by Moeller, J., Gottfried, Björn, Schlieder, Christoph and Herzog, Otthein and Friedl, P.
Abstract:
The migration of different cell types such as leukocytes and tumor cells within tissues is a fundamental process in physiologic and pathologic tissue reactions. However, the investigation and reconstruction of cell migration within 3D extracellular matrices in vitro and in vivo is limited by the lack of automated quantification methods that allow the reconstruction of cell paths from cell populations. We have developed a fully automated cell tracking device for the reconstruction of migration paths from 3D collagen matrices in front or inhomogeneous backgrounds independent of cell size and shape, sharpness of the boundary, shape change, and their migration velocity as well as cell-cell contacts. Best results for separation of cells from background were obtained by region growing from the center of preselected cells until a colour gradient was reached. Noise reduction filtering and an object matching routine comparing the pixel mass from frame to frame was used to optimize segmentation stability upon object shape change and background variation, thereby increasing path precision. Overlapping paths and cell-cell contacts were resolved by a subtraction algorithm using mass changes for path separation. Compared to manual tracking as a "gold" standard automatic tracking resulted in high precision and reproducibility of path position, step lengths and frequencies (p=0.78 to 0.96; Mann-Whitney U-Test) for migrating ameboid T cells as well as tumor cells. This automated device Separating cells with identical outlines may be of value for quantitative detection of migrating cells in 3D extracellular matrix environments in basic research and drug development.
Reference:
Moeller, J., Gottfried, Björn, Schlieder, Christoph and Herzog, Otthein and Friedl, P., "Automated Tracking of Cell Movements and Resolution of Cell-Cell Collisions in Three-dimensional Collagen Matrices", 2003.
Bibtex Entry:
@MISC{Moeller2003,
author = {Moeller, J. and Gottfried, Bj{\"o}rn and Schlieder, Christoph and
Herzog, Otthein and Friedl, P.},
title = {{Automated Tracking of Cell Movements and Resolution of Cell-Cell
Collisions in Three-dimensional Collagen Matrices}},
howpublished = {Poster presented at the Keystone-Symposium on Cell-Analysis, Breckenridge,
Colorado},
year = {2003},
abstract = {The migration of different cell types such as leukocytes and tumor
cells within tissues is a fundamental process in physiologic and
pathologic tissue reactions. However, the investigation and reconstruction
of cell migration within 3D extracellular matrices in vitro and in
vivo is limited by the lack of automated quantification methods that
allow the reconstruction of cell paths from cell populations. We
have developed a fully automated cell tracking device for the reconstruction
of migration paths from 3D collagen matrices in front or inhomogeneous
backgrounds independent of cell size and shape, sharpness of the
boundary, shape change, and their migration velocity as well as cell-cell
contacts. Best results for separation of cells from background were
obtained by region growing from the center of preselected cells until
a colour gradient was reached. Noise reduction filtering and an object
matching routine comparing the pixel mass from frame to frame was
used to optimize segmentation stability upon object shape change
and background variation, thereby increasing path precision. Overlapping
paths and cell-cell contacts were resolved by a subtraction algorithm
using mass changes for path separation. Compared to manual tracking
as a {"}gold{"} standard automatic tracking resulted in high precision
and reproducibility of path position, step lengths and frequencies
(p=0.78 to 0.96; Mann-Whitney U-Test) for migrating ameboid T cells
as well as tumor cells. This automated device Separating cells with
identical outlines may be of value for quantitative detection of
migrating cells in 3D extracellular matrix environments in basic
research and drug development.},
owner = {pmania},
timestamp = {2012.11.06}
}