Limits...
Extracting meaning from biological imaging data.

Cohen AR - Mol. Biol. Cell (2014)

Bottom Line: The summarization is validated, the results visualized, and errors corrected as needed.Finally, the customized analysis and visualization tools together with the image data and the summarization results are shared.This Perspective provides a brief guide to the mathematical ideas that rigorously quantify the notion of extracting meaning from biological image, and to the practical approaches that have been used to apply these ideas to a wide range of applications in cell and tissue optical imaging.

View Article: PubMed Central - PubMed

Affiliation: Department of Electrical and Computer Engineering, Drexel University, Philadelphia, PA 19104 acohen@coe.drexel.edu.

Show MeSH

Related in: MedlinePlus

Interactive and collaborative use of the image data, together with the summarization results and the visualization and analysis code. Left to right, single image from a 2000-frame sequence, same image with segmentation and tracking results overlaid, lineage tree with time as the vertical axis (top) and statistical model (bottom) and open source code fragment. The arrow on the bottom shows the progression from image data to source code, with summarization results falling in the gray area.
© Copyright Policy - creative-commons
Related In: Results  -  Collection


getmorefigures.php?uid=PMC4230605&req=5

Figure 1: Interactive and collaborative use of the image data, together with the summarization results and the visualization and analysis code. Left to right, single image from a 2000-frame sequence, same image with segmentation and tracking results overlaid, lineage tree with time as the vertical axis (top) and statistical model (bottom) and open source code fragment. The arrow on the bottom shows the progression from image data to source code, with summarization results falling in the gray area.

Mentions: Another challenge is the size of the data set. Current-generation time-lapse microscopes include integrated incubation and can typically acquire 100 movies or time-lapse image sequences in a single experiment. Each movie can consist of thousands of images. In our ongoing work analyzing stem cell image sequence data, a single data set of 200 movies requires 350 gigabytes (GB) of image data or more. This is obviously too much data to analyze by hand or by eye—we must turn to computational analysis. There are many software packages for working with smaller and less complex image data sets (Eliceiri et al., 2012), but here I focus on software solutions custom written for the specific characteristics of the image data in order to best summarize the data in the context of a particular biological question. One of the key challenges in biological image analysis is the lack of computational tools for interactively and collaboratively summarizing, visualizing, and validating image data. Figure 1 shows an overview of the summarization, validation, and sharing steps.


Extracting meaning from biological imaging data.

Cohen AR - Mol. Biol. Cell (2014)

Interactive and collaborative use of the image data, together with the summarization results and the visualization and analysis code. Left to right, single image from a 2000-frame sequence, same image with segmentation and tracking results overlaid, lineage tree with time as the vertical axis (top) and statistical model (bottom) and open source code fragment. The arrow on the bottom shows the progression from image data to source code, with summarization results falling in the gray area.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC4230605&req=5

Figure 1: Interactive and collaborative use of the image data, together with the summarization results and the visualization and analysis code. Left to right, single image from a 2000-frame sequence, same image with segmentation and tracking results overlaid, lineage tree with time as the vertical axis (top) and statistical model (bottom) and open source code fragment. The arrow on the bottom shows the progression from image data to source code, with summarization results falling in the gray area.
Mentions: Another challenge is the size of the data set. Current-generation time-lapse microscopes include integrated incubation and can typically acquire 100 movies or time-lapse image sequences in a single experiment. Each movie can consist of thousands of images. In our ongoing work analyzing stem cell image sequence data, a single data set of 200 movies requires 350 gigabytes (GB) of image data or more. This is obviously too much data to analyze by hand or by eye—we must turn to computational analysis. There are many software packages for working with smaller and less complex image data sets (Eliceiri et al., 2012), but here I focus on software solutions custom written for the specific characteristics of the image data in order to best summarize the data in the context of a particular biological question. One of the key challenges in biological image analysis is the lack of computational tools for interactively and collaboratively summarizing, visualizing, and validating image data. Figure 1 shows an overview of the summarization, validation, and sharing steps.

Bottom Line: The summarization is validated, the results visualized, and errors corrected as needed.Finally, the customized analysis and visualization tools together with the image data and the summarization results are shared.This Perspective provides a brief guide to the mathematical ideas that rigorously quantify the notion of extracting meaning from biological image, and to the practical approaches that have been used to apply these ideas to a wide range of applications in cell and tissue optical imaging.

View Article: PubMed Central - PubMed

Affiliation: Department of Electrical and Computer Engineering, Drexel University, Philadelphia, PA 19104 acohen@coe.drexel.edu.

Show MeSH
Related in: MedlinePlus