Hello fellow Kagglers!
\n\nDo you agree it seems odd that the competition organizers decided to share the metadata for this competition? If I understand @philculliton correctly in this discussion post, we can use this metadata for preprocessing and thus to improve our results in the competition. Curious to see how competitors will use these features in the competition!
\n\nMy question for the organizers: what are the limitations on the usage of metadata, if there are any? Are there any exploits of the metadata that can lead to disqualification?
\n\nP.S.: I've created a kernel that parses most of the features in the metadata (30 features) from the DICOM files to CSV. Feel free to download these CSV files to more conveniently access the metadata.
\n\nLink to kernel:\nhttps://www.kaggle.com/carlolepelaars/converting-dicom-metadata-to-csv-rsna-ihd-2019
", "messages": [ { "id": 633539, "postDate": "2019-09-25T05:11:26.283Z", "content": "You can refer to this link.\nhttps://www.kaggle.com/c/rsna-intracranial-hemorrhage-detection/discussion/109479#632470
\n\n@philculliton says that they're approaching metadata usage on a case-by-case basis.\n If you have a use for the metadata that falls outside of what has already been allowed, please ask, and the host will clarify. \nAll of the provided fields were included so they could be used, but the exact use should be cleared to make sure you're falling within the guidelines of the rules.
", "rawMarkdown": "You can refer to this link.\nhttps://www.kaggle.com/c/rsna-intracranial-hemorrhage-detection/discussion/109479#632470\n\n@philculliton says that they're approaching metadata usage on a case-by-case basis.\n If you have a use for the metadata that falls outside of what has already been allowed, please ask, and the host will clarify. \nAll of the provided fields were included so they could be used, but the exact use should be cleared to make sure you're falling within the guidelines of the rules.", "votes": 3 }, { "id": 633429, "postDate": "2019-09-24T23:01:48.567Z", "content": "Hello fellow Kagglers!
\n\nDo you agree it seems odd that the competition organizers decided to share the metadata for this competition? If I understand @philculliton correctly in this discussion post, we can use this metadata for preprocessing and thus to improve our results in the competition. Curious to see how competitors will use these features in the competition!
\n\nMy question for the organizers: what are the limitations on the usage of metadata, if there are any? Are there any exploits of the metadata that can lead to disqualification?
\n\nP.S.: I've created a kernel that parses most of the features in the metadata (30 features) from the DICOM files to CSV. Feel free to download these CSV files to more conveniently access the metadata.
\n\nLink to kernel:\nhttps://www.kaggle.com/carlolepelaars/converting-dicom-metadata-to-csv-rsna-ihd-2019
", "rawMarkdown": "Hello fellow Kagglers!\n\nDo you agree it seems odd that the competition organizers decided to share the metadata for this competition? If I understand @philculliton correctly in [this discussion post](https://www.kaggle.com/c/rsna-intracranial-hemorrhage-detection/discussion/109969#latest-633378), we can use this metadata for preprocessing and thus to improve our results in the competition. Curious to see how competitors will use these features in the competition!\n\nMy question for the organizers: what are the limitations on the usage of metadata, if there are any? Are there any exploits of the metadata that can lead to disqualification?\n\nP.S.: I've created a kernel that parses most of the features in the metadata (30 features) from the DICOM files to CSV. Feel free to download these CSV files to more conveniently access the metadata.\n\nLink to kernel:\nhttps://www.kaggle.com/carlolepelaars/converting-dicom-metadata-to-csv-rsna-ihd-2019\n\n", "votes": 3 } ], "comments": [ { "id": 633539, "author_name": "Zehao Zhao", "author_url": "", "post_date": "2019-09-25T05:11:26.283000", "content": "You can refer to this link.\nhttps://www.kaggle.com/c/rsna-intracranial-hemorrhage-detection/discussion/109479#632470
\n\n@philculliton says that they're approaching metadata usage on a case-by-case basis.\n If you have a use for the metadata that falls outside of what has already been allowed, please ask, and the host will clarify. \nAll of the provided fields were included so they could be used, but the exact use should be cleared to make sure you're falling within the guidelines of the rules.
", "votes": 3, "replies": [] } ], "raw_markdown_by_id": { "633539": "You can refer to this link.\nhttps://www.kaggle.com/c/rsna-intracranial-hemorrhage-detection/discussion/109479#632470\n\n@philculliton says that they're approaching metadata usage on a case-by-case basis.\n If you have a use for the metadata that falls outside of what has already been allowed, please ask, and the host will clarify. \nAll of the provided fields were included so they could be used, but the exact use should be cleared to make sure you're falling within the guidelines of the rules.", "633429": "Hello fellow Kagglers!\n\nDo you agree it seems odd that the competition organizers decided to share the metadata for this competition? If I understand @philculliton correctly in [this discussion post](https://www.kaggle.com/c/rsna-intracranial-hemorrhage-detection/discussion/109969#latest-633378), we can use this metadata for preprocessing and thus to improve our results in the competition. Curious to see how competitors will use these features in the competition!\n\nMy question for the organizers: what are the limitations on the usage of metadata, if there are any? Are there any exploits of the metadata that can lead to disqualification?\n\nP.S.: I've created a kernel that parses most of the features in the metadata (30 features) from the DICOM files to CSV. Feel free to download these CSV files to more conveniently access the metadata.\n\nLink to kernel:\nhttps://www.kaggle.com/carlolepelaars/converting-dicom-metadata-to-csv-rsna-ihd-2019\n\n" } }