Optical Coherence Tomography
Utilizing the advantages of non-invasive, fast volumetric imaging at micron-scale resolution with intrinsic contrast agents, Optical Coherence Tomography (OCT) has been one of the most powerful optical imaging modalities in the last two decades and has been widely used in ophthalmology, cardiology, dermatology, gastroenterology, and neurology. Analogous to ultrasound imaging, OCT provides depth-resolved cross-sectional image at micrometer spatial resolution with the use of low coherence interferometry. Compared to time domain OCT (TD-OCT) [Huang D, 1991, Science, Hinzenberger CK, 1991,Invest Ophthalmol Vis Sci.], Spectral/Fourier domain OCT (S/FD-OCT) [Fercher AF, 1995, Optics Communications] offers significantly improved volumetric imaging speed and sensitivity.
Relative to other widely used optical imaging technologies for functional brain imaging such as two/multi photon microscopy and confocal fluorescence microscopy, OCT possesses several advantages including, 1) it only takes a few seconds to a minute for a volumetric imaging with OCT compared to tens of minutes to a few hours using two photon microscopy; 2) OCT is capable of imaging at depths of greater than 1 mm in brain tissue; 3) since the axial resolution depends on the coherence length OCT can maintain high axial resolution with low NA objectives, allowing it image a large field of view while maintaining axial resolution; 4) OCT detects backscattered signal from intrinsic scattering avoiding the concern of cumulative dye toxicity in fluorescent imaging; and 5) direct access to the interference fringes provided by S/FD-OCT provides a wide range of novel functional application, such as SD-OCT for absorption measurements and Doppler OCT for flow velocity measurements.
In the Neurophotonics Center, we have been developing and are using various OCT imaging techniques based on S/FD-OCT for studying cerebral physiology and pathophysiology. These techniques include OCT angiography (OCTA) for cerebral vasculature visualization, Doppler OCT (D-OCT) for measuring the axial speed of red blood cells in both large vessels and capillaries, Dynamic Light Scattering-Optical Coherence Tomography (DLSOCT) for absolute blood flow speed detection and the measurement of diffusive motion, and intrinsic contrast OCT (iOCT) for neuron cell body and myelinated axon imaging.
OCTA images are constructed by a decorrelation-based method described in [Srinivasan VJ, 2010, Opt Lett]. Briefly, while conventional structural OCT imaging typically acquires one B-scan for each Y position, the decorrelation-based method repeats two B-scans and then analyzes the differences in the image intensity and phase between the two repeated B-scans. There will be no difference for repeated voxels when static tissue is imaged. In contrast, dynamic tissue, such as a blood vessel, will experience a large difference between repeated B-scans due to particle movement (e.g. flowing RBCs), and will thus appear as bright areas in the OCT angiogram image. With the 10X objective in our OCT system, the capillary bed from a depth of 0 to 1,000 µm could be imaged with an isotropic 3D resolution of 3.5 µm. For more information about OCTA please refer to Zhang A, 2015, J Biomed Opt, and Wang RK, 2007, Opt Express.
Fig 1. MIPs of OCTA images of cerebral vasculature spanning a depth of ~500 µm from brain surface. (A) Obtained with a 5X objective (0.14 NA). (B) Obtained with a 10X objective (0.28 NA). Capillaries are more readily resolved with the 10X objective. Scale: 200 µm.
Doppler OCT is capable of quantitative volumetric measurement of axial blood flow speed by detecting the Doppler frequency shift of backscattered light from a moving particle (e.g., a flowing RBC). Compared to the spectrogram-based method that used short time fast Fourier transformation or wavelet transformation [Chen Z, 1997, Opt Lett], the phase-resolved method that measures the phase change between sequential A-lines to reconstruct axial speed has become the method of choice for Doppler OCT as it possesses higher velocity sensitivity and is compatible with Fourier domain OCT [Zhao Y, 2000, Opt Lett]. For more information about D-OCT please refer to [Zhao Y, 2000, Opt Lett, Leitgeb RA, 2014, Prog Retin Eye Res, Srinivasan VJ, 2010, Opt Express].
We extended phase resolved Doppler OCT from measuring blood flow speed in large vessels to small capillaries, which is critical for studying microvascular flow regulation as is critically important for instance in the brain. This technique is capable of measuring volumetric microvasculature RBC axial speed ranging from ~0.1 mm/s to ~11 mm/s with 3.5 x 3.5 x 3.5 µm spatial resolution, with a temporal resolution up to ~45 s/volume, and as deep as ~900 µm (i.e. the whole cerebral cortex of a mouse could be imaged).
Fig 2. Phase resolved D-OCT axial speed measurement for both large vessels and small capillaries. (A) The phase resolved D-OCT blood flow velocimetry is capable of measuring the majority of capillaries; top panel: OCTA en face MIP (~160 µm stack along Z); bottom panel: D-OCT en face RBC axial speed MIP of the same region. (B) X-Z MIP (~100 µm stack along Y) of axial speed across the whole cerebral cortex obtained from an anesthetized mouse. Color bar: axial blood flow speed (mm/s); Positive value: blood flows toward brain surface; Negative value: blood flows into brain.
Dynamic Light Scattering-Optical Coherence Tomography (DLS-OCT) [Lee J, 2012 , Opt Express, Lee J, 2013, JCBFM, Tang J, 2017, J Biophotonics] takes advantage of using DLS to measure particle flow and diffusion within an OCT resolution-constrained 3D volume, enabling the simultaneous measurement of absolute RBC velocity and diffusion coefficient with high spatial resolution. This novel technique allows us to quantify, for instance, the relation between blood flow velocity and shear-induced diffusion of RBCs [Tang J, 2017, J Biophotonics], which could benefit studies of intravascular dynamics and rheology.
Fig 3. Representative DLS-OCT result. (A) Maximum Intensity Projection (MIPs) of), total blood flow velocity (left) and diffusion coefficient (right) over 0-350 µm in depth. Scale bar, 100 µm . (B) En face single plane axial velocity overlapped with diffusion coefficient at depths of 100, 110, 120, and 130 µm . Scale bar: 100 µm .
Using high spatial resolution objective (~1 lateral resolution) and appropriate data processing methods, neuronal cell body and myelinated axons can be measured with iOCT without the injection of exogenous contrast [Srinivasan VJ, 2012, Opt Express]. With iOCT the distribution of neuron cell bodies could be acquired, which could be useful for brain degeneration, damage, and ischemia studies.
Fig 4. iOCT for neuronal cell body imaging [Srinivasan VJ, 2012, Opt Express]. (A) XY en face MinIP (minimum intensity projection) image. (B) YZ MinIP image.
More background information on OCT can be found on the Wikipedia page. The OCT News updates the most recent publications and research activities of OCT as well as job opportunities in the OCT community.
The following code is available for OCT:
Matlab code for OCT angiography processing is here.
Matlab code for 3D OCT angiography for minimizing tail artifacts is here.
Phase resolved Doppler OCT for imaging blood flow velocity in capillaries is here.
Dynamic Light Scattering OCT data processing code is here.