robotic system, with its core principles of real-time deformation-aware control and iOCT guidance, holds significant promise for adaptation to other delicate surgical procedures beyond ophthalmology. Here's how:
1. Transferability of Core Principles:

Deformation-Aware Control: The fundamental concept of dynamically adjusting the surgical tool's trajectory based on real-time tissue deformation feedback is applicable across various surgical domains. Whether it's neurosurgery, cardiac surgery, or microsurgery, accounting for tissue movement during the procedure is crucial for precision and safety.
Image-Guided Surgery with iOCT: While iOCT is well-established in ophthalmology, its use is expanding to other fields. The high-resolution, real-time imaging capability of iOCT can be valuable in procedures requiring precise tissue differentiation and depth perception, such as in brain or spinal cord surgery.
2. Adapting to Specific Surgical Needs:

Surgical Tool Modification: The robotic arm and needle used in this system can be replaced with specialized tools tailored to the specific surgical procedure. For instance, micro-grippers, biopsy needles, or laser probes could be integrated.
Imaging Modality Integration: Depending on the target tissue and desired imaging depth, other imaging modalities like ultrasound or MRI could be incorporated alongside or instead of iOCT.
Control Algorithm Refinement: The control algorithms would need to be adapted and fine-tuned based on the specific biomechanical properties of the target tissue and the objectives of the surgical procedure.
3. Examples of Potential Applications:

Neurosurgery: Precise targeting of deep brain structures for biopsies or drug delivery, minimizing damage to surrounding brain tissue.
Cardiac Surgery: Performing minimally invasive procedures on beating hearts, where real-time motion compensation is critical.
Fetal Surgery: Enabling delicate in-utero interventions requiring extreme precision and minimal invasiveness.
4. Challenges and Considerations:

Tissue-Specific Imaging: Optimizing imaging modalities and algorithms for clear visualization of target structures in different tissue types.
Biomechanical Modeling: Developing accurate biomechanical models of target tissues to predict and compensate for deformation effectively.
Regulatory Approval: Navigating the regulatory landscape for medical devices and obtaining necessary approvals for new applications.

Could the reliance on ex-vivo models have skewed the results by not fully replicating the dynamic biomechanical properties of living tissue?

Yes, the reliance on ex-vivo porcine eye models could have potentially skewed the results by not fully replicating the dynamic biomechanical properties of living tissue. Here's why:
1. Differences in Biomechanical Properties:

Tissue Viability: Ex-vivo tissues lack blood flow, oxygenation, and active cellular processes, leading to changes in tissue stiffness, elasticity, and overall biomechanical behavior compared to living tissue.
Intraocular Pressure (IOP): Ex-vivo eyes lack the natural IOP present in living eyes, which influences the shape and tension of the retina, potentially affecting needle insertion dynamics.
Surgical Response: Living tissues exhibit dynamic responses to surgical manipulation, such as bleeding, inflammation, and wound healing, which are absent in ex-vivo models.
2. Potential Impact on Results:

Exaggerated Deformation: Ex-vivo tissues might exhibit exaggerated deformation or altered force feedback during needle insertion compared to living tissues, potentially influencing the robot's control algorithms.
Bounce-Back Effect: The study acknowledges the "bounce-back" effect as being more pronounced in ex-vivo eyes. This suggests that the tissue response observed in the experiments might not directly translate to in-vivo scenarios.
Bleb Formation: The characteristics and stability of blebs created in ex-vivo eyes might differ from those in living eyes due to the absence of physiological factors like intraocular pressure and fluid dynamics.
3. Importance of In-Vivo Validation:

Clinical Relevance: To ensure the system's safety and efficacy in real-world surgical settings, rigorous in-vivo validation is essential.
Fine-Tuning and Optimization: In-vivo studies will allow for fine-tuning the control algorithms, validating the accuracy of tissue deformation models, and optimizing the system's performance under realistic physiological conditions.
4. Addressing Limitations of Ex-Vivo Models:

Ex-Vivo Model Refinement: Researchers are continually working on improving ex-vivo models to better mimic living tissue properties. This includes using fresh tissues, maintaining physiological temperature, and simulating IOP.
Computational Modeling: Developing sophisticated computational models that incorporate the biomechanical properties of living tissues can help bridge the gap between ex-vivo experiments and in-vivo applications.

What ethical considerations arise from developing increasingly autonomous surgical robots, and how can they be addressed responsibly?

The development of increasingly autonomous surgical robots, while promising, raises significant ethical considerations that must be addressed responsibly:
1. Autonomy and Human Control:

Level of Autonomy: Defining appropriate levels of autonomy for surgical robots is crucial. Striking a balance between robotic assistance and surgeon control is essential to ensure patient safety and maintain the surgeon's role as the ultimate decision-maker.
Unexpected Situations: Establishing clear protocols for handling unexpected situations or complications during surgery is critical. The robot's ability to recognize and respond appropriately to unforeseen events, while allowing for seamless transition to manual control, is paramount.
2. Safety and Risk Mitigation:

Rigorous Testing and Validation: Comprehensive pre-clinical testing, including extensive simulations and in-vivo studies, is crucial to ensure the safety and reliability of autonomous surgical robots before clinical deployment.
Fail-Safe Mechanisms: Implementing robust fail-safe mechanisms and redundancy in the system's design is essential to mitigate risks and prevent harm to patients in case of technical malfunctions or errors.
3. Informed Consent and Patient Autonomy:

Transparency and Communication: Patients must be fully informed about the capabilities and limitations of autonomous surgical robots, including the level of autonomy involved in their procedure. Clear and transparent communication about potential risks and benefits is essential.
Patient Choice: Patients should have the right to choose between a surgeon-performed procedure or one involving an autonomous robot, even if the robotic option is deemed clinically superior.
4. Data Privacy and Security:

Patient Data Protection: Autonomous surgical robots often collect and analyze vast amounts of patient data. Ensuring the privacy and security of this data is paramount. Implementing robust data encryption, secure storage, and strict access controls is essential.
Algorithmic Bias: Addressing potential biases in the algorithms that control autonomous robots is crucial. These algorithms should be trained on diverse datasets and rigorously tested to ensure fairness and equitable outcomes for all patients.
5. Access and Equity:

Affordability and Availability: Efforts should be made to ensure that the benefits of autonomous surgical robotics are accessible to all patients, regardless of their socioeconomic status or geographical location.
Training and Education: Adequate training and education programs for surgeons and healthcare professionals are essential to ensure the safe and ethical implementation of autonomous surgical robots in clinical practice.
Addressing These Considerations Responsibly:

Interdisciplinary Collaboration: Fostering collaboration among engineers, surgeons, ethicists, regulators, and patient advocacy groups is crucial to address these ethical considerations comprehensively.
Ethical Guidelines and Regulations: Developing clear ethical guidelines and regulations for the development and deployment of autonomous surgical robots is essential. These guidelines should prioritize patient safety, autonomy, and well-being.
Continuous Monitoring and Evaluation: Ongoing monitoring and evaluation of the safety, efficacy, and ethical implications of autonomous surgical robots are necessary to ensure responsible innovation in this rapidly evolving field.