search

ML-Radiology / Abdominal_MRI_series_classification

Uses metadata and pixel classifiers for MRI abdomen studies
1 stars 0 forks source link

Issues with trying out the streamlit demo #1

Open deepakri201 opened 1 month ago

deepakri201 commented 1 month ago

Hi,

Thanks for this great and interesting work on DICOM series classification! Thank you for also making your notebooks and streamlit code publicly available.

I am trying to use your streamlit app to run inference on a set of DICOM images on my computer. However, I am running into a few issues.

First a few minor issues that I fixed:

Next, I am able to sucessfully load in the DICOM images: image

When I then select "Get Inference For This Image", I get the following error: image

Perhaps it is an issue with the pandas version. Thank you for your time,

Deepa

ML-Radiology commented 1 month ago

Deepa,

Thank you for your interest and your email. The requirements.txt file has been updated to fix the issues you noted. Thank you. As for your other issue, I have not come across that particular error so will have to dig a little deeper for a solution.

Regards, Chad

On Jul 15, 2024, at 12:41 PM, Deepa Krishnaswamy @.***> wrote:

Hi,

Thanks for this great and interesting work on DICOM series classification! Thank you for also making your notebooks and streamlit code publicly available.

I am trying to use your streamlit app to run inference on a set of DICOM images on my computer. However, I am running into a few issues.

First a few minor issues that I fixed:

In the requirements.txt file, the simpleITK package was missing a "=" I could not install pandas 2.0.0 as there was a dependency conflict between that and streamlit. After removing the dependency on 2.0.0, it installed pandas 1.5.3 automatically. Next, I am able to sucessfully load in the DICOM images: image.png (view on web) https://github.com/user-attachments/assets/70d5f0ac-8194-4d62-a11a-08e9c58a3e8a When I then select "Get Inference For This Image", I get the following error: image.png (view on web) https://github.com/user-attachments/assets/a21faf77-8a45-4487-ba81-597d81913794 Perhaps it is an issue with the pandas version. Thank you for your time,

Deepa

— Reply to this email directly, view it on GitHub https://github.com/ML-Radiology/Abdominal_MRI_series_classification/issues/1, or unsubscribe https://github.com/notifications/unsubscribe-auth/BEPIYVAKD6UQXNGSD5LTSJ3ZMP3STAVCNFSM6AAAAABK443VRWVHI2DSMVQWIX3LMV43ASLTON2WKOZSGQYDSMJXGUZDQMI. You are receiving this because you are subscribed to this thread.