About PSRT

PSRT stands for Progressive Stochastic Reconstruction Technique and was introduced in Turonova et al., J. Struct. Biol. 189, 195-206 (2015). It is a novel approach to 3D reconstruction out of 2D projections in Cryo Electron Tomography (CryoET) which proposes to reconstruct 3D scattering potential of a specimen using Monte Carlo random walks and Metropolis-Hastings sampling strategy.
Although designed to match specific properties of CryoET, the method can be used for any tomographic reconstruction.

PSRT Tomography

PSRT Tomography is an implementation of PSRT that has been developed at Computer Graphics Lab at the Saarland University in Germany.
It is written in C++ and distributed under an LGPL license.

Key Features

  • ⚬ Coarse-to-fine reconstruction
    • ⚬ Enhanced contrast
    • ⚬ Significant noise reduction
    • ⚬ Improvement of automatic template-based localization
  • ⚬ Importance Sampling
    • ⚬ Region-of-interest reconstruction within tomogram
  • ⚬ Memory Efficient Solution for Interior Problem (MESIP)


PSRT Tomography update

September 03, 2015

The region-of-interests can be now specified also by using binary mask. See User Guide for more details.

PSRT Tomography website was launched

August 26, 2015

The website psrt-tomography.org was launched today with precompiled binaries for Windows (64 bit) to download.