Robotics and Optical Coherence Tomography: Current Works and Future Perspectives

Automation Optics Computer Science Information Science Medical Imaging Life Sciences Physics Bioengineering Medicine
optical coherence tomography robotics medical imaging intrasurgical visualization system integration high-resolution imaging biomedical engineering

The Combination of Optical Coherence Tomography and Robotics: Current Research and Future Perspectives

Academic Background

Optical Coherence Tomography (OCT) is a non-invasive, high-resolution optical imaging technique that has been widely used in the field of biomedical imaging since its invention. It provides micrometer-level visualization of tissue structures, achieving significant success in ophthalmology, such as imaging and diagnosing diseases in corneal and retinal tissues. However, traditional OCT devices are typically designed for imaging in static environments, restricted by size, field of view (FOV), and operational flexibility. When applied in dynamic and complex medical scenarios or surgical procedures, these limitations of traditional OCT equipment become more evident, such as the inability to adapt to the movement of surgical targets or the difficulty in providing real-time, high-resolution imaging for guiding surgical operations.

Meanwhile, the rapid development of medical robotics offers new possibilities for further integration of OCT. With their high precision and multi-degree-of-freedom operational capabilities, medical robots have established themselves as indispensable tools for surgical procedures, operational assistance, and diagnostics. By combining robotics with OCT, this interdisciplinary integration leverages the flexible control capabilities of robots for large-scale three-dimensional scanning, while the imaging capabilities of OCT provide high-resolution perception. Together, they aim to overcome the limitations of traditional medical imaging and operational systems.

This review paper focuses on the integration of OCT with robotics, systematically exploring the latest advancements and technological challenges in this field. The goal of the paper is to review existing research, analyze the current state of integration technologies and application cases, and propose future research directions.

Paper Source

The paper, titled “Robotics and Optical Coherence Tomography: Current Works and Future Perspectives,” was published in the Biomedical Optics Express (Volume 16, Issue 2, February 2025). The main authors include Guangshen Ma, Morgan McCloud, Yuan Tian, Amit Narawane, Harvey Shi, Robert Trout, Ryan P. McNabb, Anthony N. Kuo, and Mark Draelos. They are affiliated with the Department of Robotics at the University of Michigan, the Department of Biomedical Engineering at Duke University, the Duke University Medical Center in Ophthalmology, and the Department of Ophthalmology and Visual Sciences at the University of Michigan Medical School. This paper, written in a review format, covers the current state and future directions of the integration of OCT with medical robotics.

Key Content

1. The Disciplinary Background and Unique Value of OCT and Robotics

The paper begins by reviewing the fundamental principles of OCT and its core advantages in biomedical imaging. OCT is based on the principle of low-coherence interferometry, utilizing the coherent signals between the sample arm and the reference arm of the light beam to produce depth-resolved tomographic images. Its intrinsic features include high spatial resolution, non-invasiveness, high-speed scanning capability, and penetration into biological tissue. These characteristics make OCT particularly suitable for tissue imaging, pathology detection, and intraoperative guidance.

Compared to traditional 3D perception systems, such as camera-based stereoscopic imaging or structured light imaging, OCT not only provides surface “2.5D” depth maps but also generates true three-dimensional volumetric data. In contrast to other medical imaging methods like MRI and CT, which, although useful for surgical navigation, are bulky and complex to operate, OCT’s compact design and high resolution make it more suitable for certain specialized surgical needs.

On the other hand, robotics has shown great potential in the medical field in recent years. Robotic systems, with their mechanical arms, control units, and real-time motion planning capabilities, have become reliable operators in complex surgical scenarios. Integrating OCT with robotic systems is expected to expand the operational range of OCT and enable multi-perspective observation of dynamic targets, effectively meeting the precision and real-time requirements of modern medical operations.

2. Four Typical Configurations of Robotics-OCT Systems

The paper categorizes robotic OCT integrated systems into four main configurations, each with distinct designs and application goals.

(1) Robot-Adjacent OCT

In this configuration, the OCT device is fixed on a tabletop, while the robot is responsible for controlling medical tools. For instance, a tabletop OCT provides imaging feedback to guide a robot for precise tool insertion or tissue manipulation. Needle insertion systems are particularly suitable for minimally invasive eye surgeries, such as Deep Anterior Lamellar Keratoplasty (DALK), retinal vein cannulation, and intraoperative drug injections. These systems use real-time B-scan or C-scan data, integrated with robotic control algorithms, to enable complete workflows from tissue segmentation and needle trajectory planning to precise operation.

Researchers have developed computer vision algorithms and deep learning models for tool and tissue localization, segmentation, and trajectory tracking. For example, utilizing the U-Net deep learning framework for real-time segmentation of OCT volumetric images to extract the relative positions between tools and target tissues and generate insertion paths. Furthermore, this configuration faces challenges such as a small surgical field of view, which can be improved through FOV expansion algorithms or multi-view stitching techniques.

(2) Robot-Mounted OCT

In this configuration, the OCT sensor is directly mounted on the robot’s end-effector for target scanning in dynamic environments. The advantage of this approach is that the robot can move the OCT sensor to change scanning perspectives or increase the scanning range. For example, some studies demonstrated the use of 6-degree-of-freedom (DOF) or 7-DOF robotic systems in conjunction with OCT for applications like ophthalmic scanning and large-area detection of organs (e.g., kidney imaging). Moreover, data mosaic techniques in this configuration allow for capturing multi-perspective OCT images from multiple positions and reconstructing high-resolution global images.

(3) Roboticized OCT Tools

This configuration integrates OCT sensors directly into medical tools, such as surgical needles or forceps, which are manipulated by robots or handheld devices for surgical operations. Such designs are often applied in spatially constrained minimally invasive surgeries or highly precise detection tasks. Researchers have developed miniaturized OCT sensors and small flexible robotic arms for guiding laser surgeries in ophthalmology and neurosurgery.

(4) Robot Endoscopic OCT

Endoscopic OCT is specifically designed for complex or narrow environments, such as the digestive tract, blood vessels, or posterior chambers of the eye. It combines flexible robotic arms with miniaturized OCT probes. The paper introduces innovative designs, such as piezoelectric-driven OCT endoscopes and robotic OCT neuroendoscopes for brain tumor detection. These technologies significantly enhance imaging efficiency and safety in minimally invasive surgeries.

3. Supporting Technologies Driving Innovation

The paper highlights key technologies that promote the integration of OCT and robotics:

  • Scan Optimization: Redesigning scanning paths or dynamically adjusting scan patterns to improve imaging efficiency. Some studies have developed robotics-based scanning strategies to effectively reduce redundant data collection.

  • Refraction Correction: Distortions in OCT images caused by light refraction at different media interfaces need to be corrected to provide accurate geometric reconstructions of surgical scenes. Real-time optical path correction algorithms are essential for such applications.

  • Machine Learning Models: Deep learning plays a crucial role in the segmentation, target detection, and motion planning of OCT images. For example, convolutional neural networks (CNNs) have been used to automatically classify OCT data and compute tissue boundaries.

4. Future Directions

The paper suggests that future robotic OCT research can develop in the following directions:

  1. Mobile Robot Integration: Developing OCT systems based on drones, wheeled and humanoid robots to expand application scenarios, such as dermatological-wide area scanning or point-to-point diagnostics.

  2. Full-Field OCT Applications: Parallel detection techniques, such as full-field OCT, could significantly improve data acquisition speed, breaking current bottlenecks in imaging rates.

  3. Interdisciplinary Collaboration: Fostering further collaboration between the fields of robotics, optics, and medicine to develop more efficient and intelligent OCT-robot integrated systems.

Significance and Value of the Paper

This review summarizes the technical innovations and application breakthroughs brought by the integration of OCT and robotics. From an academic perspective, it showcases OCT’s potential in micrometer-scale imaging while revealing the flexible operational capabilities provided by robotics. This interdisciplinary research not only has direct value in medical scenarios such as ophthalmology, neurosurgery, and gastrointestinal endoscopy but also inspires new image-guidance technologies in the field of biomedical engineering.