Extending depth range of optical coherence tomography

• Main Author: Li, Jinhan
• Other Authors: Liu Linbo
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2020
• Subjects: Engineering::Electrical and electronic engineering
• Online Access: https://hdl.handle.net/10356/137324
• Institution: Nanyang Technological University

Optical coherence tomography (OCT) is a noninvasive diagnostic imaging tool that has been established in various clinical fields. One of the most cited issue of OCT in clinical practices is the shortage of image depth range. This thesis proposes four novel techniques to significantly extend the depth range without compromising other imaging performances. The OCT depth range is fundamentally limited by two main factors, which are the axial field of view (FOV) and the system sensitivity. For standard OCT systems, the axial FOV can only be improved by decreasing the numerical aperture (NA) of the focusing optics, which unfortunately sacrifices the transverse resolution, and the sensitivity can be enhanced by increasing the incident power. To achieve a high transverse resolution over an extended depth range, variety of efforts had been made on hardware or digital refocusing approaches. For hardware approaches, such as wavefront modification and adaptive optics, the rise in system complexity is always a problem. Some of the approaches even suffer from severe intensity loss and sidelobe problems. As for the digital refocusing approaches, the high phase stability requirement and the rise in computational complexity are the two main roadblocks that remains unsolved. To properly solve the depth range problem for OCT, more efficient methods should be exploited. For B-mode imaging tasks, we changed the cross-sectional shape of the sample arm input beam from circular to elliptical. By aligning the major axis in the pupil plane with the scanning direction (X-axis), an isotropic spot was generated at the focal plane, which has higher resolution in the in-plane direction (X-axis) than that along the out-of-plane direction (Y-axis). Such an isotropic input beam as elliptical beam can mitigate the effects of the spherical aberration (including the defocus) along the Y-axis direction since the beam size is reduced along the Y-axis with respect to the circular beam producing the same in-plane transverse resolution. We considered two sample arm optical arrangements where defocus and primary spherical aberration caused significant degradation in the imaging quality, respectively. We proved both numerically and experimentally that elliptical beam can extend the axial FOV by around 50% pertaining to the defocus problem and restore the back-coupling efficiency by 25% pertaining to the primary spherical aberration problem, respectively. The proposed elliptical input beam technique can help to lower the cost of the sample arm optics since the aberration-corrected optics can be replaced by lower-cost alternatives without compromising the B-mode image quality. This method adds no system complexity to the conventional OCT system to achieve axial FOV extension and aberration mitigation. Meanwhile, this method is flexible and is promising of combining with other technologies, such as wavefront engineering techniques and digital refocusing, to further improve the performances. The elliptical input beam applies to all B-mode imaging applications, but it may not produce enough axial FOV for some of them. Given enough source spectral bandwidth, we propose to make use of the chromatics focal shift of the sample arm optics to further extend the axial FOV. In the fiber probe, the foci of the 750-1000 nm and 1100-1450 nm inputs were chromatically shifted axially. The interference signals from the two spectral bands were measured with a Si camera-based spectrometer and an InGaAs camera-based spectrometer, respectively. Therefore, two images of different focal depth are acquired simultaneously. By compounding the two acquired frames, the resulted shows an extended axial FOV which is twice as large as that of the single focus probe. The proposed method is superior over the existing axial FOV extension technologies as it can be achieved with commonly used fiber-optic OCT probes; therefore, it can be readily used for various applications. Furthermore, the proposed method experiences limited signal loss or sidelobe problems. Since the CCD with high responsivity over both the spectrum bands is un accessible, the limitation for this method is that a dual-spectrometer detection is nonavoidable, which adds to the system complexity and cost. An additional effort has been made to implement a digital refocusing technique, termed multiple aperture synthesis, in a fiber optical endoscopic probe to extend the OCT axial FOV. We designed and fabricated a novel fiber-optic probe, termed amplitude division aperture synthesis (ADAS) probe, in which a calcite beam displacer was used to create two sub-apertures. The optical aberrations in the depth domain were digitally corrected between the interferometric signals acquired through the two apertures using a pathlength encoded method. The feasibility of extending the axial FOV is demonstrated using tissue phantoms. As our approach collects all backscattered sample signal, simultaneously, the phase stability issue encountered by existing computational aberration correction methods is avoided. Which makes our design a promising way to implement in high resolution endoscopic diagnosis. In addition to the focusing engineering approach, I also seek to improve the detection sensitivity for depth range extension. Specifically, under the condition that there is excessive light source power such as the case of a supercontinuum source, a source-detector common path OCT was designed and developed, where the input light was coupled in the system via the 0-order of the spectrometer grating. We demonstrate that the source-detector common-path design is advantageous in sensitivity over the standard 50:50 fiber coupler implemented OCT by 3dB. The proposed method is superior over the existing OCT in three aspects. First, the proposed method enhances the system sensitivity, which clearly increases the image depth for high-resolution OCT. Second, a source-detection common-path OCT utilized the spectrometer grating as a beam splitter, which simplified the system structure and lower the cost. Finally, our design removes most of the bandwidth limiting optical components in the distal end, which is promising of pushing the OCT resolution one step further. Therefore, the source-detection common-path scheme is promising of implementing the highly sensitive ultrahigh-resolution OCT in the future. In conclusion, this thesis proposes four methods to significantly extend the OCT depth range while maintaining the state-of-art performance in other aspects, which have been demonstrated theoretically using analytical and numerical methods and experimentally with phantom imaging and biological tissue imaging ex vivo or in vivo.