Problems with Analyze particles

[Maria Smedh]

Hi,

I have just realized that my objects get too large when segmenting a fluorescent image in Fiji. I threshold the image and then use "Analyze particles" to create the segmented objects. My problem is that "Analyze particles" always fills holes, no matter if I have ticked the box "Include holes" or not!

I first tried to upgrade to the latest version of Fiji and then I tried to install a Life-line version, but I get the same results anyways. I have also tried with ImageJ and on a PC (my computer is a MacBook Pro).

Can anyone help me?

Best regards, Maria

------------------------------
MARIA SMEDH, PhD



Centre for Cellular Imaging
Core Facilities, the Sahlgrenska Academy
Univ. of Gothenburg, Sweden



Visiting address: Medicinareg. 7A, 41390 Gothenburg
Postal address: Box 435, 40530 Gothenburg
Delivery address: Medicinareg. 1G, 413 90 Gothenburg



Phone/Mobile: [phone]
SMS: [phone]
E-mail: [hidden email]<mailto:[hidden email]>
Web: http://www.cf.gu.se/english/Centre_for_Cellular_Imaging/






--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[gankaku]

Hi Maria,

can you potentiall post an binary example image with which you detect this
problem. This will make it easier to figure out what might be the problem.
In the current Fiji version I do not see any problem with the Analyze
Particle function if holes are in some particles.

best,
Jan

2015-03-03 16:51 GMT+01:00 Maria Smedh <[hidden email]>:

> Hi,
>
> I have just realized that my objects get too large when segmenting a
> fluorescent image in Fiji. I threshold the image and then use "Analyze
> particles" to create the segmented objects. My problem is that "Analyze
> particles" always fills holes, no matter if I have ticked the box "Include
> holes" or not!
>
> I first tried to upgrade to the latest version of Fiji and then I tried to
> install a Life-line version, but I get the same results anyways. I have
> also tried with ImageJ and on a PC (my computer is a MacBook Pro).
>
> Can anyone help me?
>
> Best regards, Maria
>
> ------------------------------
> MARIA SMEDH, PhD
>
>
>
> Centre for Cellular Imaging
> Core Facilities, the Sahlgrenska Academy
> Univ. of Gothenburg, Sweden
>
>
>
> Visiting address: Medicinareg. 7A, 41390 Gothenburg
> Postal address: Box 435, 40530 Gothenburg
> Delivery address: Medicinareg. 1G, 413 90 Gothenburg
>
>
>
> Phone/Mobile: [phone]
> SMS: [phone]
> E-mail: [hidden email]<mailto:[hidden email]>
> Web: http://www.cf.gu.se/english/Centre_for_Cellular_Imaging/
>
>
>
>
>
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html
>

--

CEO: Dr. rer. nat. Jan Brocher
phone:  [phone]
mobile: [phone]
e-mail: [hidden email]
info: [hidden email]
inquiries: [hidden email]
web: www.biovoxxel.de

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Maria Smedh]

Hi Jan,

I have uploaded the "original image" (which is not the actual original since the real original is a large tiled overview, which I start with resampling so that it will be easier to work with), the processed image that I threshold and the final binary image that I apply Analyze particles on. I have also added screen shots where you can see the settings I use in Analyze particles, the resulting objects and a zoomed in view of one of the objects where the hole has been filled. Here is the link:
https://gubox.box.com/s/y0emsnyx43t3ujtfndx8cw4kafliqngl

The processing I do on the image is to
1. remove background:
run("Subtract Background...", "rolling=100");
2. remove outliers:
run("Remove Outliers...", "radius=2 threshold=500 which=Bright");
3. smooth:
run("Gaussian Blur...", "sigma=2");

I also only look at a "tissue region" ROI, which is drawn well inside the edges of the section, since this tissue has a tendency to fold and have brighter staining around the edges even if not folded.

Any suggestions? Do I make any misstakes?

Best regards, Maria


------------------------------
MARIA SMEDH, PhD



Centre for Cellular Imaging
Core Facilities, the Sahlgrenska Academy
Univ. of Gothenburg, Sweden



Visiting address: Medicinareg. 7A, 41390 Gothenburg
Postal address: Box 435, 40530 Gothenburg
Delivery address: Medicinareg. 1G, 413 90 Gothenburg



Phone/Mobile: [phone]
SMS: [phone]
E-mail: [hidden email]<mailto:[hidden email]>
Web: http://www.cf.gu.se/english/Centre_for_Cellular_Imaging/





On 3 mar 2015, at 17:46, BioVoxxel <[hidden email]<mailto:[hidden email]>> wrote:

Hi Maria,

can you potentiall post an binary example image with which you detect this
problem. This will make it easier to figure out what might be the problem.
In the current Fiji version I do not see any problem with the Analyze
Particle function if holes are in some particles.

best,
Jan

2015-03-03 16:51 GMT+01:00 Maria Smedh <[hidden email]<mailto:[hidden email]>>:

Hi,

I have just realized that my objects get too large when segmenting a
fluorescent image in Fiji. I threshold the image and then use "Analyze
particles" to create the segmented objects. My problem is that "Analyze
particles" always fills holes, no matter if I have ticked the box "Include
holes" or not!

I first tried to upgrade to the latest version of Fiji and then I tried to
install a Life-line version, but I get the same results anyways. I have
also tried with ImageJ and on a PC (my computer is a MacBook Pro).

Can anyone help me?

Best regards, Maria

------------------------------
MARIA SMEDH, PhD



Centre for Cellular Imaging
Core Facilities, the Sahlgrenska Academy
Univ. of Gothenburg, Sweden



Visiting address: Medicinareg. 7A, 41390 Gothenburg
Postal address: Box 435, 40530 Gothenburg
Delivery address: Medicinareg. 1G, 413 90 Gothenburg



Phone/Mobile: [phone]
SMS: [phone]
E-mail: [hidden email]<mailto:[hidden email]><mailto:[hidden email]>
Web: http://www.cf.gu.se/english/Centre_for_Cellular_Imaging/






--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html




--

CEO: Dr. rer. nat. Jan Brocher
phone:  [phone]
mobile: [phone]
e-mail: [hidden email]<mailto:[hidden email]>
info: [hidden email]<mailto:[hidden email]>
inquiries: [hidden email]<mailto:[hidden email]>
web: www.biovoxxel.de<http://www.biovoxxel.de/>

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html


--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

[Anderson, Charles (DNR)]

In Analyze particles, "holes" are black regions completely enclosed by a white object.  Your Object 12 has three such holes.  Your Object 16 has 13 small holes, but the large black area is connected directly to the black background outside Object 16, and is thus part of the background.  If you rerun Analyze Particles with 'fill holes', I think you find the small holes are filled but the large 'outside' area is not.

You might experiment with Opening and Closing to cut off such regions so they are not connected outside before analyzing particles.  You may need to delete small white islands first, so they do not become connected to the large object of interest.

Hope that helps.

Charles

-----Original Message-----
From: ImageJ Interest Group [mailto:[hidden email]] On Behalf Of Maria Smedh
Sent: Wednesday, March 04, 2015 2:47 AM
To: [hidden email]
Subject: Re: Problems with Analyze particles

Hi Jan,

I have uploaded the "original image" (which is not the actual original since the real original is a large tiled overview, which I start with resampling so that it will be easier to work with), the processed image that I threshold and the final binary image that I apply Analyze particles on. I have also added screen shots where you can see the settings I use in Analyze particles, the resulting objects and a zoomed in view of one of the objects where the hole has been filled. Here is the link:
https://gubox.box.com/s/y0emsnyx43t3ujtfndx8cw4kafliqngl

The processing I do on the image is to
1. remove background:
run("Subtract Background...", "rolling=100"); 2. remove outliers:
run("Remove Outliers...", "radius=2 threshold=500 which=Bright"); 3. smooth:
run("Gaussian Blur...", "sigma=2");

I also only look at a "tissue region" ROI, which is drawn well inside the edges of the section, since this tissue has a tendency to fold and have brighter staining around the edges even if not folded.

Any suggestions? Do I make any misstakes?

Best regards, Maria


------------------------------
MARIA SMEDH, PhD



Centre for Cellular Imaging
Core Facilities, the Sahlgrenska Academy Univ. of Gothenburg, Sweden



Visiting address: Medicinareg. 7A, 41390 Gothenburg Postal address: Box 435, 40530 Gothenburg Delivery address: Medicinareg. 1G, 413 90 Gothenburg



Phone/Mobile: [phone]
SMS: [phone]
E-mail: [hidden email]<mailto:[hidden email]>
Web: http://www.cf.gu.se/english/Centre_for_Cellular_Imaging/





On 3 mar 2015, at 17:46, BioVoxxel <[hidden email]<mailto:[hidden email]>> wrote:

Hi Maria,

can you potentiall post an binary example image with which you detect this problem. This will make it easier to figure out what might be the problem.
In the current Fiji version I do not see any problem with the Analyze Particle function if holes are in some particles.

best,
Jan

2015-03-03 16:51 GMT+01:00 Maria Smedh <[hidden email]<mailto:[hidden email]>>:

Hi,

I have just realized that my objects get too large when segmenting a fluorescent image in Fiji. I threshold the image and then use "Analyze particles" to create the segmented objects. My problem is that "Analyze particles" always fills holes, no matter if I have ticked the box "Include holes" or not!

I first tried to upgrade to the latest version of Fiji and then I tried to install a Life-line version, but I get the same results anyways. I have also tried with ImageJ and on a PC (my computer is a MacBook Pro).

Can anyone help me?

Best regards, Maria

------------------------------
MARIA SMEDH, PhD



Centre for Cellular Imaging
Core Facilities, the Sahlgrenska Academy Univ. of Gothenburg, Sweden



Visiting address: Medicinareg. 7A, 41390 Gothenburg Postal address: Box 435, 40530 Gothenburg Delivery address: Medicinareg. 1G, 413 90 Gothenburg



Phone/Mobile: [phone]
SMS: [phone]
E-mail: [hidden email]<mailto:[hidden email]><mailto:[hidden email]>
Web: http://www.cf.gu.se/english/Centre_for_Cellular_Imaging/






--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html




--

CEO: Dr. rer. nat. Jan Brocher
phone:  [phone]
mobile: [phone]
e-mail: [hidden email]<mailto:[hidden email]>
info: [hidden email]<mailto:[hidden email]>
inquiries: [hidden email]<mailto:[hidden email]>
web: www.biovoxxel.de<http://www.biovoxxel.de/>

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html


--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html