Limits...
Multiscale feature analysis of salivary gland branching morphogenesis.

Bilgin CC, Ray S, Baydil B, Daley WP, Larsen M, Yener B - PLoS ONE (2012)

Bottom Line: Multiscale cell-graph analysis was most effective in classification of the tissue state.Whereas tensor analysis demonstrate that epithelial tissue was affected the most by inhibition of ROCK signaling, significant multiscale changes in mesenchymal tissue organization were identified with this analysis that were not identified in previous biological studies.We here show how to define and calculate a multiscale feature set as an effective computational approach to identify and quantify changes at multiple biological scales and to distinguish between different states in developing tissues.

View Article: PubMed Central - PubMed

Affiliation: Rensselaer Polytechnic Institute, Computer Science Department, Troy, New York, United States of America.

ABSTRACT
Pattern formation in developing tissues involves dynamic spatio-temporal changes in cellular organization and subsequent evolution of functional adult structures. Branching morphogenesis is a developmental mechanism by which patterns are generated in many developing organs, which is controlled by underlying molecular pathways. Understanding the relationship between molecular signaling, cellular behavior and resulting morphological change requires quantification and categorization of the cellular behavior. In this study, tissue-level and cellular changes in developing salivary gland in response to disruption of ROCK-mediated signaling by are modeled by building cell-graphs to compute mathematical features capturing structural properties at multiple scales. These features were used to generate multiscale cell-graph signatures of untreated and ROCK signaling disrupted salivary gland organ explants. From confocal images of mouse submandibular salivary gland organ explants in which epithelial and mesenchymal nuclei were marked, a multiscale feature set capturing global structural properties, local structural properties, spectral, and morphological properties of the tissues was derived. Six feature selection algorithms and multiway modeling of the data was performed to identify distinct subsets of cell graph features that can uniquely classify and differentiate between different cell populations. Multiscale cell-graph analysis was most effective in classification of the tissue state. Cellular and tissue organization, as defined by a multiscale subset of cell-graph features, are both quantitatively distinct in epithelial and mesenchymal cell types both in the presence and absence of ROCK inhibitors. Whereas tensor analysis demonstrate that epithelial tissue was affected the most by inhibition of ROCK signaling, significant multiscale changes in mesenchymal tissue organization were identified with this analysis that were not identified in previous biological studies. We here show how to define and calculate a multiscale feature set as an effective computational approach to identify and quantify changes at multiple biological scales and to distinguish between different states in developing tissues.

Show MeSH

Related in: MedlinePlus

Direct validations of cell-graph features using standard image analysis methods.Plots of (a) area, (b) perimeter and (c) circularity from images using conventional image analysis methods and plots of cell-graph-derived raw data pertaining to (d) area, (e) perimeter and (f) circularity are shown. Control refers to untreated epithelium and Y27632 refers to the ROCK inhibitor treatment. The same trends for control vs ROCK inhibitor treatment were observed for the features obtained using image analysis and cell-graphs. The percent differences between the conventional image analysis and our image segmentation technique are found to be 1.16% and 0.73% for the area; 5.66% and 5.94% for the perimeter; 11.0532 and 16.1463 for the circularity of the control and ROCK inhibitor-treated samples, respectively.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3293912&req=5

pone-0032906-g003: Direct validations of cell-graph features using standard image analysis methods.Plots of (a) area, (b) perimeter and (c) circularity from images using conventional image analysis methods and plots of cell-graph-derived raw data pertaining to (d) area, (e) perimeter and (f) circularity are shown. Control refers to untreated epithelium and Y27632 refers to the ROCK inhibitor treatment. The same trends for control vs ROCK inhibitor treatment were observed for the features obtained using image analysis and cell-graphs. The percent differences between the conventional image analysis and our image segmentation technique are found to be 1.16% and 0.73% for the area; 5.66% and 5.94% for the perimeter; 11.0532 and 16.1463 for the circularity of the control and ROCK inhibitor-treated samples, respectively.

Mentions: With any computational method, it is necessary to validate computational results, whenever possible, with results obtained more directly from the sample. Therefore, after extracting the full set of cell graph features, we compared the values of a subset of these cell-graph features to the corresponding values obtained using conventional image analysis methods directly on the confocal images and validated the cell-graph measurements. We calculated the average area, perimeter, and circularity and the standard error of the organ explants using standard image processing methods directly from the confocal images for each treatment, as shown in Figure 3A–C and directly compared these results with the values for the same features derived from the morphological analysis. The same trends were observed for this subset of features in control vs ROCK inhibitor-treatment for the conventional and computational analysis, calculated only for the epithelial tissue.


Multiscale feature analysis of salivary gland branching morphogenesis.

Bilgin CC, Ray S, Baydil B, Daley WP, Larsen M, Yener B - PLoS ONE (2012)

Direct validations of cell-graph features using standard image analysis methods.Plots of (a) area, (b) perimeter and (c) circularity from images using conventional image analysis methods and plots of cell-graph-derived raw data pertaining to (d) area, (e) perimeter and (f) circularity are shown. Control refers to untreated epithelium and Y27632 refers to the ROCK inhibitor treatment. The same trends for control vs ROCK inhibitor treatment were observed for the features obtained using image analysis and cell-graphs. The percent differences between the conventional image analysis and our image segmentation technique are found to be 1.16% and 0.73% for the area; 5.66% and 5.94% for the perimeter; 11.0532 and 16.1463 for the circularity of the control and ROCK inhibitor-treated samples, respectively.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3293912&req=5

pone-0032906-g003: Direct validations of cell-graph features using standard image analysis methods.Plots of (a) area, (b) perimeter and (c) circularity from images using conventional image analysis methods and plots of cell-graph-derived raw data pertaining to (d) area, (e) perimeter and (f) circularity are shown. Control refers to untreated epithelium and Y27632 refers to the ROCK inhibitor treatment. The same trends for control vs ROCK inhibitor treatment were observed for the features obtained using image analysis and cell-graphs. The percent differences between the conventional image analysis and our image segmentation technique are found to be 1.16% and 0.73% for the area; 5.66% and 5.94% for the perimeter; 11.0532 and 16.1463 for the circularity of the control and ROCK inhibitor-treated samples, respectively.
Mentions: With any computational method, it is necessary to validate computational results, whenever possible, with results obtained more directly from the sample. Therefore, after extracting the full set of cell graph features, we compared the values of a subset of these cell-graph features to the corresponding values obtained using conventional image analysis methods directly on the confocal images and validated the cell-graph measurements. We calculated the average area, perimeter, and circularity and the standard error of the organ explants using standard image processing methods directly from the confocal images for each treatment, as shown in Figure 3A–C and directly compared these results with the values for the same features derived from the morphological analysis. The same trends were observed for this subset of features in control vs ROCK inhibitor-treatment for the conventional and computational analysis, calculated only for the epithelial tissue.

Bottom Line: Multiscale cell-graph analysis was most effective in classification of the tissue state.Whereas tensor analysis demonstrate that epithelial tissue was affected the most by inhibition of ROCK signaling, significant multiscale changes in mesenchymal tissue organization were identified with this analysis that were not identified in previous biological studies.We here show how to define and calculate a multiscale feature set as an effective computational approach to identify and quantify changes at multiple biological scales and to distinguish between different states in developing tissues.

View Article: PubMed Central - PubMed

Affiliation: Rensselaer Polytechnic Institute, Computer Science Department, Troy, New York, United States of America.

ABSTRACT
Pattern formation in developing tissues involves dynamic spatio-temporal changes in cellular organization and subsequent evolution of functional adult structures. Branching morphogenesis is a developmental mechanism by which patterns are generated in many developing organs, which is controlled by underlying molecular pathways. Understanding the relationship between molecular signaling, cellular behavior and resulting morphological change requires quantification and categorization of the cellular behavior. In this study, tissue-level and cellular changes in developing salivary gland in response to disruption of ROCK-mediated signaling by are modeled by building cell-graphs to compute mathematical features capturing structural properties at multiple scales. These features were used to generate multiscale cell-graph signatures of untreated and ROCK signaling disrupted salivary gland organ explants. From confocal images of mouse submandibular salivary gland organ explants in which epithelial and mesenchymal nuclei were marked, a multiscale feature set capturing global structural properties, local structural properties, spectral, and morphological properties of the tissues was derived. Six feature selection algorithms and multiway modeling of the data was performed to identify distinct subsets of cell graph features that can uniquely classify and differentiate between different cell populations. Multiscale cell-graph analysis was most effective in classification of the tissue state. Cellular and tissue organization, as defined by a multiscale subset of cell-graph features, are both quantitatively distinct in epithelial and mesenchymal cell types both in the presence and absence of ROCK inhibitors. Whereas tensor analysis demonstrate that epithelial tissue was affected the most by inhibition of ROCK signaling, significant multiscale changes in mesenchymal tissue organization were identified with this analysis that were not identified in previous biological studies. We here show how to define and calculate a multiscale feature set as an effective computational approach to identify and quantify changes at multiple biological scales and to distinguish between different states in developing tissues.

Show MeSH
Related in: MedlinePlus