The Competition hosts decided to host 2 separate tasks. In case anyone wants to use kaggle notebooks to make segmentation predictions, this Notebook demonstrates how to acquire the Dataset from Synapse.org (host website for segmentation task). Furthermore it also demonstrates which cases(patient IDs) are missing from the original Classification Competition Dataset available on Kaggle (hence no target MGMT_value).
\nYou will need to sign-up on Synapse and use the credentials for this notebook to work in your own scripts. You can also directly add this notebook to your notebook, to use the data directly. The data is available in nii.gz (NIfTI) format.
\nHave a Great Day!
", "messages": [ { "id": "1481931", "postDate": "08/19/2021 18:33:48", "content": "The Competition hosts decided to host 2 separate tasks. In case anyone wants to use kaggle notebooks to make segmentation predictions, this Notebook demonstrates how to acquire the Dataset from Synapse.org (host website for segmentation task). Furthermore it also demonstrates which cases(patient IDs) are missing from the original Classification Competition Dataset available on Kaggle (hence no target MGMT_value).
\nYou will need to sign-up on Synapse and use the credentials for this notebook to work in your own scripts. You can also directly add this notebook to your notebook, to use the data directly. The data is available in nii.gz (NIfTI) format.
\nHave a Great Day!
", "rawMarkdown": "The Competition hosts decided to host 2 separate tasks. In case anyone wants to use kaggle notebooks to make segmentation predictions, this [Notebook](https://www.kaggle.com/snish9/get-brats-2021) demonstrates how to acquire the Dataset from Synapse.org (host website for segmentation task). Furthermore it also demonstrates which cases(patient IDs) are missing from the original Classification Competition Dataset available on Kaggle (hence no target MGMT_value). \n\nYou will need to [sign-up](https://www.synapse.org/brats2021) on Synapse and use the credentials for this notebook to work in your own scripts. You can also directly add this notebook to your notebook, to use the data directly. The data is available in nii.gz (NIfTI) format.\n\nHave a Great Day!", "votes": null }, { "id": "1482115", "postDate": "08/19/2021 22:02:44", "content": "If you do not mind answering a few questions:
\nI have compared both datasets(Task-1 & 2) by patient_id and found the cases that the host team chose to not include in the other dataset i.e. cases in task-1 dataset not available in task-2 dataset and vice versa.
\nas far as image size is concerned u can use nibabel library to check the array size
\nt2_nib = nib.load('t2.nii.gz')
\nt2_nib
\nt2_nib_array = t2_nib.get_fdata()
\nt2_nib_array.shape
have you checked? because the classification problem has a different number of images and images of different sizes: from 150 to 512
", "rawMarkdown": "have you checked? because the classification problem has a different number of images and images of different sizes: from 150 to 512", "votes": null }, { "id": "1483433", "postDate": "08/20/2021 16:22:02", "content": "No. of images for all modalities in segmentation dataset is same. I'm not sure about the image size though
", "rawMarkdown": "No. of images for all modalities in segmentation dataset is same. I'm not sure about the image size though", "votes": null } ], "comments": [ { "id": 1482115, "author_name": "zaakciiru", "author_url": "", "post_date": "08/19/2021 22:02:44", "content": "If you do not mind answering a few questions:
\nI have compared both datasets(Task-1 & 2) by patient_id and found the cases that the host team chose to not include in the other dataset i.e. cases in task-1 dataset not available in task-2 dataset and vice versa.
\nas far as image size is concerned u can use nibabel library to check the array size
\nt2_nib = nib.load('t2.nii.gz')
\nt2_nib
\nt2_nib_array = t2_nib.get_fdata()
\nt2_nib_array.shape
have you checked? because the classification problem has a different number of images and images of different sizes: from 150 to 512
", "votes": null, "replies": [] }, { "id": 1483433, "author_name": "snish9", "author_url": "", "post_date": "08/20/2021 16:22:02", "content": "No. of images for all modalities in segmentation dataset is same. I'm not sure about the image size though
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1481931": "The Competition hosts decided to host 2 separate tasks. In case anyone wants to use kaggle notebooks to make segmentation predictions, this [Notebook](https://www.kaggle.com/snish9/get-brats-2021) demonstrates how to acquire the Dataset from Synapse.org (host website for segmentation task). Furthermore it also demonstrates which cases(patient IDs) are missing from the original Classification Competition Dataset available on Kaggle (hence no target MGMT_value). \n\nYou will need to [sign-up](https://www.synapse.org/brats2021) on Synapse and use the credentials for this notebook to work in your own scripts. You can also directly add this notebook to your notebook, to use the data directly. The data is available in nii.gz (NIfTI) format.\n\nHave a Great Day!", "1482115": "If you do not mind answering a few questions:\n1. What is the size of the images in this competition?\n2. Are all images the same size?\n3. what is predicted there?", "1482506": "I have compared both datasets(Task-1 & 2) by patient_id and found the cases that the host team chose to not include in the other dataset i.e. cases in task-1 dataset not available in task-2 dataset and vice versa.\n\nas far as image size is concerned u can use nibabel library to check the array size\n`t2_nib = nib.load('t2.nii.gz')`\n`t2_nib`\n`t2_nib_array = t2_nib.get_fdata()`\n`t2_nib_array.shape`", "1482665": "have you checked? because the classification problem has a different number of images and images of different sizes: from 150 to 512", "1483433": "No. of images for all modalities in segmentation dataset is same. I'm not sure about the image size though" }, "source": "meta" }