This conference presentation was prepared for SPIE Photonics West BiOS 2024

Previous studies in optoretinography often rely on AO-OCT systems to resolve individual cells or use intensity-based image processing to extract the dynamics of the retinal layer as an ensemble. For non-AO point-scan OCT, investigating tissue dynamics from obscure speckle patterns while preserving the spatial heterogeneity of signals remains challenging. Here, we developed phase-restoring subpixel image registration and unsupervised machine learning algorithms to accurately extract spatially-resolved OCT phase signals from the outer retina in rodents. In addition to observing light-evoked deformation of the photoreceptors outer segments, we discovered an optical signature of the retinal pigment epithelium (RPE) response to visual stimuli.

Visually-evoked retinal neuronal simulation-induced transient vasodilation of retinal blood vessels is referred to as neurovascular coupling. Both systemic diseases such as Diabetes and high blood pressure, as well as potentially blinding retinal diseases such as Glaucoma have been shown to damage the elasticity of the blood vessels’ walls. In this study, a research-grade, high-resolution OCT system is combined with a commercial electroretinography (ERG) system to investigate flicker-stimulus-induced transient vasodilation in retinal blood vessels around the Optic Nerve Head and study neurovascular coupling in the human retina.

The retinal vasculature has plexuses at multiple depths, interconnected in a complex pattern. While optical coherence tomography angiography (OCTA) visualizes the vasculature at the capillary level, creating a 3D representation of the vasculature, including the connections between plexuses, remains a challenge. Here, we employ Optimally Oriented Flux (OOF) on retinal OCTA to simultaneously preserve the inter-plexus connections and suppress projection artifacts, facilitating high-contrast 3D visualization of the vasculature. Furthermore, a novel framework is developed to transform the vasculature into a graph, which enables quantification of ‘importance’ of each capillary using betweenness centrality and simulation of capillary non-perfusion.

We applied spatio-temporal optical coherence tomography (STOC-T) method, supported by computational aberration correction for high resolution in vivo imaging of the human retina. Here, we demonstrate a modified STOC-T experimental system, with an increased lateral resolution of ~3 µm, and optimized illumination to enable rendering of the retinal ganglion cells. We present high resolution en face images of the human retina NFL and GCL layers.

Optical Coherence Tomography (OCT) is a preferred technology with imaging ocular diseases. However, traditional methods require patients to be stable, preventing those with involuntary eye motion from obtaining accurate OCT images. We demonstrate a pupil tracking and aiming system that utilizes a CompactRIO embedded system for real-time image processing. We show that the pupil tracking and aiming yield an amplitude attenuation of up to 24.08 dB at stage frequencies below 0.5 Hz, outperforming our previous implementation with response latency sufficient to partially stabilize OCT images and potentially provide accurate OCT imaging for those with involuntary eye and head movement.

Optical coherence tomography (OCT) is a critical imaging technique for diagnosing diseases of the eye. Whole eye OCT acquires images of the cornea and retina simultaneously, which is believed to have diagnostic value for certain pathologies. However, patient motion during acquisition can corrupt spatial relationships in the final image. Additionally, OCT for the anterior and posterior segments of the eye require distinct instrumentation. Here we present a whole eye OCT system with polarization multiplexed independent channels allowing for simultaneous imaging of the anterior and posterior segments. Previously developed pupil tracking algorithms were integrated in our novel system to compensate for patient motion.

We report a pioneer clinical study using triple-input polarization-sensitive optical coherence tomography (TRIPS-OCT) to assess sub-macular scleral birefringence in 60 children with myopia. Results showed a significant difference in scleral birefringence between high and low myopia groups. In addition, we observed a correlation between axial length and birefringence in the low myopia group. This suggests the potential of using sub-macular scleral birefringence as a biomarker for myopia progression. Despite a high exclusion ratio due to inadequate scleral visibility and other limitations, these findings warrant further large-scale studies.

4D microscope-integrated optical coherence tomography has emerged as a valuable tool for ophthalmic surgery. However, it is still not clear how to make the best, computationally efficient use of the abundant data generated. Many techniques require accurate registration of successive volume streams. To address this need, we present a real-time machine learning method for lateral registration based on en face projections. Our proposed network predicts the lateral translational and rotational offset, which can be corrected. As ground truth, we use homography matrices from the displacements of en face images. Our pipeline thus allows for real-time, rigid registration of en face maps derived from sequentially acquired volumes, and is able to accurately correct for lateral distortions at a volume rate of 20 Hz. We therefore believe that this method could enable real-time rigid volume registration for 4D microscopy integrated optical coherence tomography.

Real-time ophthalmic surgical decision-making benefits from volumetric visualization from intraoperative optical coherence tomography and from quantification of retinal structures during surgery. Surgeons typically rely on predetermined measurements and qualitative metrics to perform delicate operations, such as subretinal microliter volume injections for drug delivery and gene therapy. Here, we present a quantitative method for volumetric estimation of subretinal injection volume based on intraocular reference features with intraoperative optical coherence tomography for use during ophthalmic surgery. We predict this method will improve the accuracy of subretinal injections and provide better ergonomics and surgical workflow over current methods for ophthalmic surgery.

Age-related macular degeneration (AMD) is a leading cause of global irreversible vision loss, with wet AMD responsible for over 90% of AMD-related vision loss. The laser-induced CNV mouse model is commonly used to study wet AMD. For this study, a functional retinal imaging system was used to investigate the pathological characteristics of laser-induced CNV in mice, providing advanced resolution to visualize structural, vascular, and molecular-specific features of the lesions and outer retinal layers.

We present a full-field Fourier-domain optical coherence tomography system capable of acquiring volumetric data at 50 MHz A-scan rate with an axial resolution of 3.3 µm in air. This high resolution is achieved by phase-correctly combining the interferograms of two independent swept lasers to achieve a total bandwidth of 145 nm at a center wavelength of 880 nm. By changing the software control of the system, it can be switched to higher acquisition speeds of more than 100 MHz A-scan rate while reducing the resolution to 5.9 µm. We show images of technical and biological samples, including the retina.

Conventional dual-input state PS-OCT incorrectly assumes that the two probing input states provide equally reliable measurements. In this work, we overcome this assumption by adapting a maximum-likelihood framework which combines all input state and spectral bin measurements to find the most likely sample Jones matrix. This processing method (MLDIPS) shows a significantly reduced retardance noise floor as well as improved qualitative characterization of white matter versus grey matter in porcine brain tissue, displaying better contrast to conventional dual-input processing.

In fields such as virtual/augmented reality, robotic vision, facial recognition, and biomedical imaging, the demand for accurate, fast coherent 3D surface imaging technology is increasing. However, current technologies like FMCW LiDAR have limitations in terms of low spectral bandwidth efficiency and high cost. We demonstrate a novel approach using synthetic wavelength phase unwrapping and line-scan off-axis holography. Our system achieved fast 3D surface imaging with a non-ambiguous depth range of 22.3mm. We validated system performance using a 3D-printed test target and a U.S. copper penny.

Optical Coherence Tomography (OCT) is a low coherence interferometric imaging modality used to obtain micron-resolution structural information in depth, however it lacks in molecular specificity, and can therefore be combined with Near Infrared Fluorescence (NIRF) imaging, using labelled monoclonal antibodies to provide molecular contrast. Here, a novel capsule endoscope design is presented, which facilitates circumferential Immuno-OCT imaging in the gastrointestinal tract. The capsule endoscope is fitted with a double clad fiber to achieve dual modality imaging. We present preliminary endoscopic data for both imaging modalities, for which we have used phantoms to validate the imaging capabilities of the capsule endoscope.

Optical coherence tomography (OCT) is a powerful label-free approach for volumetric morphological imaging, especially within biomedicine. However, the penetration depth of OCT is limited. We postulate that in OCT, multiple scattering can enhance image contrast at depth. We demonstrate this using spatially offset OCT (SO-OCT), an original collection geometry that leads to a preferential collection of multiply scattered light. SO-OCT images have improved contrast at depth, especially in highly scattering tissues. We show that by merging images from several offsets, we can generate a composite image with reduced speckle and improved contrast over a large depth range. Finally, we investigate opportunities for using SO-OCT to improve segmentation of tissues, an important step in image interpretation. Our results show the high potential of SO-OCT as a new tool in biomedical imaging and diagnosis.

We have used Full Field OCT (FFOCT) in its static (morphology contrast) and dynamic (metabolic contrast) modes mostly for tissue diagnosis. Last year we introduced a new way of optical sectioning 3D structures. For this purpose, we use Full Field Optical Transmission Tomography (FFOTT) a Gouy phase shift interference approach that takes place close to the focus of a microscope objective. We will show striking differences between FFOCT and FFOTT associated to the scattering anisotropy of tissue structures. Moreover, we will pay a particular attention to the requirements of FFOTT in term of spatial coherence and to the speckle appearance for both techniques.

Imaging speed of OCT is a crucial factor for various applications. In swept source OCT, the imaging speed is directly determined by the sweep repetition speed of the wavelength-swept laser. Recently, a stretched-pulse mode-locked (SPML) wavelength-swept laser has been developed, utilizing chromatic dispersion to generate the wavelength-swept output with repetition rates from a few MHz to over 10 MHz without using any mechanical filter. In this work, we investigated SPML lasers with various repetition rates of 10 MHz, 20 MHz, and 80 MHz utilizing 10 m (930 ps/nm), 5 m (465 ps/nm), and 1 m (93 ps/nm) long CFBGs, respectively. We present coherence length and relative intensity noise (RIN) performances of the SPML lasers with various speeds and show the comparison of images acquired using the SPML lasers of 10 MHz and 80 MHz repetition rates.

This work focuses on making a novel ophthalmic Optical transmission tomography (OTT) device at the lowest possible cost and size. OTT prototype demonstrates images from all the layers in anterior human eye, while also benefiting from the cost-efficient design solutions: common-path architecture, mass-market CMOS cameras, latest USB data transfer standards, Arduino electronic control, etc. Notably, we show that the large degree of noise degradation (due to the use of low-cost optics/cameras) can be corrected with denoised neural networks. Moreover, the model trained on one type of camera (global shutter) can be used to improve signal in another camera (rolling shutter).

For OCTA to become a widespread tool in the ophthalmic clinic, the significant increase in acquisition time must be kept at a minimum. Therefore, we developed a new OCTA processing pipeline that enables 4D coherent averaging and leverages from it to improve angiography contrast without using more repetitions. We first present a new metric for computing angiography contrast that exploits both spatial and time coherence, g1C, which significantly reduce phase noise while using minimal repetitions for averaging. We then show, for the first time, how to perform advanced image registration to correct for motion between repetitions while maintaining phase coherency.

Point-scanning OCT systems often use a pair of sequential, single-axis galvanometer scanners to acquire volumetric data. This can introduce uncertainty in the beam position at the ocular pupil plane, an effect known as beam wander or pupil wobble, which can distort the resulting images. We propose a new approach to characterize and optically correct the pupil wobble with an additional scanning mirror placed anti-conjugate to the pupil plane. We evaluate this method by modeling the pattern of pupil wobble present in a research OCT system both theoretically and experimentally, and correcting for it with the proposed method.

We presented a method to study the influence of static and dynamic ocular aberrations on FFOCT image quality when imaging the retina at high-resolution. The results presented here give important information to infer the best AO design to be used in FFOCT.

OCT speckle carries information on sub-cellular tissue structures, and speckle statistics have been shown to be potential biomarkers in tissue characterization for disease detection and monitoring. Current methods for estimating speckle parameters use simple methods in which speckle statistics are determined inside a fixed kernel, which makes them unsuitable in heterogeneous tissue and have a clear trade-off between accuracy and spatial resolution. These limitations make them unsuitable for automatically detecting spatially-resolved differences in cellular microstructure that occurs in a diseased tissue. To address this unmet need, we have developed an algorithm based on a probabilistic approach to automatically select kernels consisting of pixels that have a high probability of sharing the same speckle probability density function and use them to estimate spatially-resolved speckle parameters using likelihood-based estimation. Our proposed method enables new capabilities in producing speckle parametric images, providing information on spatial variability of speckle distribution throughout OCT volumes and additional information to structural OCT imaging.

This study presents an exploratory approach to performing optical coherence tomographic angiography (OCTA) using motion contrast between repeated volumetric scans captured at 200 ms intervals, as opposed to conventional OCTA which measure motion contrast between repeat B-scans captured at shorter intervals. The proposed inter-volumetric OCTA (IV-OCTA) is made possible through an advanced image registration algorithm based on Mattes mutual information metric, and it is implemented on an ultra-high-resolution spectral domain 840 nm OCT system with a 250 KHz A-scan rate. IV-OCTA demonstrates a high detection sensitivity for microvessels with slow blood flow and reduces bulk motion artifacts using the embedded volume registration algorithm. Meanwhile, averaging repeated volumes can substantially reduce the speckle noise for clearer structural imaging compared to the averaged repeated B-scans.

Montaging multiple optical coherence tomography (OCT) volumes can expand the field of view of OCT. This is often necessary when the biological sample is too large or high-density scanning is desired. We developed a method for montaging OCT volumes acquired from different locations and at different angles relative to the incident light. Our method involves representing OCT data as a point cloud and determining the transformation necessary to align intersecting point clouds with each other. The effectiveness of our algorithm is demonstrated in reconstructing the anterior segment of the eye.

Bidirectional scanning patterns in OCT allow for faster data acquisition with less mechanical strain on scanning components. However, the alternation in scanning direction could lead to phase drift between adjacent B-scans, due to hysteresis of the galvanometer or changes in the incident beam angle, causing errors for digital adaptive optics (DAO) methods. To overcome this limitation, we employed a phase bias estimation method, using bulk phase estimation (BPE) on de-interlaced bidirectional scans to calculate and compensate the systemic phase offset in alternating frames. The corrected scans are then re-interlaced, providing higher-quality DAO correction.

Time-domain Full-Field OCT (FFOCT) acquires en-face images of a sample in a given depth. One of the biggest challenge to use FFOCT for in-vivo retinal imaging is the presence of retinal axial motion during image acquisition. To address this challenge, we previously proposed to couple a spectral-domain OCT (SD-OCT), where axial motion can be measured and later corrected by moving the FFOCT reference arm accordingly in a control loop fashion. However, due to the inherent temporal delay of the control-loop (typically 2-frame delay and 50Hz loop rate), the achieved precision was only around 10µm rms, against ideal 4µm rms (coherence gate of 8µm). Here, we propose to use learning-based prediction methods to enhance the precision of the axial retinal motion correction in real-time, thus improving the FFOCT robustness for in-vivo retinal imaging.

This work presents a clustering approach for Time-domain Full-field optical coherence tomography (TD-FFOCT) retinal images.TD-FFOCT is an efficient method for cellular-level analysis of retinal structures, with fast acquisition and wide field-of-view. However, clinical use faces challenges from involuntary axial retinal motion due to breathing, heartbeat, and pulsation. Despite real-time axial motion compensation, achieved precision is around 10µm rms, below the ideal 4µm for an 8µm coherence gate, impacting system robustness and image depth selection. One way to overcome this is to group together images featuring similar retinal structures and acquired at the same depth. We propose a comprehensive clustering approach using learning-based and non-learning-based methods for feature extraction and clustering. Results show that clustering can help mitigate the effects of motion on the acquired image data, improving imaging accuracy and robustness.

Polarization-sensitive optical coherence tomography (PS-OCT) can measure the depth-resolved birefringence of tissue. However, determining the depth-resolved optic axis orientation is challenging due to the round-trip propagation of the probe beam in tissue layers. This paper analyzed the optic axes of a multi-layer retarder with Stokes representation and proposed a computationally effective reconstruction algorithm of optic axes. Filaments of the 3D printer were measured by Jones-matrix OCT. The angles between the filaments measured from the reconstructed relative optic axis have a maximum difference of 21° from the estimation.

In this presentation we will evaluate the signal processing steps required to obtain stimulus-evoked nm-scale changes in the photoreceptor outer segment (OS) length using full field swept source optical coherence tomography (FF-SS-OCT). Evaluation of stimulus-evoked OS length changes is the basis of photoreceptor optoretinography (ORG), an emerging test of retinal function. To allow cellular resolution, an adaptive optics (AO) enhanced FF-SS-OCT system was constructed. This system allows up to kHz volume acquisition rates, which greatly reduce retinal motion artifacts. Specifically, the effects of wave number calibration, dispersion compensation, and application of short-time Fourier transform (STFT) on image quality metrics, SNR of individual OCT volumes, and extracted phase noise will be presented.

We propose a B-scan-wise- multi-focus average method (B-MFA) to suppress multiple scattering noise for in-vivo imaging. This method uses only multiple B-scans with modulated focus position for complex averaging, and a one-dimensional computational refocusing is applied to cancel the defocus in each B-scan. To validate the proposed method, ten medaka fish were measured in-vivo without anesthesia. The results show that B-MFA can enhance the OCT image contrast in the deep regions.

The proper guidance of the epidural needle is crucial for safe and effective epidural anesthesia. In this research, we developed an innovative endoscopic system based on polarization-sensitive optical coherence tomography (PS-OCT). To assess its feasibility, we conducted experiments using ex vivo human epidural specimens. During the experiments, we imaged and analyzed different spinal tissue layers that the epidural needle passes through, including subcutaneous fat, supraspinous and interspinous ligament, ligamentum flavum, epidural space, dura, and spinal cord. Each of these tissue layers exhibited distinct imaging patterns. In addition, we used deep learning for automatic tissue recognition.

A representative selection of works is discussed where attempts are reported at achieving axial super-resolution in Optical Coherence Tomography (OCT). The presented approaches – the algorithms of Spectral Estimation OCT, the algorithms of multi-rate OCT, and the experimental configurations of quantum OCT and our own quantum-mimic OCT – are considered in terms of the mechanisms underlying the depth image formation that they use to achieve the improvement. The specific mechanisms are identified in each research, described more broadly in relation to OCT imaging and characterized in terms of their limitations.

In Fourier-domain optical coherence tomography (FD-OCT), image reconstruction has been extensively studied. This paper addresses the trade-off between reconstruction time and image quality of the optimization methods by proposing an unsupervised deep learning-based approach. Different from the existing learning-based methods, the proposed unsupervised learning method incorporates a neural network as an inverse solver and eliminates the need for large training pairs. A proof-of-concept simulation was conducted, comparing our method with an iterative optimization technique using stochastic gradient descent (SGD). Results show that the proposed method achieves real-time reconstruction with a small decrease in image quality compared to SGD, while enabling real-time reconstruction at a speed of 0.008s per B-scan (125 frames per second). In contrast, the SGD method took 0.32s per B-scan, making it 40 times slower. This deep learning-based method has significant potential for real-time image reconstruction and display in future FD-OCT.

Speckle is a well-known and inevitable phenomenon exhibiting in optical coherence tomography (OCT). However, conventional OCT imaging formula cannot isolate the speckles from a meaningful OCT image. Here we introduce a new formulation of OCT to mathematically explicitly represent the meaningful image and speckle. We represent an intensity OCT image as a collection of complex contributions from scatterers. This formulation clarified that the intensity OCT image is the summation of a meaningful OCT image and speckles. In addition, it was found that the meaningful OCT image is equivalent to the incoherent image of the sample. This equation also recapitulates well-known properties of speckle.

We investigate the dependency of dynamic optical coherence tomography (DOCT) signal on wavelength and resolution in two systems. One is 840-nm spectral domain OCT, and the other is 1310-nm swept-source OCT. For wavelength dependency, we made the resolution roughly the same among two systems by computationally reducing the axial resolution of 840-nm system. For resolution dependency, we compared LIV images with and without reduction of the resolution. In addition, we performed numerical simulation by modeling the intracellular scatterer dynamics by a diffusion. Experiment and simulation suggested that LIV is highly affected by the wavelength, where shorter wavelength gives higher LIV.

This work presents a novel approach to the use of bidirectional swept sources. A MEMS-VCSEL light source with a bidirectional sweeping rate of 1.6MHz is used. Bidirectional sweeping poses a challenge as the tuning curve is different for each sweep. To exploit the full performance of the swept source, both sweeps within the bidirectional sweeping must be utilized and compensated. The proposed approach involves a software solution based on the Master-Slave protocol, where sets of theoretically inferred channelled spectra (masks) are prepared in advance for each sweep. The phase retrieval method called CMS (Complex Master Slave) is employed to handle bidirectional sweeping SS-OCT, involving the calculation of two sets of functions for each sweeping direction. The correct set of masks is used for each sweep direction to accurately represent A-Scan peaks. Differences in the sweep are analysed. Finally, the proposed method is successfully tested on posterior and anterior chambers.

Optical Coherence Tomography (OCT) is a non-invasive imaging technique, essential in medical diagnostics due to its ability to produce high-resolution images of internal structures of biological tissues. One of the unique features of the FDML based MHz-OCT is the optical buffering that increases the A-scan rate by creating successive time-delayed copies of the original sweep. However, due to the optical buffering, numerous studies have reported that A-lines originates from different buffer can have different amplitude and phase. Another challenge associated with high A-scan laser source is to pair with the high-speed mechanical scanning protocol to avoid oversampling. Most of the FDML based OCT system is oversampled due to the mechanical limitation of the galvanometer. In this paper, an optimization method is applied to the backward scanning data to eliminate the distortions. Moreover, the phase and amplitude misalignment issues are also numerically corrected. The amplitude inconsistencies in the acquired interferogram are also addressed and solved.

Simulated data plays an important role in the process of developing and testing new processing methods in optical coherence tomography (OCT); they provide ground truths, and enable the generation of large amounts of data with high diversity in terms of tissue and system properties, without the burdens of experimental acquisitions. Here, we present an open-source MATLAB simulation tool that allows the generation of synthetic OCT data in an efficient and versatile way, while modeling the most relevant characteristics of the OCT signal, by combining the well-known forward model of OCT imaging with scattering and polarization properties of tissue.

A novel method is proposed for correcting aberrations and diffraction-induced artifacts in optical coherence tomography (OCT) images. The method leverages light backpropagation models combined with region-based despeckling and sharpness optimization algorithms to improve overall OCT image quality across a variety of sample types. The algorithm was applied to data acquired using a Line-Field OCT (LF-OCT) system with high numerical aperture (NA) and short depth-of-focus (DOF) Significant improvements were made to images acquired at different depths within various sample types. Such improvement holds the promise of providing ultra-high resolution volumetric OCT data without the need for depth scanning.

Currently, no diagnostic method exists to visualize and measure the airway smooth muscle (ASM) in vivo. Endoscopic Polarization Sensitive Optical Coherence Tomography (PS-OCT) has the potential to detect and quantify the ASM, by assessing tissue birefringence. We performed in vivo PS-OCT in patients (n=17) with airway diseases (asthma and interstitial lung disease) and one healthy volunteer. PS-OCT images were acquired with an in-house built distal scanning catheter (1.35mm), which allowed circumferential scan of the airways at 52 fps B-scan rate. We demonstrated PS-OCT to be a minimally invasive technique to assess ASM thickness in diseased airways relative to healthy subjects.

A framework presented in this study demonstrates an advanced high-speed and long-range wide-field swept-source OCT/OCTA system for imaging the oral cavity. By employing a k-clock calibration strategy, the system's point spread function performance is significantly improved. The achieved wide FOV of 42×42 mm² for OCT and OCTA imaging demonstrates a feasibility for imaging of oral cavity morphology and vascular distribution in a clinical setting. This advanced system provides promising potential for efficient and comprehensive in vivo oral cavity imaging, further expanding OCT's clinical applications in dentistry.

Fractional CO2 laser therapy holds potential for treating the Genitourinary Syndrome of Menopause (GSM). However, it remains controversial in female pelvic medicine. Here, we developed a non-invasive intravaginal OCT/OCT angiography (OCTA) endoscope. This system can simultaneously obtain both structural and vascular information during vaginal laser procedures. Our clinical research involved 25 post-menopausal GSM patients and took OCT/OCTA scans from four distinct vaginal regions throughout treatment sessions. With the help of deep learning, we measured Vaginal Epithelial Thickness (VET). We used OCTA to monitor Blood Vessel Density (BVD) and track vascular densification. We also conducted long-term follow-up experiments to assess the subsequential efficacy of laser treatment. Additionally, involving 45 women of varied menopausal stages, we analyzed their VET, BVD, and self-test scores to quantify the correlation between objective and subjective parameters. Our OCT/OCTA endoscope provides a noninvasive biopsy tool for gynecological research, aiding clinicians in evaluating tissue responses to laser treatments.

We present a multifunctional tethered capsule endomicroscopy (TCE) system for surveillance and potential treatment of the upper gastrointestinal (GI) tract. The system integrates dual-wavelength OCT imaging operating at 800 nm and 1300 nm, an ultracompact endoscope camera, and an ablation laser. The dual-wavelength OCT enables ultrahigh-resolution imaging of superficial layers (800 nm) as well as deep tissue layers (1300 nm), while the miniature camera provides real-time visualization of the OCT imaged and laser ablation area. The multifunctional imaging and ablation capability of the capsule was demonstrated in vivo on swine models. By combining these functions in a single device, the system offers an effective solution for endoscopic screening, diagnosis, and potential ablation treatment in the GI tract.

Lung transplantation is the last therapy option for patients with advanced lung diseases but the shortage of donor lungs currently is a large challenge. We used a polarization-sensitive optical coherence tomography (PS-OCT) system to obtain the distribution and quantification of alveolus and fibrous tissues. Our results showed that the distribution of alveolus and fibrosis was various at different locations on the lung. PS-OCT was able to provide effective quantifications of the alveolus density and size, and the fibrous tissues verified by the histology results. PS-OCT could serve as a promising tool for assessing the quality of donor lungs.

Quantification of total hemoglobin concentrations [tHb] is imperative to determine blood-related diseases such as anemia. In this work, we validate [tHb] estimates by visible-light spectroscopic optical coherence tomography (vis-sOCT) with the gold standard, intravenous blood sample analysis from a commercial avoximeter. We recruited a group of 27 volunteers and retrieved [tHb] with vis-sOCT data from the microcirculation in the skin of the left posterior forearm. Expected [tHb] were obtained from blood sample analysis by a commercial analyzer. We compare the [tHb] estimated by vis-sOCT with the expected [tHb]. Additionally, the effect of multiple factors of influence such as gender, skin tone, and epidermal thickness is investigated.

Extending OCT into meter-scale working distances has potential applications in robotic vision, surgical planning and assistance, and non-invasive medical imaging. However, long-range imaging significantly decreases the numerical aperture (NA) of the system unless a much larger aperture is used. This diminishes the signal and requires the use of Gaussian beam theory for accurate determination of the beam profile and focusing behavior. We demonstrate and experimentally validate theory for focusing a laser source to optimize both resolution and signal and discuss the impact of the “focal shift” effect in which the geometric focus of the objective and Gaussian beam waist diverge at low NA.

Calibrating intensity values of optical coherence tomography (OCT) images helps to extract valuable tissue information, such as the attenuation coefficient. To achieve this, knowledge of focus position and Rayleigh length is crucial. We study the potential of obtaining the focus position from an OCT scan acquired with a single focus setting using the system’s longitudinal chromatic aberration. Different focus positions for sub-spectrum reconstructed OCT images are realized by taking advantage of the chromatic focal shift. Ratios of these sub-spectral images are used to estimate the focus position. High-resolution B-scan reconstruction is demonstrated by coherently combining sub-spectrum confocal function corrected B-scans, followed by OCT attenuation coefficient imaging. Additionally, we present an approach to experimentally identify the chromatic focal shift from OCT data itself.

Polarization-sensitive optical coherence tomography (PS-OCT) allows the visualization of biological tissue microstructure by measuring the pathlength difference, amplitude, and polarization of backscattered light. Speckle grains complicate the visualizations due to scattering structures in tissue smaller than the PS-OCT resolution. We developed an angular compounding system to reduce speckle by rotating samples and collecting tomograms at multiple imaging angles, without modifying PS-OCT hardware or optical pathways. Tomograms were acquired, aligned with affine transformations, and averaged. This method successfully reduced the speckle and improved visualization of intensity and birefringence images.

The last few decades have seen the emergence of a huge number of optical imaging techniques. Among them, full-field optical coherence tomography (FF-OCT) has become valuable for many biomedical applications. Indeed, FF-OCT is a noninvasive and label-free imaging technique that produces high-resolution 3D microscopic images of scattering biological samples. Using FF-OCT approach for in-vivo imaging would enable the observation of cell-scale structures in living samples. Moreover, living samples have an active vascularization that can therefore be observed using Doppler imaging. We propose in this study a new FF-OCT approach that enables single-shot acquisitions which is suitable for in-vivo and Doppler imaging.

The progress made in optical coherence tomography (OCT) has facilitated the noninvasive observation of biological structures. While Fourier-Domain Full-Field OCT (FD-FF-OCT) enables rapid collection of volumetric data, it encounters challenges arising from coherent noise. To tackle this issue, we introduced STOC-T ľ spatial-temporal optical coherence tomography, which utilizes spatial phase modulation alleviate crosstalk interference. This research thoroughly investigates the characteristics of optical fibers in the context of STOC-T, crafted for volumetric imaging and the mitigation of coherent crosstalk noise. We assess the effects of multimode fibers and execute a performance analysis employing a purpose-designed scattering phantom. Our presentation encompasses detailed insights into the experimental arrangement and outcomes, emphasizing noise reduction. We also explore hurdles like determining the ideal fiber length, analyzing speckle patterns, and ensuring signal quality. Through a meticulous selection process of multimode fibers, we amplify the potential of STOC-T for structural and functional retinal imaging.

We have developed a novel OCT needle probe appropriate for imaging deep within muscle and assessed its ability to quantify intramuscular fat. The OCT needle probe is designed to operate in common-path mode, removing the need for a pathlength-matched reference arm. Scanning is performed by manually inserting and removing the needle from the muscle, using a speckle decorrelation algorithm to correct for variation in the speed at which the needle is moved. Four OCT needle probes were integrated into a single handpiece for simultaneous measurement at different locations in the muscle and multiplexed to a single OCT scanner. The OCT data is segmented using a convolutional neural network, automatically quantifying the percentage of intramuscular fat. The system was assessed on muscle from 275 lambs carcases and measured the percentage of intramuscular fat with an average absolute error of 0.65%.

We present a multiscale imaging system in which a conventional microscope stand is inserted in the sample arm of a Spectral-Domain Optical Coherence Microscopy (SD-OCM) system. The instrument was designed to facilitate localisation and maturity assessment of oocytes and early-stage embryos to be used for in vitro fertilisation (IVF), by a versatile and easy switch of microscope objectives. The dispersion variation due to change of the microscope objectives is dealt with by employing the Complex Master Slave procedure that enables correct operation without the use of matching glass in the reference arm for dispersion compensation.

We present a novel optical coherence microscopy (OCM) setup capable of generating dynamic intensity, phase, and pseudo-spectroscopic contrast with single-shot full-field video-rate imaging called BiTe (Bichromatic Tetraphasic) full-field OCM. BiTe OCM uniquely uses the phase-shifting properties of AR coatings outside their rated bandwidths to create four distinct phase shifts between the sample and reference. BiTe OCM utilizes the raw data effectively to generate three complementary contrasts. We demonstrate BiTe OCM to observe cellular dynamics, capture the spectroscopic hemodynamics of tumors, and image the microstructure of fall foliage in two different color channels.

Optical coherence elastography has emerged as a powerful non-invasive tool for quantifying tissue biomechanics; however, traditional elastography techniques have been limited by a maximum wave frequency below 10 kHz, resulting in compromised temporal and spatial resolution. Here, we present an optical coherence elastography technique capable of extending the frequency range to the MHz regime. Our system enables the measurement of materials across a broad spectrum of stiffness, from the hydrogel to the bone. The rich broadband dispersion of surface waves allowed us to profile the depth-dependent shear modulus in cartilages ex vivo and human skin in vivo with sub-mm anatomical resolution.

Dynamic OCT (dOCT) method uses time-depended OCT signal fluctuations arising from cellular motion in biological tissues to generate additional contrast in OCT images. So far dOCT studies have been conducted with Point-scanning and Full-field dOCT platforms, each of them offering advantages and limitations such as FOV, volumetric scan and phase stability. Here we present the first line-field dOCT platform (LF-dOCT), which allows low jitter B-scan dOCT signal collections. A Powell lens is utilized for uniform line-direction system sensitivity. Various biological tissues have been imaged with the LF-dOCT platform to validate its ability to detect cellular and intracellular dynamics.

Strain retrieval from maps of displacement due to mechanical loading is central to optical coherence elastography. However, displacement data is usually highly oscillatory since it is derived from the phase difference between OCT images. Oscillations limit sensitivity and signal to noise ratio of retrieved strain. We present a novel strain retrieval method which determines the unique spectral domain transformation and transverse displacement compensation that maps the unloaded A-scan to the loaded A-scan, exactly, for regions of constant strain. Our novel method of strain retrieval has a higher sensitivity and signal to noise ratio than existing approaches.

Dynamic optical coherence tomography (dOCT) uses signal fluctuation for contrasting different cellular and acellular components in living biological tissue. The autocorrelation or Fourier transform of time series of OCT measurements are converted to a color contrast. However, a quantitative analysis is still challenging. Here we investigate theoretically, how noise of the OCT measurement influences the fluctuation spectra. Probability functions are derived for the different components in the spectra and validated by numerical simulation. With an appropriate calibration of the OCT device a separation of OCT noise and a quantification the dynamic OCT should be feasible.

Dynamic Optical Coherence Tomography combines high resolution tomographic imagery with a cell specific contrast by Fourier analysis. However, the conversion from frequency space into RGB images by binning requires a priori knowledge and artifacts due to global movements provide another obstacle for in vivo application. We could show that an automated binning based on the Neural Gas algorithm can yield the highest spectral contrast without a priori knowledge and that motion artifacts can be reduced with shorter sequence lengths. Imaging murine airways, we observed that even just 6 frames are enough to generate dOCT images without losing important image information.

Myopia is a major cause of visual impairment worldwide. A reduction in scleral rigidity is hypothesized as an effective biomarker of degenerative myopia. However, no current technique is able to characterize posterior rigidity in-vivo and non-invasively. We present a novel methodology using phase-sensitive optical coherence tomography (PhS-OCT) to measure scleral rigidity. The rigidity is reflected by the magnitude of pulsatile fundus motion that is measured by PhS-OCT. A preliminary clinical experiment confirmed our methodology. Our findings can offer potential new diagnostic methods for degenerative myopia and related diseases, as well as applications for evaluating therapies that alter scleral mechanical properties.

Intraocular pressure is the main risk factor for glaucoma and is primarily regulated by resistance in the conventional outflow pathway. Although visualizing the anatomical structure of the outflow pathway has the potential to guide glaucoma treatment, noninvasive imaging of the outflow pathway is difficult owing to its small size and deep position within the sclera. To address challenges in imaging the outflow pathway, we developed a robotic visible-light optical coherence tomography system. We reconstruct the full 360 degrees of the pathway, finding segmental anatomical differences in the anatomical structure.

Owing to its high axial resolution, optical coherence tomography (OCT) has the capacity to reconstruct 3-D structures with micron precision. However, due to tissue scattering, limited field-of-view per scan, and limited depth of focus, it is challenging to reconstruct large, complex volumetric objects that do not fit within one OCT scan. To address this problem, we developed novel montaging schemes to combine OCT volumes acquired from different angles and positions. We demonstrate the effectiveness of this method by measuring the aqueous volume of the anterior and posterior chambers of the mouse eye.

Poor-fitting contact lenses are a common cause of eye irritation, potentially leading to keratitis. We seek to improve custom contact lens manufacturing by integrating visible-light optical coherence tomography (vis-OCT) imaging and 3D printing with micrometer resolution and nanometer surface smoothness. We show that our customized contact lens fits well on various corneas and improves the quality of OCT imaging of the rodent retina.

This study reports the results of our recent in vivo study using attenuation-based quantitative OCT for intra-operative mapping of brain cancer in patients. A total of 34 brain cancer patients were recruited, and an optimal optical attenuation threshold of 4.3 mm-1 was established with cancer of lower values. We achieved an excellent specificity of over 98%, with a corresponding sensitivity of over 95% for both high-grade and low-grade cancers. Applying the established optical attenuation threshold to infiltrated tissues, we were able to identify regions of cancers qualitatively matching the neuropathologist’s assessment. The study suggests that optical attenuation-based quantitative OCT represents a promising technology for intraoperative brain cancer detection and

Reconstructed depth-resolved optic axis orientation obtained by catheter based PSOCT in a tissue volume informs on the orientation of the white matter fiber bundles in the brain, owing to the birefringence of myelinated axons. The physical organization of white matter also leads to anisotropic diffusion of water molecules, which is the basis of dMRI for non-invasive imaging of the three-dimensional orientation of white matter fiber bundles. Having access to fiber orientation in both imaging modalities, we are trying to map the depth-resolved birefringence and optic axis orientation to the larger scale dMRI as well as an atlas of brain anatomy.

In this study, we present a refined xf-irOCM system and post-processing pipeline for detailed investigation of cerebral vessel structure and function. Our method uses deep learning for 3D segmentation of high- resolution angiograms and accurately estimates flow velocities across the cerebral vasculature. Our graph- based approach uniquely enables multiscale assessments, capturing data from intricate capillaries to broad network relationships. Specifically, it aids in understanding vascular alterations in neurovascular pathologies, such as stroke. Our approach will pave the way for future microvasculature studies, offering promising avenues for further research into neurovascular diseases.

Multicellular tumour spheroids have recently become important tools to investigate different stages of cancer development due to their 3D nature. We propose dynamic optical coherence microscopy (OCM) as a label-free low coherence interferometric technique for 3D characterization of morphology and cell motility during cell migration in multicellular cancer cell spheroids. We integrate dynamic OCM with confocal fluorescence microscopy (CFM) to validate and co-register the subcellular-scale endogenous contrast generated by dynamic OCM signal with sub-cellular features such as cell nucleus and membrane. We apply dynamic OCM integrated with CFM to scan metastatic and non-metastatic breast cancer cell spheroids embedded in gelatin-methacryloyl (GelMA) hydrogel and demonstrate that dynamic OCM provides high-contrast morphological imaging equivalent to that of confocal fluorescence in cancer cell spheroids. We use dynamic OCM to visualize different phases of cell migration such as invadopodia formation, cells breaking off from the primary tumour model, and migrating cells presenting a spindle-like shape, and to characterize cell motility at different stages.

Prostate cancer is the most prevalent cancer among Canadian men. Historically, cancer has been studied using 2D cell culture methods, but recently 3D cell culture methods like tumor spheroids have been become popular to better replicate the physiological environment of the body. We developed ultrahigh resolution line-field OCT (LF-OCT) technology and a dynamic OCT (dOCT) method to image tumor spheroids and identify areas of necrosis, for comparison with computer simulation. The LF-OCT system has sufficient spatial resolution to identify individual cells and capture the cellular level of activity while the dOCT algorithm can resolve the dynamic activity accurately. The study will be extended to investigate the spatial and temporal development of necrosis in PC3 prostate tumor spheroids under different conditions like such as temperature, pressure, and culture medium type. These results will be used to inform and extend the model to an in vivo environment.

Airway organoid is a three-dimensional (3D) epithelial models for human airway where ciliary behaviors are important to indicate their health state. A real-time noninvasive imaging technique to assess the ciliary movement in airway organoids in vitro is highly demanded. Here we propose a new imaging approach to monitor the ciliary beating of airway organoids noninvasively using dynamic contrast optical coherence tomography (DyC-OCT). We employed DyC-OCT to measure the ciliary beating frequency (CBF) and to highlight ciliary beating clusters distribution in 3D. With DyC-OCT, we monitored the morphology change in ciliary beating clusters as well as the change in CBF from the same organoids during the respiratory syncytial virus (RSV) infection. The capability of DyC-OCT in ciliary beating monitoring shows its potential for respiratory disease diagnosis and pathogen assay.

The COVID-19 pandemic increased interest in large-scale disinfection of public spaces using UV-C germicidal lights. However, excessive UV-C exposure harms the skin and cornea. Although novel Far UV-C (180 – 230 nm) sources are widely considered safe, an easy and non-invasive experimental assessment of UV damage could help to exclude any health risks of extensive Far UV-C exposure. Dynamic-microscopic optical coherence tomography (dmOCT) is a non-invasive, label-free technique providing sub-cellular resolution images. This study used dmOCT to confirm Far UV-C's skin safety with in-vitro models. Results showed no signs of keratinocyte damage, while UV-B exposed samples express immediate structural and functional changes.

We proposed a neural network to generate volumetric dynamic optical coherence tomography (DOCT) from small-number OCT frames. In this study, we used a DOCT method (i.e., logarithmic OCT intensity variance; LIV) and it is applied to tumor spheroid samples. A U-Net-based NN model was trained to generate a LIV image from only 4 OCT frames. The NN-generated LIV was subjectively and objectively compared with conventional LIV images generated from 32 frames. The comparison showed a high similarity between the NN-generated LIV and the conventional LIV. This NN-based method enabled volumetric DOCT with only 6.55 s acquisition time.

Conventional Optical Coherence Tomography (OCT) suffers from the frame-rate/resolution tradeoff, whereby increasing image resolution leads to decreases in the maximum achievable frame rate. We extended the conventional probabilistic adaptive scanning technique that overcomes this tradeoff with machine-learning-based scene prediction and kinodynamic path planning based on the Clustered Traveling Salesperson Problem. In online imaging, we found that our new technique produces an equivalent frame rate speed-up as previously reported while creating higher quality output OCT images. These results generalized across scenes of varying types, including those of intrasurgical relevance.

Removing artifacts, reducing noise, and shaping the point spread function (PSF) in Fourier-domain optical coherence tomography (FD-OCT) are critical for optimal image quality. Existing methods require frequent calibration, fail for certain background signals, or introduce various artifacts such as side-lobes for the lateral PSF. Here, we present methods for numerical optimization of FD-OCT image quality and demonstrate them for different OCT systems. Based on simulations, we also show that suitable algorithms for line-field and full-field FD-OCT can perform self-balancing, effectively eliminating relative intensity noise (RIN) and thus eliminating the need for balanced detection with a second detector or camera.

Optical coherence tomography (OCT) is capable of angstrom-scale vibrometry of particular interest to researchers of auditory mechanics. We develop a method for compressed sensing vibrometry using OCT that significantly reduces acquisition time for dense motion maps. Our method, based on total generalized variation with uniform sub-sampling, can reduce the number of samples needed to measure motion maps by a factor of ten with less than 5% normalized mean square error when tested on a diverse set of in vivo measurements from the gerbil cochlea. This opens up the possibility for more complex in vivo experiments for cochlear mechanics.

Parametric imaging of attenuation using optical coherence tomography (OCT) is a powerful tool for assessing changes in tissue morphology associated with disease. Several models are available for extracting the optical properties of tissue from OCT images. However, the accuracy of these models and their dependence on the tissue optical properties has yet to be established. Here, we investigate the accuracy of several OCT models and assess their suitability for extracting optical properties. We establish that the single scattering models produce more precise results (lower variance), but the EHF model is more accurate. Furthermore, the accuracy of the single scattering model degrades as the scattering coefficient and thickness of the tissue increase. We intend that the results of this study will aid in the development of standardized protocols for extracting optical properties from OCT images.

We present the development of an optical coherence tomography (OCT) catheter designed for in vivo subsurface imaging during colonoscopy, along with the results of a clinical pilot study involving 36 subjects to assess its ability to characterize colorectal polyps real-time. High-resolution cross-sectional OCT imaging of polyp microsctructure revealed distinct morphological structures that correlated with histological findings, including tubular adenoma, tubulovillous adenoma, sessile serrated polyps, and cancer. To enhance the in vivo diagnostic capabilities, we integrated a Vision Transformer (ViT) based deep learning classifier to differentiate between cancerous and complex benign polyps, and achieved a 100% accuracy for 5 test cases. Our findings suggest that the OCT catheter combined with deep learning complements standard-of-care imaging and has the potential to enhance real-time polyp characterization and improve clinical decision-making.