Deep learning has transformed computational imaging, but traditional pixel-based representations limit their ability to capture continuous multiscale object features. Addressing this gap, we introduce a local conditional neural field (LCNF) framework, which leverages a continuous neural representation to provide flexible object representations. LCNF’s unique capabilities are demonstrated in solving the highly ill-posed phase retrieval problem of multiplexed Fourier ptychographic microscopy. Our network, termed neural phase retrieval (NeuPh), enables continuous-domain resolution-enhanced phase reconstruction, offering scalability, robustness, accuracy, and generalizability that outperform existing methods. NeuPh integrates a local conditional neural representation and a coordinate-based training strategy. We show that NeuPh can accurately reconstruct high-resolution phase images from low-resolution intensity measurements. Furthermore, NeuPh consistently applies continuous object priors and effectively eliminates various phase artifacts, demonstrating robustness even when trained on imperfect datasets. Moreover, NeuPh improves accuracy and generalization compared with existing deep learning models. We further investigate a hybrid training strategy combining both experimental and simulated datasets, elucidating the impact of domain shift between experiment and simulation. Our work underscores the potential of the LCNF framework in solving complex large-scale inverse problems, opening up new possibilities for deep-learning-based imaging techniques.
HiLo microscopy is a widefield optical sectioning technique that involves computational reconstruction from two images, one with structured illumination and the other with uniform illumination. A variety of methods, including speckle and periodic grids, can be employed to achieve structured illumination. In this study, we introduce a novel HiLo strategy that utilizes an off-the-shelf holographic diffuser and a low-coherence LED source to generate random caustic patterns. This method offers several benefits over existing ones, such as simplicity and cost-effectiveness. We achieve 4.5 µm optical sectioning capability with a 20x 0.75 NA objective and demonstrate the performance of our method by imaging a 400 µm thick, highly scattering brain section. We anticipate that our caustic-based structured illumination approach will augment the versatility of HiLo microscopy and extend to various imaging applications.
To uncover the pathological mechanism of neurodegenerative diseases, simple and cost-effective analytical methods are highly desirable to obtain biochemistry and morphology information on Tau protein aggregation. Here, we demonstrate the fluorescence-guided computational mid-infrared photothermal microscope, termed Fluorescence-guided Bond-selective Intensity Diffraction Tomography (FBS-IDT). FBS-IDT enables hyperspectral 3D bond-selective refractive index map recovery on Tau fibrils in cells. Depth-resolved mid-infrared fingerprint spectra extraction and related secondary protein structure analysis are demonstrated. 3D reconstruction of β sheet in Tau fibril is realized mainly based on the mid-infrared photothermal effects.
Volumetric chemical imaging is highly desired for investigating biochemical processes at the sub-cellular level. Here, we report bond-selective intensity diffraction tomography (BS-IDT) based on 3D quantitative phase detection of the mid-infrared photothermal effect. BS-IDT demonstrates volumetric chemical imaging with incoherent diffraction-limited resolution and a high speed up to ~6 Hz per volume. The mid-IR spectrum extracted from BS-IDT shows high fidelity compared with ground truth measured by an FTIR spectrometer. The 3D chemical imaging results from cancer cells and Caenorhabditis elegans validate BS-IDT’s superior performance.
Quantitative 3D analysis of brain vasculature is a fundamental problem with important applications, for which vessel segmentation is a first step. Traditional segmentation methods based on parametric models have limited accuracy. More recent techniques based on machine learning have promising results but limited generalization capability. We present a deep-learning based segmentation method that overcomes limitations of existing systems and demonstrates the ability to generalize to various imaging setups, samples including both in-vivo/ex-vivo data, with state-of-the-art results. We achieve so by exploiting several novel methods in deep learning, such as semi-supervised learning. We believe that our work will be another step forward towards improved large-scale neurovascular analysis.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.