Search

Automatic detection and segmentation of objects in 2D and 3D microscopy data is important for countless biomedical applications. In the natural image domain, spatial embedding-based instance segmentation methods are known to yield high-quality results, but their utility for segmenting microscopy data is currently little researched. Here we introduce EmbedSeg, an embedding-based instance segmentation method which outperforms existing state-of-the-art baselines on 2D as well as 3D microscopy datasets. Additionally, we show that EmbedSeg has a GPU memory footprint small enough to train even on laptop GPUs, making it accessible to virtually everyone. Finally, we introduce four new 3D microscopy datasets, which we make publicly available alongside ground truth training labels. Our open-source implementation is available at https://github.com/juglab/EmbedSeg.

Image denoising is the first step in many biomedical image analysis pipelines and Deep Learning (DL) based methods are currently best performing. A new category of DL methods such as Noise2Void or Noise2Self can be used fully unsupervised, requiring nothing but the noisy data. However, this comes at the price of reduced reconstruction quality. The recently proposed Probabilistic Noise2Void (PN2V) improves results, but requires an additional noise model for which calibration data needs to be acquired. Here, we present improvements to PN2V that (i) replace histogram based noise models by parametric noise models, and (ii) show how suitable noise models can be created even in the absence of calibration data. This is a major step since it actually renders PN2V fully unsupervised. We demonstrate that all proposed improvements are not only academic but indeed relevant., Comment: Accepted at ISBI 2020

Deep learning (DL) has arguably emerged as the method of choice for the detection and segmentation of biological structures in microscopy images. However, DL typically needs copious amounts of annotated training data that is for biomedical projects typically not available and excessively expensive to generate. Additionally, tasks become harder in the presence of noise, requiring even more high-quality training data. Hence, we propose to use denoising networks to improve the performance of other DL-based image segmentation methods. More specifically, we present ideas on how state-of-the-art self-supervised CARE networks can improve cell/nuclei segmentation in microscopy data. Using two state-of-the-art baseline methods, U-Net and StarDist, we show that our ideas consistently improve the quality of resulting segmentations, especially when only limited training data for noisy micrographs are available., Comment: accepted at ISBI 2020

Life science today involves computational analysis of a large amount and variety of data, such as volumetric data acquired by state-of-the-art microscopes, or mesh data from analysis of such data or simulations. Visualization is often the first step in making sense of data, and a crucial part of building and debugging analysis pipelines. It is therefore important that visualizations can be quickly prototyped, as well as developed or embedded into full applications. In order to better judge spatiotemporal relationships, immersive hardware, such as Virtual or Augmented Reality (VR/AR) headsets and associated controllers are becoming invaluable tools. In this work we introduce scenery, a flexible VR/AR visualization framework for the Java VM that can handle mesh and large volumetric data, containing multiple views, timepoints, and color channels. scenery is free and open-source software, works on all major platforms, and uses the Vulkan or OpenGL rendering APIs. We introduce scenery's main features and example applications, such as its use in VR for microscopy, in the biomedical image analysis software Fiji, or for visualizing agent-based simulations., Comment: Added IEEE DOI, version published at VIS 2019

Multiview light sheet fluorescence microscopy (LSFM) allows to image developing organisms in 3D at unprecedented temporal resolution over long periods of time. The resulting massive amounts of raw image data requires extensive processing interactively via dedicated graphical user interface (GUI) applications. The consecutive processing steps can be easily automated and the individual time points can be processed independently, which lends itself to trivial parallelization on a high performance cluster (HPC). Here we introduce an automated workflow for processing large multiview, multi-channel, multi-illumination time-lapse LSFM data on a single workstation or in parallel on a HPC. The pipeline relies on snakemake to resolve dependencies among consecutive processing steps and can be easily adapted to any cluster environment for processing LSFM data in a fraction of the time required to collect it., Comment: 13 pages with supplement, LATEX; 1 table, 1 figure, 2 supplementary figures, 2 supplementary lists, 2 supplementary tables; corrected error in results table, results unchanged

The increasingly popular light sheet microscopy techniques generate very large 3D time-lapse recordings of living biological specimen. The necessity to make large volumetric datasets available for interactive visualization and analysis has been widely recognized. However, existing solutions build on dedicated servers to generate virtual slices that are transferred to the client applications, practically leading to insufficient frame rates (less than 10 frames per second) for truly interactive experience. An easily accessible open source solution for interactive arbitrary virtual re-slicing of very large volumes and time series of volumes has yet been missing. We fill this gap with BigDataViewer, a Fiji plugin to interactively navigate and visualize large image sequences from both local and remote data sources., Comment: 38 pages, 1 main figure, 27 supplementary figures, under review at Nature Methods

Light sheet fluorescence microscopy is able to image large specimen with high resolution by imaging the sam- ples from multiple angles. Multi-view deconvolution can significantly improve the resolution and contrast of the images, but its application has been limited due to the large size of the datasets. Here we present a Bayesian- based derivation of multi-view deconvolution that drastically improves the convergence time and provide a fast implementation utilizing graphics hardware., Comment: 48 pages, 20 figures, 1 table, under review at Nature Methods

Expansion of the neocortex is a hallmark of human evolution. However, it remains an open question what adaptive mechanisms facilitated its expansion. Here we show, using gyrencephaly index (GI) and other physiological and life-history data for 102 mammalian species, that gyrencephaly is an ancestral mammalian trait. We provide evidence that the evolution of a highly folded neocortex, as observed in humans, requires the traversal of a threshold of 10^9 neurons, and that species above and below the threshold exhibit a bimodal distribution of physiological and life-history traits, establishing two phenotypic groups. We identify, using discrete mathematical models, proliferative divisions of progenitors in the basal compartment of the developing neocortex as evolutionarily necessary and sufficient for generating a fourteen-fold increase in daily prenatal neuron production and thus traversal of the neuronal threshold. We demonstrate that length of neurogenic period, rather than any novel progenitor-type, is sufficient to distinguish cortical neuron number between species within the same phenotypic group., Comment: Currently under review; 38 pages, 5 Figures, 13 Supplementary Figures, 2 Tables

Light sheet microscopy promises to revolutionize developmental biology by enabling live in toto imaging of entire embryos with minimal phototoxicity. We present detailed instructions for building a compact and customizable Selective Plane Illumination Microscopy (SPIM) system. The integrated OpenSPIM hardware and software platform is shared with the scientific community through a public website, thereby making light sheet microscopy accessible for widespread use and optimization to various applications., Comment: 7 pages, 3 figures, 6 supplementary videos, submitted to Nature Methods, associated public website http://openspim.org

The X chromosome is present as a single copy in the heterogametic sex, and this hemizygosity is expected to drive unusual patterns of evolution on the X relative to the autosomes. For example, the hemizgosity of the X may lead to a lower chromosomal effective population size compared to the autosomes suggesting that the X might be more strongly affected by genetic drift. However, the X may also experience stronger positive selection than the autosomes because recessive beneficial mutations will be more visible to selection on the X where they will spend less time being masked by the dominant, less beneficial allele - a proposal known as the faster-X hypothesis. Thus, empirical studies demonstrating increased genetic divergence on the X chromosome could be indicative of either adaptive or non-adaptive evolution. We measured gene expression in Drosophila species and in D. melanogaster inbred strains for both embryos and adults. In the embryos we found that expression divergence is on average more than 20% higher for genes on the X chromosome relative to the autosomes, but in contrast, in the inbred strains gene expression variation is significantly lower on the X chromosome. Furthermore, expression divergence of genes on Muller's D element is significantly greater along the branch leading to the obscura sub-group, in which this element segregates as a neo-X chromosome. In the adults, divergence is greatest on the X chromosome for males, but not for females, yet in both sexes inbred strains harbour the lowest level of gene expression variation on the X chromosome. We consider different explanations for our results and conclude that they are most consistent within the framework of the faster-X hypothesis., Comment: 75 pages, 8 Figures, 26 Supplementary Figures, 13 Supplementary Tables