An official website of the United States government
Confocal microscopy is a standard approach for obtaining volumetric images of a sample with high axial and lateral resolution, especially when dealing with scattering samples. Unfortunately, a confocal microscope is quite expensive compared to traditional microscopes. In addition, the point scanning in confocal microscopy leads to slow imaging speed and photobleaching due to the high dose of laser energy. In this paper, we demonstrate how the advances in machine learning can be exploited to teach a traditional wide-field microscope, one that’s available in every lab, into producing 3D volumetric images like a confocal microscope. The key idea is to obtain multiple images with different focus settings using a wide-field microscope and use a 3D generative adversarial network (GAN) based neural network to learn the mapping between the blurry low-contrast image stacks obtained using a wide-field microscope and the sharp, high-contrast image stacks obtained using a confocal microscope. After training the network with widefield-confocal stack pairs, the network can reliably and accurately reconstruct 3D volumetric images that rival confocal images in terms of its lateral resolution, z-sectioning and image contrast. Our experimental results demonstrate generalization ability to handle unseen data, stability in the reconstruction results, high spatial resolution even when imaging thick (∼40 microns) highly-scattering samples. We believe that such learning-based microscopes have the potential to bring confocal imaging quality to every lab that has a wide-field microscope.
more »
« less
Learning-based model using unpaired datasets for super-resolution confocal microscopy
One of the major drawbacks of confocal microscopy is its limited spatial resolution. This work assesses the performance of an unpaired learning-based model to provide confocal images with improved resolution.
more »
« less
- Award ID(s):
- 2404769
- PAR ID:
- 10570617
- Publisher / Repository:
- Optica Publishing Group
- Date Published:
- ISBN:
- 978-1-957171-37-1
- Page Range / eLocation ID:
- DTh4H.5
- Format(s):
- Medium: X
- Location:
- Toulouse
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
We demonstrate hyperspectral confocal microscopy in the short-wave infrared (SWIR) range of 1100–1600 nm using a wavelength-scanning laser in tandem with laser scanning confocal microscopy. Confocal microscopy in the SWIR range allows for high-resolution inspection of an integrated circuit (IC) chip, while hyperspectral imaging, together with a chemometric analysis, enables us to identify functional circuit block groups in the acquired image. With the extended capability, the developed instrument can be potentially used for inline inspection and non-invasive failure analysis of IC chips.more » « less
-
Unpaired data training enables super-resolution confocal microscopy from low-resolution acquisitionsSupervised deep-learning models have enabled super-resolution imaging in several microscopic imaging modalities, increasing the spatial lateral bandwidth of the original input images beyond the diffraction limit. Despite their success, their practical application poses several challenges in terms of the amount of training data and its quality, requiring the experimental acquisition of large, paired databases to generate an accurate generalized model whose performance remains invariant to unseen data. Cycle-consistent generative adversarial networks (cycleGANs) are unsupervised models for image-to-image translation tasks that are trained on unpaired datasets. This paper introduces a cycleGAN framework specifically designed to increase the lateral resolution limit in confocal microscopy by training a cycleGAN model using low- and high-resolution unpaired confocal images of human glioblastoma cells. Training and testing performances of the cycleGAN model have been assessed by measuring specific metrics such as background standard deviation, peak-to-noise ratio, and a customized frequency content measure. Our cycleGAN model has been evaluated in terms of image fidelity and resolution improvement using a paired dataset, showing superior performance than other reported methods. This work highlights the efficacy and promise of cycleGAN models in tackling super-resolution microscopic imaging without paired training, paving the path for turning home-built low-resolution microscopic systems into low-cost super-resolution instruments by means of unsupervised deep learning.more » « less
-
Abstract We demonstrate the use of tip-enhanced Raman spectroscopy (TERS) on polymeric microstructures fabricated by two-photon polymerization direct laser writing (TPP-DLW). Compared to the signal intensity obtained in confocal Raman microscopy, a linear enhancement of almost two times is measured when using TERS. Because the probing volume is much smaller in TERS than in confocal Raman microscopy, the effective signal enhancement is estimated to be ca. 104. We obtain chemical maps of TPP microstructures using TERS with relatively short acquisition times and with high spatial resolution as defined by the metallic tip apex radius of curvature. We take advantage of this high resolution to study the homogeneity of the polymer network in TPP microstructures printed in an acrylic-based resin. We find that the polymer degree of conversion varies by about 30% within a distance of only 100 nm. The combination of high resolution topographical and chemical data delivered by TERS provides an effective analytical tool for studying TPP-DLW materials in a non-destructive way.more » « less
-
Abstract Confocal microscopy is an indispensable tool for biological imaging due to its high resolution and optical sectioning capability. However, its slow imaging speed and severe photobleaching have largely prevented further applications. Here, we present dual inclined beam line‐scanning (LS) confocal microscopy. The reduced excitation intensity of our imaging method enabled a 2‐fold longer observation time of fluorescence compared to traditional LS microscopy while maintaining a good sectioning capability and single‐molecule sensitivity. We characterized the performance of our method and applied it to subcellular imaging and three‐dimensional single‐molecule RNA imaging in mammalian cells.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1364/3D.2024.DTh4H.5
