Well-ordered cell motility is necessary for normal development, tissue healing and immune defense. On the other hand, uncontrolled cell invasion is involved in life-threatening diseases, such as in tumor growth and metastasis. Understanding of cellular and molecular mechanism regulating cell motility is necessary to speed up development of better treatments.
The most common techniques to investigate cell motility are based on two dimensional (2D) (Video 1) cultures in which cells are growing as a monolayer. The advance is feasibility of experimental settings and microscopic imaging of cellular events is relatively easy. Dynamic analysis can reveal how cells response to drug treatments and genetic manipulations relevant to migration parameters. Unfortunately, the experiments on plastic culture dishes are artificial in many ways and assays on planar surfaces poorly represent conditions in tissues.
Organotypic cultures (Video 2) mimic more closely tissue environment than plastic dishes. The challenges are establishment of sophisticated culturing conditions and high resolution imaging of thick specimens. In addition, imaging of cellular events occurring in three dimensional matrixes over the time results in complex and large image files that pose a great challenge to existing image analysis software. Therefore, there is a great need for computationally effective image analysis techniques to help recognize cellular phenotypes and to reduce the burden of time consuming manual work.
In our project we are developing new customizable image analysis tools based on state-of-art machine vision techniques to analyze various forms, structures and dynamics of cellular events in experimental animal models, in human tissues resected from patients, and in experimental matrixes that mimic human tissues.
We are working on computer vision - based automatic approaches to detect often overlapping single cells among other cells, to track individual cells in time-lapse movies, and to automatically quantify the cellular characteristics of interest such as migration velocity and migration modes (Figure 1). We are also developing methodologies to combine multidimensional image information to find out the relationship between cellular behavior in 3D matrixes with other characteristics such as tissue elasticity, the interactions of different cell types, and effects of molecular composition and macromolecular architecture of tissue environments.
The project is implemented as an interdisciplinary and collaborative effort of two spearhead units of University of Oulu; the Center for Machine Vision Research (CMV) at Infotech Oulu and Biocenter Oulu Tissue Imaging Center (BCO-TIC), as well as our collaborators.
- Academy Research Fellow Lauri Eklund
- Ph.D. Mika Kaakinen
- Professor Janne Heikkilä
- DSc (Tech) Sami Huttunen
- DSc (Tech) Juho Kannala
- Dr Neslihan Bayramoglu
- Saad Ullah Akram (CMV)
- Professor Tuula Salo
- PhD Matthias Nees
- PhD Varpu Marjomäki
Last updated: 27/6/2016