OCT as both a shape sensor and a tomographic imager for large-scale freeform robotic scanning

Area of Science:

• Biomedical Engineering
• Robotics
• Optical Imaging

Background:

• Optical coherence tomography (OCT) has limitations in imaging large-scale, freeform objects due to artifacts and surface irregularities.
• Existing methods struggle to achieve high-resolution, large-scale tomographic imaging of complex geometries.

Purpose of the Study:

• To develop a novel methodological framework for large-scale freeform object imaging using OCT.
• To enhance OCT's capabilities by integrating robotic scanning with advanced algorithms.
• To overcome limitations of traditional OCT in complex object characterization.

Main Methods:

• Utilized OCT as both a shape sensor and a tomographic imager within a robotic scanning system.
• Integrated a deep-learning-based surface detection algorithm to mitigate OCT artifacts.
• Employed an adaptive robotic arm pose adjustment algorithm for uneven object sensing and imaging.

Main Results:

• Demonstrated a robust framework for high-resolution, large-scale tomographic imaging of diverse objects.
• Successfully managed OCT artifacts and surface irregularities in complex freeform object scans.
• Achieved superior imaging performance compared to conventional OCT methods.

Conclusions:

• The proposed robotic OCT scanning framework significantly expands OCT's applicability for large-scale, freeform object imaging.
• This method offers a promising solution for both medical and industrial applications requiring detailed 3D structural information.
• The integration of AI and robotics enhances OCT's precision and adaptability in challenging imaging scenarios.