Probabilistic volumetric speckle suppression in OCT using deep learning.

We present a deep learning framework for volumetric speckle reduction in optical coherence tomography (OCT) based on a conditional generative adversarial network (cGAN) that leverages the volumetric nature of OCT data. In order to utilize the volumetric nature of OCT data, our network takes partial OCT volumes as input, resulting in artifact-free despeckled volumes that exhibit excellent speckle reduction and resolution preservation in all three dimensions. Furthermore, we address the ongoing challenge of generating ground truth data for supervised speckle suppression deep learning frameworks by using volumetric non-local means despeckling-TNode to generate training data. We show that, while TNode processing is computationally demanding, it serves as a convenient, accessible gold-standard source for training data; our cGAN replicates efficient suppression of speckle while preserving tissue structures with dimensions approaching the system resolution of non-local means despeckling while being two orders of magnitude faster than TNode. We demonstrate fast, effective, and high-quality despeckling of the proposed network in different tissue types acquired with three different OCT systems compared to existing deep learning methods. The open-source nature of our work facilitates re-training and deployment in any OCT system with an all-software implementation, working around the challenge of generating high-quality, speckle-free training data.

Mapping optical scattering properties to physical particle information in singly and multiply scattering samples.

Optical coherence tomography (OCT) leverages light scattering by biological tissues as endogenous contrast to form structural images. Light scattering behavior is dictated by the optical properties of the tissue, which depend on microstructural details at the cellular or sub-cellular level. Methods to measure these properties from OCT intensity data have been explored in the context of a number of biomedical applications seeking to access this sub-resolution tissue microstructure and thereby increase the diagnostic impact of OCT. Most commonly, the optical attenuation coefficient, an analogue of the scattering coefficient, has been used as a surrogate metric linking OCT intensity to subcellular particle characteristics. To record attenuation coefficient data that is accurately representative of the underlying physical properties of a given sample, it is necessary to account for the impact of the OCT imaging system itself on the distribution of light intensity in the sample, including the numerical aperture (NA) of the system and the location of the focal plane with respect to the sample surface, as well as the potential contribution of multiple scattering to the reconstructed intensity signal. Although these considerations complicate attenuation coefficient measurement and interpretation, a suitably calibrated system may potentiate a powerful strategy for gaining additional information about the scattering behavior and microstructure of samples. In this work, we experimentally show that altering the OCT system geometry minimally impacts measured attenuation coefficients in samples presumed to be singly scattering, but changes these measurements in more highly scattering samples. Using both depth-resolved attenuation coefficient data and layer-resolved backscattering coefficients, we demonstrate the retrieval of scattering particle diameter and concentration in tissue-mimicking phantoms, and the impact of presumed multiple scattering on these calculations. We further extend our approach to characterize a murine brain tissue sample and highlight a tumor-bearing region based on increased scattering particle density. Through these methods, we not only enhance conventional OCT attenuation coefficient analysis by decoupling the independent effects of particle size and concentration, but also discriminate areas of strong multiple scattering through minor changes to system topology to provide a framework for assessing the accuracy of these measurements.

Computational refocusing in phase-unstable polarization-sensitive optical coherence tomography.

We present computational refocusing in polarization-sensitive optical coherence tomography (PS-OCT) to improve spatial resolution in the calculated polarimetric parameters and extend the depth-of-field in phase-unstable, fiber-based PS-OCT systems. To achieve this, we successfully adapted short A-line range phase-stability adaptive optics (SHARP), a computational aberration correction technique compatible with phase-unstable systems, into a Stokes-based PS-OCT system with inter-A-line polarization modulation. Together with the spectral binning technique to mitigate system-induced chromatic polarization effects, we show that computational refocusing improves image quality in tissue polarimetry of swine eye anterior segment ex vivo with PS-OCT. The benefits, drawbacks, and potential applications of computational refocusing in anterior segment imaging are discussed.

Measuring collagen injury depth for burn severity determination using polarization sensitive optical coherence tomography.

Determining the optimal treatment course for a dermatologic burn wound requires knowledge of the wound's severity, as quantified by the depth of thermal damage. In current clinical practice, burn depth is inferred based exclusively on superficial visual assessment, a method which is subject to substantial error rates in the classification of partial thickness (second degree) burns. Here, we present methods for direct, quantitative determination of the depth extent of injury to the dermal collagen matrix using polarization-sensitive optical coherence tomography (PS-OCT). By visualizing the depth-dependence of the degree of polarization of light in the tissue, rather than cumulative retardation, we enable direct and volumetric assessment of local collagen status. We further augment our PS-OCT measurements by visualizing adnexal structures such as hair follicles to relay overall dermal viability in the wounded region. Our methods, which we have validated ex vivo with matched histology, offer an information-rich tool for precise interrogation of burn wound severity and healing potential in both research and clinical settings.

Layer-based, depth-resolved computation of attenuation coefficients and backscattering fractions in tissue using optical coherence tomography.

Structural optical coherence tomography (OCT) images of tissue stand to benefit from greater functionalization and quantitative interpretation. The OCT attenuation coefficient µ, an analogue of the imaged sample's scattering coefficient, offers potential functional contrast based on the relationship of µ to sub-resolution physical properties of the sample. Attenuation coefficients are computed either by fitting a representative µ over several depth-wise pixels of a sample's intensity decay, or by using previously-developed depth-resolved attenuation algorithms by Girard et al. [Invest. Ophthalmol. Vis. Sci.52, 7738 (2011). 10.1167/iovs.10-6925] and Vermeer et al. [Biomed. Opt. Express5, 322 (2014). 10.1364/BOE.5.000322]. However, the former method sacrifices axial information in the tomogram, while the latter relies on the stringent assumption that the sample's backscattering fraction, another optical property, does not vary along depth. This assumption may be violated by layered tissues commonly observed in clinical imaging applications. Our approach preserves the full depth resolution of the attenuation map but removes its dependence on backscattering fraction by performing signal analysis inside individual discrete layers over which the scattering properties (e.g., attenuation and backscattering fraction) vary minimally. Although this approach necessitates the detection of these layers, it removes the constant-backscattering-fraction assumption that has constrained quantitative attenuation coefficient analysis in the past, and additionally yields a layer-resolved backscattering fraction, providing complementary scattering information to the attenuation coefficient. We validate our approach using automated layer detection in layered phantoms, for which the measured optical properties were in good agreement with theoretical values calculated with Mie theory, and show preliminary results in tissue alongside corresponding histological analysis. Together, accurate backscattering fraction and attenuation coefficient measurements enable the estimation of both particle density and size, which is not possible from attenuation measurements alone. We hope that this improvement to depth-resolved attenuation coefficient measurement, augmented by a layer-resolved backscattering fraction, will increase the diagnostic power of quantitative OCT imaging.

Computational adaptive optics in phase-unstable optical coherence tomography.

We present a scheme for correction of x-y-separable aberrations in optical coherence tomography (OCT) designed to work with phase unstable systems with no hardware modifications. Our approach, termed SHARP, is based on computational adaptive optics and numerical phase correction and follows from the fact that local phase stability is sufficient for the deconvolution of optical aberrations. We demonstrate its applicability in a raster-scan polygon-laser OCT system with strong phase-jitter noise, achieving successful refocusing at depths up to 4 times the Rayleigh range. We also present in vivo endoscopic and ex vivo anterior segment OCT data, showing significant enhancement of image quality, particularly when combining SHARP results with a resolution-preserving despeckling technique like TNode.

Advanced iterative algorithm for phase calibration of spatial light modulators integrated in optical instrumentation in a vibration environment.

We present a method to obtain the phase modulation characteristic curve of a spatial light modulator (SLM) under severe vibration conditions. The procedure is based on the well-known advanced iterative algorithm (AIA), which allows wavefront extraction from unknown phase-shifted interferograms. Generally, AIA is used to determine the wavefront and the determined phase shifts are of little interest. In contrast, in our method, the main goal of using AIA is to determine the unknown phase shifts induced by an SLM during the calibration procedure. Using a segmented approach to calibration, AIA enables successful calibration even in the presence of additional random phase shifts due to environmental changes. This method has the potential to calibrate SLMs integrated in complex optical instruments with little to no modifications to the optical setup, no matter the environmental conditions. We demonstrate our technique by calibrating an SLM under vacuum conditions (10-5mbar) in a common-path configuration compatible with usage of an SLM as a wavefront modulator at the pupil plane of an instrument. Our technique compensates for the vibrations produced by the vacuum pumps and reduces an order of magnitude the root-mean-squared error of the calibration curve evaluated with vibration errors. Our technique enhances the potential use of SLMs in complex optical systems, including aerospace optical instrumentation.

Real-time co-localized OCT surveillance of laser therapy using motion corrected speckle decorrelation.

We present a system capable of real-time delivery and monitoring of laser therapy by imaging with optical coherence tomography (OCT) through a double-clad fiber (DCF). A double-clad fiber coupler is used to inject and collect OCT light into the core of a DCF and inject the therapy light into its larger inner cladding, allowing for both imaging and therapy to be perfectly coregistered. Monitoring of treatment depth is achieved by calculating the speckle intensity decorrelation occurring during tissue coagulation. Furthermore, an analytical noise correction was used on the correlation to extend the maximum monitoring depth. We also present a method for correcting motion-induced decorrelation using a lookup table. Using the value of the noise- and motion-corrected correlation coefficient in a novel approach, our system is capable of identifying the depth of thermal coagulation in real time and automatically shut the therapy laser off when the targeted depth is reached. The process is demonstrated ex vivo in rat tongue and abdominal muscles for depths ranging from 500 µm to 1000 µm with induced motion in real time.

A Neural Network Approach to Quantify Blood Flow from Retinal OCT Intensity Time-Series Measurements.

Many diseases of the eye are associated with alterations in the retinal vasculature that are possibly preceded by undetected changes in blood flow. In this work, a robust blood flow quantification framework is presented based on optical coherence tomography (OCT) angiography imaging and deep learning. The analysis used a forward signal model to simulate OCT blood flow data for training of a neural network (NN). The NN was combined with pre- and post-processing steps to create an analysis framework for measuring flow rates from individual blood vessels. The framework's accuracy was validated using both blood flow phantoms and human subject imaging, and across flow speed, vessel angle, hematocrit levels, and signal-to-noise ratio. The reported flow rate of the calibrated NN framework was measured to be largely independent of vessel angle, hematocrit levels, and measurement signal-to-noise ratio. In vivo retinal flow rate measurements were self-consistent across vascular branch points, and approximately followed a predicted power-law dependence on the vessel diameter. The presented OCT-based NN flow rate estimation framework addresses the need for a robust, deployable, and label-free quantitative retinal blood flow mapping technique.

Forward multiple scattering dominates speckle decorrelation in whole-blood flowmetry using optical coherence tomography.

Quantitative blood flow measurements using optical coherence tomography (OCT) have a wide potential range of medical research and clinical applications. Flowmetry based on the temporal dynamics of the OCT signal may have the ability to measure three-dimensional flow profiles regardless of the flow direction. State-of-the-art models describing the OCT signal temporal statistics are based on dynamic light scattering (DLS), a model which is inherently limited to single scattering regimes. DLS methods continue to be applied to OCT despite the knowledge that red blood cells produce strong forward multiple scattering. Here, we postulate that forward multiple scattering is the primary mechanism causing the rate of speckle-decorrelation derived from data acquired in vivo to deviate from the rate of decorrelation determined in phantom experiments. We also postulate that multiple scattering contributions to decorrelation are only present when the sample exhibits velocity field inhomogeneities larger than the scale of a resolution volume and are thus absent in rigid bulk motion. To test these hypotheses, we performed a systematic study of the effects of forward multiple scattering on OCT signal decorrelation with phantom experiments under physiologically relevant flow conditions and relative bulk motion. Our experimental results confirm that the amount of forward multiple scattering affects the proportionality between lateral flow and decorrelation. We propose that multiply scattered light carries information from different locations in the sample and each location imprints scattering dynamics on the scattered light causing increased decorrelation rates. Our analysis confirms that the detection of forward scattered light inside the vessel lumen causes an increase in the rate of decorrelation which results in an overestimation of blood flow velocities at depths as shallow as 40 µm into whole blood for OCT systems with typical numerical apertures used in retinal imaging.

Noise and bias in optical coherence tomography intensity signal decorrelation.

Functional optical coherence tomography (OCT) imaging based on the decorrelation of the intensity signal has been used extensively in angiography and is finding use in flowmetry and therapy monitoring. In this work, we present a rigorous analysis of the autocorrelation function, introduce the concepts of contrast bias, statistical bias and variability, and identify the optimal definition of the second-order autocorrelation function (ACF) g (2) to improve its estimation from limited data. We benchmark different averaging strategies in reducing statistical bias and variability. We also developed an analytical correction for the noise contributions to the decorrelation of the ACF in OCT that extends the signal-to-noise ratio range in which ACF analysis can be used. We demonstrate the use of all the tools developed in the experimental determination of the lateral speckle size depth dependence in a rotational endoscopic probe with low NA, and we show the ability to more accurately determine the rotational speed of an endoscopic probe to implement NURD detection. We finally present g (2)-based angiography of the finger nailbed, demonstrating the improved results from noise correction and the optimal bias mitigation strategies.

Balloon catheter-based radiofrequency ablation monitoring in porcine esophagus using optical coherence tomography.

We present a microscopic image guidance platform for radiofrequency ablation (RFA) using a clinical balloon-catheter-based optical coherence tomography (OCT) system, currently used in the surveillance of Barrett's esophagus patients. Our integrated thermal therapy delivery and monitoring platform consists of a flexible, customized bipolar RFA electrode array designed for use with a clinical balloon OCT catheter and a processing algorithm to accurately map the thermal coagulation process. Non-uniform rotation distortion was corrected using a feature tracking-based technique, which enables robust, frame-to-frame analysis of the temporal fluctuation of the complex OCT signal. With proper noise calibration, precise delineation of the thermal therapy zone was demonstrated using cumulative complex differential variance in porcine esophagus ex vivo with the integrated OCT-RFA system, as validated by nitroblue tetrazolium chloride (NBTC) histology. The ability to directly and accurately visualize the thermal coagulation process at high resolution is critical to the precise delivery of thermal energy to a wide range of epithelial lesions.

Volumetric non-local-means based speckle reduction for optical coherence tomography.

We present a novel tomographic non-local-means based despeckling technique, TNode, for optical coherence tomography. TNode is built upon a weighting similarity criterion derived for speckle in a three-dimensional similarity window. We present an implementation using a two-dimensional search window, enabling the despeckling of volumes in the presence of motion artifacts, and an implementation using a three-dimensional window with improved performance in motion-free volumes. We show that our technique provides effective speckle reduction, comparable with B-scan compounding or out-of-plane averaging, while preserving isotropic resolution, even to the level of speckle-sized structures. We demonstrate its superior despeckling performance in a phantom data set, and in an ophthalmic data set we show that small, speckle-sized retinal vessels are clearly preserved in intensity images en-face and in two orthogonal, cross-sectional views. TNode does not rely on dictionaries or segmentation and therefore can readily be applied to arbitrary optical coherence tomography volumes. We show that despeckled esophageal volumes exhibit improved image quality and detail, even in the presence of significant motion artifacts.

Robust wavenumber and dispersion calibration for Fourier-domain optical coherence tomography.

Many Fourier-domain optical coherence tomography (FD-OCT) systems sample the interference fringes with a non-uniform wavenumber (k) interval, introducing a chirp to the signal that depends on the path length difference underlying each fringe. A dispersion imbalance between sample and reference arms also generates a chirp in the fringe signal which, in contrast, is independent of depth. Fringe interpolation to obtain a signal linear in k and compensate dispersion imbalance is critical to achieving bandwidth-limited axial resolution. In this work, we propose an optimization-based algorithm to perform robust and automated calibration of FD-OCT systems, recovering both the interpolation function and the dispersion imbalance. Our technique relies on the fact that the unique function that correctly linearizes the fringe data in k space produces a depth-independent chirp. The calibration procedure requires experimental data corresponding to a single reflector at various depth locations, which can easily be obtained by acquiring data while moving a sample mirror in depth. We have tested both spectral domain OCT and swept source OCT systems with various nonlinearities. Results indicate that the proposed calibration method has excellent performance on a wide range of data sets and enables nearly constant resolution at all imaging depths. An implementation of the algorithm is available online.

Extended bandwidth wavelength swept laser source for high resolution optical frequency domain imaging.

Improving the axial resolution by providing wider bandwidth wavelength swept lasers remains an important issue for optical frequency domain imaging (OFDI). Here, we demonstrate a wide tuning range, all-fiber wavelength swept laser at a center wavelength of 1250 nm by combining two ring cavities that share a single Fabry-Perot tunable filter. The two cavities contain semiconductor optical amplifiers with central wavelengths of 1190 nm and 1292 nm, respectively. To avoid disturbing interference effects in the overlapping spectral region, we modulated the amplifiers in order to obtain consecutive wavelength sweeps in the two spectral regions. The two sweeps were fused together in post-processing to achieve a total scanning range of 223 nm, corresponding to 3.3 µm axial resolution in air. We confirm improved image quality and reduced speckle size in tomograms of swine esophagus ex vivo, and human skin and nailbed in vivo.

Laser thermal therapy monitoring using complex differential variance in optical coherence tomography.

Conventional thermal therapy monitoring techniques based on temperature are often invasive, limited by point sampling, and are indirect measures of tissue injury, while techniques such as magnetic resonance and ultrasound thermometry are limited by their spatial resolution.  The visualization of the thermal coagulation zone at high spatial resolution is particularly critical to the precise delivery of thermal energy to epithelial lesions. In this work, an integrated thulium laser thermal therapy monitoring system was developed based on complex differential variance (CDV), which enables the 2D visualization of the dynamics of the thermal coagulation process at high spatial and temporal resolution with an optical frequency domain imaging system. With proper calibration to correct for noise, the CDV-based technique was shown to accurately delineate the thermal coagulation zone, which is marked by the transition from high CDV upon heating to a significantly reduced CDV once the tissue is coagulated, in 3 different tissue types ex vivo: skin, retina, and esophagus. The ability to delineate thermal lesions in multiple tissue types at high resolution opens up the possibility of performing microscopic image-guided procedures in a vast array of epithelial applications ranging from dermatology, ophthalmology, to gastroenterology and beyond.

Velocity gradients in spatially resolved laser Doppler flowmetry and dynamic light scattering with confocal and coherence gating.

Dynamic light scattering (DLS) is widely used to characterize diffusive motion to obtain precise information on colloidal suspensions by calculating the autocorrelation function of the signal from a heterodyne optical system. DLS can also be used to determine the flow velocity field in systems that exhibit mass transport by incorporating the effects of the deterministic motion of scatterers on the autocorrelation function, a technique commonly known as laser Doppler flowmetry. DLS measurements can be localized with confocal and coherence gating techniques such as confocal microscopy and optical coherence tomography, thereby enabling the determination of the spatially resolved velocity field in three dimensions. It has been thought that spatially resolved DLS can determine the axial velocity as well as the lateral speed in a single measurement. We demonstrate, however, that gradients in the axial velocity of scatterers exert a fundamental influence on the autocorrelation function even in well-behaved, nonturbulent flow. By obtaining the explicit functional relation between axial-velocity gradients and the autocorrelation function, we show that the velocity field and its derivatives are intimately related and their contributions cannot be separated. Therefore, a single DLS measurement cannot univocally determine the velocity field. Our extended theoretical model was found to be in good agreement with experimental measurements.

Vortex-enhanced coherent-illumination phase diversity for phase retrieval in coherent imaging systems.

We propose a phase-retrieval method based on the numerical optimization of a new objective function using coherent phase-diversity images as inputs for the characterization of aberrations in coherent imaging systems. By employing a spatial light modulator to generate multiple-order spiral phase masks as diversities, we obtain an increase in the accuracy of the retrieved phase compared with similar state-of-the-art phase-retrieval techniques that use the same number of input images. We present simulations that show a consistent advantage of our technique, and experimental validation where our implementation is used to characterize a highly aberrated 4F optical system.

Rotational distortion correction in endoscopic optical coherence tomography based on speckle decorrelation.

We present a new technique for the correction of nonuniform rotation distortion in catheter-based optical coherence tomography (OCT), based on the statistics of speckle between A-lines using intensity-based dynamic light scattering. This technique does not rely on tissue features and can be performed on single frames of data, thereby enabling real-time image correction. We demonstrate its suitability in a gastrointestinal (GI) balloon-catheter OCT system, determining the actual rotational speed with high temporal resolution, and present corrected cross-sectional and en face views showing significant enhancement of image quality.

Quantitative technique for robust and noise-tolerant speed measurements based on speckle decorrelation in optical coherence tomography.

Intensity-based techniques in optical coherence tomography (OCT), such as those based on speckle decorrelation, have attracted great interest for biomedical and industrial applications requiring speed or flow information. In this work we present a rigorous analysis of the effects of noise on speckle decorrelation, demonstrate that these effects frustrate accurate speed quantitation, and propose new techniques that achieve quantitative and repeatable measurements. First, we derive the effect of background noise on the speckle autocorrelation function, finding two detrimental effects of noise. We propose a new autocorrelation function that is immune to the main effect of background noise and permits quantitative measurements at high and moderate signal-to-noise ratios. At the same time, this autocorrelation function is able to provide motion contrast information that accurately identifies areas with movement, similar to speckle variance techniques. In order to extend the SNR range, we quantify and model the second effect of background noise on the autocorrelation function through a calibration. By obtaining an explicit expression for the decorrelation time as a function of speed and diffusion, we show how to use our autocorrelation function and noise calibration to measure a flowing liquid. We obtain accurate results, which are validated by Doppler OCT, and demonstrate a very high dynamic range (> 600 mm/s) compared to that of Doppler OCT (±25 mm/s). We also derive the behavior for low flows, and show that there is an inherent non-linearity in speed measurements in the presence of diffusion due to statistical fluctuations of speckle. Our technique allows quantitative and robust measurements of speeds using OCT, and this work delimits precisely the conditions in which it is accurate.