Even though all basic-science use creative language to argue that their work will eventually help cure cancer, I find this portion of description to be a big-time overreach:
\n\n**We’ve all seen people suffer from diseases like cancer, heart disease, chronic obstructive pulmonary disease, Alzheimer’s, and diabetes. Many have seen their loved ones pass away. Think how many lives would be transformed if cures came faster.
\n\nBy automating nucleus detection, you could help unlock cures faster—from rare disorders to the common cold.**
\n\nSure, everything will help cure cancer eventually, but some things are more likely to help than others. To be clear: I am not saying that finding nuclei automatically is not worth doing. But if you are into curing cancer, this and this may get you there a little bit faster. The latter also happens to be a game.
", "messages": [ { "id": "269540", "postDate": "01/17/2018 01:12:01", "content": "Even though all basic-science use creative language to argue that their work will eventually help cure cancer, I find this portion of description to be a big-time overreach:
\n\n**We’ve all seen people suffer from diseases like cancer, heart disease, chronic obstructive pulmonary disease, Alzheimer’s, and diabetes. Many have seen their loved ones pass away. Think how many lives would be transformed if cures came faster.
\n\nBy automating nucleus detection, you could help unlock cures faster—from rare disorders to the common cold.**
\n\nSure, everything will help cure cancer eventually, but some things are more likely to help than others. To be clear: I am not saying that finding nuclei automatically is not worth doing. But if you are into curing cancer, this and this may get you there a little bit faster. The latter also happens to be a game.
", "rawMarkdown": "Even though all basic-science use creative language to argue that their work will eventually help cure cancer, I find this portion of description to be a big-time overreach:\n\n**We’ve all seen people suffer from diseases like cancer, heart disease, chronic obstructive pulmonary disease, Alzheimer’s, and diabetes. Many have seen their loved ones pass away. Think how many lives would be transformed if cures came faster.\n\nBy automating nucleus detection, you could help unlock cures faster—from rare disorders to the common cold.**\n\nSure, everything will help cure cancer eventually, but some things are more likely to help than others. To be clear: I am not saying that finding nuclei automatically is not worth doing. But if you are into curing cancer, [__this__](https://en.wikipedia.org/wiki/Rosetta@home) and [__this__](https://fold.it/) may get you there a little bit faster. The latter also happens to be a game.", "votes": null }, { "id": "269545", "postDate": "01/17/2018 01:18:47", "content": "I agree, I feel that they have overreached in their attempt to hype up the competition. The jump from nucleus detection -> cure formation is rather big. Further, curing cancer is a great goal given that cancer isn't a single disease but a classification for a plethora of diseases.
\n\nThat being said, boring problems occasionally produce unexpected breakthroughs. I'm excited to see how far we can go as a community and what we might possibly find.
", "rawMarkdown": "I agree, I feel that they have overreached in their attempt to hype up the competition. The jump from nucleus detection -> cure formation is rather big. Further, curing cancer is a great goal given that cancer isn't a single disease but a classification for a plethora of diseases.\n\n That being said, boring problems occasionally produce unexpected breakthroughs. I'm excited to see how far we can go as a community and what we might possibly find.", "votes": null }, { "id": "269850", "postDate": "01/17/2018 11:56:47", "content": "I agree with both of you. For some people it is hyped up and for (the majority) it's the right level of motivation.\nEven fold.it attracts asks users to \"solve puzzles for science\" which is quite general (although each individual puzzle is more specific).
", "rawMarkdown": "I agree with both of you. For some people it is hyped up and for (the majority) it's the right level of motivation.\nEven fold.it attracts asks users to \"solve puzzles for science\" which is quite general (although each individual puzzle is more specific).", "votes": null }, { "id": "269859", "postDate": "01/17/2018 12:16:25", "content": "You pose a very fair question! It is a challenge to describe the challenge in an understandable but detailed way. But I can assure you solving this problem will accelerate research across all disease areas.
\n\nIf you would like to dig a little deeper, I made a 5 minute video that might be helpful to explain more: https://www.youtube.com/watch?v=Dbiq6l50zO8
\n\nAs well, you can look at successes resulting from classical image processing techniques; a more accurate/easier solution would simply accelerate these: http://cellprofiler.org/impact Of course, these are just a few examples we are aware of that result from the use of our lab's open-source software but there are dozens of examples of drugs that came from microscopy experiments where nucleus detection is the first step.
\n\nHope this clarifies!
", "rawMarkdown": "You pose a very fair question! It is a challenge to describe the challenge in an understandable but detailed way. But I can assure you solving this problem will accelerate research across all disease areas.\n\nIf you would like to dig a little deeper, I made a 5 minute video that might be helpful to explain more: https://www.youtube.com/watch?v=Dbiq6l50zO8\n\nAs well, you can look at successes resulting from classical image processing techniques; a more accurate/easier solution would simply accelerate these: http://cellprofiler.org/impact Of course, these are just a few examples we are aware of that result from the use of our lab's open-source software but there are dozens of examples of drugs that came from microscopy experiments where nucleus detection is the first step.\n\nHope this clarifies!", "votes": null }, { "id": "269905", "postDate": "01/17/2018 14:04:06", "content": "Thanks for your kind video description. This video has given me enough motivation to compete in this competition.
", "rawMarkdown": "Thanks for your kind video description. This video has given me enough motivation to compete in this competition.", "votes": null }, { "id": "269931", "postDate": "01/17/2018 14:43:25", "content": "Thank you for the detailed explanation.i feel more connected to the problem now.
", "rawMarkdown": "Thank you for the detailed explanation.i feel more connected to the problem now.", "votes": null }, { "id": "270136", "postDate": "01/17/2018 19:42:27", "content": "Thank you for the video, it is very helpful.
", "rawMarkdown": "Thank you for the video, it is very helpful.", "votes": null }, { "id": "270147", "postDate": "01/17/2018 20:30:54", "content": "@Anne Carpenter All NIH-funded research is biomedical and aimed at improving human health, but you and I both know that it takes a lot of extrapolation to make connections between a good chunk of that research and human health. Nothing wrong with that - I am also partial to basic science!
\n\nMy point was that the competition description oversells the immediate impact of nuclei detection on cancer research. Nothing wrong with that either if one has proper background to appreciate the well-intended hyperbole in those statements.
", "rawMarkdown": "Anne Carpenter All NIH-funded research is biomedical and aimed at improving human health, but you and I both know that it takes a lot of extrapolation to make connections between a good chunk of that research and human health. Nothing wrong with that - I am also partial to basic science!\n\nMy point was that the competition description oversells the immediate impact of nuclei detection on cancer research. Nothing wrong with that either if one has proper background to appreciate the well-intended hyperbole in those statements.", "votes": null }, { "id": "270166", "postDate": "01/17/2018 21:21:20", "content": "Hi Anne, can you add your video to sticky External Data thread? So it will be always available and easy to find.
", "rawMarkdown": "Hi Anne, can you add your video to sticky External Data thread? So it will be always available and easy to find.", "votes": null }, { "id": "270304", "postDate": "01/18/2018 03:20:16", "content": "A more appropriate human health outcome might be, \"By automating nucleus detection, you could help sad and tired graduate students get out of the lab faster, improving their quality of life and reducing depression rates!\"
", "rawMarkdown": "A more appropriate human health outcome might be, \"By automating nucleus detection, you could help sad and tired graduate students get out of the lab faster, improving their quality of life and reducing depression rates!\"", "votes": null }, { "id": "270388", "postDate": "01/18/2018 07:12:40", "content": "@Anne Carpenter I think your video should be linked from the Description page of this competition. It really clarifies some of its goals.
", "rawMarkdown": "Anne Carpenter I think your video should be linked from the Description page of this competition. It really clarifies some of its goals.", "votes": null }, { "id": "270391", "postDate": "01/18/2018 07:14:12", "content": "Thanks for your explanation, Dr. Carpenter!
", "rawMarkdown": "Thanks for your explanation, Dr. Carpenter!", "votes": null }, { "id": "271047", "postDate": "01/19/2018 15:31:39", "content": "That will absolutely be the most immediate outcome! \nA generation of students offers their thanks in advance for your work on this competition :)
", "rawMarkdown": "That will *absolutely* be the most immediate outcome! \nA generation of students offers their thanks in advance for your work on this competition :)", "votes": null }, { "id": "271049", "postDate": "01/19/2018 15:33:10", "content": "Thanks for the suggestion, the team plans to post the video somewhere more obvious, I suspect the Description page.
", "rawMarkdown": "Thanks for the suggestion, the team plans to post the video somewhere more obvious, I suspect the Description page.", "votes": null }, { "id": "271051", "postDate": "01/19/2018 15:38:26", "content": "But I should emphasize that the cancer connection is not a strained very-long-term, basic-research one (though I agree with Tilii that basic research is crucial for our future).
\n\nFor example, our current software CellProfiler is being used in a clinical trial that has already apparently saved a few patients:
\n\nPaper (OA): http://www.thelancet.com/journals/lanhae/article/PIIS2352-3026(17)30208-9/fulltext\nComment: http://www.thelancet.com/journals/lanhae/article/PIIS2352-3026(17)30213-2/fulltext
\n\nImagine the impact of an improved version! On this and tens of thousands of other biomedical projects around the world.
", "rawMarkdown": "But I should emphasize that the cancer connection is not a strained very-long-term, basic-research one (though I agree with Tilii that basic research is crucial for our future). \n\nFor example, our current software CellProfiler is being used in a clinical trial that has already apparently saved a few patients: \n\nPaper (OA): http://www.thelancet.com/journals/lanhae/article/PIIS2352-3026(17)30208-9/fulltext\nComment: http://www.thelancet.com/journals/lanhae/article/PIIS2352-3026(17)30213-2/fulltext\n\nImagine the impact of an improved version! On this and tens of thousands of other biomedical projects around the world.", "votes": null }, { "id": "271172", "postDate": "01/19/2018 20:33:42", "content": "completely agree. There is also another gap: are people more interested in identifying cells or nuclei?
", "rawMarkdown": "completely agree. There is also another gap: are people more interested in identifying cells or nuclei?", "votes": null }, { "id": "526436", "postDate": "05/03/2019 04:06:23", "content": "I actually have a much simpler way of curing cancer using machine learning, but we lack the specific dataset needed unfortunately. Also I don't feel like dying any time soon, and everyone who finds a \"cure for cancer\" seems to mysteriously die soon after. People can disagree but facts are facts and shouldn't be ignored.
", "rawMarkdown": "I actually have a much simpler way of curing cancer using machine learning, but we lack the specific dataset needed unfortunately. Also I don't feel like dying any time soon, and everyone who finds a \"cure for cancer\" seems to mysteriously die soon after. People can disagree but facts are facts and shouldn't be ignored.", "votes": null }, { "id": "609617", "postDate": "08/27/2019 23:35:06", "content": "Hi Anne, i know this is an old thread and maybe no one will read it. But i just wanted to say thank you for sharing those link. I'm still new to kaggle and cancer research (i come from an cloud architecture background and have some average experience with deep learning). This will help me understand cancer research better. Thanks for sharing those links.
\n\ncheers.
", "rawMarkdown": "Hi Anne, i know this is an old thread and maybe no one will read it. But i just wanted to say thank you for sharing those link. I'm still new to kaggle and cancer research (i come from an cloud architecture background and have some average experience with deep learning). This will help me understand cancer research better. Thanks for sharing those links.\n\ncheers.", "votes": null }, { "id": "609653", "postDate": "08/28/2019 01:04:32", "content": "glad to hear it!
", "rawMarkdown": "glad to hear it!", "votes": null } ], "comments": [ { "id": 269545, "author_name": "jzerox2", "author_url": "", "post_date": "01/17/2018 01:18:47", "content": "I agree, I feel that they have overreached in their attempt to hype up the competition. The jump from nucleus detection -> cure formation is rather big. Further, curing cancer is a great goal given that cancer isn't a single disease but a classification for a plethora of diseases.
\n\nThat being said, boring problems occasionally produce unexpected breakthroughs. I'm excited to see how far we can go as a community and what we might possibly find.
", "votes": null, "replies": [] }, { "id": 269850, "author_name": "fabsta", "author_url": "", "post_date": "01/17/2018 11:56:47", "content": "I agree with both of you. For some people it is hyped up and for (the majority) it's the right level of motivation.\nEven fold.it attracts asks users to \"solve puzzles for science\" which is quite general (although each individual puzzle is more specific).
", "votes": null, "replies": [] }, { "id": 269859, "author_name": "drannecarpenter", "author_url": "", "post_date": "01/17/2018 12:16:25", "content": "You pose a very fair question! It is a challenge to describe the challenge in an understandable but detailed way. But I can assure you solving this problem will accelerate research across all disease areas.
\n\nIf you would like to dig a little deeper, I made a 5 minute video that might be helpful to explain more: https://www.youtube.com/watch?v=Dbiq6l50zO8
\n\nAs well, you can look at successes resulting from classical image processing techniques; a more accurate/easier solution would simply accelerate these: http://cellprofiler.org/impact Of course, these are just a few examples we are aware of that result from the use of our lab's open-source software but there are dozens of examples of drugs that came from microscopy experiments where nucleus detection is the first step.
\n\nHope this clarifies!
", "votes": null, "replies": [ { "id": 269905, "author_name": "ttagu99", "author_url": "", "post_date": "01/17/2018 14:04:06", "content": "Thanks for your kind video description. This video has given me enough motivation to compete in this competition.
", "votes": null, "replies": [] }, { "id": 269931, "author_name": "ravivadapalli", "author_url": "", "post_date": "01/17/2018 14:43:25", "content": "Thank you for the detailed explanation.i feel more connected to the problem now.
", "votes": null, "replies": [] }, { "id": 270136, "author_name": "serhiy", "author_url": "", "post_date": "01/17/2018 19:42:27", "content": "Thank you for the video, it is very helpful.
", "votes": null, "replies": [] }, { "id": 270147, "author_name": "tilii7", "author_url": "", "post_date": "01/17/2018 20:30:54", "content": "@Anne Carpenter All NIH-funded research is biomedical and aimed at improving human health, but you and I both know that it takes a lot of extrapolation to make connections between a good chunk of that research and human health. Nothing wrong with that - I am also partial to basic science!
\n\nMy point was that the competition description oversells the immediate impact of nuclei detection on cancer research. Nothing wrong with that either if one has proper background to appreciate the well-intended hyperbole in those statements.
", "votes": null, "replies": [] }, { "id": 270166, "author_name": "serhiy", "author_url": "", "post_date": "01/17/2018 21:21:20", "content": "Hi Anne, can you add your video to sticky External Data thread? So it will be always available and easy to find.
", "votes": null, "replies": [] }, { "id": 270388, "author_name": "pauljurczak", "author_url": "", "post_date": "01/18/2018 07:12:40", "content": "@Anne Carpenter I think your video should be linked from the Description page of this competition. It really clarifies some of its goals.
", "votes": null, "replies": [] }, { "id": 270391, "author_name": "u39kun", "author_url": "", "post_date": "01/18/2018 07:14:12", "content": "Thanks for your explanation, Dr. Carpenter!
", "votes": null, "replies": [] }, { "id": 271049, "author_name": "drannecarpenter", "author_url": "", "post_date": "01/19/2018 15:33:10", "content": "Thanks for the suggestion, the team plans to post the video somewhere more obvious, I suspect the Description page.
", "votes": null, "replies": [] }, { "id": 609617, "author_name": "manuklinkz", "author_url": "", "post_date": "08/27/2019 23:35:06", "content": "Hi Anne, i know this is an old thread and maybe no one will read it. But i just wanted to say thank you for sharing those link. I'm still new to kaggle and cancer research (i come from an cloud architecture background and have some average experience with deep learning). This will help me understand cancer research better. Thanks for sharing those links.
\n\ncheers.
", "votes": null, "replies": [] }, { "id": 609653, "author_name": "drannecarpenter", "author_url": "", "post_date": "08/28/2019 01:04:32", "content": "glad to hear it!
", "votes": null, "replies": [] } ] }, { "id": 270304, "author_name": "stkbailey", "author_url": "", "post_date": "01/18/2018 03:20:16", "content": "A more appropriate human health outcome might be, \"By automating nucleus detection, you could help sad and tired graduate students get out of the lab faster, improving their quality of life and reducing depression rates!\"
", "votes": null, "replies": [ { "id": 271047, "author_name": "drannecarpenter", "author_url": "", "post_date": "01/19/2018 15:31:39", "content": "That will absolutely be the most immediate outcome! \nA generation of students offers their thanks in advance for your work on this competition :)
", "votes": null, "replies": [] }, { "id": 271051, "author_name": "drannecarpenter", "author_url": "", "post_date": "01/19/2018 15:38:26", "content": "But I should emphasize that the cancer connection is not a strained very-long-term, basic-research one (though I agree with Tilii that basic research is crucial for our future).
\n\nFor example, our current software CellProfiler is being used in a clinical trial that has already apparently saved a few patients:
\n\nPaper (OA): http://www.thelancet.com/journals/lanhae/article/PIIS2352-3026(17)30208-9/fulltext\nComment: http://www.thelancet.com/journals/lanhae/article/PIIS2352-3026(17)30213-2/fulltext
\n\nImagine the impact of an improved version! On this and tens of thousands of other biomedical projects around the world.
", "votes": null, "replies": [] } ] }, { "id": 271172, "author_name": "jeffluo", "author_url": "", "post_date": "01/19/2018 20:33:42", "content": "completely agree. There is also another gap: are people more interested in identifying cells or nuclei?
", "votes": null, "replies": [] }, { "id": 526436, "author_name": "wfilkins", "author_url": "", "post_date": "05/03/2019 04:06:23", "content": "I actually have a much simpler way of curing cancer using machine learning, but we lack the specific dataset needed unfortunately. Also I don't feel like dying any time soon, and everyone who finds a \"cure for cancer\" seems to mysteriously die soon after. People can disagree but facts are facts and shouldn't be ignored.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "269540": "Even though all basic-science use creative language to argue that their work will eventually help cure cancer, I find this portion of description to be a big-time overreach:\n\n**We’ve all seen people suffer from diseases like cancer, heart disease, chronic obstructive pulmonary disease, Alzheimer’s, and diabetes. Many have seen their loved ones pass away. Think how many lives would be transformed if cures came faster.\n\nBy automating nucleus detection, you could help unlock cures faster—from rare disorders to the common cold.**\n\nSure, everything will help cure cancer eventually, but some things are more likely to help than others. To be clear: I am not saying that finding nuclei automatically is not worth doing. But if you are into curing cancer, [__this__](https://en.wikipedia.org/wiki/Rosetta@home) and [__this__](https://fold.it/) may get you there a little bit faster. The latter also happens to be a game.", "269545": "I agree, I feel that they have overreached in their attempt to hype up the competition. The jump from nucleus detection -> cure formation is rather big. Further, curing cancer is a great goal given that cancer isn't a single disease but a classification for a plethora of diseases.\n\n That being said, boring problems occasionally produce unexpected breakthroughs. I'm excited to see how far we can go as a community and what we might possibly find.", "269850": "I agree with both of you. For some people it is hyped up and for (the majority) it's the right level of motivation.\nEven fold.it attracts asks users to \"solve puzzles for science\" which is quite general (although each individual puzzle is more specific).", "269859": "You pose a very fair question! It is a challenge to describe the challenge in an understandable but detailed way. But I can assure you solving this problem will accelerate research across all disease areas.\n\nIf you would like to dig a little deeper, I made a 5 minute video that might be helpful to explain more: https://www.youtube.com/watch?v=Dbiq6l50zO8\n\nAs well, you can look at successes resulting from classical image processing techniques; a more accurate/easier solution would simply accelerate these: http://cellprofiler.org/impact Of course, these are just a few examples we are aware of that result from the use of our lab's open-source software but there are dozens of examples of drugs that came from microscopy experiments where nucleus detection is the first step.\n\nHope this clarifies!", "269905": "Thanks for your kind video description. This video has given me enough motivation to compete in this competition.", "269931": "Thank you for the detailed explanation.i feel more connected to the problem now.", "270136": "Thank you for the video, it is very helpful.", "270147": "Anne Carpenter All NIH-funded research is biomedical and aimed at improving human health, but you and I both know that it takes a lot of extrapolation to make connections between a good chunk of that research and human health. Nothing wrong with that - I am also partial to basic science!\n\nMy point was that the competition description oversells the immediate impact of nuclei detection on cancer research. Nothing wrong with that either if one has proper background to appreciate the well-intended hyperbole in those statements.", "270166": "Hi Anne, can you add your video to sticky External Data thread? So it will be always available and easy to find.", "270304": "A more appropriate human health outcome might be, \"By automating nucleus detection, you could help sad and tired graduate students get out of the lab faster, improving their quality of life and reducing depression rates!\"", "270388": "Anne Carpenter I think your video should be linked from the Description page of this competition. It really clarifies some of its goals.", "270391": "Thanks for your explanation, Dr. Carpenter!", "271047": "That will *absolutely* be the most immediate outcome! \nA generation of students offers their thanks in advance for your work on this competition :)", "271049": "Thanks for the suggestion, the team plans to post the video somewhere more obvious, I suspect the Description page.", "271051": "But I should emphasize that the cancer connection is not a strained very-long-term, basic-research one (though I agree with Tilii that basic research is crucial for our future). \n\nFor example, our current software CellProfiler is being used in a clinical trial that has already apparently saved a few patients: \n\nPaper (OA): http://www.thelancet.com/journals/lanhae/article/PIIS2352-3026(17)30208-9/fulltext\nComment: http://www.thelancet.com/journals/lanhae/article/PIIS2352-3026(17)30213-2/fulltext\n\nImagine the impact of an improved version! On this and tens of thousands of other biomedical projects around the world.", "271172": "completely agree. There is also another gap: are people more interested in identifying cells or nuclei?", "526436": "I actually have a much simpler way of curing cancer using machine learning, but we lack the specific dataset needed unfortunately. Also I don't feel like dying any time soon, and everyone who finds a \"cure for cancer\" seems to mysteriously die soon after. People can disagree but facts are facts and shouldn't be ignored.", "609617": "Hi Anne, i know this is an old thread and maybe no one will read it. But i just wanted to say thank you for sharing those link. I'm still new to kaggle and cancer research (i come from an cloud architecture background and have some average experience with deep learning). This will help me understand cancer research better. Thanks for sharing those links.\n\ncheers.", "609653": "glad to hear it!" }, "source": "meta" }