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
wFLFM: enhancing the resolution of Fourier light-field microscopy using a hybrid wide-field image
Abstract We introduce wFLFM, an approach that enhances the resolution of Fourier light-field microscopy (FLFM) through a hybrid wide-field image. The system exploits the intrinsic compatibility of image formation between the on-axis FLFM elemental image and the wide-field image, allowing for minimal instrumental and computational complexity. The numerical and experimental results of wFLFM present a two- to three-fold improvement in the lateral resolution without compromising the 3D imaging capability in comparison with conventional FLFM.
more »
« less
- Award ID(s):
- 1830941
- PAR ID:
- 10209221
- Publisher / Repository:
- DOI PREFIX: 10.35848
- Date Published:
- Journal Name:
- Applied Physics Express
- Volume:
- 14
- Issue:
- 1
- ISSN:
- 1882-0778
- Page Range / eLocation ID:
- Article No. 012007
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Optical imaging with nanoscale resolution and a large field of view is highly desirable in many research areas. Unfortunately, it is challenging to achieve these two features simultaneously while using a conventional microscope. An objective lens with a low numerical aperture (NA) has a large field of view but poor resolution. In contrast, a high NA objective lens will have a higher resolution but reduced field of view. In an effort to close the gap between these trade-offs, we introduce an acoustofluidic scanning nanoscope (AS-nanoscope) that can simultaneously achieve high resolution with a large field of view. The AS-nanoscope relies on acoustofluidic-assisted scanning of multiple microsized particles. A scanned 2D image is then compiled by processing the microparticle images using an automated big-data image algorithm. The AS-nanoscope has the potential to be integrated into a conventional microscope or could serve as a stand-alone instrument for a wide range of applications where both high resolution and large field of view are required.more » « less
-
Abstract The Hobby–Eberly Telescope (HET) Dark Energy Experiment (HETDEX) is undertaking a blind wide-field low-resolution spectroscopic survey of 540 deg2of sky to identify and derive redshifts for a million Lyα-emitting galaxies in the redshift range 1.9 <z< 3.5. The ultimate goal is to measure the expansion rate of the universe at this epoch, to sharply constrain cosmological parameters and thus the nature of dark energy. A major multiyear Wide-Field Upgrade (WFU) of the HET was completed in 2016 that substantially increased the field of view to 22′ diameter and the pupil to 10 m, by replacing the optical corrector, tracker, and Prime Focus Instrument Package and by developing a new telescope control system. The new, wide-field HET now feeds the Visible Integral-field Replicable Unit Spectrograph (VIRUS), a new low-resolution integral-field spectrograph (LRS2), and the Habitable Zone Planet Finder, a precision near-infrared radial velocity spectrograph. VIRUS consists of 156 identical spectrographs fed by almost 35,000 fibers in 78 integral-field units arrayed at the focus of the upgraded HET. VIRUS operates in a bandpass of 3500−5500 Å with resolving powerR≃ 800. VIRUS is the first example of large-scale replication applied to instrumentation in optical astronomy to achieve spectroscopic surveys of very large areas of sky. This paper presents technical details of the HET WFU and VIRUS, as flowed down from the HETDEX science requirements, along with experience from commissioning this major telescope upgrade and the innovative instrumentation suite for HETDEX.more » « less
-
Abstract Light-field fluorescence microscopy uniquely provides fast, synchronous volumetric imaging by capturing an extended volume in one snapshot, but often suffers from low contrast due to the background signal generated by its wide-field illumination strategy. We implemented light-field-based selective volume illumination microscopy (SVIM), where illumination is confined to only the volume of interest, removing the background generated from the extraneous sample volume, and dramatically enhancing the image contrast. We demonstrate the capabilities of SVIM by capturing cellular-resolution 3D movies of flowing bacteria in seawater as they colonize their squid symbiotic partner, as well as of the beating heart and brain-wide neural activity in larval zebrafish. These applications demonstrate the breadth of imaging applications that we envision SVIM will enable, in capturing tissue-scale 3D dynamic biological systems at single-cell resolution, fast volumetric rates, and high contrast to reveal the underlying biology.more » « less
-
Abstract BackgroundMaize (Zea mays ssp. mays) is 1 of 3 crops, along with rice and wheat, responsible for more than one-half of all calories consumed around the world. Increasing the yield and stress tolerance of these crops is essential to meet the growing need for food. The cost and speed of plant phenotyping are currently the largest constraints on plant breeding efforts. Datasets linking new types of high-throughput phenotyping data collected from plants to the performance of the same genotypes under agronomic conditions across a wide range of environments are essential for developing new statistical approaches and computer vision–based tools. FindingsA set of maize inbreds—primarily recently off patent lines—were phenotyped using a high-throughput platform at University of Nebraska-Lincoln. These lines have been previously subjected to high-density genotyping and scored for a core set of 13 phenotypes in field trials across 13 North American states in 2 years by the Genomes 2 Fields Consortium. A total of 485 GB of image data including RGB, hyperspectral, fluorescence, and thermal infrared photos has been released. ConclusionsCorrelations between image-based measurements and manual measurements demonstrated the feasibility of quantifying variation in plant architecture using image data. However, naive approaches to measuring traits such as biomass can introduce nonrandom measurement errors confounded with genotype variation. Analysis of hyperspectral image data demonstrated unique signatures from stem tissue. Integrating heritable phenotypes from high-throughput phenotyping data with field data from different environments can reveal previously unknown factors that influence yield plasticity.more » « less
