More Like this

1. Optimal decoding of neural dynamics occurs at mesoscale spatial and temporal resolutions

   https://doi.org/10.3389/fncel.2024.1287123  

   Samiei, Toktam ; Zou, Zhuowen ; Imani, Mohsen ; Nozari, Erfan ( February 2024 , Frontiers in Cellular Neuroscience)

   Jonathan R. Whitlock (Ed.)

   Understanding the neural code has been one of the central aims of neuroscience research for decades. Spikes are commonly referred to as the units of information transfer, but multi-unit activity (MUA) recordings are routinely analyzed in aggregate forms such as binned spike counts, peri-stimulus time histograms, firing rates, or population codes. Various forms of averaging also occur in the brain, from the spatial averaging of spikes within dendritic trees to their temporal averaging through synaptic dynamics. However, how these forms of averaging are related to each other or to the spatial and temporal units of information representation within the neural code has remained poorly understood.

   In this work we developed NeuroPixelHD, a symbolic hyperdimensional model of MUA, and used it to decode the spatial location and identity of static images shown ton= 9 mice in the Allen Institute Visual Coding—NeuroPixels dataset from large-scale MUA recordings. We parametrically varied the spatial and temporal resolutions of the MUA data provided to the model, and compared its resulting decoding accuracy.

   For almost all subjects, we found 125ms temporal resolution to maximize decoding accuracy for both the spatial location of Gabor patches (81 classes for patches presented over a 9×9 grid) as well as the identity of natural images (118 classes corresponding to 118 images) across the whole brain. This optimal temporal resolution nevertheless varied greatly between different regions, followed a sensory-associate hierarchy, and was significantly modulated by the central frequency of theta-band oscillations across different regions. Spatially, the optimal resolution was at either of two mesoscale levels for almost all mice: the area level, where the spiking activity of all neurons within each brain area are combined, and the population level, where neuronal spikes within each area are combined across fast spiking (putatively inhibitory) and regular spiking (putatively excitatory) neurons, respectively. We also observed an expected interplay between optimal spatial and temporal resolutions, whereby increasing the amount of averaging across one dimension (space or time) decreases the amount of averaging that is optimal across the other dimension, and vice versa.

   Our findings corroborate existing empirical practices of spatiotemporal binning and averaging in MUA data analysis, and provide a rigorous computational framework for optimizing the level of such aggregations. Our findings can also synthesize these empirical practices with existing knowledge of the various sources of biological averaging in the brain into a new theory of neural information processing in which theunit of informationvaries dynamically based on neuronal signal and noise correlations across space and time.

    

   more » « less
2. Deep Denoising for Scientific Discovery: A Case Study in Electron Microscopy

   Mohan, Sreyas ; Manzorro, Ramon ; Vincent, Joshua L. ; Tang, Binh ; Sheth, Dev Y. ; Simoncelli, Eero P. ; Matteson, David S. ; Crozier, Peter A. ; Fernandez-Granda, Carlos ( March 2021 , ArXivorg)

   null (Ed.)

   Denoising is a fundamental challenge in scientific imaging. Deep convolutional neural networks (CNNs) provide the current state of the art in denoising natural images, where they produce impressive results. However, their potential has been inadequately explored in the context of scientific imaging. Denoising CNNs are typically trained on real natural images artificially corrupted with simulated noise. In contrast, in scientific applications, noiseless ground-truth images are usually not available. To address this issue, we propose a simulation-based denoising (SBD) framework, in which CNNs are trained on simulated images. We test the framework on data obtained from transmission electron microscopy (TEM), an imaging technique with widespread applications in material science, biology, and medicine. SBD outperforms existing techniques by a wide margin on a simulated benchmark dataset, as well as on real data. We analyze the generalization capability of SBD, demonstrating that the trained networks are robust to variations of imaging parameters and of the underlying signal structure. Our results reveal that state-of-the-art architectures for denoising photographic images may not be well adapted to scientific-imaging data. For instance, substantially increasing their field-of-view dramatically improves their performance on TEM images acquired at low signal-to-noise ratios. We also demonstrate that standard performance metrics for photographs (such as PSNR and SSIM) may fail to produce scientifically meaningful evaluation. We propose several metrics to remedy this issue for the case of atomic resolution electron microscope images. In addition, we propose a technique, based on likelihood computations, to visualize the agreement between the structure of the denoised images and the observed data. Finally, we release a publicly available benchmark dataset of TEM images, containing 18,000 examples. 

   more » « less
3. Dislocation structure of deformed olivine single crystals from conventional EBSD maps

   https://doi.org/10.1007/s00269-021-01157-3  

   Faul, Ulrich ( September 2021 , Physics and Chemistry of Minerals)

   Dislocations, linear defects in a crystalline lattice characterized by their slip systems, can provide a record of grain internal deformation. Comprehensive examination of this record has been limited by intrinsic limitations of the observational methods. Transmission electron microscopy reveals individual dislocations, but images only a few square$$\upmu$$$\mu$m of sample. Oxidative decoration requires involved sample preparation and has uncertainties in detection of all dislocations and their types. The possibility of mapping dislocation density and slip systems by conventional (Hough-transform based) EBSD is investigated here with naturally and experimentally deformed San Carlos olivine single crystals. Geometry and dislocation structures of crystals deformed in orientations designed to activate particular slip systems were previously analyzed by TEM and oxidative decoration. A curvature tensor is calculated from changes in orientation of the crystal lattice, which is inverted to calculate density of geometrically necessary dislocations with the Matlab Toolbox MTEX. Densities of individual dislocation types along with misorientation axes are compared to orientation change measured on the deformed crystals. After filtering (denoising), noise floor and calculated dislocation densities are comparable to those reported from high resolution EBSD mapping. For samples deformed in [110]c and [011]c orientations EBSD mapping confirms [100](010) and [001](010), respectively, as the dominant slip systems. EBSD mapping thus enables relatively efficient observation of dislocation structures associated with intracrystalline deformation, both distributed, and localized at sub-boundaries, over substantially larger areas than has previously been possible. This will enable mapping of dislocation structures in both naturally and experimentally deformed polycrystals, with potentially new insights into deformation processes in Earth’s upper mantle.

    

   more » « less
4. Asymmetric nanoparticle oxidation observed in-situ by the evolution of diffraction contrast

   https://doi.org/10.1088/2515-7639/ad025f  

   Poerwoprajitno, Agus R. ; Baradwaj, Nitish ; Singh, Manish Kumar ; Carter, C. Barry ; Huber, Dale L. ; Kalia, Rajiv ; Watt, John ( October 2023 , Journal of Physics: Materials)

   The use of transmission electron microscopy (TEM) to observe real-time structural and compositional changes has proven to be a valuable tool for understanding the dynamic behavior of nanomaterials. However, identifying the nanoparticles of interest typically require an obvious change in position, size, or structure, as compositional changes may not be noticeable during the experiment. Oxidation or reduction can often result in subtle volume changes only, so elucidating mechanisms in real-time requires atomic-scale resolution orin-situelectron energy loss spectroscopy, which may not be widely accessible. Here, by monitoring the evolution of diffraction contrast, we can observe both structural and compositional changes in iron oxide nanoparticles, specifically the oxidation from a wüstite-magnetite (FeO@Fe₃O₄) core–shell nanoparticle to single crystalline magnetite, Fe₃O₄nanoparticle. Thein-situTEM images reveal a distinctive light and dark contrast known as the ‘Ashby-Brown contrast’, which is a result of coherent strain across the core–shell interface. As the nanoparticles fully oxidize to Fe₃O₄, the diffraction contrast evolves and then disappears completely, which is then confirmed by modeling and simulation of TEM images. This represents a new, simplified approach to tracking the oxidation or reduction mechanisms of nanoparticles usingin-situTEM experiments.

    

   more » « less
5. Direct Visualization of the Earliest Stages of Crystallization

   https://doi.org/10.1017/S1431927621000441  

   Singh, Manish Kumar ; Ghosh, Chanchal ; Miller, Benjamin ; Carter, C. Barry ( January 2021 , Microscopy and Microanalysis)

   null (Ed.)

   Investigating the earliest stages of crystallization requires the transmission electron microscope (TEM) and is particularly challenging for materials which can be affected by the electron beam. Typically, when imaging at magnifications high enough to observe local crystallinity, the electron beam's current density must be high to produce adequate image contrast. Yet, minimizing the electron dose is necessary to reduce the changes caused by the beam. With the advent of a sensitive, high-speed, direct-detection camera for a TEM that is corrected for spherical aberration, it is possible to probe the early stages of crystallization at the atomic scale. High-quality images with low contrast can now be analyzed using new computing methods. In the present paper, this approach is illustrated for crystallization in a Ge 2 Sb 2 Te 5 (GST-225) phase-change material which can undergo particularly rapid phase transformations and is sensitive to the electron beam. A thin (20 nm) film of GST-225 has been directly imaged in the TEM and the low-dose images processed using Python scripting to extract details of the nanoscale nuclei. Quantitative analysis of the processed images in a video sequence also allows the growth of such nuclei to be followed. 

   more » « less