search

Steven-N-Hart / dicom_wsi

Converts whole slide images to DICOM format
https://dicom-wsi.readthedocs.io
MIT License
17 stars 8 forks source link

Concatenation of magnifications #20

Open maxfscher opened 2 years ago

maxfscher commented 2 years ago

Hi,

I noticed some strange behavior when converting tiff files from the Camelyon16 data set. I tried to convert one file from the publicly available data set to DICOM WSI and when viewing the images in SLIM Viewer (https://herrmannlab.github.io/slim/) I am seeing an offset between the levels that are rendered. While this works for each other files and data types, this continues to fail for the Camelyon tiffs. While zooming through the layers, another position from the higher magnification is loaded than the position that is expected.

Steven-N-Hart commented 2 years ago

Thanks Max.

Can you send me a specific TIFF and the config file you are using so I can reproduce the issue?

Steven Hart, Ph.D. Senior Associate Consultant - Artificial Intelligence Associate Professor of Biomedical Informatics Division of Computational Pathology and AI Department of Laboratory Medicine and Pathology Mayo College of Medicine Phone: 507-538-5288 | @.**@.> Mayo Clinic | 200 First Street SW | Rochester, MN 55905 | www.mayoclinic.orghttp://www.mayoclinic.org/

From: Maximilian Fischer @.> Sent: Wednesday, November 16, 2022 4:13 AM To: Steven-N-Hart/dicom_wsi @.> Cc: Subscribed @.***> Subject: [EXTERNAL] [Steven-N-Hart/dicom_wsi] Concatenation of magnifications (Issue #20)

Hi,

I noticed some strange behavior when converting tiff files from the Camelyon16 data set. I tried to convert one file from the publicly available data set to DICOM WSI and when viewing the images in SLIM Viewer (https://herrmannlab.github.io/slim/https://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fherrmannlab.github.io%2Fslim%2F&data=05%7C01%7CHart.Steven%40mayo.edu%7C57af1a63e0834d2771c308dac7bb249f%7Ca25fff9c3f634fb29a8ad9bdd0321f9a%7C0%7C0%7C638041903866176487%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=tHyhzc2ON3zQh6306piGGIZpqTeDRSfi%2B7I98goW7VM%3D&reserved=0) I am seeing an offset between the levels that are rendered. While this works for each other files and data types, this continues to fail for the Camelyon tiffs. While zooming through the layers, another position from the higher magnification is loaded than the position that is expected.

- Reply to this email directly, view it on GitHubhttps://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2FSteven-N-Hart%2Fdicom_wsi%2Fissues%2F20&data=05%7C01%7CHart.Steven%40mayo.edu%7C57af1a63e0834d2771c308dac7bb249f%7Ca25fff9c3f634fb29a8ad9bdd0321f9a%7C0%7C0%7C638041903866176487%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=i2KEEEhtdeOHRj0FyPVm1YtJMPLt6SRP2YFFFNK6RWk%3D&reserved=0, or unsubscribehttps://nam12.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FACBKURIQBNK7VZ5IY6C2EEDWISXSRANCNFSM6AAAAAASCBMZWY&data=05%7C01%7CHart.Steven%40mayo.edu%7C57af1a63e0834d2771c308dac7bb249f%7Ca25fff9c3f634fb29a8ad9bdd0321f9a%7C0%7C0%7C638041903866176487%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=8cv4hcbepFEWvdnEaxdi0gi2GBC0rmreD78iRhPsnpQ%3D&reserved=0. You are receiving this because you are subscribed to this thread.Message ID: @.**@.>>

maxfscher commented 2 years ago

Hi Steven, thanks for your reply. Here is the yaml file and the path where you can download the sample file. It is the path that is provided via the challenge Homepage. I used the file /training/normal/normal_001.tif

https://pan.baidu.com/s/1UW_HLXXjjw5hUvBIUYPgbA#list/path=%2FCAMELYON16%2Ftraining%2Fnormal base.txt

Steven-N-Hart commented 2 years ago

Can you verify for me that your conversion ran without error? I noticed when trying to convert normal_001.tif, I got an error form the VIPS library:

Just trying to rule out simple issues first.