Real-time display with large field of view on fourier domain optical coherence tomography at 1310 nm wavelength for dermatology

Abstract

A Fourier domain optical coherence tomography (OCT) system with 1310 nm light was demon strated to study inflammatory human skin and the skin coated with a moisturizer in vivo. By using a graphics processing unit (GPU), the display rate could reach 20 frames/s with 1000 A scans contained in one image. The field of view (FOV) of the cross sectional image is 7 mm in the lateral direction and the penetration depth is ~1 mm in skin. The result shows that, in inflammatory skin, the epidermis became thicker and had a decreased scattering; furthermore, the region of the severe lesion present an uneven thickness of the epider mis compared with the peripheral area. For the result of a finger tip coated with the moisturizer, the anti reflection effect was significant and the stratum corneum became more transparent. In this letter, we demon strated that real time display with a large FOV could enable screening of a large tissue area; thereby increasing the dermatologic diagnostic potential of the method by permitting a comparison of the lesion and the normal peripheral region.

Optimization of compute unified device architecture for real-time ultrahigh-resolution optical coherence tomography

“…In addition, if the acquisition speed is relatively slow, motion artifacts could degrade the reconstructed image. To prevent reduction in the frame rate during large area scanning [22], we fixed the A-lines and increased the step size of the galvanometer. We used full speed spectrum acquisition of the line scan camera, which had a 6.96 μs exposure time with a nominal focal plane array sensitivity of 450 eÀ /counts.…”

Adaptive optimization of reference intensity for optical coherence imaging using galvanometric mirror tilting method

“…We programmed hardware control routine such as beam scanning, triggering, data storing, and camera control with LabVIEW. Core OCT signal processing including IFFT, cubic spline interpolation, and fixed pattern noise reduction was executed on graphic processing unit by loading functions from dynamic link library written in C þ þ with compute unified device architecture (CUDA) [14][15][16][17]. CUDA enabled realtime image reconstruction with handling the full speed measurement of our SD-OCT system [18].…”

Graphics processing unit aided highly stable real-time spectral-domain optical coherence tomography at 1375 nm based on dual-coupled-line subtraction

“…Data acquisition and hardware control were performed by a LabVIEW-based custom program on a desktop computer with an Intel i7-950 central processing unit. Principle signal processing for image reconstruction was executed by a GTX460 graphics processing unit (GPU) manufactured by nVidia [23].…”