Journal of Biomedical Optics最新文献

Significance: Metastasis is a leading cause of cancer-related deaths. Disseminated circulating tumor cells (CTCs) through the bloodstream seed metastatic tumors at distant sites. Most methods for enumerating CTCs in humans clinically rely on drawing and analyzing small blood samples, but these may yield inaccurate estimates of CTC burden and cannot measure CTC changes over time. Identification and enumeration of CTCs for experimental or clinical purposes largely rely on marker-driven analyses by flow cytometry.

Aim: In principle, noninvasive fluorescence enumeration of CTCs directly in vivo could provide a more accurate method for enumerating CTCs. However, this will require a specific contrast agent for CTCs. The goal of this work is to define characteristics of useful CTC contrast agents and perform preliminary testing of candidate contrast agents used for fluorescence-guided surgery (FGS).

Approach: We evaluated a clinical small-molecule folate receptor-targeted molecular imaging agent (OTL38, pafolacianine), a fluorogenic pan-cathepsin imaging agent (VGT-309, abenacianine), and a set of custom-designed, small-molecule prostate-specific membrane antigen (PSMA)-targeted fluorophores. We tested these contrast agents using in vitro cell culture models and in in vivo murine models.

Results: All tested contrast agents showed not only high uptake and labeling by target cell lines but also small but significant labeling of non-cancer blood cells. Contrast agents that exhibited rapid clearance from circulation and the fluorogenic approach resulted in significantly reduced non-specific interfering background fluorescence signals.

Conclusions: Although all of the tested targeted fluorescence contrast agents have properties useful for labeling of CTCs, thus far, none has exhibited the required high specificity. This resulted in some labeling of non-cancer blood cells, which presented false-positive CTC counts. Improved contrast agent design and multiplexed use of more than one contrast agent may improve this specificity.

Significance: Speckle contrast optical spectroscopy (SCOS) and speckle plethysmography (SPG) are increasingly used to measure deep tissue blood flow from tissues. However these methods derive flow from measurements at a single exposure duration, which could introduce errors due to inefficient choice of a single integration time, changes in speckle averaging factors ( $\beta$ ), and noise.

Aim: The aims are to compare and evaluate the robustness of single- and multi-exposure speckle contrast methods in estimating blood flow changes under experimental conditions such as $\beta$ mismatch, noise, and pulsatile flow.

Approach: Speckle visibility was simulated at 10,000 logarithmically spaced exposure durations (0.01 to 10 ms), incorporating $\pm 5$ to 15% noise, using a semi-infinite diffusion model at different physiologically relevant steady state and pulsatile flow rates. Speckle visibility was used to estimate blood flow changes using both the multi-exposure approach (non-linear fitting of the full curve) and conventional single exposure approach at 0.1, 1, and 5 ms durations (SCOS, SPG, and look-up-table), under different noise conditions and mismatches in $\beta$ .

Results: Multi-exposure speckle imaging (MESI) maintained $<1\%$ mean error despite $\beta$ mismatch or noise while maintaining $\le 12\%$ pulsatile-flow error at $\pm 5\%$ noise. Single-exposure methods showed errors up to 90% for $\beta$ mismatch/noise and $\ge 22\%$ pulsatile-flow errors.

Conclusions: MESI outperformed single-exposure approaches in estimating blood flow changes by reducing $\beta$ bias, minimizing noise sensitivity, and accurately tracking pulsatile dynamics.

Significance: Freehand optical ultrasound (OpUS) is an emerging imaging modality that utilizes arrays of fiber-optic, photoacoustical ultrasound sources, and a single fiber-optic detector to perform pulse-echo ultrasound imaging in real time and at video rate. The freehand OpUS imaging probes presented to date have utilized either flat-cleaved optical fibers to generate an array of small, circular sources, or featured an additional array of eccentric optical waveguides to generate elliptical sources that improve ultrasound directivity. However, the incorporation of waveguides imposed severe restrictions on the size of the sources, the source pitch within the array, and the overall size of the imaging probe, and ultimately limited achievable image quality and clinical utility of freehand OpUS probes.

Aim: We present an alternative method for generating eccentric fiber-optic OpUS sources, which incorporates an anisotropic holographic diffuser element (HDE) to shape the profile of fiber-delivered excitation light.

Approach: A commercially available HDE was selected and imprinted into a UV-curable adhesive to improve optical resilience. The monolithic and near-uniform structure of the resulting adhesive-imprinted HDE structure greatly facilitated alignment with, and allowed for arbitrary positioning of, arrays of optical fibers, without adding significant bulk.

Results: The use of an HDE enabled dense, cladding-to-cladding packing of optical fibers, resulting in the first freehand OpUS probe featuring an acoustic source pitch below the spatial Nyquist limit (thus eliminating grating lobe artifacts), high channel count (144 sources), and endoluminal-scale physical dimensions (diameter: 18.5 mm).

Conclusion: This HDE-enabled freehand OpUS imaging probe achieved video-rate, real-time imaging at increased image quality and depth and allowed for the first endoluminal freehand OpUS imaging within a realistic imaging phantom.

Significance: Bacterial contamination poses significant risks, particularly in medical settings, where timely identification is crucial for effective treatment. Traditional diagnostic methods often fail to provide rapid results, underlining the need for alternative approaches.

Aim: We investigate the potential of fluorescence lifetime imaging microscopy (FLIM) for fast label-free classification of bacterial species based on their growth states and intrinsic fluorescence characteristics.

Approach: Utilizing two-photon excitation microscopy, we examined the fluorescence lifetimes of the endogenous fluorophores NAD(P)H and FAD in Escherichia coli K12, Staphylococcus aureus, and Pseudomonas fluorescens. Cells were studied in different growth states-exponential, cooled, and dead-and measurements were taken at excitation wavelengths of 740 and 900 nm. The FLIM data were processed using a multi-component exponential decay model and visualized using phasor plots and kernel density estimation.

Results: Results revealed distinct fluorescence lifetime profiles for each species, with clear differences observed in the exponential growth phase. Notably, E. coli K12 and S. aureus could be distinguished in mixed samples using a single excitation wavelength and emission channel.

Conclusion: These findings demonstrate that FLIM-based metabolic imaging can distinguish between bacterial species without labels, providing a promising basis for rapid, high-resolution diagnostic tools in clinical and research settings. We underscore the potential of FLIM as a powerful tool for bacterial classification, offering advantages in speed and spatial resolution over traditional methods.

Significance: Understanding microbial growth and morphology at the single-cell level is essential for studying microbial physiology, providing valuable insights for research and biotechnology. However, existing workflows often rely on manual cell annotation, which limits efficiency and scalability. Therefore, developing an automated workflow for quantitative analysis of cell growth and morphology is highly desirable.

Aim: We aim to develop an AI-driven analysis system that efficiently segments and indexes microbial cells, and quantitatively analyzes individual cellular features without requiring expensive annotations, for automated monitoring of cell counts and morphological characteristics.

Approach: The automated system consists of four modular components: denoising, zero-shot segmentation using the Segment Anything Model (SAM), structured post-processing, and a quantitative feature extraction. To evaluate the effectiveness of each component, we conducted ablation experiments and systematically studied their impact on the overall system performance.

Results: Denoising and post-processing improved segmentation accuracy by 12.10% and 2.30%, respectively. Among the evaluated SAM variants, the SAM-H model achieved the best performance, with an average error rate of only 3.0% across 1162 manually annotated Escherichia coli cells. Using the optimized SAM-H pipeline, the system efficiently extracted morphometric and intensity features from Escherichia coli cells and nuclei of the yeast and cancer cell lines.

Conclusions: This framework automates quantitative analysis of microbial cells in high-resolution microscopy images. It will enable advanced research on microbial adaptations, with the potential to accelerate studies of extremophiles under harsh environments.

Significance: Vascular abnormalities may contribute to amyloid-beta accumulation and neurotoxicity in Alzheimer's disease (AD). Monitoring vascular degeneration as AD progresses is essential. Three-photon fluorescence microscopy enables high-resolution deep tissue imaging with minimal invasiveness and photodamage.

Aim: In this proof-of-concept study, we established a longitudinal 3P imaging pipeline to quantify vascular and amyloid plaque changes in the ${\mathrm{APP}}^{\mathrm{NL}\text{-}G\text{-}F}$ mouse model.

Approach: A cranial window allowed repeated 3P imaging at 4-week intervals beginning at 5 weeks after surgery. Vessels labeled with Texas-Red were segmented using DeepVess, whereas plaques labeled with methoxy-XO4 were segmented using custom scripts. Quantitative analyses assessed vascular parameters (diameter, tortuosity, length, inter-vessel distance, total volume) and plaque metrics (radius, total volume).

Results: We imaged the same field over 4 weeks, quantifying an overall decrease in vasculature and an increase in amyloid plaques between two sessions. Significant changes in vessel diameter, inter-vessel distance, and alterations in vessel length and plaque radius were observed. Changes in vessel tortuosity were not significant.

Conclusions: We demonstrate the potential of three-photon imaging to track vascular and amyloid-related changes in deep cortical structures. It offers a tool for studying the interplay between vascular and amyloid pathologies in AD, supporting future research into disease mechanisms and therapeutic strategies.

Significance: Accurate transfer of annotations from histological images to fluorescence images is essential in developing deep learning (DL)-based optical imaging systems for intraoperative assessment of tumor margins. Manual annotation is time-consuming, prone to interobserver variability, and impractical for large-scale datasets.

Aim: We present a semi-automated method that can effectively transfer tumor annotations from pathologist-annotated hematoxylin and eosin (H&E) images to fluorescence images captured using microscopy with ultraviolet surface excitation (MUSE). This method is not intended for intraoperative use but rather to facilitate the creation of annotated datasets for DL model development.

Approach: Our semi-automated method consists of nonrigid image registration, outline extraction and refinement, and annotation transfer. The method was applied to H&E and MUSE image pairs from 35 breast and lung tissue samples. Manual annotations in MUSE images were used as the ground truth for evaluation.

Results: The proposed method achieved a Dice score coefficient of $0.87\pm 0.07$ , convolutional-neural-network-based feature similarity of $0.94\pm 0.04$ , and a normalized Hausdorff distance of $0.15\pm 0.06$ across the dataset.

Conclusion: These results demonstrate that the method provides a fast and accurate solution for generating annotated MUSE datasets necessary for training DL algorithms for intraoperative tumor margin detection.

Significance: Collagen, the main load-bearing component in tissue, is present in all animals and forms a variety of networks from the fibrils, fibers, bundles, and lamellae into which it self-assembles. The collagen microstructure is different among tissue types, and the different microstructures give rise to tissue-specific mechanical properties. Therefore, methods for visualizing collagen fibers and their orientation are essential for understanding the biomechanical properties of tissue.

Aim: Our aim in this review is to provide the basis for understanding the methodology of polarized light imaging methods and how they can be used to characterize collagen microstructure.

Approach: We begin with a description of collagen microstructure and its relationship to tissue biomechanics, a basic formalism of polarized light, and how collagen interacts with polarized light. We then describe polarized light microscopy and its various forms, particularly instant polarized light microscopy, then polarization-sensitive optical coherence tomography, and last, polarization-resolved second-harmonic generation microscopy.

Results: We describe methods for imaging collagen microstructure with polarized light from in vivo methods to high-resolution volumetric imaging of tissue sections.

Conclusions: We intend to help those interested in using polarized light to image and understand the relationship between collagen microstructure and biomechanics.

Significance: Traditional optical-property image reconstruction techniques are often constrained by artifacts arising from suboptimal source-detector configurations, the amplification of measurement noise during inversion, and limited depth sensitivity, which particularly impacts the accurate reconstruction of deep-seated anomalies such as tumors.

Aim: To overcome these challenges, this research proposes and implements an end-to-end deep learning framework, Channel Attention Fusion Network (CAFNet).

Approach: CAFNet employs AUTOMAP for domain transformation, feature extraction modules for multi-scale feature learning, and channel attention mechanisms to prioritize critical features. The proposed model is trained and tested on simulated and experimental datasets, utilizing metrics such as mean squared error (MSE), peak signal-to-noise ratio (PSNR), and structural similarity index (SSIM) for evaluating model performance.

Results: CAFNet outperforms traditional and state-of-the-art models, achieving the highest SSIM and PSNR values with the lowest MSE. It effectively reconstructs optical properties with high precision, showcasing its ability to detect and localize inclusions in experimental phantoms. An ablation study is performed to highlight the importance of channel attention in CAFNet.

Conclusions: CAFNet demonstrates a significant advancement in diffuse optical imaging, addressing challenges with noise and domain variability issues. Its robust performance highlights the potential in practical medical imaging applications, offering a reliable solution for reconstructing optical properties in complex scenarios.

Significance: The lymphatic system, crucial for immune function and fluid balance, is difficult to study due to its nearly invisible channels and passive movements. Despite its importance, real-time, noninvasive lymphatic imaging is limited and often relies on exogenous agents. A largely overlooked feature of the lymphatic system is its natural hypoxia, arising because lymphatic vessels and nodes are distant from oxygen-rich blood and serve as a waste reservoir and from amplifying factors such as inflammation. This hypoxia has been linked to lymphangiogenesis and cancer metastasis.

Aim: We introduce a method for real-time macroscopic imaging of lymphatic function and response to inflammation, through the intrinsic hypoxia transients that occur in lymphatic structures.

Approach: The naturally hypoxic environment of the lymphatic system was imaged in vivo in mice through the delayed fluorescence (DF) of metabolized endogenous protoporphyrin IX (PpIX), induced by 5-aminolevulinic acid. PpIX localizes to inflamed regions; when inflammation is cleared by lymphatics, the low oxygen conditions cause the DF signal to be amplified. DF imaging and lymphatic kinetics were characterized in wound, inflammation, and pancreatic tumor models. High-intensity popliteal and sentinel nodes, wounds, and tumors were excised for immunohistochemistry (IHC) and hematoxylin and eosin analysis.

Results: Lymphatic pumping frequency changed with increasing wound severity, and hypoxia appeared in sentinel nodes near tumors. Cyclical pumping occurred at edema sites and in wound and tumor-adjacent nodes. Uninjured anesthetized mice showed little contrast, whereas awake mice exhibited hypoxia localized to lymph nodes. Microscopy and IHC confirmed PpIX and hypoxia presence in nodes, tumors, and wounds, localized to macrophages and T cells.

Conclusions: Unlike injection-based regional lymph node mapping, DF hypoxia imaging appears to provide a natural whole-body contrast mechanism, highlighting its potential for visualizing lymphatic function and associated hypoxia dynamics. This original documentation of lymphatic hypoxia has potential applications in surgical guidance, tracking of metastatic tumors, and immune response tracking.