July 2010
Spotlight Summary by Robert J. Zawadzki
Multi-Megahertz OCT: High quality 3D imaging at 20 million A-scans and 4.5 GVoxels per second
In this latest paper from Robert Huber’s group, Wolfgang Wieser describes an optical coherence tomography (OCT) instrument with unprecedented acquisition speed—about 100 times faster than most other competing OCT designs. It is also capable of acquiring high-quality OCT B-scans and volumes. Although this was achieved by optimization of the OCT system components, it clearly sets a new benchmark for usable OCT acquisition speeds. Its importance can be compared to progress achieved by the introduction of Fourier-domain (Fd) detection to OCT. Several factors are presented regarding a Fd mode-locked (FDML) laser and detection channel to allow Multi-Megahertz OCT.
It’s worth noting that the paper is well crafted and includes all the necessary details for the non-expert to understand without referring to other literature. In the introduction, the authors review ultra-high-speed OCT acquisition instruments. Separate paragraphs clarify key parameters of every Fd-OCT system: its sensitivity and roll-off, dynamic range, axial resolution, and imaging speed, with special focus on scan, pixel, and voxel rates. Because the acquisition speed limits available electronics, special consideration is devoted to assessing requirements of analog-to-digital converter (ADC) bit depth on OCT image quality. Interestingly, the authors prove that an effective bit depth above 7 is suitable for OCT imaging. Three FDML laser sources were used, and each of them is described in detail so that the reader can appreciate the many technological challenges and choices made by the authors. These include an ultra-high-speed bulk optics Fabry- Perot filter and buffering and dual output configuration. The custom design four-channel detector is also described. The design for a novel multispot interferometer that allows multibeam scanning and simultaneous data acquisition is followed by examples of multispot scanning optics approaches. Finally, the authors present their 3D data-acquisition procedure, removal of bidirectional scanning artifacts and merging data sets from multiple spots. In the Results section, one finds details pertaining to sensitivity roll-off performance of all three tested FDML Lasers. Several examples of high-quality OCT B-scans and volumes acquired on human skin in vivo as well as cellular structures of kiwi and cucumber allow better appreciation of the technological advancement reported in this paper. Volume averaging as a method to reduce speckle contrast is illustrated.
In summary, the Multi-Megahertz OCT instrument presented in this paper, despite some limitations—including its predefined spectral window, beam power, and small size of data storage—may open new doors for OCT technology. Its impressive acquisition speed may make new OCT applications possible. As a bonus, several useful metrics are presented that can be directly implemented to characterize performance of any OCT instrument. These metrics may allow new OCT developers to characterize their systems and to identify some limiting factors that are often overlooked.
You must log in to add comments.
It’s worth noting that the paper is well crafted and includes all the necessary details for the non-expert to understand without referring to other literature. In the introduction, the authors review ultra-high-speed OCT acquisition instruments. Separate paragraphs clarify key parameters of every Fd-OCT system: its sensitivity and roll-off, dynamic range, axial resolution, and imaging speed, with special focus on scan, pixel, and voxel rates. Because the acquisition speed limits available electronics, special consideration is devoted to assessing requirements of analog-to-digital converter (ADC) bit depth on OCT image quality. Interestingly, the authors prove that an effective bit depth above 7 is suitable for OCT imaging. Three FDML laser sources were used, and each of them is described in detail so that the reader can appreciate the many technological challenges and choices made by the authors. These include an ultra-high-speed bulk optics Fabry- Perot filter and buffering and dual output configuration. The custom design four-channel detector is also described. The design for a novel multispot interferometer that allows multibeam scanning and simultaneous data acquisition is followed by examples of multispot scanning optics approaches. Finally, the authors present their 3D data-acquisition procedure, removal of bidirectional scanning artifacts and merging data sets from multiple spots. In the Results section, one finds details pertaining to sensitivity roll-off performance of all three tested FDML Lasers. Several examples of high-quality OCT B-scans and volumes acquired on human skin in vivo as well as cellular structures of kiwi and cucumber allow better appreciation of the technological advancement reported in this paper. Volume averaging as a method to reduce speckle contrast is illustrated.
In summary, the Multi-Megahertz OCT instrument presented in this paper, despite some limitations—including its predefined spectral window, beam power, and small size of data storage—may open new doors for OCT technology. Its impressive acquisition speed may make new OCT applications possible. As a bonus, several useful metrics are presented that can be directly implemented to characterize performance of any OCT instrument. These metrics may allow new OCT developers to characterize their systems and to identify some limiting factors that are often overlooked.
Add Comment
You must log in to add comments.
Article Information
Multi-Megahertz OCT: High quality 3D imaging at 20 million A-scans and 4.5 GVoxels per second
Wolfgang Wieser, Benjamin R. Biedermann, Thomas Klein, Christoph M. Eigenwillig, and Robert Huber
Opt. Express 18(14) 14685-14704 (2010) View: Abstract | HTML | PDF