Local adaptive wiener filtering for class averaging in single particle reconstruction

Ali Abdollahzadeh, Erman Acar, Sari Peltonen, Ulla Ruotsalainen

    Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

    Abstract

    In cryo-electron microscopy (cryo-EM), the Wiener filter is the optimal operation – in the least-squares sense – of merging a set of aligned low signal-to-noise ratio (SNR) micrographs to obtain a class average image with higher SNR. However, the condition for the optimal behavior of the Wiener filter is that the signal of interest shows stationary characteristic thoroughly, which cannot always be satisfied. In this paper, we propose substituting the conventional Wiener filter, which encompasses the whole image for denoising, with its local adaptive implementation, which denoises the signal locally. We compare our proposed local adaptive Wiener filter (LA-Wiener filter) with the conventional class averaging method using a simulated dataset and an experimental cryo-EM dataset. The visual and numerical analyses of the results indicate that LA-Wiener filter is superior to the conventional approach in single particle reconstruction (SPR) applications.

    Original languageEnglish
    Title of host publicationImage Analysis - 20th Scandinavian Conference, SCIA 2017, Proceedings
    PublisherSpringer Verlag
    Pages233-244
    Number of pages12
    ISBN (Print)9783319591285
    DOIs
    Publication statusPublished - 2017
    Publication typeA4 Article in conference proceedings
    EventScandinavian Conference on Image Analysis -
    Duration: 1 Jan 1900 → …

    Publication series

    NameLecture Notes in Computer Science
    Volume10270
    ISSN (Print)0302-9743
    ISSN (Electronic)1611-3349

    Conference

    ConferenceScandinavian Conference on Image Analysis
    Period1/01/00 → …

    Keywords

    • Class averaging
    • Electron microscopy
    • Local adaptive Wiener filter
    • Single particle reconstruction
    • Spectral signal-to-noise ratio

    Publication forum classification

    • Publication forum level 1

    ASJC Scopus subject areas

    • Theoretical Computer Science
    • Computer Science(all)

    Fingerprint

    Dive into the research topics of 'Local adaptive wiener filtering for class averaging in single particle reconstruction'. Together they form a unique fingerprint.

    Cite this