WP4 - Point processes in bioimaging

First funding period

During the first funding period of CSGB, a number of important applications of point processes in bioimaging have been studied, including

  • the minicolumn hypothesis in the cerebral cortex of the brain
  • protein interactions inferred from FRET data
  • the density and level of repulsion of vesicles in the presynaptic area

In the second funding period, these projects will be continued.

WP4 - Subprojects

WP4.1: Nodule identification and classification from confocal microscopy
WP4.2: Characterization of moving protein clusters using TIRF imaging
WP4.3: Synaptic vesicle sampling, distribution and shape from electron microscopy
WP4.4: Improved variance prediction for stereological estimators of total number

Second funding period

New challenges relating to the previously mentioned projects are

  • minicolumn hypothesis: the aim is to extend our models and methods to more complicated data such as two-point pattern data sets of neurons.
  • protein interactions from FRET data: more sophisticated Bayesian priors will be tested, implying substantial computational challenges.
  • vesicles in the presynaptic area: the possibilities for sampling vesicles by Focused Ion Beam Scanning Electron Microscopy (FIB-SEM) will be investigated. We want to use the superior modelling capability of spatial point process models to investigate the variation in shape, size and, in case of ellipsoidal vesicles, orientation.

Two new projects in confocal microscopy and total internal reflection fluorescence (TIRF) microscopy, respectively, concern

  • nodule identification by confocal microscopy: the interest is in early stage formation of root nodules (nodule primordia) where nodules are difficult to distinguish from structures that later will develop into lateral roots that do not house rhizobia.
  • moving protein clusters by TIRF microscopy: the aim is to study the localization and dynamic movement and interaction of plasma membrane transport proteins, using spatio-temporal point processes.

Finally, a long-standing question of improved variance prediction for particle counts is taken up in the second funding period. The question will be considered in a stereological setting where the particle counting is the last sampling step in a hierarchical sampling design.

Selected references

  • Andersen, I.T., Hahn, U., Arnspang, E.C., Nejsum, L.N. & Jensen, E.B.V. (2018): Double Cox cluster processes – with applications to photoactivated localization microscopy. Spat. Stat. 27, 58–73.
  • Hooghoudt, J., Barroso, M. & Waagepetersen, R. (2017): Towards Bayesian inference of the spatial distribution of proteins from three-cube Förster resonance energy transfer data. Ann. Appl. Stat.11, 1711-1737.
  • Khanmohammadi, M., Darkner, S., Nava, N., Nyengaard, J.R., Wegener, G., Popoli, M. & Sporring, J. (2017): 3D analysis of synaptic vesicle density and distribution after acute foot-shock stress by using serial section transmission electron microscopy. J. Microsc. 265, 101-110.
  • Kiderlen, M. & Dorph-Petersen, K.-A. (2017): The Cavalieri estimator with unequal section spacing revisited. Image Anal. Stereol. 36, 135-141.
  • Rasmussen, J.G. & Christensen, H.S. (2018): Point processes on directed linear networks. CSGB Research Report 2018-11.