ariel-waisman / LabelsToROIs

A Fiji/ImageJ plugin to generate ROIs from label images, allowing ROI erosion and quantification
GNU General Public License v3.0
21 stars 9 forks source link

I can't Run Lois to ROI on my image. #22

Open Momo-MLB opened 2 months ago

Momo-MLB commented 2 months ago

Greeting, I would love to use this amazing tool LOIStoROI to segment and quantify sections of myofiber. I followed the steps one by one (cellpose and Lois to Roi plugin on Fiji, both on windows and on mac) but it doesn't work. First : the label png image that I got from cellpose was just a black image : is that normal ? (I don't know how should the image look like). I attach herein the two images, the original and the label output image. Second, (I assumed the black label image is normal so I proceed to the Lois to Roi plugin on Fiji (updated to the last version). I choose single image, and load the original TIF and the label image PNG. when I click GO two windows appear but they are blank nothing is going on (I tried again and again but invane, I also wated two hours, thought maybe it's a processor thing but also invane). When I tried that on windows, the Fiji app closes abruptly when I click on GO.

Please help me resolve this issue please. Thank you all in advance !!!!! Here is the label image i got from the Colab-Cellpose Snap01-c1_label and I converted the TIF image to JPEG so I can upload it here. Snap01-c1

ariel-waisman commented 2 months ago

Hi!

The problem is that you were trying to segment and RGB image and not a grayscale 8 bit image. If you open your original image in ImageJ, go to Image / Type / 8-bit and save it in that format, you can run the Cellpose Colab and it should work.

I attach the label image after running the colab with the grayscale image. (just for visualization, I color coded the labels using MorpholibJ label images plugin in ImageJ), but this is just a visualization tool, it works the same as the output image from the colab.

Snap01-c1_label with color code

And this is an overlay of the ROIs and your original image. Snap01-c1-1 overlay

Hope that helps!

Good luck, Ariel

On Fri, Jul 5, 2024 at 9:51 AM Momo-MLB @.***> wrote:

Greeting, I would love to use this amazing tool LOIStoROI to segment and quantify sections of myofiber. I followed the steps one by one (cellpose and Lois to Roi plugin on Fiji, both on windows and on mac) but it doesn't work. First : the label png image that I got from cellpose was just a black image : is that normal ? (I don't know how should the image look like). I attach herein the two images, the original and the label output image. Second, (I assumed the black label image is normal so I proceed to the Lois to Roi plugin on Fiji (updated to the last version). I choose single image, and load the original TIF and the label image PNG. when I click GO two windows appear but they are blank nothing is going on (I tried again and again but invane, I also wated two hours, thought maybe it's a processor thing but also invane). When I tried that on windows, the Fiji app closes abruptly when I click on GO.

Please help me resolve this issue please. Thank you all in advance !!!!! Here is the label image i got from the Colab-Cellpose Snap01-c1_label.png (view on web) https://github.com/ariel-waisman/LabelsToROIs/assets/174803747/62e618ee-9270-44b6-b6ee-5ef360edd676 and I converted the TIF image to JPEG so I can upload it here. Snap01-c1.jpeg (view on web) https://github.com/ariel-waisman/LabelsToROIs/assets/174803747/63ce53d2-ae4c-4330-b72f-45b11c46fcb0

— Reply to this email directly, view it on GitHub https://github.com/ariel-waisman/LabelsToROIs/issues/22, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFAA54H7TCETVOUV6BMM7WTZK2JENAVCNFSM6AAAAABKNEZS4WVHI2DSMVQWIX3LMV43ASLTON2WKOZSGM4TENJVGIZTENI . You are receiving this because you are subscribed to this thread.Message ID: @.***>

Momo-MLB commented 2 months ago

Greetings DOCTOR !!!!!

THANK YOU SOOOOOOOO MUCH THANK YOUUUU SO MUCH !!!

It worked THANK YOUUUU SO MUCH , you saved hours of my times THANK YOUUUUUUU SO MUCH DOCTOR Ariel. I exteremly appreciate your HELP and the EFFORT you went through to make accessible this GREAT TOOL !!! THANK YOU ETERNELLY!!!!

WISH YOU ALL THE BEST ! Momo