TL;DR
This paper introduces a tele-operated robotic system (TOOS) designed for safe and effective oropharyngeal (OP) swab sampling, a critical step in COVID-19 diagnosis. The system features a novel soft robotic hand (TSS hand) that mimics human dexterity and a hybrid virtual fixture controller to ensure safe force application, addressing the high infection risk faced by healthcare workers during manual sampling.
What Problem Does This Paper Address?
The COVID-19 pandemic highlighted the critical need for rapid and accurate viral testing, with oropharyngeal swab (OP SWAB) sampling being a primary method for collecting viral nucleic acid (VNA) specimens. However, manual OP swab collection poses a significant infection risk to healthcare professionals due to close-contact exposure. While robotic solutions exist, they often suffer from safety concerns (due to rigid components in delicate oral cavities), high costs, and complex control, hindering widespread adoption. This paper aims to bridge this gap by developing a safe, cost-effective, and dexterous robotic system that leverages soft robotics to perform OP swab sampling with minimal risk to medical staff.
Methods
The TOOS robot is composed of two main components:
- TStone Soft (TSS) Hand: This novel soft robotic hand is bio-inspired by human hand movements and designed for OP swab manipulation. It consists of:
- Soft Wrist: A multi-backbone continuum soft robot capable of omnidirectional bending and adjustable stiffness. This allows for agile movements within the confined oral space and can be tuned for safety or forceful interaction.
- Soft Gripper: A hybrid soft-rigid structure designed to securely pinch and release disposable swabs, facilitating easy replacement.
- 3-DOF RCM Platform: This platform acts as the "arm" of the robot, providing global positioning of the TSS hand to bring the swab to the correct location within the patient's mouth.
- Teleoperation Framework: The system utilizes a teleoperation setup where a medical professional (operator) controls the robot from a separate room. This involves:
- Visual Feedback: A camera mounted on the TSS hand provides real-time video of the sampling site.
- Haptic Feedback: A Geomagic touch device provides haptic feedback to the operator, allowing them to "feel" the interaction forces.
- Hybrid Motion/Stiffness Virtual Fixture (H-VF) Controller: This is a key innovation. It combines motion constraints (to keep the swab within safe boundaries in the X-Y plane) and stiffness constraints (to limit force application, especially in the Z direction) to ensure the sampling force remains within a comfortable and safe range (around 60g).
Datasets
Not discussed in the paper. The experiments involved a nasopharyngology model and an oral cavity model for testing.
Key Insights and Ideas
- Soft Robotics for Oral Cavity: The paper strongly advocates for soft robotics as a superior approach for tasks involving delicate human anatomy like the oral cavity, emphasizing safety and compliance over rigid robotics.
- Bio-inspiration for Dexterity: Designing the TSS hand based on human sampling observations led to a compact, dexterous end-effector that rivals human capabilities in this specific task.
- Hybrid Virtual Fixtures are Crucial: The H-VF controller is a critical component for ensuring safety. By combining motion and stiffness constraints, it effectively prevents excessive force application and movement, which is paramount in a sensitive area like the throat.
- Adjustable Stiffness for Versatility: The ability to adjust the wrist's stiffness is a significant advantage, allowing the system to adapt to different sampling requirements and maintain stability.
- Teleoperation with Haptic Feedback Enhances Intuitiveness: The combination of visual and haptic feedback makes the teleoperation intuitive, allowing operators to quickly adapt and perform the task effectively.
Key Results
- Dexterity and Workspace: The TSS hand demonstrated comparable dexterity and workspace to a human hand, facilitating intuitive teleoperation.
- Teleoperation Accuracy: The teleoperation system achieved an acceptable accuracy with root mean square errors of approximately 1-2 mm across axes.
- Safe Sampling Force: The H-VF controller effectively kept the sampling force within the comfortable range (0-60g), preventing excessive force application.
- High Sampling Success Rate: The system achieved a 100% success rate in collecting samples from a phantom model, confirming the effectiveness of the vision-guided teleoperation and dexterous hand.
- Durability: The prototype TSS hand showed good durability after 50 sampling cycles and maintained its performance over two months.
- User Acceptance: Human users, including those with no prior OP sampling experience, found the system intuitive and the VF constraint acceptable.
- Sampling Time: The robotic system achieved an OP sampling time of around 30 seconds, comparable to manual methods.
Limitations and Future Work
- Limitations:
- The current H-VF system relies on approximate registration between the testee and robot, which is compensated by the soft hand's compliance.
- The speed of insertion and extraction by the RCM platform is limited and could be enhanced.
- The paper acknowledges that while the current system is promising, further refinement is needed for widespread clinical adoption.
- Future Work:
- Seeking ethics approval for in-vivo testing.
- Further refining the TSS hand design.
- Integrating sensors for specimen contact detection and real-time oral configuration mapping.
- Improving the hybrid virtual fixture controller.
- Developing a dual-arm manipulation system.
- Exploring disposable shielding covers and UVC sterilization for enhanced disinfection.
- Integrating computer vision assistance and AI for human behavior learning to further improve teleoperation.
Summary updated