What will we do when complete the section 8 in order to have the classification results?

[tphankr]

Thank you @gallegi. I trained successfully all the section 8. And the next, please help me how do I do the next step, I also saw the section 9, as below following, but I don't know how do I use it in order to have the classification results. Thank you. .

Thanks.

[gallegi]

Oh I forgot to add the training cmd. i will update it soon

On Fri, 29 Oct 2021 at 04:09 tphankr @.***> wrote:

Thank you @gallegi https://github.com/gallegi. I trained successfully all the section 8. And the next, please help me how do I do the next step, I also saw the section 9, as below following, but I don't know how do I use it in order to have the classification results. Thank you. [image: image] https://user-images.githubusercontent.com/37217354/139335670-bbf6122e-b691-4754-a3c5-d93fac3e6ffd.png .

[image: image] https://user-images.githubusercontent.com/37217354/139335923-baa40ce3-366a-4840-bed3-2ca381d1a356.png

Thanks.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/gallegi/T4E_MICCAI_BrainTumor/issues/2, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGVECU2DBOOTXDEOSUVRNSDUJG3ZLANCNFSM5G52UVFA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

[gallegi]

I have updated the guide on README. Please check if it works.

[tphankr]

Thank you @gallegi, I ran completely your update. Now, I had a .pth file in each FLAIR, T1w, T1wCE and T2w folder in Models/eca_nfnet_l0_2d_classification. A example, the .pth file in T1w, as following: . And the next section 9, it will be created the submission.csv file that the result of this task, right? how do we use the section 9? When I open the section 9, it is the kaggle.com website.

Please guide me in order to have the final classification results, thank you @gallegi.

[gallegi]

If you have completed training all the model, you have 2 option:

1. Run inference on your local machine. Please refer to section 8.1. It will use your trained models to predict the MGMT value on test set. Note that, it requires you to download all the dicom data as described in section 3.
2. Run inference on Kaggle. You can refer to the notebook link on section
3. Upload your own trained weights to new Kaggle datasets named brain2dsefbest (for segmentation weight) and brain2dclflstm (for classification weights). And create a copy of my notebook and replace the 2 mentioned datasets with your own ones. Because Kaggle has dicom data for you, you are able to run inference now. Set variable FAST_SUB = False to get all the result of the public test.

On Sat, Oct 30, 2021 at 7:54 AM tphankr @.***> wrote:

Thank you @gallegi https://github.com/gallegi, I ran completely your update. Now, I had a .pth file in each FLAIR, T1w, T1wCE and T2w folder in Models/eca_nfnet_l0_2d_classification. A example, the .pth file in T1w, as following: [image: image] https://user-images.githubusercontent.com/37217354/139514142-b8c09a40-0996-4daf-893c-06f239f68847.png . And the next section 9, it will be created the submission.csv file that the result of this task, right? how do we use the section 9? When I open the section 9, it is kaggle.com website. [image: image] https://user-images.githubusercontent.com/37217354/139514285-266ece8a-9585-44e6-9f1a-6e48192795a6.png [image: image] https://user-images.githubusercontent.com/37217354/139514375-3ed55d60-d9c9-4a96-8fff-86acb81cd9ca.png

Please guide me in order to have the final classification results, thank you @gallegi https://github.com/gallegi.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/gallegi/T4E_MICCAI_BrainTumor/issues/2#issuecomment-955117092, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGVECU64U6JP7X4ZKY32XDLUJM65NANCNFSM5G52UVFA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

[tphankr]

Thank you for your reply. I did it by the first option. Thanks. And now, I'm wondering how do we have the raw_jpg_data. Can you share the code that convert the dicom file to the jpg file? I also tried to convert them but have a problem in background, the background is not black like this:

Thank you @gallegi.

[gallegi]

Please refer to this notebook: https://www.kaggle.com/namgalielei/brainsavedata/notebook

On Mon, Nov 1, 2021 at 3:23 PM tphankr @.***> wrote:

Thank you for your reply. I did it by the first option. And now, I'm wondering how do we have the raw_jpg_data. Can you share the code convert the dicom file to the jpg file? I also tried to convert them but have a problem in background, the background is not black like this: [image: image] https://user-images.githubusercontent.com/37217354/139642577-7462f8c1-d4b8-401c-9505-73da3e7dccca.png

Thank you @gallegi https://github.com/gallegi.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/gallegi/T4E_MICCAI_BrainTumor/issues/2#issuecomment-956030878, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGVECU2GUMKNNQUSVT2PQDLUJZFARANCNFSM5G52UVFA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

[tphankr]

Thank you @gallegi. I will follow.