search

mrvturan96 / Brain-Tumor-Detection-and-Segmentation-using-Deep-Learning

This repo includes Glioma Segmentation with Mask R-CNN and U-Net.
62 stars 22 forks source link

data processing #1

Open zhao706 opened 4 years ago

zhao706 commented 4 years ago

Hello,are you using the via tool to mark comments that manually label the images in the t1ce folder?

mrvturan96 commented 4 years ago

Hi, Yes, I used via tool for annotations and labeled manually.

zhao706 notifications@github.com, 1 Mar 2020 Paz, 12:53 tarihinde şunu yazdı:

Hello,are you using the via tool to mark comments that manually label the images in the t1ce folder?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/mrvturan96/Brain-Tumor-Detection-and-Segmentation-using-Deep-Learning/issues/1?email_source=notifications&email_token=AJFAJBZAKJMRJLUNZVHYXW3RFIWB5A5CNFSM4K7DIGP2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4IRPF2HA, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJFAJB2M7JDYWKKENDC5FXLRFIWB5ANCNFSM4K7DIGPQ .

zhao706 commented 4 years ago

Thank you. But if you mark the location of brain tumor manually, the final segmentation results will be affected by the accuracy of people's labeling. @mrvturan96

mrvturan96 commented 4 years ago

You're right. Actually, we labeled only t1ce manually for Mask-RCNN. For this, we convert NIFTI images to .jpg images. Tumors very pronounced on t1ce sequence. (We used only the most obvious slide.) But we didn't manually labeled on U-Net. Because NIFTI images labeled by radiologists.

zhao706 commented 4 years ago

OK, I see.Thank you very much.

------------------ Original ------------------ From: Merve TURAN <notifications@github.com> Date: Wed,Mar 4,2020 2:26 PM To: mrvturan96/Brain-Tumor-Detection-and-Segmentation-using-Deep-Learning <Brain-Tumor-Detection-and-Segmentation-using-Deep-Learning@noreply.github.com> Cc: zhao706 <1043314579@qq.com>, Author <author@noreply.github.com> Subject: Re: [mrvturan96/Brain-Tumor-Detection-and-Segmentation-using-Deep-Learning] data processing (#1)

You're right. Actually, we labeled only t1ce manually for Mask-RCNN. For this, we convert NIFTI images to .jpg images. Tumors very pronounced on t1ce sequence. (We used only the most obvious slide.) But we didn't manually labeled on U-Net. Because NIFTI images labeled by radiologists.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

javaidiqbal11 commented 3 years ago

Hello, Thanks for sharing this work, I want some help to execute the code and also worked on the same project to add some work with you. Please reply to me soon. jtechofficial198@gmail.com

javaidiqbal11 commented 3 years ago

OK, I see.Thank you very much. … ------------------ Original ------------------ From: Merve TURAN <notifications@github.com> Date: Wed,Mar 4,2020 2:26 PM To: mrvturan96/Brain-Tumor-Detection-and-Segmentation-using-Deep-Learning <Brain-Tumor-Detection-and-Segmentation-using-Deep-Learning@noreply.github.com> Cc: zhao706 <1043314579@qq.com>, Author <author@noreply.github.com> Subject: Re: [mrvturan96/Brain-Tumor-Detection-and-Segmentation-using-Deep-Learning] data processing (#1) You're right. Actually, we labeled only t1ce manually for Mask-RCNN. For this, we convert NIFTI images to .jpg images. Tumors very pronounced on t1ce sequence. (We used only the most obvious slide.) But we didn't manually labeled on U-Net. Because NIFTI images labeled by radiologists. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe.

Can you guide me on what steps you follow to execute the complete work now. please

javaidiqbal11 commented 3 years ago

What the research methodology you proposed and can you please share the research paper for this work.