University of Virginia School of Medicine researchers and their collaborators have solved a decades-old mystery about how E. coli and other bacteria are able to move. Bacteria push themselves forward by coiling long, threadlike appendages into corkscrew shapes that act as makeshift propellers. But how exactly they do this has baffled scientists, because the "propellers" are made of a single protein.
An international team led by UVA's Edward H. Egelman, PhD, a leader in the field of high-tech cryo-electron microscopy (cryo-EM), has cracked the case. The researchers used cryo-EM and advanced computer modeling to reveal what no traditional light microscope could see: the strange structure of these propellers at the level of individual atoms.
"While models have existed for 50 years for how these filaments might form such regular coiled shapes, we have now determined the structure of these filaments in atomic detail," said Egelman, of UVA's Department of Biochemistry and Molecular Genetics.
"We can show that these models were wrong, and our new understanding will help pave the way for technologies that could be based upon such miniature propellers."
Blueprints for Bacteria's 'Supercoils'
Different bacteria have one or many appendages known as a flagellum, or, in the plural, flagella. A flagellum is made of thousands of subunits, but all these subunits are exactly the same. You might think that such a tail would be straight, or at best a bit flexible, but that would leave the bacteria unable to move. That's because such shapes can't generate thrust. It takes a rotating, corkscrew-like propeller to push a bacterium forward. Scientists call the formation of this shape "supercoiling," and now, after more than 50 years, they understand how bacteria do it.
Using cryo-EM, Egelman and his team found that the protein that makes up the flagellum can exist in 11 different states. It is the precise mixture of these states that causes the corkscrew shape to form. It has been known that the propeller in bacteria is quite different than similar propellers used by hearty one-celled organisms called archaea. Archaea are found in some of the most extreme environments on Earth, such as in nearly boiling pools of acid, the very bottom of the ocean and in petroleum deposits deep in the ground.
Egelman and colleagues used cryo-EM to examine the flagella of one form of archaea, Saccharolobus islandicus, and found that the protein forming its flagellum exists in 10 different states. While the details were quite different than what the researchers saw in bacteria, the result was the same, with the filaments forming regular corkscrews. They conclude that this is an example of "convergent evolution" -- when nature arrives at similar solutions via very different means. This shows that even though bacteria and archaea's propellers are similar in form and function, the organisms evolved those traits independently.
"As with birds, bats and bees, which have all independently evolved wings for flying, the evolution of bacteria and archaea has converged on a similar solution for swimming in both," said Egelman, whose prior imaging work saw him inducted into the National Academy of Sciences, one of the highest honors a scientist can receive. "Since these biological structures emerged on Earth billions of years ago, the 50 years that it has taken to understand them may not seem that long."
If you open a biology textbook and run through the images depicting how DNA is organized in the cell's nucleus, chances are you'll start feeling hungry; the chains of DNA would seem like a bowl of ramen: long strings floating in liquid. However, according to two new studies—one experimental and the other theoretical—that are the outcome of the collaboration between the groups of Prof. Talila Volk of the Molecular Genetics Department and Prof. Sam Safran of the Chemical and Biological Physics Department at the Weizmann Institute of Science, this image should be reconsidered. Clarifying it is essential since DNA's spatial arrangement in the nucleus can affect the expression of genes contained within the DNA molecule, and hence the proteins found in the cell.
This story began when Volk was studying how mechanical forces influence cell nuclei in the muscle and found evidence that muscle contractions had an immediate effect on gene expression patterns. "We couldn't explore this further because existing methods relied on imaging of chemically preserved cells, so they failed to capture what happens in the cell nuclei of an actual working muscle," she says.
To address this issue, Dr. Dana Lorber, a research associate in Volk's group, led the design of a device that makes it possible to study muscle nuclei in live fruit fly larvae. The device holds the tiny, translucent larva within a groove that allows it to contract and relax its muscles but keeps its movement constrained so that it can be scanned by a fluorescence microscope. Using the device, the researchers obtained images of the internal, linearly-organized complexes of DNA and its proteins (known as chromatin), surrounded by the membrane of the muscle nuclei.
Expecting a bowl full of ramen, Lorber and Dr. Daria Amiad-Pavlov, a postdoctoral fellow in Volk's group, were in for a surprise. Rather than filling up the entire volume of the nucleus, the "noodles," or long chromatin molecules, were organized as a relatively thin layer, attached to its inner walls. Similar to the outcome of the interaction between oil and water, what is known as "phase separation," the chromatin separated itself from the bulk of the liquid inside of the nucleus and found its place at its outskirts, while most of the fluid medium remained at the center.
The researchers realized that they were on their way to addressing a fundamental biological question, that is—how is chromatin, and hence DNA, organized in the nucleus in a living organism. "But the findings were so unexpected, we had to make sure no error had crept in and that this organization was universal," Lorber says.
In a major scientific "quantum" leap, University of Queensland researchers have created a quantum microscope that can reveal biological structures that would otherwise be impossible to see. This paves the way for applications in biotechnology, and could extend far beyond this into areas ranging from navigation to medical imaging. The microscope is powered by the science of quantum entanglement, an effect Einstein described as "spooky interactions at a distance."
Professor Warwick Bowen, from UQ's Quantum Optics Lab and the ARC Centre of Excellence for Engineered Quantum Systems (EQUS), said it was the first entanglement-based sensor with performance beyond the best possible existing technology. "This breakthrough will spark all sorts of new technologies -- from better navigation systems to better MRI machines, you name it," Professor Bowen said. "Entanglement is thought to lie at the heart of a quantum revolution. We've finally demonstrated that sensors that use it can supersede existing, non-quantum technology. This is exciting -- it's the first proof of the paradigm-changing potential of entanglement for sensing."
Australia's Quantum Technologies Roadmap sees quantum sensors spurring a new wave of technological innovation in healthcare, engineering, transport and resources. A major success of the team's quantum microscope was its ability to catapult over a 'hard barrier' in traditional light-based microscopy. "The best light microscopes use bright lasers that are billions of times brighter than the sun," Professor Bowen said. "Fragile biological systems like a human cell can only survive a short time in them and this is a major roadblock. The quantum entanglement in our microscope provides 35 per cent improved clarity without destroying the cell, allowing us to see minute biological structures that would otherwise be invisible. The benefits are obvious -- from a better understanding of living systems, to improved diagnostic technologies."
Professor Bowen said there were potentially boundless opportunities for quantum entanglement in technology. "Entanglement is set to revolutionize computing, communication and sensing," he said. "Absolutely secure communication was demonstrated some decades ago as the first demonstration of absolute quantum advantage over conventional technologies. Computing faster than any possible conventional computer was demonstrated by Google two years ago, as the first demonstration of absolute advantage in computing. The last piece of the puzzle was sensing, and we've now closed that gap."
A fast, low-cost technique to see and count viruses or proteins from a sample in real time, without any chemicals or dyes, could underpin a new class of devices for rapid diagnostics and viral load monitoring, including HIV and the virus that causes COVID-19.
Researchers at the University of Illinois Urbana-Champaign described the technique, called Photonic Resonator Interferometric Scattering Microscopy, or PRISM, in the journal Nature Communications.
“We have developed a new form of microscopy that amplifies the interaction between light and biological materials. We can use it for very rapid and sensitive forms of diagnostic testing, and also as a very powerful tool for understanding biological processes at the scale of individual items, like counting individual proteins or recording individual protein interactions,” said study leader Brian Cunningham, the Intel Alumni Endowed Chair of electrical and computer engineering at Illinois.
In optical microscopes, light bounces off any molecules or viruses it encounters on a slide, creating a signal. Instead of a regular glass slide, the PRISM technique uses a photonic crystal — a nanostructured glass surface that brilliantly reflects only one specific wavelength of light. Cunningham’s group designed and fabricated a photonic crystal that reflects red light, so that the light from a red laser would be amplified.
“The molecules we are looking at – in this study, viruses and small proteins – are extremely small. They cannot scatter enough light to create a signal that can be detected by a conventional optical microscope,” said graduate student Nantao Li, the first author of the paper. “The benefit of using the photonic crystal is that it amplifies the light’s intensity so it’s easier to detect those signals and enables us to study these proteins and viruses without any chemical labels or dyes that might modify their natural state or hinder their activity – we can just use the intrinsic scattering signal as the gauge for determining if those molecules are present.”
The researchers verified their technique by detecting the virus that causes COVID-19. PRISM detected individual coronaviruses as they traveled across the slide’s surface. The researchers also used PRISM to detect individual proteins such as ferritin and fibrinogen. The technique could allow researchers to study such biological targets in their natural states – watching as proteins interact, for example – or researchers could seed the surface of the photonic crystal slide with antibodies or other molecules to capture the targeted items and hold them in place.
“It takes 10 seconds to get a measurement, and in that time we can count the number of viruses captured on the sensor,” Cunningham said. “It’s a single-step detection method that works at room temperature. It is also fast, very sensitive and low cost. It’s very different from the standard way we do viral testing now, which involves breaking open the viruses, extracting their genetic material and putting it through a chemical amplification process so we can detect it. That method, called PCR, is accurate and sensitive, but it requires time, specialized equipment and trained technicians.”
Researchers at the Dutch Advanced Research Center for Nanolithography (ARCNL) and Vrije Universiteit Amsterdam (VU), developed an advanced microscope capable of super-resolution microscopy through an ultra-thin fiber.
Up until now, it was generally the case that the higher the resolution of a microscope, the larger the device needed to be, making it virtually impossible to look inside the human body in real-time. Although some methods that enable researchers to look inside living animals already exist, their resolution is very limited, and it takes a long time to generate an acceptable image.
With the use of smart signal processing, the researchers are able to beat the theoretical limits of resolution and speed. With this newly developed compact setup, scientists are finally able to, for example, look inside the brain in real-time and high resolution, using an ultra-thin fiber. Because the method does not require any unique fluorescent labeling, it is promising for both medical uses and characterization of 3D structures in nano-lithography!
Pathology typically involves cutting tissue samples by hand, placing each sample between two pieces of glass, and studying it under a microscope. While microscopes have improved with time, the method has remained largely unchanged for more than 150 years. A human can typically process about 12 sample slices per hour.
3Scan speeds this process up considerably. Its KESM tool uses an automated diamond knife to cut samples at 1,000 slices per hour while simultaneously scanning an image of each slice. Those scans are layered to create a 3D tissue model with micron-scale resolution.
3Scan’s platform has the potential to illuminate the mechanisms by which biological processes become abnormal, which could improve diagnostics, Daniel says: “There’s only so much tissue one person can see in their lifetime, and if we can build something that looks at pathology across many different demographics, across many different cases and diseases, we can get better insights.
There are also reports that show that pathologists achieve consensus on a case 80% of the time. That’s a pretty high success rate—unless it’s your diagnosis, in which case it’s very scary.” Once 3Scan’s tools do the imaging and initial analytics, the data is sent back to pathologists, who explore and translate the findings. “We like to imagine the pathologist as the conductor of an orchestra of robots that can go out there and image vast fields of biology,” Daniel, one of the inventors says. “The pathologist plays a crucial role in being the informed human perspective, differentiating what is pathologic versus normal within that biology.”
University of British Columbia researchers have developed a specialized microscope that has the potential ability to both diagnose diseases that include skin cancer and perform incredibly precise surgery—all without cutting skin. The researchers describe the technology in a study published today in Science Advances.
“Our technology allows us to scan tissue quickly, and when we see a suspicious or abnormal cell structure, we can perform ultra-precise surgery and selectively treat the unwanted or diseased structure within the tissue—without cutting into the skin,” said Yimei Huang, co-lead author of the study and a former postdoctoral fellow at the department of dermatology and skin science at UBC and BC Cancer.
Huang co-led the study with Zhenguo Wu, a UBC PhD student.
The device is a specialized type of multiphoton excitation microscope that allows imaging of living tissue up to about one millimeter in depth using an ultrafast infrared laser beam. What sets the researchers’ microscope apart from previous technology is that it’s capable of not only digitally scanning living tissue, but also treating the tissue by intensifying the heat produced by the laser.
When applied to treating diseases of the skin, the microscope allows medical professionals to pinpoint the exact location of the abnormality, diagnose it and treat it instantly. It could be used to treat any structure of the body that is reached by light and that requires extremely precise treatment, including nerves or blood vessels in the skin, eye, brain or other vital structures.
“We can alter the pathway of blood vessels without impacting any of the surrounding vessels or tissues,” said study co-author Harvey Lui, professor at the department of dermatology and skin science at UBC and the Vancouver Coastal Health Research Institute, and a dermatologist at BC Cancer. “For diagnosing and scanning diseases like skin cancer, this could be revolutionary.”
Single molecule localization microscopy can generate 3D super-resolution images without scanning by leveraging the axial variations of normal or engineered point spread functions (PSF). Successful implementation of these approaches for extended axial ranges remains, however, challenging.
A group of engineers and scientists present Zernike Optimized Localization Approach in 3D (ZOLA-3D), an easy-to-use computational and optical solution that achieves optimal resolution over a tunable axial range. They use ZOLA-3D to demonstrate 3D super-resolution imaging of mitochondria, nuclear pores and microtubules in entire nuclei or cells up to ~5 μm deep.
In summary, ZOLA-3D is a versatile, easy-to-use optical and computational imaging system for 3D SMLM that achieves theoretically optimal resolution over adjustable axial ranges from <1 μm up to at least 5 μm. Even larger axial ranges can in principle be achieved using different PSFs, although their larger lateral extent implies that the density of active fluorophores must be kept very low to maintain high localization precision.
ZOLA may be used to track single particles or molecules in 3D in live cells and that the use of reflective, rather than refractive optics will facilitate applications to multicolor imaging. The presented software is freely available from https://github.com/imodpasteur/ZOLA-3D, together with sample data and instructions. The researchers anticipate that ZOLA-3D will greatly facilitate 3D imaging of entire nuclei and cells with super-resolution.
Deep learning, which uses multi-layered artificial neural networks, is a form of machine learning that has demonstrated significant advances in many fields, including natural language processing, image/video labeling and captioning. In image processing, deep learning demonstrates significant potential for automated identification and labeling of features of interest, such as abnormal regions in a medical image.
UCLA researchers have demonstrated an innovative application of deep learning to significantly extend the imaging depth of a hologram. In holography, image reconstruction requires performing autofocusing and phase recovery, which are in general cumbersome and time-consuming to perform over a large sample volume. In a recent article published in Optica, a journal of the Optical Society of America, UCLA researchers have demonstrated a new approach they termed HIDEF based on a convolutional neural network that simultaneously performs autofocusing and phase recovery to significantly extend the image depth of field and the reconstruction speed in holography.
This research was led by Dr. Aydogan Ozcan, the Chancellor's Professor of electrical and computer engineering at UCLA and an HHMI Professor with the Howard Hughes Medical Institute, along with Yichen Wu, a graduate student, and Dr. Yair Rivenson, a postdoctoral scholar, both at the UCLA electrical and computer engineering department.
The authors validated this deep learning based approach by successfully reconstructing holograms of aerosols and human tissue samples. Overall, this approach significantly boosts the computational efficiency and the reconstruction speed of high-resolution holographic imaging by simultaneously performing autofocusing and phase recovery, which also increases the robustness of the image reconstruction process to potential misalignments in the optical setup by extending the depth of the reconstructed images.
Scientists have developed a new way to see inside individual cells, and study how they move and operate inside the human body. The improved understanding of cell-level activity could give researchers extra insight and tools to tackle cancers and other diseases.
The researchers are using a lab-made protein called an Affimer that binds to the F-actin protein. F-actin is part of the network within cells which gives them their shape and helps them move and divide. By seeing the cells move and change, scientists can begin to develop chemical compounds to target them, over time becoming new drug candidates.
Affimers were first developed at the University of Leeds and are a man-made alternative to animal-derived antibodies. This has the important extra benefit of reducing the numbers of animals used in research work.
This latest type of Affimer technology which recognises the F-actin protein is an important step forward in giving scientists the tools to combat diseases. Details are published in the Scientific Reports journal.
This review gives an overview of recent technology for imaging cells and viruses by light microscopy, in particular fluorescence microscopy in static and live-cell modes. The review lays out guidelines for how novel fluorescent chemical probes and proteins can be used in light microscopy to illuminate cells, and how they can be used to study virus infections. Discussed are advantages and opportunities of confocal and multi-photon microscopy, selective plane illumination microscopy, and super-resolution microscopy. The authors emphasize the prevalent concepts in image processing and data analyses, and provide an outlook into label-free digital holographic microscopy for virus research.
The new technique produces better images than current methods, and it’s easier to implement because it requires fewer measurements and performs computation.
A form of machine learning called deep learning is one of the key technologies behind recent advances in applications like real-time speech recognition and automated image and video labeling.
The approach, which uses multi-layered artificial neural networks to automate data analysis, also has shown significant promise for health care: It could be used, for example, to automatically identify abnormalities in patients’ X-rays, CT scans and other medical images and data.
In two new papers, UCLA researchers report that they have developed new uses for deep learning: reconstructing a hologram to form a microscopic image of an object and improving optical microscopy. Their new holographic imaging technique produces better images than current methods that use multiple holograms, and it’s easier to implement because it requires fewer measurements and performs computations faster.
For one study (PDF), published in Light: Science and Applications, the researchers produced holograms of Pap smears, which are used to screen for cervical cancer, and blood samples, as well as breast tissue samples. In each case, the neural network learned to extract and separate the features of the true image of the object from undesired light interference and from other physical byproducts of the image reconstruction process.
“These results are broadly applicable to any phase recovery and holographic imaging problem, and this deep-learning–based framework opens up myriad opportunities to design fundamentally new coherent imaging systems, spanning different parts of the electromagnetic spectrum, including visible wavelengths and even X-rays,” said Ozcan, who also is an HHMI Professor at the Howard Hughes Medical Institute.
A research team led by a scientist from the U.S. Department of Energy's Ames Laboratory has demonstrated for the first time that the magnetic fields of bacterial cells and magnetic nano-objects in liquid can be studied at high resolution using electron microscopy. This proof-of-principle capability allows first-hand observation of liquid environment phenomena, and has the potential to vastly increase knowledge in a number of scientific fields, including many areas of physics, nanotechnology, biofuels conversion, biomedical engineering, catalysis, batteries and pharmacology.
"It is much like being able to travel to a Jurassic Park and witness dinosaurs walking around, instead of trying to guess how they walked by examining a fossilized skeleton," said Tanya Prozorov, an associate scientist in Ames Laboratory's Division of Materials Sciences and Engineering. Prozorov works with biological and bioinspired magnetic nanomaterials, and faced what initially seemed to be an insurmountable challenge of observing them in their native liquid environment. She studies a model system, magnetotactic bacteria, which form perfect nanocrystals of magnetite. In order to best learn how bacteria do this, she needed an alternative to the typical electron microscopy process of handling solid samples in vacuum, where soft matter is studied in prepared, dried, or vitrified form.
For this work, Prozorov received DOE recognition through an Office of Science Early Career Research Program grant to use cutting-edge electron microscopy techniques with a liquid cell insert to learn how the individual magnetic nanocrystals form and grow with the help of biological molecules, which is critical for making artificial magnetic nanomaterials with useful properties.
To study magnetism in bacteria, she applied off-axis electron holography, a specialized technique that is used for the characterization of magnetic nanostructures in the transmission electron microscope, in combination with the liquid cell. "When we look at samples prepared in the conventional way, we have to make many assumptions about their properties based on their final state, but with the new technique, we can now observe these processes first-hand," said Prozorov. "It can help us understand the dynamics of macromolecule aggregation, nanoparticle self-assembly, and the effects of electric and magnetic fields on that process."
"This method allows us to obtain large amounts of new information," said Prozorov. "It is a first step, proving that the mapping of magnetic fields in liquid at the nanometer scale with electron microscopy could be done; I am eager to see the discoveries it could foster in other areas of science."
Measuring nascent macromolecular synthesis in vivo is key to understanding how cells and tissues progress through development and respond to external cues. Researchers recently performed in vivo injection of alkyne- or azide-modified analogs of thymidine, uridine, methionine, and glucosamine to label nascent synthesis of DNA, RNA, protein, and glycosylation. Three-dimensional volumetric imaging of nascent macromolecule synthesis was performed in axolotl salamander tissue using whole-mount click chemistry-based fluorescent staining followed by light sheet fluorescent microscopy. They also developed an image processing pipeline for segmentation and classification of morphological regions of interest and individual cells, and applied this pipeline to the regenerating humerus. They were able to demonstrate that this approach is sensitive to biological perturbations by measuring changes in DNA synthesis after limb denervation. Taken together, this method provides a powerful means to quantitatively interrogate macromolecule synthesis in heterogenous tissues at the organ, cellular, and molecular levels of organization.
Ultrafast electron microscope opens up new avenues for the development of sensors and quantum devices. Everyone who has ever been to the Grand Canyon can relate to having strong feelings from being close to one of nature's edges. Similarly, scientists at the U.S. Department of Energy's (DOE) Argonne National Laboratory have discovered that nanoparticles of gold act unusually when close to the edge of a one-atom thick sheet of carbon, called graphene. This could have big implications for the development of new sensors and quantum devices.
This new discovery was made possible with a newly established ultrafast electron microscope (UEM) at Argonne's Center for Nanoscale Materials (CNM), a DOE Office of Science User Facility. The UEM enables the visualization and investigation of phenomena at the nanoscale and on time frames of less than a trillionth of a second. This discovery could make a splash in the growing field of plasmonics, which involves light striking a material surface and triggering waves of electrons, known as plasmonic fields.
"With these ultrafast capabilities, there is no telling what we might see as we tweak different materials and their properties." -- Haihua Liu, Argonne nanoscientist. For years, scientists have been pursuing development of plasmonic devices with a wide range of applications -- from quantum information processing to optoelectronics (which combine light-based and electronic components) to sensors for biological and medical purposes. To do so, they couple two-dimensional materials with atomic-level thickness, such as graphene, with nanosized metal particles. Understanding the combined plasmonic behavior of these two different types of materials requires understanding exactly how they are coupled.
In a recent study from Argonne, researchers used ultrafast electron microscopy to look directly at the coupling between gold nanoparticles and graphene. "Surface plasmons are light-induced electron oscillations on the surface of a nanoparticle or at an interface of a nanoparticle and another material," said Argonne nanoscientist Haihua Liu. "When we shine a light on the nanoparticle, it creates a short-lived plasmonic field. The pulsed electrons in our UEM interact with this short-lived field when the two overlap, and the electrons either gain or lose energy. Then, we collect those electrons that gain energy using an energy filter to map the plasmonic field distributions around the nanoparticle."
In studying the gold nanoparticles, Liu and his colleagues discovered an unusual phenomenon. When the nanoparticle sat on a flat sheet of graphene, the plasmonic field was symmetric. But when the nanoparticle was positioned close to a graphene edge, the plasmonic field concentrated much more strongly near the edge region. "It's a remarkable new way of thinking about how we can manipulate charge in the form of a plasmonic field and other phenomena using light at the nanoscale," Liu said. "With ultrafast capabilities, there's no telling what we might see as we tweak different materials and their properties."
Collaboration between deep learning experts and microscopy experts leads to a significantly improved data-intensive light-field microscopy method by using AI and pairing it with light-sheet microscopy. The result is the power of light-field microscopy available to biologists in near real time vs. days or weeks, AND the expansion of biologists' ability to use this microscopy for many things more things requiring the most detailed observation.
To observe the swift neuronal signals in a fish brain, scientists have started to use a technique called light-field microscopy, which makes it possible to image such fast biological processes in 3D. But the images are often lacking in quality, and it takes hours or days for massive amounts of data to be converted into 3D volumes and movies.
Now, EMBL scientists have combined artificial intelligence (AI) algorithms with two cutting-edge microscopy techniques - an advance that shortens the time for image processing from days to mere seconds, while ensuring that the resulting images are crisp and accurate. The findings are published in Nature Methods.
"Ultimately, we were able to take 'the best of both worlds' in this approach," says Nils Wagner, one of the paper's two lead authors and now a PhD student at the Technical University of Munich. "AI enabled us to combine different microscopy techniques, so that we could image as fast as light-field microscopy allows and get close to the image resolution of light-sheet microscopy."
Although light-sheet microscopy and light-field microscopy sound similar, these techniques have different advantages and challenges. Light-field microscopy captures large 3D images that allow researchers to track and measure remarkably fine movements, such as a fish larva's beating heart, at very high speeds. But this technique produces massive amounts of data, which can take days to process, and the final images usually lack resolution.
Light-sheet microscopy homes in on a single 2D plane of a given sample at one time, so researchers can image samples at higher resolution. Compared with light-field microscopy, light-sheet microscopy produces images that are quicker to process, but the data are not as comprehensive, since they only capture information from a single 2D plane at a time. To take advantage of the benefits of each technique, EMBL researchers developed an approach that uses light-field microscopy to image large 3D samples and light-sheet microscopy to train the AI algorithms, which then create an accurate 3D picture of the sample.
"If you build algorithms that produce an image, you need to check that these algorithms are constructing the right image," explains Anna Kreshuk, the EMBL group leader whose team brought machine learning expertise to the project. In the new study, the researchers used light-sheet microscopy to make sure the AI algorithms were working, Anna says. "This makes our research stand out from what has been done in the past."
Robert Prevedel, the EMBL group leader whose group contributed the novel hybrid microscopy platform, notes that the real bottleneck in building better microscopes often isn't optics technology, but computation. That's why, back in 2018, he and Anna decided to join forces. "Our method will be really key for people who want to study how brains compute. Our method can image an entire brain of a fish larva, in real time," Robert says.
Unprecedented details of enamel structure may point to new ways to prevent or halt cavities.
Scientists used a combination of advanced microscopy and chemical detection techniques to uncover the structural makeup of human tooth enamel at unprecedented atomic resolution, revealing lattice patterns and unexpected irregularities. The findings could lead to a better understanding of how tooth decay develops and might be prevented. The research was supported in part by the National Institute of Dental and Craniofacial Research (NIDCR) at the National Institutes of Health. The findings appear in Nature.
“This work provides much more detailed information about the atomic makeup of enamel than we previously knew,” said Jason Wan, Ph.D., a program officer at NIDCR. “These findings can broaden our thinking and approach to strengthening teeth against mechanical forces, as well as repairing damage due to erosion and decay.”
Your teeth are remarkably resilient, despite enduring the stress and strain of biting, chewing, and eating for a lifetime. Enamel — the hardest substance in the human body — is largely responsible for this endurance. Its high mineral content gives it strength. Enamel forms the outer covering of teeth and helps prevent tooth decay, or caries.
Tooth decay is one of the most common chronic diseases, affecting up to 90% of children and the vast majority of adults worldwide, according to the World Health Organization. Left untreated, tooth decay can lead to painful abscesses, bone infection, and bone loss.
Tooth decay starts when excess acid in the mouth erodes the enamel covering. Scientists have long sought a more complete picture of enamel’s chemical and mechanical properties at the atomic level to better understand—and potentially prevent or reverse—enamel loss. To survey enamel at the tiniest scales, researchers use microscopy methods such as scanning transmission electron microscopy (STEM), which directs a beam of electrons through a material to map its atomic makeup.
OpenAI's Microscope project visualizes popular machine learning models at the neuron level to enable advances like reverse engineering neural networks.
OpenAI Microscope is a collection of visualizations of every significant layer and neuron of several common “model organisms” which are often studied in interpretability. Microscope makes it easier to analyze the features that form inside these neural networks, and we hope it will help the research community as we move towards understanding these complicated systems.
Who is this for?
While we’re making this available to anyone who’s interested in exploring how neural networks work, we think the primary value is in providing persistent, shared artifacts to facilitate long-term comparative study of these models. We also hope that researchers with adjacent expertise — neuroscience, for instance — will find value in being able to more easily approach the internal workings of these vision models.
How do I use it? Can you point me at some interesting things?
The OpenAI Microscope is based on two concepts, a location in a model and a technique. Metaphorically, the location is where you point the microscope, the technique is what lens you affix to it.
Our models are composed of a graph of “nodes” (the neural network layers), which are connected to each other through “edges.” Each op contains hundreds of “units”, which are roughly analogous to neurons. Most of the techniques we use are useful only at a specific resolution. For instance, feature visualization can only be pointed at a “unit”, not its parent “node”. The circuits collaboration is a great place to find interesting units to look at, from low-level features like curve detectors (eg. 3b:379,406,385,343,342,388, ...) to high level features like pose-invariant dog detectors (eg.4b:409,418) or car detectors (4c:447).
What visualizations are available in Microscope?
We’re currently including a couple visualizations we find particularly helpful in the microscope, including feature visualizations, DeepDream, dataset examples (images that cause neurons to fire strongly), and synthetic tuning curves (neuroscience-inspired plots of how units respond to synthetic image families). We expect to add more over time. Each technique is explained when it is shown.
Rather than relying on optics, the DNA microscopy system offers a chemically encoded way to map biomolecules’ relative positions.
Microscopy just got reinvented – again. Traditionally, scientists have used light, x-rays, and electrons to peer inside tissues and cells. Today, scientists can trace thread-like fibers of nerves throughout the brain and even watch living mouse embryos conjure the beating cells of a rudimentary heart. But there’s one thing these microscopes can’t see: what’s happening in cells at the genomic level.
Now, biophysicist Joshua Weinstein and colleagues have invented an unorthodox type of imaging dubbed “DNA microscopy” that can do just that. Instead of relying on light (or any kind of optics at all), the team uses DNA “bar codes” to help pinpoint molecules’ relative positions within a sample.
With DNA microscopy, scientists can build a picture of cells and simultaneously amass enormous amounts of genomic information, Weinstein says. “This gives us another layer of biology that we haven’t been able to see.”
Weinstein, Howard Hughes Medical Institute (HHMI) Investigator Aviv Regev, and molecular biologist Feng Zhang, who was selected as an HHMI investigator in 2018, report the work June 20, 2019, in the journal Cell.
“It’s an entirely new category of microscopy,” Regev says. “It’s not just a new technique, it’s a way of doing things that we haven’t ever considered doing before.”
Applications of deep learning to medical disciplines including ophthalmology, dermatology, radiology, and pathology have recently shown great promise to increase both the accuracy and availability of high-quality healthcare to patients around the world. At Google, we have also published results showing that a convolutional neural network is able to detect breast cancer metastases in lymph nodes at a level of accuracy comparable to a trained pathologist. However, because direct tissue visualization using a compound light microscope remains the predominant means by which a pathologist diagnoses illness, a critical barrier to the widespread adoption of deep learning in pathology is the dependence on having a digital representation of the microscopic tissue.
Using a combination of photonic time stretch and deep learning, the device is capable of analyzing 36 million images per second while leaving blood samples undamaged so they may be used for other testing purposes. The new photonic time stretch technology, invented by Barham Jalali who also led the study, takes pictures of blood cells using flashing lasers occurring in nanosecond bursts.
“Each frame is slowed down in time and optically amplified so it can be digitized,” said Ata Mahjoubfar, a UCLA postdoctoral fellow who worked on the project. “This lets us perform fast cell imaging that the artificial intelligence component can distinguish.”
While such a quick burst of capturing photos would typically require intense illumination, which could destroy live cells, UCLA doctoral student Clair Lifan Chen explains how Jalali’s new technique eliminates that problem. “The photonic time stretch technique allows us to identify rogue cells in a short time with low-level illumination,” Chen said.
The photos are processed using deep learning, which runs data through a mass of algorithms to efficiently and accurately “read” the information. Deep learning has also been used to analyze patients’ genes, allowing identification of diseases or cancer that may otherwise go undetected, and has the potential to further understand cancer-forming mutations.
So far, the newly created process has shown 95% accuracy in differentiating healthy and cancer-riddled cells — a 17% improvement over current techniques. Existent techniques for detecting cancerous cells include identifying cancer cells based on physical characteristics, a method that oftentimes misidentifies regular cells as damaged, or using fluorescent staining that binds to cancerous cells allowing cell detection yet effectively damaging the blood sample.
In January 2018, a group of specialists in Wales, Germany, London, the United States, and Newcastle accomplished the same technique utilizing technology similar to face and fingerprint recognition software. In addition, the researchers found that their new method could also determine a cell’s age, which plays an important factor in the effectiveness of treatments.
Continuing discoveries and advancements from scientists fighting cancer using cutting edge technology has, according to Dr. Fabian Theis of the Helmholtz Zentrum Munchen in Germany, “open[ed] a whole new perspective that could also be used for entirely different research questions, not only for cell analysis.”
More than 350 years ago, the English natural philosopher Robert Hooke looked through a microscope at a thin slice of cork and discovered that it was made of small, box-like compartments, which he named “cells.” From that moment on, Hooke and countless inquisitive minds after him strove for a better view of these fundamental building blocks of life. And now, the window into the cellular world has become a lot clearer.
In a new study in the April 20, 2018 issue of Science, researchers from Howard Hughes Medical Institute’s (HHMI) Janelia Research Campus, Harvard Medical School, and collaborating institutions report the development of a microscope capable of capturing, in unprecedented detail, 3-D images and videos of cells inside living organisms. Adapting a technique used by astronomers to study distant stars, the research team, led by Nobel laureate and Janelia group leader Eric Betzig, showcased the new technology by generating a series of stunning movies: cancer cells crawling through blood vessels, spinal nerve cells wiring up into circuits, immune cells cruising through a zebrafish’s inner ear, and much more.
The resolution of the microscope is stunning and so powerful it can even capture subcellular details such as the dynamics of miniscule bubbles known as vesicles, which transport molecular cargo through to the cell. “This is the miracle of being able to see what we have never been able to see before. It’s simply incredible,” said study co-author Tomas Kirchhausen, HMS professor of cell biology, and the Springer Family Chair of pediatrics and a senior investigator at Boston Children’s Hospital.
Fast live-cell 3D phase imaging of cellular dynamics. (Left) Human fibroblast migrating on a glass substrate, showing first frame of a 25-second movie imaged.
Scientists at the Laboratory of Biomedical Optics (LOB) at EPFL (École Polytechnique Fédérale de Lausanne) in Switzerland have developed the first microscope platform that can perform “super-resolution” imaging in both space and time — capturing unprecedented “4D” views inside living cells. The landmark paper is published in Nature Photonics and on open-access ArXiv.
But super-resolution microscopy only offers improved spatial resolution. That might suffice for static samples, like solid materials or fixed cells, but living cells are highly dynamic and depend on a complex set of constantly changing biological processes that occur across sub-second timescales. So to visualize and understand how living cells function in health and disease, high “temporal” (time) resolution is also required.
Enter the 4D microscope. A team led by Professor Theo Lasser, head of the LOB, has developed a “4D microscope” that they dubbed PRISM (Phase Retrieval Instrument with Super-resolution Microscopy). A simple add-on to existing widefield microscopes, it combines 3D super-resolution microscopy (for high spatial resolution) with fast 3D phase (time) imaging in a single instrument. Phase imaging translates phase changes (changes over time) of light — caused by changes in cells and their organelles — into conventional spatial maps of the cells.
In a new study in the April 20 issue of Science, researchers from Howard Hughes Medical Institute's (HHMI) Janelia Research Campus, Harvard Medical School and collaborating institutions report the development of a microscope capable of capturing 3-D images and videos of cells inside living organisms in unprecedented detail.
“It’s like ‘Star Trek.’ It’s the age of exploration again." - Gokul Upadhyayula, HMS instructor of pediatrics said. Adapting a technique used by astronomers to study distant stars, the research team, led by Nobel laureate and Janelia group leader Eric Betzig, showcased the new technology by generating a series of stunning movies: cancer cells crawling through blood vessels, spinal nerve cells wiring up into circuits, immune cells cruising through a zebrafish’s inner ear and much more.
The resolution of the microscope is powerful enough to even capture subcellular details such as the dynamics of miniscule bubbles known as vesicles, which transport molecular cargo through to the cell. “This is the miracle of being able to see what we have never been able to see before. It’s simply incredible,” said study co-author Tomas Kirchhausen, HMS professor of cell biology, the Springer Family Chair of pediatrics and a senior investigator at Boston Children’s Hospital.
Researchers at the UCLA Samueli School of Engineering have demonstrated that deep learning, a powerful form of artificial intelligence, can discern and enhance microscopic details in photos taken by smartphones. The technique improves the resolution and color details of smartphone images so much that they approach the quality of images from laboratory-grade microscopes.
The advance could help bring high-quality medical diagnostics into resource-poor regions, where people otherwise do not have access to high-end diagnostic technologies. And the technique uses attachments that can be inexpensively produced with a 3-D printer, at less than $100 a piece, versus the thousands of dollars it would cost to buy laboratory-grade equipment that produces images of similar quality.
Cameras on today's smartphones are designed to photograph people and scenery, not to produce high-resolution microscopic images. So the researchers developed an attachment that can be placed over the smartphone lens to increase the resolution and the visibility of tiny details of the images they take, down to a scale of approximately one millionth of a meter.
But that only solved part of the challenge, because no attachment would be enough to compensate for the difference in quality between smartphone cameras' image sensors and lenses and those of high-end lab equipment. The new technique compensates for the difference by using artificial intelligence to reproduce the level of resolution and color details needed for a laboratory analysis.The research was led by Aydogan Ozcan, Chancellor's Professor of Electrical and Computer Engineering and Bioengineering, and Yair Rivenson, a UCLA postdoctoral scholar. Ozcan's research group has introduced several innovations in mobile microscopy and sensing, and it maintains a particular focus on developing field-portable medical diagnostics and sensors for resource-poor areas.
"Using deep learning, we set out to bridge the gap in image quality between inexpensive mobile phone-based microscopes and gold-standard bench-top microscopes that use high-end lenses," Ozcan said.
New research published in Nature Methods will dramatically improve how scientists "see inside" molecular structures in solution, allowing for much more precise ways to image data in various fields, from astronomy to drug discovery.
The new method will allow for the visualization of many more biological molecules, providing critical information about what is inside molecules to scientists who currently can only access their outer shape or envelope. Such information could be a major boost to studies of viruses, for example.
"With existing techniques, you can only see the outline of the virus," said author Thomas D. Grant, PhD, research assistant professor in the Department of Structural Biology in the Jacobs School of Medicine and Biomedical Sciences at the University at Buffalo and the Department of Materials, Design and Innovation in the UB School of Engineering and Applied Sciences and Hauptman-Woodward Medical Research Institute. "This new method allows us to see inside the virus molecule to understand how the genetic information is arranged, potentially giving new insight into how the virus injects this genetic information into its host."
Grant is the sole author of the paper, a rarity among papers published in this journal. He is a scientist with BioXFEL (Biology with X-ray Free Electron Lasers), a National Science Foundation Science and Technology Center composed of eight U.S. research universities that is headquartered at UB. Its mission is to address fundamental questions in biology at the molecular level using cutting-edge techniques, including X-ray laser science.
Grant's method has solved the phase problem for a particular molecular determination technique called solution scattering. The phase problem is where critical information about the phase of a molecule is lost during the experimental process of making a physical measurement.
He explained that most molecular structures today are solved using X-ray crystallography, where the structures scatter intense X-rays in patterns consisting of hundreds of thousands of unique pieces of information, which are used to ultimately reveal the structure at high-resolution.
"The problem is that more than 75 percent of molecular structures do not readily form the ordered crystals that diffract well," explained Grant. "That means many molecules are difficult to visualize in three dimensions."
In addition, he said, biological molecules can exhibit dynamic motions that have an impact on how they function but those motions are missing when structures crystallize, resulting in the loss of important biological information. One way around this obstacle is to use a technique called solution scattering in which X-rays scatter off of molecules floating in solution instead of arranged in a crystal.
"Solution scattering allows the molecules to move dynamically in their natural states, enabling the visualization of large-scale conformational dynamics important for biological function," said Grant. "However, as the molecules tumble in solution, they scatter the X-rays in many different orientations, losing most of the information, typically yielding only 10 to 20 unique pieces of data." Until now, such little information only yielded low-resolution outlines of the particle shape.
To get content containing either thought or leadership enter:
To get content containing both thought and leadership enter:
To get content containing the expression thought leadership enter:
You can enter several keywords and you can refine them whenever you want. Our suggestion engine uses more signals but entering a few keywords here will rapidly give you great content to curate.
Your new post is loading...
