Preferencje help
Widoczny [Schowaj] Abstrakt
Liczba wyników

Znaleziono wyników: 7

Liczba wyników na stronie
Pierwsza strona wyników Pięć stron wyników wstecz Poprzednia strona wyników Strona / 1 Następna strona wyników Pięć stron wyników wprzód Ostatnia strona wyników

Wyniki wyszukiwania

help Sortuj według:

help Ogranicz wyniki do:
Pierwsza strona wyników Pięć stron wyników wstecz Poprzednia strona wyników Strona / 1 Następna strona wyników Pięć stron wyników wprzód Ostatnia strona wyników
Spatial integration of multimodal imaging data is a common denominator of all whole brain mapping projects. This process requires robust image registration pipelines, high‑ quality 3D brain atlases, as well as, scalable methods for quantitative image analysis. During the talk, I will discuss the computational challenges behind these components, exemplify ways of addressing them, and discuss requirements for setting up one’s own computational pipeline. I will also demonstrate how these novel computational methods and approaches could provide a deeper understanding of the structure and function of the central nervous system, especially in the context of high-throughput and large‑scale experiments. The talk will be complemented with examples of specific neuroscientific findings arising from whole brain mapping projects in rodents and primates, highlighting synergy between the computational and experimental aspects in these projects.
AIM: Images of brain tissue are common outcome of a great number of neuroscience experiments; however, their impact on scientific development is limited since they are usually not published directly. Internet technology facilitates making the images available online, nonetheless, raw image files of good quality are very large and therefore inconvenient to provide and analyze online directly. There is still no convenient way to share annotated images of neuroscientific specimens. To fill that gap we developed BrainSlices software – a user-friendly tool dedicated for that purpose. METHODS: The software is built in the client-server model and consist of a web application at the client side, and server software running on a Linux system. Thanks to this design there is no need for installation and the only thing a BrainSlices user must do is to open a website. This approach also makes BrainSlices a cross-platform tool. Handling of large images is solved with image pyramid technique. RESULTS: We used BrainSlices software to set up an online repository (http://brainslices.org) of annotated high quality images of brain tissue. The repository combines the power of the image pyramid technique with convenience of image upload. Every image can be easily annotated with exhaustive metadata which facilitate search. The user interface of the repository and the provided facilities were developed with typical neuroscience use in mind. Every image receives a unique identifier and a permalink which allows direct access and citing. CONCLUSION: We provide the community with a user friendly tool for multiple image storage, viewing, sharing, and annotation. The tool may be used to store one’s own collection of slice images to share high quality specimen images with collaborators, to share them with the whole community, or to provide the images online as supplementary material for publications.
Local field potentials (LFP), low-frequency part of extracellular electric potentials, seem to reflect dendritic processing of incoming activity to neural populations. Long-range nature of electric field leads to correlations even between remote recordings showing sources from millimeters away which complicates analysis of LFP. To get more insight it is convenient whenever possible to look for current sources of the potentials or to decompose the signals into meaningful components using statistical techniques. In Łęski and coauthors (2010) we have combined inverse current source density method with independent component analysis (ICA) to decompose 140 recordings in rat forebrain obtaining physiologically meaningful components across a group of seven animals. To find out what can be really observed with such an approach experimentally we simulated local field potentials generated in a single cortical column in a model of 3560 cells with non-trivial morphologies. Having both the current source density (CSD) and LFP generated by twelve cortical populations included we compared it with independent components obtained in the decomposition of data generated by the whole network. We assumed a set of potential measurements on a regular grid, low-pass filtered it temporally under 500 Hz, reconstructed the sources using kernel current source density and performed the ICA. We found that the recorded evoked activity was dominated by two populations of pyramidal neurons, which were well separated by ICA. Other populations could not be clearly distinguished in the simulated potentials nor in the ICA. Supported by grants POIG.02.03.00-00-003/09, POIG.02.03.00-00-018/08.
INTRODUCTION: Processes such as perception, action and cognition are determined by the connectivity between different neuronal groups. Understanding the principles of this network is a core objective of present-day neuroscience. Several animal models are used to investigate this relationship between structure and function, among them marmosets, which recently came to prominence. They are small monkeys (300–400 g) but their brain retains all defining features of the primate brain. AIM(S): The aim is to create a publicly available, the world’s most comprehensive repository of the afferent cortico-cortical connectivity of any primate species, enabling a new level of analysis and modelling. The connectome will be publicly available on‑line making it possible to flexibly access all the data via a graphical front-end or via an application programming interface. METHOD(S): The already available body of data comprises results of over 100 monosynaptic retrograde tracer injections in marmosets. The brains were cut in 40 µm sections. The sections were plotted using an epifluorescence microscope, and stained for Nissl substance. To map individual injections into the atlas space, a previously established pipeline was used. RESULTS: The current version of the portal is available at http://marmoset.braincircuits.org. It allows one to access unprocessed experimental data, mostly injections in dorsal prefrontal cortex, parietal and occipital lobes. Additionally, the locations of individual cells are expressed in atlas-based stereotaxic coordinates which allows one to perform either area-based or parcellation-free connectivity analyses. CONCLUSIONS: The release of open access connectomes is known for triggering numerous follow-up modelling and theoretical studies. In a longer perspective, the unique nature of data in our project will help to understand how the highly complex network of neuronal connections enable brain functions in primates, and, in general, in mammals. FINANCIAL SUPPORT: The project is supported by the Australian Research Council grant (DP140101968) and International Neuroinformatics Coordinating Facility Seed Funding grant.
The Monodelphis opossum became an important laboratory animal and is often used in biomedical research. However, data on the brain anatomy are scarce and there is no reliable brain anatomy reference. The aim of this study is to present neuroanatomical delineation of basic brain structures. Data which served for construction of the 3-dimensional atlas were magnetic resonance images (MRI) and stained brain sections. MRI was obtained 48 h after perfusion of the animal with 4% paraformaldehyde and gadotheridol contrast (ProHance 20:1 v:v). The second MRI was performed 30 days after perfusion of the same animal. Both MRIs were aquired using Bruker Biospin system with voxel reolution of 50 µm3. For Nissland myelin staining, coronal brain sections were cut in cryostat at 40 µm thickness. To minimize tissue deformation, sections were transferred from the cutting blade to slides using the Tape-Transfer System. Then brain sections stained either with Nissl or for myelin were imaged with a high resolution scanner and were transformed to three-dimensional form. By superimposing all three-dimensional data, several brain structures were delineated, e.g., the olfactory bulb, cerebral cortex, hippocampus, white matter and other. Supported by grant from the Polish Ministry of Regional Development POIG.02.03.00-00-003/09.
INTRODUCTION: Imaging of entire brains at cellular resolution, enabled by light-sheet fluorescence microscopy (LSFM) and optical tissue clearing, offers insights into neural activity at a high magnification while preserving the brain‑wide context. AIM(S): We propose a set of open-source computational tools that address three fundamental challenges associated with the analysis of LSFM images of entire rodent brains, namely: management of voluminous imaging data, alignment to a reference atlas, and object detection and localization. METHOD(S): The data for each brain, such as multichannel acquisitions and spatial information, are compressed and stored in an HDF5-based container as a pyramid of resolutions to facilitate and standardize data access and manipulation. Unlike most other alignment approaches, our pipeline is not only guided by standard similarity metrics such as mutual information, but also utilizes Deep Convolutional Neural Networks to generate label maps corresponding to specific brain structures such as main white matter tracts or dentate gyrus. This step significantly increases the accuracy and robustness of the registration procedure. The c‑Fos‑positive nuclei are identified and quantified with the help of another neural network trained on synthetic data, generated to simulate the original nuclei which eliminated the laborious process of manual image annotation. The software was applied to investigate c-Fos-mediated neuroplasticity in iDISCO-cleared brains in experimental paradigms of appetitive and aversive learning and alcohol addiction. RESULTS: Voxel-wise statistical analysis revealed brain areas involved in the neuroplasticity of alcohol addiction and appetitive or aversive learning in mice. CONCLUSIONS: We demonstrate the ability of our software to combine efficient data management, accurate atlas alignment, and object detection to facilitate LSFM analyses. FINANCIAL SUPPORT: ERA‑NET NEURON/17/2017 grant from NCBR, G2631 grant from NCN.
Pierwsza strona wyników Pięć stron wyników wstecz Poprzednia strona wyników Strona / 1 Następna strona wyników Pięć stron wyników wprzód Ostatnia strona wyników
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.