Hessian-based quantitative image analysis of host-pathogen confrontation assays.

Abstract:

Host-fungus interactions have gained a lot of interest in the past few decades, mainly due to an increasing number of fungal infections that are often associated with a high mortality rate in the absence of effective therapies. These interactions can be studied at the genetic level or at the functional level via imaging. Here, we introduce a new image processing method that quantifies the interaction between host cells and fungal invaders, for example, alveolar macrophages and the conidia of Aspergillus fumigatus. The new technique relies on the information content of transmitted light bright field microscopy images, utilizing the Hessian matrix eigenvalues to distinguish between unstained macrophages and the background, as well as between macrophages and fungal conidia. The performance of the new algorithm was measured by comparing the results of our method with that of an alternative approach that was based on fluorescence images from the same dataset. The comparison shows that the new algorithm performs very similarly to the fluorescence-based version. Consequently, the new algorithm is able to segment and characterize unlabeled cells, thus reducing the time and expense that would be spent on the fluorescent labeling in preparation for phagocytosis assays. By extending the proposed method to the label-free segmentation of fungal conidia, we will be able to reduce the need for fluorescence-based imaging even further. Our approach should thus help to minimize the possible side effects of fluorescence labeling on biological functions. (c) 2017 International Society for Advancement of Cytometry.

SEEK ID: https://funginet.hki-jena.de/publications/152

PubMed ID: 28914994

Projects: B4

Journal: Cytometry A

Citation: Cytometry A. 2018 Mar;93(3):346-356. doi: 10.1002/cyto.a.23201. Epub 2017 Sep 15.

Date Published: 16th Sep 2017

Authors: Z. Cseresnyes, Kaswara Kraibooj, Marc Thilo Figge

Help
help Creator
Activity

Views: 322

Created: 16th Feb 2021 at 15:33

help Attributions

None

Related items

Powered by
(v.1.9.1)
Copyright © 2008 - 2019 The University of Manchester and HITS gGmbH