SlideShare a Scribd company logo
1 of 3
• Less than 5% error when
comparing nuclei count
from MATLAB with the
nuclei count from the
trained user
• Goals:
1. Minimize error
2. Decrease manual
hours computing
nuclei count
2D Spatial Filtering Images of Cardiac Cells for Nuclei Isolation
Jessica E. Herrmann, Ha D.H. Le, Theresa C. Rizk
Efficiently and accurately measuring nuclei count can provide a better understanding of changes in the
heart due to disease progression
GENERAL PROCEDURE:
1. Optimize 2D spatial filter in ImageJ
2. Run the filter and use MATLAB to count the nuclei
3. Have a trained user manually count the nuclei
4. Perform statistical analysis to determine error
ASSUMPTIONS AND LIMITATIONS:
• Limitation or assumption.
• Another limitation or assumption.
• Image analysis can provide
quantifiable data (e.g.: cell density)
• It is deterred by low quality images
that are often varied across samples
• How can we develop a filter to isolate
nuclei from an image of cardiac cells
in order to enhance quantification of
the nuclei?
EXPECTEDOUTCOMEIMPACT
METHODOLOGY
PROBLEMCONCEPT
Hypothesis: Fourier transforms
and image processing software
can be used to develop a 2D
spatial filter that will isolate the
nuclei in images of cardiac cells.
Signs of
inflammatio
n
Cell thickness
Ventricular
wall thickness
How does chronic kidney
disease affect the heart?
Counting nuclei
more accurately
Cell
Density
Convolution Mask
Potential Filters
Laplacian of the Gaussian
1. Gaussian filter to remove noise
2. Fourier Transform to sharpen
3. Laplacian to detect edges
Gaussian Edge Detection
1. Gaussian filter to remove noise
2. Fourier Transform to sharpen
3. Gradient
Threshold Filtering of Binary Image
1. Convert RGB to grey scale
2. Threshold to produce binary image
3. Erosion and dilation
h[i, j]
x[i, j] y[i, j]
?
Countx
Color deconvolution
Updates and Progress
• Color deconvolution of Image
• Laplacian of the Gaussian
1. Apply “despeckle” filter to de-noise
2. Remove outliers
3. Gaussian filter
4. Apply Laplacian filter to the image [this filter only works in Fiji]
5. Run MATLAB image analysis
• Gaussian Edge Detection
1. Steps 1-3 above
2. Apply “find edges” (Sobel edge detector,
uses gradient of image intensity function)
3. Run MATLAB image analysis
• Threshold Filtering
1. Convert convoluted image to binary
2. Erode and dilate to enhance image
3. Apply Sobel edge detector
4. Run MATLAB image analysis
Laplacian Results: 778 nuclei counted
Gaussian Results: 395 nuclei counted
Threshold Filtering Results:
Gaussian edge Detection
MATLAB script to count the objects
First increases the contrast of the
image
Converts grayscale to binary image by
using
thresholding
Further removes background noise
Counts objects in the image
Obstacles
• No method for testing whether nuclei are being eroded through
filtering
• Nuclei touching in image can
be perceived as single entity
• Balance between decreasing
noise and maintaining integrity of object edges
• Altering size of nuclei – misleading data
• Some noise still remains after filtering and can be counted as nuclei

More Related Content

Similar to Pentachart_AllGraphics

Image enhancement
Image enhancementImage enhancement
Image enhancementAyaelshiwi
 
SIRG-BSU_3_used-important.pdf
SIRG-BSU_3_used-important.pdfSIRG-BSU_3_used-important.pdf
SIRG-BSU_3_used-important.pdfDrAhmedElngar
 
Image processing sw & hw
Image processing sw & hwImage processing sw & hw
Image processing sw & hwamalalhait
 
Image enhancement lecture
Image enhancement lectureImage enhancement lecture
Image enhancement lectureISRAR HUSSAIN
 
Cahall Final Intern Presentation
Cahall Final Intern PresentationCahall Final Intern Presentation
Cahall Final Intern PresentationDaniel Cahall
 
Using Image Acquisition Is The Input Text Document
Using Image Acquisition Is The Input Text DocumentUsing Image Acquisition Is The Input Text Document
Using Image Acquisition Is The Input Text DocumentLisa Williams
 
SINGLE IMAGE SUPER RESOLUTION IN SPATIAL AND WAVELET DOMAIN
SINGLE IMAGE SUPER RESOLUTION IN SPATIAL AND WAVELET DOMAINSINGLE IMAGE SUPER RESOLUTION IN SPATIAL AND WAVELET DOMAIN
SINGLE IMAGE SUPER RESOLUTION IN SPATIAL AND WAVELET DOMAINijma
 
Image enhancement ppt nal2
Image enhancement ppt nal2Image enhancement ppt nal2
Image enhancement ppt nal2Surabhi Ks
 
An Approach for Image Deblurring: Based on Sparse Representation and Regulari...
An Approach for Image Deblurring: Based on Sparse Representation and Regulari...An Approach for Image Deblurring: Based on Sparse Representation and Regulari...
An Approach for Image Deblurring: Based on Sparse Representation and Regulari...IRJET Journal
 
PES ncetec conference
PES ncetec conferencePES ncetec conference
PES ncetec conferenceAvinash P M
 
study-and-development-of-digital-image-processing-tool-for-application-of-dia...
study-and-development-of-digital-image-processing-tool-for-application-of-dia...study-and-development-of-digital-image-processing-tool-for-application-of-dia...
study-and-development-of-digital-image-processing-tool-for-application-of-dia...Jyoti Patil
 
X-Ray Image Enhancement using CLAHE Method
X-Ray Image Enhancement using CLAHE MethodX-Ray Image Enhancement using CLAHE Method
X-Ray Image Enhancement using CLAHE MethodIRJET Journal
 
30thSep2014
30thSep201430thSep2014
30thSep2014Mia liu
 
Paper on image processing
Paper on image processingPaper on image processing
Paper on image processingSaloni Bhatia
 
IRJET - An Enhanced Approach for Extraction of Text from an Image using Fuzzy...
IRJET - An Enhanced Approach for Extraction of Text from an Image using Fuzzy...IRJET - An Enhanced Approach for Extraction of Text from an Image using Fuzzy...
IRJET - An Enhanced Approach for Extraction of Text from an Image using Fuzzy...IRJET Journal
 

Similar to Pentachart_AllGraphics (20)

Image enhancement
Image enhancementImage enhancement
Image enhancement
 
SIRG-BSU_3_used-important.pdf
SIRG-BSU_3_used-important.pdfSIRG-BSU_3_used-important.pdf
SIRG-BSU_3_used-important.pdf
 
Image processing sw & hw
Image processing sw & hwImage processing sw & hw
Image processing sw & hw
 
Project Poster
Project PosterProject Poster
Project Poster
 
Image enhancement lecture
Image enhancement lectureImage enhancement lecture
Image enhancement lecture
 
Cahall Final Intern Presentation
Cahall Final Intern PresentationCahall Final Intern Presentation
Cahall Final Intern Presentation
 
Brain tumor detection
Brain tumor detectionBrain tumor detection
Brain tumor detection
 
image processing
image processingimage processing
image processing
 
Using Image Acquisition Is The Input Text Document
Using Image Acquisition Is The Input Text DocumentUsing Image Acquisition Is The Input Text Document
Using Image Acquisition Is The Input Text Document
 
SINGLE IMAGE SUPER RESOLUTION IN SPATIAL AND WAVELET DOMAIN
SINGLE IMAGE SUPER RESOLUTION IN SPATIAL AND WAVELET DOMAINSINGLE IMAGE SUPER RESOLUTION IN SPATIAL AND WAVELET DOMAIN
SINGLE IMAGE SUPER RESOLUTION IN SPATIAL AND WAVELET DOMAIN
 
Image enhancement ppt nal2
Image enhancement ppt nal2Image enhancement ppt nal2
Image enhancement ppt nal2
 
An Approach for Image Deblurring: Based on Sparse Representation and Regulari...
An Approach for Image Deblurring: Based on Sparse Representation and Regulari...An Approach for Image Deblurring: Based on Sparse Representation and Regulari...
An Approach for Image Deblurring: Based on Sparse Representation and Regulari...
 
PES ncetec conference
PES ncetec conferencePES ncetec conference
PES ncetec conference
 
study-and-development-of-digital-image-processing-tool-for-application-of-dia...
study-and-development-of-digital-image-processing-tool-for-application-of-dia...study-and-development-of-digital-image-processing-tool-for-application-of-dia...
study-and-development-of-digital-image-processing-tool-for-application-of-dia...
 
X-Ray Image Enhancement using CLAHE Method
X-Ray Image Enhancement using CLAHE MethodX-Ray Image Enhancement using CLAHE Method
X-Ray Image Enhancement using CLAHE Method
 
WT in IP.ppt
WT in IP.pptWT in IP.ppt
WT in IP.ppt
 
Deeplearning
Deeplearning Deeplearning
Deeplearning
 
30thSep2014
30thSep201430thSep2014
30thSep2014
 
Paper on image processing
Paper on image processingPaper on image processing
Paper on image processing
 
IRJET - An Enhanced Approach for Extraction of Text from an Image using Fuzzy...
IRJET - An Enhanced Approach for Extraction of Text from an Image using Fuzzy...IRJET - An Enhanced Approach for Extraction of Text from an Image using Fuzzy...
IRJET - An Enhanced Approach for Extraction of Text from an Image using Fuzzy...
 

Pentachart_AllGraphics

  • 1. • Less than 5% error when comparing nuclei count from MATLAB with the nuclei count from the trained user • Goals: 1. Minimize error 2. Decrease manual hours computing nuclei count 2D Spatial Filtering Images of Cardiac Cells for Nuclei Isolation Jessica E. Herrmann, Ha D.H. Le, Theresa C. Rizk Efficiently and accurately measuring nuclei count can provide a better understanding of changes in the heart due to disease progression GENERAL PROCEDURE: 1. Optimize 2D spatial filter in ImageJ 2. Run the filter and use MATLAB to count the nuclei 3. Have a trained user manually count the nuclei 4. Perform statistical analysis to determine error ASSUMPTIONS AND LIMITATIONS: • Limitation or assumption. • Another limitation or assumption. • Image analysis can provide quantifiable data (e.g.: cell density) • It is deterred by low quality images that are often varied across samples • How can we develop a filter to isolate nuclei from an image of cardiac cells in order to enhance quantification of the nuclei? EXPECTEDOUTCOMEIMPACT METHODOLOGY PROBLEMCONCEPT Hypothesis: Fourier transforms and image processing software can be used to develop a 2D spatial filter that will isolate the nuclei in images of cardiac cells. Signs of inflammatio n Cell thickness Ventricular wall thickness How does chronic kidney disease affect the heart? Counting nuclei more accurately Cell Density Convolution Mask Potential Filters Laplacian of the Gaussian 1. Gaussian filter to remove noise 2. Fourier Transform to sharpen 3. Laplacian to detect edges Gaussian Edge Detection 1. Gaussian filter to remove noise 2. Fourier Transform to sharpen 3. Gradient Threshold Filtering of Binary Image 1. Convert RGB to grey scale 2. Threshold to produce binary image 3. Erosion and dilation h[i, j] x[i, j] y[i, j] ? Countx
  • 2. Color deconvolution Updates and Progress • Color deconvolution of Image • Laplacian of the Gaussian 1. Apply “despeckle” filter to de-noise 2. Remove outliers 3. Gaussian filter 4. Apply Laplacian filter to the image [this filter only works in Fiji] 5. Run MATLAB image analysis • Gaussian Edge Detection 1. Steps 1-3 above 2. Apply “find edges” (Sobel edge detector, uses gradient of image intensity function) 3. Run MATLAB image analysis • Threshold Filtering 1. Convert convoluted image to binary 2. Erode and dilate to enhance image 3. Apply Sobel edge detector 4. Run MATLAB image analysis Laplacian Results: 778 nuclei counted Gaussian Results: 395 nuclei counted Threshold Filtering Results: Gaussian edge Detection MATLAB script to count the objects First increases the contrast of the image Converts grayscale to binary image by using thresholding Further removes background noise Counts objects in the image
  • 3. Obstacles • No method for testing whether nuclei are being eroded through filtering • Nuclei touching in image can be perceived as single entity • Balance between decreasing noise and maintaining integrity of object edges • Altering size of nuclei – misleading data • Some noise still remains after filtering and can be counted as nuclei