Broadband wavelength-swept Cr4+:YAG crystal fiber laser.

We present a broadband wavelength-swept laser using a 16-µm-core-diameter Cr4+:YAG crystal fiber as the gain medium. The laser-diode-pumped crystal fiber laser has a threshold of only 102 mW due to the low propagation loss and high heat dissipation efficiency. The laser achieves a sweeping wavelength range of 134 nm, centered around 1425 nm, with a scanning speed of 163 k nm/s. Notably, the cross-polarization-coupled excited state absorption of the signal wavelength constrained the long-wavelength lasing limit. This laser has the potential for swept source optical coherence tomography applications, providing an axial resolution of 11.4 µm.

Ultra-broadband wavelength-swept Ti:sapphire crystal fiber laser.

An ultra-broadband wavelength-swept laser (WSL) was generated using glass-clad Ti:sapphire crystal fiber as the gain media. Due to the low signal propagation loss of the crystal fiber, the swept laser has a tuning bandwidth of 250 nm (i.e., 683 nm to 933 nm) at a repetition rate of 1200 Hz. The steady-state and pulsed dynamics of the WSL were analyzed. The 0.018-nm instantaneous linewidth corresponds to a 3-dB coherence roll-off of 7 mm. When using the laser for swept-source optical coherence tomography, an estimated axial resolution of 1.8 µm can be achieved.

Spectroscopic characterization of Si/Mo thin-film stack at extreme ultraviolet range.

A noninvasive method for characterizing Si/Mo thin-film stack thickness and its complex transfer function using common-path optical coherence tomography is proposed, analyzed, and experimentally demonstrated. A laser-produced plasma (LPP)-based extreme ultraviolet (EUV) source was excited by a four-stage nanosecond Yb:fiber laser amplifier with a pulse energy of 1.01 mJ. The tabletop LPP EUV source was compact and stable for generating the EUV interference fringes. The measured complex transfer function of the Si/Mo stack was verified near the pristine 13.5-nm wavelength range.

Compact millijoule diode-seeded two-stage fiber master oscillator power amplifier using a multipass and forward pumping scheme.

The multipass scheme for a diode-seeded fiber master oscillator power amplifier with a nanojoule-to-millijoule output energy level at a repetition rate of <100 kHz is numerically analyzed for comparison to an experimental benchmark. For a 6/125 single-mode preamplifier with a small input energy (<1 nJ), there is a significant improvement in the output energy from 0.7% to 80% and 95% of the maximum extractable energy using the double-pass and four-pass schemes, respectively. For a 30/250 large-mode-area power amplifier using the double-pass and forward pumping scheme, the required input energy is decreased from 100 µJ to 18 µJ for millijoule energy extraction with accompanying Stokes waves of less than 10% of the total energy. The system based on the full master oscillator power amplifier configuration with an output energy exceeding millijoule level can be optimally simplified to two stages for commercialization.

Laser-diode pumped glass-clad Ti:sapphire crystal fiber laser.

Efficient glass-clad crystal fiber (CF) lasers were demonstrated using a Ti:sapphire crystalline core as the gain medium. With a core diameter of 18 µm, the laser diode (LD) pump source can be effectively coupled and guided throughout the crystal fiber for a low threshold and high slope efficiency laser operation. The advantage of high heat dissipation efficiency of the fiber structure can be derived from the low core temperature rising measurement (i.e., 17 K/W) with passive cooling. At an output transmittance of 23%, the lowest absorbed threshold of 118.2 mW and highest slope efficiency of 29.6% were achieved, with linear laser polarization.

Broadband Ce/Cr-doped crystal fibers for high axial resolution OCT light source.

The fabrication and characteristics of Ce/Cr-doped crystal fibers employing drawing tower technique are reported. The fluorescence spectrum of the Ce/Cr fibers at the core diameter ranging from 10 to 21 µm exhibited a 200-nm near-Gaussian broadband emission which enabled to provide an axial resolution of 1.8-µm and a power density of 79.1 nW/nm. The proposed broadband Ce/Cr-doped crystal fibers may be provided as a high-resolution light source for the use in optical coherence tomography system as well as industrial inspection and biomedical imaging applications.

Broadband and high-brightness light source: glass-clad Ti:sapphire crystal fiber.

High-brightness near-infrared broadband amplified spontaneous emission (ASE) was generated by glass-clad Ti:sapphire crystal fibers, which were developed using the co-drawing laser-heated pedestal growth method. As much as 29.2 mW of ASE power was generated using 520 nm laser diodes as the excitation source on an a-cut, 18 µm core-diameter Ti:sapphire crystal fiber (CF). The 3 dB bandwidth was 163.8 nm, and the radiance was 53.94 W·mm(-2) sr(-1). The propagation loss of the glass-clad sapphire CF measured using the cutback method was 0.017 cm(-1) at 780 nm. For single-mode applications, more than 100 µW of power was coupled into a SM600 single-mode fiber.

Spectroscopic measurement of absorptive thin films by spectral-domain optical coherence tomography.

A non-invasive method for measuring the refractive index, extinction coefficient and film thickness of absorptive thin films using spectral-domain optical coherent tomography is proposed, analyzed and experimentally demonstrated. Such an optical system employing a normal-incident beam of light exhibits a high spatial resolution. There are no mechanical moving parts involved for the measurement except the transversal scanning module for the measurement at various transversal locations. The method was experimentally demonstrated on two absorptive thin-film samples coated on transparent glass substrates. The refractive index and extinction coefficient spectra from 510 to 580 nm wavelength range and film thickness were simultaneously measured. The results are presented and discussed.

High-resolution full-field optical coherence tomography microscope for the evaluation of freshly excised skin specimens during Mohs surgery: A feasibility study.

Histopathology for tumor margin assessment is time-consuming and expensive. High-resolution full-field optical coherence tomography (FF-OCT) images fresh tissues rapidly at cellular resolution and potentially facilitates evaluation. Here, we define FF-OCT features of normal and neoplastic skin lesions in fresh ex vivo tissues and assess its diagnostic accuracy for malignancies. For this, normal and neoplastic tissues were obtained from Mohs surgery, imaged using FF-OCT, and their features were described. Two expert OCT readers conducted a blinded analysis to evaluate their diagnostic accuracies, using histopathology as the ground truth. A convolutional neural network was built to distinguish and outline normal structures and tumors. Of the 113 tissues imaged, 95 (84%) had a tumor (75 basal cell carcinomas [BCCs] and 17 squamous cell carcinomas [SCCs]). The average reader diagnostic accuracy was 88.1%, with a sensitivity of 93.7%, and a specificity of 58.3%. The artificial intelligence (AI) model achieved a diagnostic accuracy of 87.6 ± 5.9%, sensitivity of 93.2 ± 2.1%, and specificity of 81.2 ± 9.2%. A mean intersection-over-union of 60.3 ± 10.1% was achieved when delineating the nodular BCC from normal structures. Limitation of the study was the small sample size for all tumors, especially SCCs. However, based on our preliminary results, we envision FF-OCT to rapidly image fresh tissues, facilitating surgical margin assessment. AI algorithms can aid in automated tumor detection, enabling widespread adoption of this technique.

Training With Uncertain Annotations for Semantic Segmentation of Basal Cell Carcinoma From Full-Field OCT Images.

Semantic segmentation of basal cell carcinoma (BCC) from full-field optical coherence tomography (FF-OCT) images of human skin has received considerable attention in medical imaging. However, it is challenging for dermatopathologists to annotate the training data due to OCT's lack of color specificity. Very often, they are uncertain about the correctness of the annotations they made. In practice, annotations fraught with uncertainty profoundly impact the effectiveness of model training and hence the performance of BCC segmentation. To address this issue, we propose an approach to model training with uncertain annotations. The proposed approach includes a data selection strategy to mitigate the uncertainty of training data, a class expansion to consider sebaceous gland and hair follicle as additional classes to enhance the performance of BCC segmentation, and a self-supervised pre-training procedure to improve the initial weights of the segmentation model parameters. Furthermore, we develop three post-processing techniques to reduce the impact of speckle noise and image discontinuities on BCC segmentation. The mean Dice score of BCC of our model reaches 0.503±0.003, which, to the best of our knowledge, is the best performance to date for semantic segmentation of BCC from FF-OCT images.

Fluorescence enhancement in broadband Cr-doped fibers fabricated by drawing tower.

The fluorescence enhancement in broadband Cr-doped fibers (CDFs) fabricated by a drawing tower with a redrawn powder-in-tube preform is proposed and demonstrated. The CDFs after heat treatment exhibited Cr4⁺ emission enhancement with spectral density of 200 pW/nm, verified by the formation of α-Mg2SiO4 nanocrystalline structures in the core of CDFs. The high fluorescence achievement in the CDFs is essential to develop a broadband CDF amplifier for next-generation optical communication systems.

Toward cell nuclei precision between OCT and H&E images translation using signal-to-noise ratio cycle-consistency.

Medical image-to-image translation is often difficult and of limited effectiveness due to the differences in image acquisition mechanisms and the diverse structure of biological tissues. This work presents an unpaired image translation model between in-vivo optical coherence tomography (OCT) and ex-vivo Hematoxylin and eosin (H&E) stained images without the need for image stacking, registration, post-processing, and annotation. The model can generate high-quality and highly accurate virtual medical images, and is robust and bidirectional. Our framework introduces random noise to (1) blur redundant features, (2) defend against self-adversarial attacks, (3) stabilize inverse conversion, and (4) mitigate the impact of OCT speckles. We also demonstrate that our model can be pre-trained and then fine-tuned using images from different OCT systems in just a few epochs. Qualitative and quantitative comparisons with traditional image-to-image translation models show the robustness of our proposed signal-to-noise ratio (SNR) cycle-consistency method.

Integration of cellular-resolution optical coherence tomography and Raman spectroscopy for discrimination of skin cancer cells with machine learning.

Significance: An integrated cellular-resolution optical coherence tomography (OCT) module with near-infrared Raman spectroscopy was developed on the discrimination of various skin cancer cells and normal cells. Micron-level three-dimensional (3D) spatial resolution and the spectroscopic capability on chemical component determination can be obtained simultaneously. Aim: We experimentally verified the effectiveness of morphology, intensity, and spectroscopy features for discriminating skin cells. Approach: Both spatial and spectroscopic features were employed for the discrimination of five types of skin cells, including keratinocytes (HaCaT), the cell line of squamous cell carcinoma (A431), the cell line of basal cell carcinoma (BCC-1/KMC), primary melanocytes, and the cell line of melanoma (A375). The cell volume, compactness, surface roughness, average intensity, and internal intensity standard deviation were extracted from the 3D OCT images. After removing the fluorescence components from the acquired Raman spectra, the entire spectra (600 to 2100 cm-1) were used. Results: An accuracy of 85% in classifying five types of skin cells was achieved. The cellular-resolution OCT images effectively differentiate cancer and normal cells, whereas Raman spectroscopy can distinguish the cancer cells with nearly 100% accuracy. Conclusions: Among the OCT image features, cell surface roughness, internal average intensity, and standard deviation of internal intensity distribution effectively differentiate the cancerous and normal cells. The three features also worked well in sorting the keratinocyte and melanocyte. Using the full Raman spectra, the melanoma and keratinocyte-based cell carcinoma cancer cells can be discriminated effectively.

Depth-dependent in vivo human skin backscattering spectra extraction from full-field optical coherence tomography.

With homemade active crystalline fibers, we generated bright and broadband light sources for full-field optical coherence tomography, offering deep penetration into skin tissues with cellular resolution at a high frame rate. Extraction of backscattered spectra from the tissue has potential applications in biomedicine. The hysteresis nonlinearity of the piezoelectric transducer actuating the Mirau interferometer has been greatly reduced by a feedforward compensation approach. The linearized hysteresis response enables us to extract depth-dependent spectra accurately. To validate, the complex dispersion of a fused silica plate was characterized with 2% error. Further validation on an in vitro setting, the backscattered spectra from indocyanine green pigment and nonpigmented microspheres were obtained and verified. For in vivo skin measurement, the backscattered spectra show depth-dependent spectral shift and bandwidth variation due to the complex skin anatomy and pigment absorption. Such a high-speed spectra acquisition of in vivo deep tissue backscattering could lead to disease diagnosis in clinical settings.

Identification of Sex and Age from Macular Optical Coherence Tomography and Feature Analysis Using Deep Learning.

PURPOSE: To develop deep learning models for identification of sex and age from macular optical coherence tomography (OCT) and to analyze the features for differentiation of sex and age. DESIGN: Algorithm development using database of macular OCT. METHODS: We reviewed 6147 sets of macular OCT images from the healthy eyes of 3134 individuals from a single eye center in Taiwan. Deep learning-based algorithms were used to develop models for the identification of sex and age, and 10-fold cross-validation was applied. Gradient-weighted class activation mapping was used for feature analysis. RESULTS: The accuracy for sex prediction using deep learning from macular OCT was 85.6% ± 2.1% compared with accuracy of 61.9% using macular thickness and 61.4% ± 4.0% using deep learning from infrared fundus photography (P < .001 for both). The mean absolute error for age prediction using deep learning from macular OCT was 5.78 ± 0.29 years. A thorough analysis of the prediction accuracy and the gradient-weighted class activation mapping showed that the cross-sectional foveal contour lead to a better sex distinction than macular thickness or fundus photography, and the age-related characteristics of macula were on the whole layers of retina rather than the choroid. CONCLUSIONS: Sex and age could be identified from macular OCT using deep learning with good accuracy. The main sexual difference of macula lies in the foveal contour, and the whole layers of retina differ with aging. These novel findings provide useful information for further investigation in the pathogenesis of sex- and age-related macular structural diseases.

Non-invasive characterization of the domain boundary and structure properties of periodically poled ferroelectrics.

Shaping the ferroelectric domains as waveguide, grating, lens, and prism are key to the successful penetration of periodically-poled ferroelectrics on various wavelength conversion applications. The complicated structures are, however, difficult to be fully characterized, especially the unexpected index contrast at the anti-parallel domain boundaries are typical in the order of 10(-4) or less. An ultrahigh resolution optical coherence tomography was employed to fully characterize the domain boundary and structure properties of a periodically-poled lithium niobate (PPLN) waveguide with an axial resolution of 0.68 µm, an transversal resolution of 3.2 µm, and an index contrast sensitivity of 4x10(-7). The anti-parallel domain uniformity can clearly be seen non-invasively. Dispersion of the ferroelectric material was also obtained from 500 to 750 nm.

Microstructural and microspectral characterization of a vertically aligned liquid crystal display panel.

The microstructural and microspectral characteristics of a vertically aligned liquid crystal display (VA-LCD) panel were obtained noninvasively for the first time. With 1 µm axial and 2 µm transversal resolutions, the cell gap profile beneath the patterned thin-film transistor of the VA-LCD panel can clearly be resolved. The thicknesses of the multiple thin-film layers and the embedded defects can also be unveiled. As far as spectral response is concerned, the light transmittance at the layer boundaries can be estimated from the measured reflectance, which is crucial information for the design of a highly transmissive panel. The color shift of the VA-LCD panel due to fabrication error was evaluated.

Efficient and low-threshold Cr4+:YAG double-clad crystal fiber laser.

A significant advancement of cw lasing in Cr4+:Y3Al5O12 (Cr4+:YAG) double-clad crystal fiber grown by the codrawing laser-heated pedestal growth technique was demonstrated at RT. The optical-to-optical slope efficiency of 33.9% is the highest, to the best of our knowledge, among all Cr4+:YAG lasers, whether they are in bulk or fiber forms. The low-threshold lasing of 78.2 mW and high efficiency are in good agreement with the simulation. The keys to the high laser efficiency are twofold: one is the improved Cr4+ emission cross section and fluorescence lifetime due to release of the strain on the distorted Cr4+ tetrahedron, which also mitigates photobleaching in Cr4+:YAG; the other is the improved core uniformity at long fiber lengths. In addition, because of the low threshold, the impact of excited state absorption of the pump light is significantly reduced. The effects of crystal-orientation, self-selected, and pump-dependent linear polarization states were also addressed.

Dermal epidermal junction detection for full-field optical coherence tomography data of human skin by deep learning.

Full-field optical coherence tomography (FF-OCT) has been developed to obtain three-dimensional (3D) OCT data of human skin for early diagnosis of skin cancer. Detection of dermal epidermal junction (DEJ), where melanomas and basal cell carcinomas originate, is an essential step for skin cancer diagnosis. However, most existing DEJ detection methods consider each cross-sectional frame of the 3D OCT data independently, leaving the relationship between neighboring frames unexplored. In this paper, we exploit the continuity of 3D OCT data to enhance DEJ detection. In particular, we propose a method for noise reduction of the training data and a multi-directional convolutional neural network to predict the probability of epidermal pixels in the 3D OCT data, which is more stable than one-directional convolutional neural network for DEJ detection. Our crosscheck refinement method also exploits the domain knowledge to generate a smooth DEJ surface. The average mean error of the entire DEJ detection system is approximately 6 µm.

Mirau-type full-field optical coherence tomography with switchable partially spatially coherent illumination modes.

A crystalline-fiber-based Mirau-type full-field optical coherence tomography (FF-OCT) system utilizing two partially coherent illumination modes is presented. Using a diode-pumped Ti:sapphire crystalline fiber with a high numerical aperture, spatially-incoherent broadband emission can be generated with high radiance. With two modes of different spatial coherence settings, either deeper penetration depth or higher B-scan rate can be achieved. In a wide-field illumination mode, the system functions like FF-OCT with partially coherent illumination to improve the penetration depth. In a strip-field illumination mode, a compressed field is generated on the sample, and a low-speckle B-scan can be acquired by compounding pixel lines within.