Biomedical optics express最新文献

Corneal neovascularization (CNV) is characterized by abnormal vessel growth into the cornea, often impairing vision. Conventional laser therapies for CNV lack precision and risk collateral damage. We developed a multiphoton photothermolysis (MPP) approach using femtosecond (fs) laser-based two-photon absorption guided by real-time reflectance confocal microscopy. Two MPP laser treatment configurations were implemented: dual lasers for imaging (785 nm CW) and treatment (830 nm fs), and a single fs laser for both functions. We demonstrated that in the mouse eye limbus, MPP achieved selective vessel closure without collateral tissue damage. Results also indicate therapeutic effects are mediated by two- rather than one-photon absorption, highlighting MPP's potential as a precise and safe CNV treatment.

We present a low-cost 3D-printing method of fabricating optical quality lenslet arrays for integration in a multifocal structured illumination microscope (mSIM), achieving super-resolution fluorescence imaging using 3D-printed optics for the first time. We detail the design and manufacturing processes to produce high-quality 3D-printed optics, showing their comparable surface roughness of 30 ± 2.5 nm for the 3D-printed elements compared to 37 ± 1.4 nm for commercial glass optics. A 3D-printed lenslet array with a 'honeycomb' geometry and 1.2 mm lenslet diameter was compared to a high-end glass commercial lenslet array with 250 µm lenslet diameter and a lower cost commercial lenslet array with a 1 mm by 1.4 mm lenslet footprint. The imaging performance of the different optics was benchmarked using a custom mSIM setup by quantifying the beam profile homogeneity and the experimental lateral resolution. The mSIM setup incorporating the different microlens arrays was tested using a commercial bovine pulmonary artery endothelial cell specimen, highlighting an achievable resolution enhancement from 237 nm ± 12 nm with laser-scanning illumination to 151 ± 12 nm using the high-end commercial microlens array and from 232 ± 18 nm with laser-scanning illumination to 151 nm ± 12 nm using the 3D-printed honeycomb lenslet array. Advantages of improved background rejection through the custom lenslet geometry are discussed, highlighting the super-resolution microscope performance achievable using custom 3D-printed optics costing as low as £0.50 to produce.

Optical coherence tomography (OCT) images have long been plagued by speckle noise, which has both multiplicative characteristics and spatial correlation, making noise removal extremely difficult. Traditional methods based on the Noisy2Noisy framework have poor denoising performance due to their "additive-independent" assumption. Therefore, we propose a two-stage unsupervised network, Pixel Shuffle-Blindspot Dual-Stage Network (PSB-DSN), which requires no clean data during training. In the first stage, pixel shuffle downsampling (PD) is used to cut off the noise correlation, while a global masker is introduced to select 1/4 of the pixels as blind spots. The self-replacement mechanism substitutes masked pixels with blind-spot convolution outputs, ensuring J-invariance while allowing the remaining visible pixels to provide intact contextual information. Additionally, parallel training with multiple masks is used to achieve rapid convergence and information supplementation. In the second stage, a random mask refinement network is designed to fuse the denoised output from the first stage with the original noisy image, further dispersing the checkerboard artifacts and achieving secondary enhancement of global contextual information. Evaluation on six OCT datasets shows that PSB-DSN achieves superior denoising enhancement performance over state-of-the-art unsupervised approaches. Furthermore, downstream retinal layer segmentation experiments confirm that images processed by PSB-DSN achieve consistent and stable improvements in segmentation accuracy.

While cancer characteristics can vary significantly across types, methods that distinguish malignant cells from normal ones hold promise by targeting shared cellular anomalies. Among these, morphological differences play a key role in driving the aggressive behavior and altered function typical of cancer cells. Detecting and analyzing such cells within complex, densely packed tissue environments requires advanced imaging techniques. Polarization-resolved fluorescence microscopy offers rich insights into cellular composition, molecular binding affinities, and structural organization, particularly in revealing biomolecular order and subcellular polarity loss. In this work, we study polarization-resolved two-photon excitation fluorescence tissue imaging microscopy in vitro to investigate ordered versus disordered chromatin organization within cell nuclei. We employ an innovative phasor map analysis to facilitate quick interpretation, using colorectal cancer identification and liquid crystal as a case study and baseline, respectively. Our method aims to identify cancer within tissue by adding polarimetric contrast to fluorescence due to the anisotropic feature of fluorescent molecular probes. Accordingly, the proposed phasor map provides a graphically transformed representation of polarization-based fluorescence imaging for histopathological tissue identification on a pixel-wise basis, facilitating comprehensive classification of diverse tissue samples. This study presents initial steps toward showing the potential for cancer identification and lays a foundation for future diagnostic strategies.

The corneoscleral limbus contains essential aqueous channels and veins that form the primary aqueous outflow pathway, making it a critical focus of glaucoma research. Optical coherence tomography (OCT) has emerged as a powerful tool for investigating this pathway, but conventional imaging is typically performed in segments and restricted to a single functional contrast. Such approaches are limited in evaluating the overall condition of the outflow system, which contributes circumferentially asymmetrically to the aqueous outflow and is composed of multiple tissues requiring different OCT functional extensions. To address these limitations, we propose a circumferential, multi-contrast mapping framework for in vivo analysis of limbal vasculature and aqueous outflow structures/functions. The workflow begins with a rapid panoramic OCT angiography scan using a circular-scan anterior segment OCT (CircAS-OCT) system to generate a global limbal reference map. This reference is then used to guide the stitching of sectional scans acquired from high-resolution OCT devices and protocols. In two normal subjects, the method successfully enabled circumferential, multi-contrast mapping of selected outflow-related structures and processes, including depth-resolved limbal vasculature, collector channel orifice positions, trabecular meshwork regions, and trabecular motion strength, with results aligning well with established anatomical knowledge. By combining circumferential coverage and cross-function integration, our approach provides a versatile imaging platform with strong potential for comprehensive assessment of the aqueous outflow pathway, supporting diagnostic and surgical strategies in glaucoma management.

Axial resolution in swept-source-based Fourier-domain optical coherence tomography (FD-OCT) is limited by the sweep range of the source. However, broad swept-sources are not readily available, and passively combining two laser sources is not straightforward. In this paper, we develop a framework to overcome this limit by computationally combining independently sweeping sources to increase the bandwidth and subsequently the axial resolution of full-field Fourier-domain optical coherence tomography (FF-FD-OCT) systems. To this end, we demonstrate a dual-laser full-field FD-OCT system that uses two lasers, which sweep sequentially and are stitched phase-correctly in post-processing to obtain high-bandwidth spectra. After combining, we achieve an effective bandwidth of 145 nm at a central wavelength of 878 nm. The system has a high axial resolution of 3.1 μm and can operate at an A-scan rate of 50 MHz. Our method requires a one-time calibration measurement to determine the non-linear sweeps from the lasers and the wavelength overlap, as well as a volume-by-volume phase matching procedure to compensate for sample motion. We demonstrate this for ex-vivo phantoms as well as in-vivo retinal data. Overall, the framework allows for extension to multiple lasers to further improve the axial resolution.

This study investigates relative peripheral refraction (RPR) in emmetropic and myopic eyes in the 25° nasal and temporal visual fields under far and near fixation, with control for any fluctuations in accommodation. Additional analysis of axial length and comparison with recently published eye models are also presented, constituting complementary adult data to the Stockholm Myopia Study. In the ten emmetropes, a pronounced nasal-temporal asymmetry was observed, with significantly more myopic RPR nasally and less myopic / more hyperopic RPR temporally at both accommodative states (p = 0.005). The nine myopes, in contrast, exhibited more symmetric peripheral profiles, with no significant nasal-temporal differences. Accommodation induced systematic shifts in both groups, producing increased relative myopia nasally and relative hyperopia temporally (p < 0.001). Axial length was significantly correlated with temporal hyperopic shifts during accommodation in myopes (p = 0.005), suggesting a structural contribution of ocular growth to peripheral optics. Comparison with eye models showed partial agreement, though experimental results revealed greater asymmetry than predicted in emmetropes and a weaker nasal-temporal distinction in myopes. Our findings indicate that variations in relative peripheral refraction over the horizontal visual field and with accommodation might be linked to ocular growth and are important for optical myopia control.

We describe a refractive adaptive optics scanning light ophthalmoscope designed for small animal imaging through a 1.8 mm diameter pupil. The optical setup, based on a search of achromatic doublets through multiple lens catalogs, consists of a sequence of modified afocal relays that deliver diffraction-limited imaging in pupil and retina conjugates. Real ray tracing is used to compare imaging performance when correcting large focus errors using a pupil conjugate wavefront corrector, a traditional Badal optometer, and a modified Badal optometer. Polarization control, focal length selection, and systematic lens tilting are explored for mitigating reflections with minimal imaging performance degradation. A 2-dimensional optical scanner with a 29.2 kHz resonant frequency around one axis and low dynamic surface distortion allows doubling the frame rate of prior instruments and simplifies the optical setup. Scanner orientation and trigger electrical signals are used to correct sinusoidal image warping and line sampling jitter. The instrument is demonstrated by imaging mice under 800 nm illumination with two reflectance detection modalities: confocal and quadrant non-confocal.

Optical clearing combined with light-sheet microscopy enables high-resolution imaging of extended tissue at scale. However, standard mesoSPIM systems are optimised for intact organs and are not suited to thin tissue slices. We present an oblique compensation scanning method using obliquely mounted samples held between refractive-index-matched slides in a 3D-printed frame. This enables mechanical de-skewing during acquisition, minimising post-processing requirements. We demonstrate feasibility in fluorescent bead phantoms and rabbit heart tissue, achieving a 4.8 × reduction in processing time and a 1.5 × improvement in axial resolution ((13.15±1.36) μm to (8.72±1.80) μm) compared to conventional z scan. The oblique compensation acquisition method extends mesoSPIM's utility to fragile, laterally extended tissue sections.

Retinal diseases, including myopic and diabetic retinopathies, require early detection through precise retinal-layer segmentation in optical coherence tomography images. Existing deep-learning models generalize poorly across devices (domain shifts, noise, and high annotation costs). We propose dual-level pseudo-label learning for segmentation (DPLSeg), an unsupervised segmentation model with a dual-level pseudo-label learning strategy and a hierarchical transformer encoder to enhance feature representation and domain adaptability. Validated on 850 optical coherence tomography images from three devices, DPLSeg achieves a mean intersection over union of 79.9%, surpassing DeepLab (75.2%) and DAFormer, reducing annotation needs by 80% and providing a scalable clinical diagnostic tool.