Before many of us wasting our precious drive storage and TPU runtime, can the host disclose if external datasets like CheXpert, chest14 or MIMIC was allowed for this competition? Thanks in advance!
", "messages": [ { "id": "1149696", "postDate": "01/12/2021 04:29:31", "content": "Before many of us wasting our precious drive storage and TPU runtime, can the host disclose if external datasets like CheXpert, chest14 or MIMIC was allowed for this competition? Thanks in advance!
", "rawMarkdown": "Before many of us wasting our precious drive storage and TPU runtime, can the host disclose if external datasets like CheXpert, chest14 or MIMIC was allowed for this competition? Thanks in advance!", "votes": null }, { "id": "1149981", "postDate": "01/12/2021 09:22:17", "content": "from the rules:
\nchest14 is freely available with the best licence there is (CC0) - you can do anything you want with it.
\nMIMIC/CheXpert on the other hand has more strict licences, and I would personally leave them out.
on the other hand, chest14 is the perfect external data and could be sufficient, as it contains more unhealthy patients - more than half of the photos have cvc/ett/ngt present
", "rawMarkdown": "from the rules:\n\n2. EXTERNAL DATA.\nPublicly, freely available external data is permitted. Entrants may re-annotate images in the training set, however Entrants will (i) ensure the re-annotated data is available to use by all participants of the competition for purposes of the competition at no cost to the other participants and (ii) post such access to the re-annotated data for the participants to the official competition forum prior to the Entry Deadline. Entrants may not hand-label predictions in the test data set, including having human observers rate and evaluate the test data set.\n\nchest14 is freely available with the best licence there is (CC0) - you can do anything you want with it.\nMIMIC/CheXpert on the other hand has more strict licences, and I would personally leave them out.\n\non the other hand, chest14 is the perfect external data and could be sufficient, as it contains more unhealthy patients - more than half of the photos have cvc/ett/ngt present", "votes": null }, { "id": "1150016", "postDate": "01/12/2021 09:48:27", "content": "yes. but in some other competition like SIIM-ACR the rules was:
\n\n\nEXTERNAL DATA: Publicly, freely available external data is permitted. No hand-labeling of the test set.
\n
but the host still allowed chexpert dataset in that competition. so I guess it's still nothing to lose to ask first :D otherwise, i will be happy with chest14
", "rawMarkdown": "yes. but in some other competition like SIIM-ACR the rules was:\n> EXTERNAL DATA: Publicly, freely available external data is permitted. No hand-labeling of the test set.\n\n\nbut the host still allowed chexpert dataset in that competition. so I guess it's still nothing to lose to ask first :D otherwise, i will be happy with chest14", "votes": null }, { "id": "1150132", "postDate": "01/12/2021 11:57:23", "content": "i think it is good to confirm with the host.
\ni recalled some of the previous chest x-ray competitions are using \"external data\" as private data.
sure. but if the host doesnt provide any further information, leave chexpert and mimic aside might a safe move
", "rawMarkdown": "sure. but if the host doesnt provide any further information, leave chexpert and mimic aside might a safe move", "votes": null }, { "id": "1152230", "postDate": "01/14/2021 00:35:28", "content": "Hi, are these links right for these datasets?
\nChest14: https://www.kaggle.com/nih-chest-xrays/data
\nChexpert: https://stanfordmlgroup.github.io/competitions/chexpert/
Yes i download chest14 from the kaggle link above
", "rawMarkdown": "Yes i download chest14 from the kaggle link above", "votes": null }, { "id": "1152268", "postDate": "01/14/2021 01:59:12", "content": "Oh, Thanks :)
", "rawMarkdown": "Oh, Thanks :)", "votes": null }, { "id": "1176303", "postDate": "01/29/2021 15:13:59", "content": "Can we use the chest14 data?
", "rawMarkdown": "Can we use the chest14 data?", "votes": null }, { "id": "1176499", "postDate": "01/29/2021 16:50:56", "content": "Yes yo can
", "rawMarkdown": "Yes yo can", "votes": null }, { "id": "1184189", "postDate": "02/03/2021 12:31:13", "content": "@moewie94 Do you use the chest14 data?Thanks!
", "rawMarkdown": "moewie94 Do you use the chest14 data?Thanks!", "votes": null }, { "id": "1220157", "postDate": "02/27/2021 16:37:42", "content": "@moewie94 does these images contain any of catheter tubes positioned ?
", "rawMarkdown": "moewie94 does these images contain any of catheter tubes positioned ?", "votes": null }, { "id": "1220304", "postDate": "02/27/2021 20:38:22", "content": "It was discussed in some other thread. All of the images for this competition come from the chest14 dataset. It is very large. The chest14 dataset was relabeled for this competition. They must have selected for catheter presence and then ran some labeling on it.
", "rawMarkdown": "It was discussed in some other thread. All of the images for this competition come from the chest14 dataset. It is very large. The chest14 dataset was relabeled for this competition. They must have selected for catheter presence and then ran some labeling on it.", "votes": null }, { "id": "1220305", "postDate": "02/27/2021 20:39:33", "content": "discussed here https://www.kaggle.com/c/ranzcr-clip-catheter-line-classification/discussion/221808
", "rawMarkdown": "discussed here https://www.kaggle.com/c/ranzcr-clip-catheter-line-classification/discussion/221808", "votes": null } ], "comments": [ { "id": 1149981, "author_name": "raddar", "author_url": "", "post_date": "01/12/2021 09:22:17", "content": "from the rules:
\nchest14 is freely available with the best licence there is (CC0) - you can do anything you want with it.
\nMIMIC/CheXpert on the other hand has more strict licences, and I would personally leave them out.
on the other hand, chest14 is the perfect external data and could be sufficient, as it contains more unhealthy patients - more than half of the photos have cvc/ett/ngt present
", "votes": null, "replies": [ { "id": 1150016, "author_name": "moewie94", "author_url": "", "post_date": "01/12/2021 09:48:27", "content": "yes. but in some other competition like SIIM-ACR the rules was:
\n\n\nEXTERNAL DATA: Publicly, freely available external data is permitted. No hand-labeling of the test set.
\n
but the host still allowed chexpert dataset in that competition. so I guess it's still nothing to lose to ask first :D otherwise, i will be happy with chest14
", "votes": null, "replies": [] }, { "id": 1152230, "author_name": "youhanlee", "author_url": "", "post_date": "01/14/2021 00:35:28", "content": "Hi, are these links right for these datasets?
\nChest14: https://www.kaggle.com/nih-chest-xrays/data
\nChexpert: https://stanfordmlgroup.github.io/competitions/chexpert/
Yes i download chest14 from the kaggle link above
", "votes": null, "replies": [] }, { "id": 1152268, "author_name": "youhanlee", "author_url": "", "post_date": "01/14/2021 01:59:12", "content": "Oh, Thanks :)
", "votes": null, "replies": [] }, { "id": 1176303, "author_name": "bcwang", "author_url": "", "post_date": "01/29/2021 15:13:59", "content": "Can we use the chest14 data?
", "votes": null, "replies": [] }, { "id": 1176499, "author_name": "raddar", "author_url": "", "post_date": "01/29/2021 16:50:56", "content": "Yes yo can
", "votes": null, "replies": [] }, { "id": 1184189, "author_name": "bcwang", "author_url": "", "post_date": "02/03/2021 12:31:13", "content": "@moewie94 Do you use the chest14 data?Thanks!
", "votes": null, "replies": [] }, { "id": 1220157, "author_name": "jaideepvalani", "author_url": "", "post_date": "02/27/2021 16:37:42", "content": "@moewie94 does these images contain any of catheter tubes positioned ?
", "votes": null, "replies": [] }, { "id": 1220304, "author_name": "ryches", "author_url": "", "post_date": "02/27/2021 20:38:22", "content": "It was discussed in some other thread. All of the images for this competition come from the chest14 dataset. It is very large. The chest14 dataset was relabeled for this competition. They must have selected for catheter presence and then ran some labeling on it.
", "votes": null, "replies": [] }, { "id": 1220305, "author_name": "ryches", "author_url": "", "post_date": "02/27/2021 20:39:33", "content": "discussed here https://www.kaggle.com/c/ranzcr-clip-catheter-line-classification/discussion/221808
", "votes": null, "replies": [] } ] }, { "id": 1150132, "author_name": "hengck23", "author_url": "", "post_date": "01/12/2021 11:57:23", "content": "i think it is good to confirm with the host.
\ni recalled some of the previous chest x-ray competitions are using \"external data\" as private data.
sure. but if the host doesnt provide any further information, leave chexpert and mimic aside might a safe move
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "1149696": "Before many of us wasting our precious drive storage and TPU runtime, can the host disclose if external datasets like CheXpert, chest14 or MIMIC was allowed for this competition? Thanks in advance!", "1149981": "from the rules:\n\n2. EXTERNAL DATA.\nPublicly, freely available external data is permitted. Entrants may re-annotate images in the training set, however Entrants will (i) ensure the re-annotated data is available to use by all participants of the competition for purposes of the competition at no cost to the other participants and (ii) post such access to the re-annotated data for the participants to the official competition forum prior to the Entry Deadline. Entrants may not hand-label predictions in the test data set, including having human observers rate and evaluate the test data set.\n\nchest14 is freely available with the best licence there is (CC0) - you can do anything you want with it.\nMIMIC/CheXpert on the other hand has more strict licences, and I would personally leave them out.\n\non the other hand, chest14 is the perfect external data and could be sufficient, as it contains more unhealthy patients - more than half of the photos have cvc/ett/ngt present", "1150016": "yes. but in some other competition like SIIM-ACR the rules was:\n> EXTERNAL DATA: Publicly, freely available external data is permitted. No hand-labeling of the test set.\n\n\nbut the host still allowed chexpert dataset in that competition. so I guess it's still nothing to lose to ask first :D otherwise, i will be happy with chest14", "1150132": "i think it is good to confirm with the host.\ni recalled some of the previous chest x-ray competitions are using \"external data\" as private data.", "1151846": "sure. but if the host doesnt provide any further information, leave chexpert and mimic aside might a safe move", "1152230": "Hi, are these links right for these datasets?\n\nChest14: https://www.kaggle.com/nih-chest-xrays/data\nChexpert: https://stanfordmlgroup.github.io/competitions/chexpert/", "1152239": "Yes i download chest14 from the kaggle link above", "1152268": "Oh, Thanks :)", "1176303": "Can we use the chest14 data?", "1176499": "Yes yo can", "1184189": "moewie94 Do you use the chest14 data?Thanks!", "1220157": "moewie94 does these images contain any of catheter tubes positioned ?", "1220304": "It was discussed in some other thread. All of the images for this competition come from the chest14 dataset. It is very large. The chest14 dataset was relabeled for this competition. They must have selected for catheter presence and then ran some labeling on it.", "1220305": "discussed here https://www.kaggle.com/c/ranzcr-clip-catheter-line-classification/discussion/221808" }, "source": "meta" }