need clarification about generating `roi.tif`

[xuefei-wang]

Thanks for making the great tool & writing the detailed tutorial!

I am now using it to help count cells. But I have a question about how to generate the roi.tif in Analyze_ROIs.m. This is the process I inferred: first I generate an ROI channel and combine it into the raw tif image as one single file. After the whole process, I will obtain the transformed tif file, then I split the roi channel out of the transformed tif file. This roi channel itself would be roi.tif. Is it correct? If so, I wonder whether there is available scripts to do this? Thank you!

https://github.com/cortex-lab/allenCCF/blob/4fa077d91051c716a76986c9deedb19b43ec15e6/SHARP-Track/Analyze_ROIs.m#L14

[philshams]

Yes, that is one way to do it. To merge the ROI channel with histology image, you can either use something like ImageJ or do it in matlab with something like:

multi_channel_image = zeros(dim_1_of_original_image, dim_2_of_original_image, 2);
multi_channel_image(:,:,1) = original_image;
multi_channel_image(:,:,2) = ROI_image;

If you then want to extract the roi image without the histology for analysis, it should be just one line:

roi_location = 'C:\Drive\Histology\for tutorial - sample data\SS096_1_1_ROIs.tif';
roi_location = roi_location(:,:,2)

(here, assuming 2 is the color channel that the ROI data are found on)