Loading presentation...

Present Remotely

Send the link below via email or IM


Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.


Cell-tacular: Matlab Group Project

No description

Andrew Stinson

on 22 April 2013

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of Cell-tacular: Matlab Group Project

Translate the green layer of different intensities, to a logical matrix. Task Delegation perfect collaboration Functions
and Processes Output Counting cells by hand no longer necessary! TIFF files
RGB bleed
Connected Components
creating binary images
Cycling through images within Gui
Printout from Gui
Gui Guide for users Celltacular M. Allred, J. Bruns, C. Martell, A. Staples, A. Stinson Problem Definition Dr. Patel's Lab Solution(s) MATLAB Project Proposal Deliver a program which will replace tedious manual counting with a MATLAB script. This script will accurately and efficiently count the number of living cells and dead nuclei. Living Cells Dead Cells Algorithm Development (pt. 2) Thank You for Your Attention Sponsoring Faculty: Dr. Amit Patel Method(s) Used Background Current Methods:
watershed function applied using ImageJ software
count by hand In cell culturing labs, images taken from a microscope must be processed. It is important to be able to accurately process and count cell proliferation Process Import and read the file.
MATLAB converts the .tiff image into a matrix 3 layers deep, one for each red, green, and blue.
Each layer of the image is mapped based on pixels, where each pixel is a number that has a value from 0 to 256 depending on the intensity of the specific color of that pixel in the original image. Obstacles Bleed-through: red layer seemed to pick up some of green layer, distorting the counts
Picking a default threshold.
Green count enlarged unrealistically by size difference between green-stained cells and red-stained nuclei
Dealing with large cell masses
Overlapping RGB layers
Creating a Matlab script that counts cells within an acceptable error Mitigation Bleed-through
choose intensity range that filters intensities below what the naked eye can see
the intensities chosen are 44 for the green layer, and 88 for the red layer (out of 256 possible intensity)
Green count enlargement
eliminated green cells smaller than the average size of a red nucleus
Dealing with large cell masses
separate large masses using connected components, divide large components up using the average cell/nucleus size per image
Overlapping RGB layers
image thresholding done by eye: user interface for maximum accuracy
counted selected areas by hand to test algorithms
developed multiple methods of counting to ensure accuracy Red and Green Layers Intensities below the threshold are to be set to zero. Default: 88 Translate the red layer to a logical matrix. Distinguish the different clumps of red nuclei. (Mean+Median)/2 Intensities below the threshold are to be set to zero. Default: 48 Define the different clumps of green cells. Agenda Background
Problem Definition
MATLAB Project Proposal
Algorithm Development
Method Used
Outcome Take-Aways accurate results to report in scientific presentations
realistic idea of viability of scaffolds
measurable, dependable method of counting cells
easy to use interactive gui for image processing
technical manual for future modification and in-depth understanding of cell-counting functions
easy printout capturing overall live/dead cell counts group work can be fun
importance of setting goals and deadlines
always going back to the problem Cell Scaffolds 10 um slices 50 um apart
green stain-calcein am
stains entire cell body
red stain-propidium iodide
only stains nuclei No effective and efficient methods for counting the number of cells:
Image watershed method inaccurate
Size differences between red and green cells creates bias when counting
Relative comparison between images the only way to tell increased viability
Green cells smaller than the average red nucleus should be excluded but are not Living
Cells Dead Cells Two Methods First method uses average cell size to break up total area of red/green

Second method counts cells uses the average cell size to break up large cell chunks

--> when combined, results match cell numbers counted and estimated by hand ...and now a demonstration! Algorithm Development (pt. 1) Original Method Method(s) Used determining a constant coefficient
e.g. (total cell pixel count)X(coefficient)=number of cells.
method is rather reliable for images of consistent cell sizes

the second way of counting uses an average cell size.
e.g. (total cell pixel count)/(average cell size)=number of cells M. Allred, J. Bruns, C. Martell, A. Staples, A. Stinson
Full transcript