TY - GEN
T1 - A quantitative analysis of F-actin features and distribution in fluorescence microscopy images to distinguish cells with different modes of motility
AU - Cheng, Jie
AU - Zhu, Xiaoping
AU - Cheng, Hao
AU - Zhao, Hong
AU - Wong, Stephen T C
N1 - Copyright:
Copyright 2013 Elsevier B.V., All rights reserved.
PY - 2013
Y1 - 2013
N2 - Actin is one of the most abundant proteins in eukaryote cells, playing a key role in cell dynamic morphological alterations and tumor metastatic spread. To investigate the relationship between the distribution patterns of actin and the aggressiveness of cancer cells, we developed an image analysis framework for quantifying cell F-actin distributions examined with fluorescence microscopy. The images are first segmented with multichannel information of both F-actin and nuclear staining. Using the watershed method and Voronoi tessellation, the cells can be well segmented based on both F-actin and nuclear information. Altogether, sixteen F-actin distribution features are calculated for each individual cell. A linear Support Vector Machine (SVM) is then applied in the feature space to separate cells with different modes of motility. Our results show that cells with different modes of motility can be distinguished by F-actin distributions. To our knowledge, this is the first study managing to distinguish cancer cells with different aggressiveness based on quantitative analysis of F-actin distribution patterns.
AB - Actin is one of the most abundant proteins in eukaryote cells, playing a key role in cell dynamic morphological alterations and tumor metastatic spread. To investigate the relationship between the distribution patterns of actin and the aggressiveness of cancer cells, we developed an image analysis framework for quantifying cell F-actin distributions examined with fluorescence microscopy. The images are first segmented with multichannel information of both F-actin and nuclear staining. Using the watershed method and Voronoi tessellation, the cells can be well segmented based on both F-actin and nuclear information. Altogether, sixteen F-actin distribution features are calculated for each individual cell. A linear Support Vector Machine (SVM) is then applied in the feature space to separate cells with different modes of motility. Our results show that cells with different modes of motility can be distinguished by F-actin distributions. To our knowledge, this is the first study managing to distinguish cancer cells with different aggressiveness based on quantitative analysis of F-actin distribution patterns.
KW - Image segmentation
KW - Cancer
KW - Support vector machines
KW - Feature extraction
KW - Microscopy
KW - Biomedical imaging
KW - Nuclear measurements
UR - http://www.scopus.com/inward/record.url?scp=84886506770&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84886506770&partnerID=8YFLogxK
U2 - 10.1109/EMBC.2013.6609456
DO - 10.1109/EMBC.2013.6609456
M3 - Conference contribution
C2 - 24109643
AN - SCOPUS:84886506770
SN - 9781457702167
VL - 2013
T3 - Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS
SP - 136
EP - 139
BT - 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2013
T2 - 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2013
Y2 - 3 July 2013 through 7 July 2013
ER -