Real-time speckle variance swept-source optical coherence tomography using a graphics processing unit.

Area of Science:

• Biomedical Optics
• Medical Imaging Technology
• Optical Coherence Tomography

Background:

• Swept-source optical coherence tomography (SS-OCT) systems are advancing rapidly due to faster laser technologies.
• High-speed imaging is crucial for microvascular detection using speckle variance (SV) but is limited by computational demands.
• Interframe motion in SS-OCT and SV imaging can lead to artifacts and reduced vascular contrast.

Purpose of the Study:

• To develop a high-speed SS-OCT platform capable of real-time data acquisition, processing, display, and saving.
• To overcome computational bottlenecks in SS-OCT and SV analysis for improved microvascular imaging.
• To demonstrate the system's effectiveness in challenging imaging conditions with significant tissue motion.

Main Methods:

• Utilized a commercial graphics processing unit (GPU) optimized for parallel processing.
• Implemented real-time data acquisition, processing, and display at 108,000 lines per second.
• Incorporated subpixel image registration for structural images before SV calculations to mitigate motion artifacts.

Main Results:

• Achieved a complete high-speed SS-OCT platform operating at 108,000 lines/sec in real-time.
• Successfully reduced decorrelation from bulk tissue motion using real-time subpixel image registration.
• Demonstrated the system's viability in high-motion scenarios, displaying SV images (512x512, n=4) at 54 frames per second in human fingernail root imaging.

Conclusions:

• The developed high-speed SS-OCT platform effectively addresses computational limitations for real-time microvascular imaging.
• Real-time image registration significantly improves vascular contrast and reduces artifacts in high-motion environments.
• This technology holds promise for advanced in vivo imaging applications, particularly in challenging biological tissues.