Deep learning-based photodamage reduction on harmonic generation microscope at low-level optical power.

The trade-off between high-quality images and cellular health in optical bioimaging is a crucial problem. We demonstrated a deep-learning-based power-enhancement (PE) model in a harmonic generation microscope (HGM), including second harmonic generation (SHG) and third harmonic generation (THG). Our model can predict high-power HGM images from low-power images, greatly reducing the risk of phototoxicity and photodamage. Furthermore, the PE model trained only on normal skin data can also be used to predict abnormal skin data, enabling the dermatopathologist to successfully identify and label cancer cells. The PE model shows potential for in-vivo and ex-vivo HGM imaging.

Quantitative Evaluation of Caries and Calculus with Ultrahigh-Resolution Optical Coherence Tomography.

Dental caries on the crown's surface is caused by the interaction of bacteria and carbohydrates, which then gradually alter the tooth's structure. In addition, calculus is the root of periodontal disease. Optical coherence tomography (OCT) has been considered to be a promising tool for identifying dental caries; however, diagnosing dental caries in the early stage still remains challenging. In this study, we proposed an ultrahigh-resolution OCT (UHR-OCT) system with axial and transverse resolutions of 2.6 and 1.8 µm for differentiating the early-stage dental caries and calculus. The same teeth were also scanned by a conventional spectral-domain OCT (SD-OCT) system with an axial resolution of 7 µm. The results indicated that early-stage carious structures such as small cavities can be observed using UHR-OCT; however, the SD-OCT system with a lower resolution had difficulty identifying it. Moreover, the estimated surface roughness and the scattering coefficient of enamel were proposed for quantitatively differentiating the different stages of caries. Furthermore, the thickness of the calculus can be estimated from the UHR-OCT results. The results have demonstrated that UHR-OCT can detect caries and calculus in their early stages, showing that the proposed method for the quantitative evaluation of caries and calculus is potentially promising.

Improving Individual Brain Age Prediction Using an Ensemble Deep Learning Framework.

Brain age is an imaging-based biomarker with excellent feasibility for characterizing individual brain health and may serve as a single quantitative index for clinical and domain-specific usage. Brain age has been successfully estimated using extensive neuroimaging data from healthy participants with various feature extraction and conventional machine learning (ML) approaches. Recently, several end-to-end deep learning (DL) analytical frameworks have been proposed as alternative approaches to predict individual brain age with higher accuracy. However, the optimal approach to select and assemble appropriate input feature sets for DL analytical frameworks remains to be determined. In the Predictive Analytics Competition 2019, we proposed a hierarchical analytical framework which first used ML algorithms to investigate the potential contribution of different input features for predicting individual brain age. The obtained information then served as a priori knowledge for determining the input feature sets of the final ensemble DL prediction model. Systematic evaluation revealed that ML approaches with multiple concurrent input features, including tissue volume and density, achieved higher prediction accuracy when compared with approaches with a single input feature set [Ridge regression: mean absolute error (MAE) = 4.51 years, R 2 = 0.88; support vector regression, MAE = 4.42 years, R 2 = 0.88]. Based on this evaluation, a final ensemble DL brain age prediction model integrating multiple feature sets was constructed with reasonable computation capacity and achieved higher prediction accuracy when compared with ML approaches in the training dataset (MAE = 3.77 years; R 2 = 0.90). Furthermore, the proposed ensemble DL brain age prediction model also demonstrated sufficient generalizability in the testing dataset (MAE = 3.33 years). In summary, this study provides initial evidence of how-to efficiency for integrating ML and advanced DL approaches into a unified analytical framework for predicting individual brain age with higher accuracy. With the increase in large open multiple-modality neuroimaging datasets, ensemble DL strategies with appropriate input feature sets serve as a candidate approach for predicting individual brain age in the future.

3-D Res-CapsNet convolutional neural network on automated breast ultrasound tumor diagnosis.

PURPOSE: We propose a 3-D tumor computer-aided diagnosis (CADx) system with U-net and a residual-capsule neural network (Res-CapsNet) for ABUS images and provide a reference for early tumor diagnosis, especially non-mass lesions. METHODS: A total of 396 patients with 444 tumors (226 malignant and 218 benign) were retrospectively enrolled from Sun Yat-sen University Cancer Center. In our CADx, preprocessing was performed first to crop and resize the tumor volumes of interest (VOIs). Then, a 3-D U-net and postprocessing were applied to the VOIs to obtain tumor masks. Finally, a 3-D Res-CapsNet classification model was executed with the VOIs and the corresponding masks to diagnose the tumors. Finally, the diagnostic performance, including accuracy, sensitivity, specificity, and area under the curve (AUC), was compared with other classification models and among three readers with different years of experience in ABUS review. RESULTS: For all tumors, the accuracy, sensitivity, specificity, and AUC of the proposed CADx were 84.9 %, 87.2 %, 82.6 %, and 0.9122, respectively, outperforming other models and junior reader. Next, the tumors were subdivided into mass and non-mass tumors to validate the system performance. For mass tumors, our CADx achieved an accuracy, sensitivity, specificity, and AUC of 85.2 %, 88.2 %, 82.3 %, and 0.9147, respectively, which was higher than that of other models and junior reader. For non-mass tumors, our CADx achieved an accuracy, sensitivity, specificity, and AUC of 81.6 %, 78.3 %, 86.7 %, and 0.8654, respectively, outperforming the two readers. CONCLUSION: The proposed CADx with 3-D U-net and 3-D Res-CapsNet models has the potential to reduce misdiagnosis, especially for non-mass lesions.

Multiscale Molecular Simulation of Solution Processing of SMDPPEH: PCBM Small-Molecule Organic Solar Cells.

Solution-processed small-molecule organic solar cells are a promising renewable energy source because of their low production cost, mechanical flexibility, and light weight relative to their pure inorganic counterparts. In this work, we developed a coarse-grained (CG) Gay-Berne ellipsoid molecular simulation model based on atomistic trajectories from all-atom molecular dynamics simulations of smaller system sizes to systematically study the nanomorphology of the SMDPPEH/PCBM/solvent ternary blend during solution processing, including the blade-coating process by applying external shear to the solution. With the significantly reduced overall system degrees of freedom and computational acceleration from GPU, we were able to go well beyond the limitation of conventional all-atom molecular simulations with a system size on the order of hundreds of nanometers with mesoscale molecular detail. Our simulations indicate that, similar to polymer solar cells, the optimal blending ratio in small-molecule organic solar cells must provide the highest specific interfacial area for efficient exciton dissociation, while retaining balanced hole/electron transport pathway percolation. We also reveal that blade-coating processes have a significant impact on nanomorphology. For given donor/acceptor blending ratios, applying an external shear force can effectively promote donor/acceptor phase segregation and stacking in the SMDPPEH domains. The present study demonstrated the capability of an ellipsoid-based coarse-grained model for studying the nanomorphology evolution of small-molecule organic solar cells during solution processing/blade-coating and provided links between fabrication protocols and device nanomorphologies.

Optical coherence tomography-guided laser microsurgery for blood coagulation with continuous-wave laser diode.

Blood coagulation is the clotting and subsequent dissolution of the clot following repair to the damaged tissue. However, inducing blood coagulation is difficult for some patients with homeostasis dysfunction or during surgery. In this study, we proposed a method to develop an integrated system that combines optical coherence tomography (OCT) and laser microsurgery for blood coagulation. Also, an algorithm for positioning of the treatment location from OCT images was developed. With OCT scanning, 2D/3D OCT images and angiography of tissue can be obtained simultaneously, enabling to noninvasively reconstruct the morphological and microvascular structures for real-time monitoring of changes in biological tissues during laser microsurgery. Instead of high-cost pulsed lasers, continuous-wave laser diodes (CW-LDs) with the central wavelengths of 450 nm and 532 nm are used for blood coagulation, corresponding to higher absorption coefficients of oxyhemoglobin and deoxyhemoglobin. Experimental results showed that the location of laser exposure can be accurately controlled with the proposed approach of imaging-based feedback positioning. Moreover, blood coagulation can be efficiently induced by CW-LDs and the coagulation process can be monitored in real-time with OCT. This technology enables to potentially provide accurate positioning for laser microsurgery and control the laser exposure to avoid extra damage by real-time OCT imaging.

PSII-LHCII supercomplex organizations in photosynthetic membrane by coarse-grained simulation.

Green plant photosystem II (PSII) and light-harvesting complex II (LHCII) in the stacked grana regions of thylakoid membranes can self-organize into various PSII-LHCII supercomplexes with crystalline or fluid-like supramolecular structures to adjust themselves with external stimuli such as high/low light and temperatures, rendering tunable solar light absorption spectrum and photosynthesis efficiencies. However, the mechanisms controlling the PSII-LHCII supercomplex organizations remain elusive. In this work, we constructed a coarse-grained (CG) model of the thylakoid membrane including lipid molecules and a PSII-LHCII supercomplex considering association/dissociation of moderately bound-LHCIIs. The CG interaction between CG beads were constructed based on electron microscope (EM) experimental results, and we were able to simulate the PSII-LHCII supramolecular organization of a 500 × 500 nm(2) thylakoid membrane, which is compatible with experiments. Our CGMD simulations can successfully reproduce order structures of PSII-LHCII supercomplexes under various protein packing fractions, free-LHCII:PSII ratios, and temperatures, thereby providing insights into mechanisms leading to PSII-LHCII supercomplex organizations in photosynthetic membranes.

Electrode materials, thermal annealing sequences, and lateral/vertical phase separation of polymer solar cells from multiscale molecular simulations.

The nanomorphologies of the bulk heterojunction (BHJ) layer of polymer solar cells are extremely sensitive to the electrode materials and thermal annealing conditions. In this work, the correlations of electrode materials, thermal annealing sequences, and resultant BHJ nanomorphological details of P3HT:PCBM BHJ polymer solar cell are studied by a series of large-scale, coarse-grained (CG) molecular simulations of system comprised of PEDOT:PSS/P3HT:PCBM/Al layers. Simulations are performed for various configurations of electrode materials as well as processing temperature. The complex CG molecular data are characterized using a novel extension of our graph-based framework to quantify morphology and establish a link between morphology and processing conditions. Our analysis indicates that vertical phase segregation of P3HT:PCBM blend strongly depends on the electrode material and thermal annealing schedule. A thin P3HT-rich film is formed on the top, regardless of bottom electrode material, when the BHJ layer is exposed to the free surface during thermal annealing. In addition, preferential segregation of P3HT chains and PCBM molecules toward PEDOT:PSS and Al electrodes, respectively, is observed. Detailed morphology analysis indicated that, surprisingly, vertical phase segregation does not affect the connectivity of donor/acceptor domains with respective electrodes. However, the formation of P3HT/PCBM depletion zones next to the P3HT/PCBM-rich zones can be a potential bottleneck for electron/hole transport due to increase in transport pathway length. Analysis in terms of fraction of intra- and interchain charge transports revealed that processing schedule affects the average vertical orientation of polymer chains, which may be crucial for enhanced charge transport, nongeminate recombination, and charge collection. The present study establishes a more detailed link between processing and morphology by combining multiscale molecular simulation framework with an extensive morphology feature analysis, providing a quantitative means for process optimization.

Two-level optical coherence tomography scheme for suppressing spectral saturation artifacts.

We demonstrate a novel method for reducing saturation artifacts in spectral-domain optical coherence tomography (SD-OCT) systems. This method is based on a two-level SD-OCT system with a dual-line charge-coupled device (CCD) camera. We compensate the saturated signal detected by the first line using the unsaturated signal detected by the second line. The Fourier transform of the compensated spectrum shows effective suppression of saturation artifacts. This method was also successfully performed on phantom material and skin on a human finger. Our method causes neither back-scattering power loss nor signal-to-noise ratio (SNR) degradation. The only difference between the traditional system and our two-level system is our utilization of the dual-line CCD camera; no additional devices or complex designs are needed.

Investigation of temporal vascular effects induced by focused ultrasound treatment with speckle-variance optical coherence tomography.

Focused ultrasound (FUS) can be used to locally and temporally enhance vascular permeability, improving the efficiency of drug delivery from the blood vessels into the surrounding tissue. However, it is difficult to evaluate in real time the effect induced by FUS and to noninvasively observe the permeability enhancement. In this study, speckle-variance optical coherence tomography (SVOCT) was implemented for the investigation of temporal effects on vessels induced by FUS treatment. With OCT scanning, the dynamic change in vessels during FUS exposure can be observed and studied. Moreover, the vascular effects induced by FUS treatment with and without the presence of microbubbles were investigated and quantitatively compared. Additionally, 2D and 3D speckle-variance images were used for quantitative observation of blood leakage from vessels due to the permeability enhancement caused by FUS, which could be an indicator that can be used to determine the influence of FUS power exposure. In conclusion, SVOCT can be a useful tool for monitoring FUS treatment in real time, facilitating the dynamic observation of temporal effects and helping to determine the optimal FUS power.

Morphology, molecular stacking, dynamics and device performance correlations of vacuum-deposited small-molecule organic solar cells.

The "all carbon" organic solar cells (OSCs) based on the homocyclic molecule tetraphenyldibenzoperiflanthene (DBP) as a donor and C60 as an acceptor were comprehensively characterized. The optimized planar-mixed heterojunction device with a DBP:C60 mixture ratio of DBP : C60 (1 : 2) exhibited a power conversion efficiency of 4.47%. To understand why DBP possesses such advantageous characteristics, the correlations of the morphology, molecular stacking, carrier dynamics and performance of DBP:fullerene-based devices have been systematically studied. First, the face-on stacked DBP molecules could enhance both the absorption of light and the charge carrier mobility. Second, DBP : C60 (1 : 2) thin films with optimized domain sizes and partially interconnected acceptor grains led to the most balanced carrier mobility and the lowest bimolecular recombination in devices. Finally, the DBP molecules were found to stack closely using grazing incidence wide-angle X-ray scattering measurements, with a π-π stacking spacing of 4.58 Å, indicating an effective molecular orbital overlap in DBP. The study not only reveals the promising characteristics of DBP as a donor in OSCs but the clear correlations of the thin-film nano-morphology, molecular stacking, carrier mobility and charge recombination found here could also provide insights into the characterization methodology and optimization of the small molecule OSCs.

Quantitative observation of focused-ultrasound-induced vascular leakage and deformation via fluorescein angiography and optical coherence tomography.

Focused ultrasound (FUS) is a recently discovered noninvasive technique for local and temporal enhancement of vascular permeability, which facilitates drug delivery from the vessels into the surrounding tissue. However, exposure to FUS at a high intensity may cause permanent damage. To investigate the effects of the FUS treatment on blood vessels, we propose to use fluorescein angiography (FA) and optical coherence tomography (OCT) for real-time observation of the diffusion of fluorescence dye from blood vessels and to evaluate the morphological changes of the vessels in vivo. With time-resolved FA imaging, the relationship between the exposed power and the improved permeability of the vessels can be assessed according to the enhancement of the fluorescent intensity due to the dye leakage. Furthermore, the variation of the time-resolved fluorescent intensities can be used to identify the occurrence of dye leakage. In contrast, OCT can be implemented for the reconstruction of tissue microstructures. To quantitatively evaluate the morphological changes of the vessels after the FUS exposure with OCT, a new algorithm was proposed to estimate the vessel area based on the comparison of backscattering properties resulting from the tissue and vascular structures. Results showed that the vessel area increased as the exposed power increased, and the area became significantly larger at a higher FUS exposure power of 10 W. In conclusion, integrated FA and OCT observation can be potentially effective for monitoring the outcome and investigating the effects of FUS treatment.

Evaluation of moisture-related attenuation coefficient and water diffusion velocity in human skin using optical coherence tomography.

In this study, time-resolved optical coherence tomography (OCT) scanning images of the process of water diffusion in the skin that illustrate the enhancement in the backscattered intensities due to the increased water concentration are presented. In our experiments, the water concentration in the skin was increased by soaking the hand in water, and the same region of the skin was scanned and measured with the OCT system and a commercial moisture monitor every three minutes. To quantitatively analyze the moisture-related optical properties and the velocity of water diffusion in human skin, the attenuation coefficients of the skin, including the epidermis and dermis layers, were evaluated. Furthermore, the evaluated attenuation coefficients were compared with the measurements made using the commercial moisture monitor. The results demonstrate that the attenuation coefficient increases as the water concentration increases. Furthermore, by evaluating the positions of center-of mass of the backscattered intensities from OCT images, the diffusion velocity can be estimated. In contrast to the commercial moisture monitor, OCT can provide three-dimensional structural images of the skin and characterize its optical property, which together can be used to observe morphological changes and quantitatively evaluate the moisture-related attenuation coefficients in different skin layers.

Noninvasive imaging of heart chamber in Drosophila with dual-beam optical coherence tomography.

The heart chamber of an adult Drosophila is approximately 2 mm long and 0.5 mm wide, and the interwall separation of different heart portions during systole and diastole range from tens of micrometers to hundreds of micrometers. Furthermore, the heart chamber has a curved structure, which results in the larger differences in depth between the different heart portions. However, applying the wavelength calibration process before Fourier transform in an optical coherence tomography (OCT) system may cause degradation in system sensitivity and longitudinal resolution when the optical path difference between the reference and sample arms increases, which makes imaging the entire heart chamber difficult with OCT system. Additionally, since the heartbeat rate of Drosophila is approximately 6 beats/s, a high-speed OCT system is necessary to record the dynamics of the heat beats. In this study, we propose a new approach to visualize the entire heart chamber including the conical chamber and four ostia portions, and to observe the retrograde and anterograde beats. A buffered Fourier-domain mode-locked (FDML) laser is implemented to provide a high imaging speed. Two output ports of the buffered FDML laser are used simultaneously to scan the different heart portions of Drosophila, and the effective A-scan rate of the OCT system can be doubled. Then, the two scanned images are merged into a single B-mode scan. Furthermore, with dual-beam OCT system, the beating behaviors of the different heart portions from 7-day-old and 21-day-old flies are compared.

Diagnosis of oral precancer with optical coherence tomography.

A procedure for computer analyzing an optical coherence tomography (OCT) image of normal and precancerous oral mucosae is demonstrated to reasonably plot the boundary between epithelium (EP) and lamina propria (LP) layers, determine the EP thickness, and estimate the range of dysplastic cell distribution based on standard deviation (SD) mapping. In this study, 54 normal oral mucosa, 39 oral mild dysplasia, and 44 oral moderate dysplasia OCT images are processed for evaluating the diagnosis statistics. Based on SD mapping in an OCT image, it is found that the laterally average range percentages of 70% SD maximum level in the EP layer is a reasonably good threshold for differentiating moderate dysplasia from mild dysplasia oral lesion based on the OCT image analysis. The sensitivity and specificity in diagnosis statistics can reach 82 and 90%, respectively.

Method for suppressing the mirror image in Fourier-domain optical coherence tomography.

A method, novel to our knowledge, for effective mirror image suppression in Fourier-domain optical coherence tomography based on a phase shift between neighboring A-mode scans is demonstrated. By realizing that the phase shifts of the real and mirror images are mutually reversed and assuming that the real image intensities of the two successive A-mode scans are the same, we can solve a set of two coupled equations to obtain the real image signals. The images based on the scanning of a high-resolution spectral-domain optical coherence tomography system are processed to show effective mirror image suppression results. Compared with a similar method of broad application, our approach has the advantages of shorter process time and higher flexibility in selecting the concerned image portions for processing.

Observations of cardiac beating behaviors of wild-type and mutant Drosophilae with optical coherence tomography.

Time-resolved optical coherence tomography (OCT) scanning images of wild-type and mutant fruit flies (Drosophila melanogaster), illustrating the heartbeat patterns for evaluating their cardiac functions, are demonstrated. Based on the heartbeat patterns, the beat rate and the relative phase between the first two heart segments can be evaluated. The OCT scanning results of mutant flies with impaired proteasome function in cardiac muscles show irregular heartbeat patterns and systematically decreased average beat rates, when compared with the regular patterns of ~4.97 beats/s in average beat rate of the wild-type. In both wild-type and proteasome mutant flies, the beatings at different locations in the same heart segment are essentially synchronized. However, between different heart segments, although the beating in the second segment shows a lag in phase behind that of the first segment in a wild-type, in a proteasome mutant, the beating in the second segment becomes significantly leading that of the first segment. Besides the comparison between the wild-type and proteasomal mutant flies, the influences of using different methods for immobilizing flies during OCT scanning on the heart functions are demonstrated.

Defect detection and property evaluation of indium tin oxide conducting glass using optical coherence tomography.

This study demonstrates a new approach for evaluating the properties of indium tin oxide (ITO) conducting glass and identifying defects using optical coherence tomography (OCT). A swept-source OCT system was implemented to scan the ITO conducting glass to enable two-dimensional or three-dimensional imaging. With OCT scanning, the defects can be clearly identified at various depths. Several parameters in addition to morphological information can be estimated simultaneously, including the thickness of the glass substrate, the refractive index, reflection coefficient, and transmission coefficient, all of which can be used to evaluate the quality of ITO conducting glass. This study developed a modified method for evaluating the refractive index of glass substrates without having to perform multiple scans as well as a segmentation algorithm to separate the interfaces. The results show the potential of OCT as an imaging tool for the inspection of defects in ITO conducting glass.

Motion-insensitive optical coherence tomography based micro-angiography.

An improved image processing procedure for suppressing the phase noise due to a motion artifact acquired during optical coherence tomography scanning and effectively illustrating the blood vessel distribution in a living tissue is demonstrated. This new processing procedure and the widely used procedure for micro-angiography application are based on the selection of high-frequency components in the spatial-frequency spectrum of B-mode scanning (x-space), which are contributed from the image portions of moving objects. However, by switching the processing order between the x-space and k-space, the new processing procedure shows the superior function of effectively suppressing the phase noise due to a motion artifact. After the blood vessel positions are precisely acquired based on the new processing procedure, the projected blood flow speed can be more accurately calibrated based on a previously reported method. The demonstrated new procedure is useful for clinical micro-angiography application, in which a stepping motor of generating motion artifacts is usually used in the scanning probe.

Au nanorings for enhancing absorption and backscattering monitored with optical coherence tomography.

Preparation of a high-concentration Au nanoring (NR) water solution and its applications to the enhancement of image contrast in optical coherence tomography (OCT) and the generation of the photothermal effect in a bio-sample through localized surface plasmon (LSP) resonance are demonstrated. Au NRs are first fabricated on a sapphire substrate with colloidal lithography and secondary sputtering of Au, and then transferred into a water solution through a liftoff process. By controlling the NR geometry, the LSP dipole resonance wavelength in tissue can cover a spectral range of 1300 nm for OCT scanning of deep tissue penetration. The extinction cross sections of the fabricated Au NRs in water are estimated to give levels of 10(-10)-10(-9) cm(2) near their LSP resonance wavelengths. The fabricated Au NRs are then delivered into pig adipose samples for OCT scanning. It is observed that, when resonant Au NRs are delivered into such a sample, LSP resonance-induced Au NR absorption results in a photothermal effect, making the opaque pig adipose cells transparent. Also, the delivered Au NRs in the intercellular substance enhance the image contrast of OCT scanning through LSP resonance-enhanced scattering. By continuously OCT scanning a sample, both photothermal and image contrast enhancement effects are observed. However, by continually scanning a sample with a low scan frequency, only the image contrast enhancement effect is observed.