{
  "id": 161595,
  "title": "Additional prizes, scientific paper, and MICCAI workshop",
  "url": "/competitions/prostate-cancer-grade-assessment/discussion/161595",
  "author_name": "Wouter Bulten",
  "post_date": "2020-06-25T10:49:32.149000",
  "votes": 35,
  "comment_count": 19,
  "views": 0,
  "content": "<p>With the PANDA challenge well underway, we would like to update you on the scientific part of this project. This competition is part of an ongoing research project and has an associated workshop at <a href=\"https://www.miccai2020.org/en/\">MICCAI</a>. Our goal in hosting this competition was to stimulate new ideas and focus on the clinical applicability of methods. Extend your work on this competition and you may be a co-author on our scientific paper, a presenter at MICCAI, and/or you may win additional prize money for your ideas! </p>\n\n<h1>Three new awards</h1>\n\n<p>As part of the PANDA workshop, we are introducing three additional awards on top of the competition prizes. The three new awards are:</p>\n\n<p><strong>Most innovative method award (€ 1000, runner-up € 250)</strong>\nWe want to incentivize new and innovative methods for automatically grading prostate cancer. Based on the write-ups and code we will pick one winner and a runner-up. This award is not necessarily linked to the leaderboard score. </p>\n\n<p><strong>Best video presentation award (€ 500, runner-up € 250)</strong>\nDuring the PANDA workshops, presentations will be graded based on overall presentation quality, clarity, presentation skills, etc. A jury will decide on the best presentation and a runner-up.</p>\n\n<p><strong>Best generalization performance award (€ 1000, runner-up € 250)</strong>\nAs part of the scientific paper, methods will be evaluated on external unseen data. In the end, the efforts of this competition should go towards building models that can work in clinical practice. The method that achieves the best performance across all datasets will win. This unseen data will be in the same format as the competition’s test set but differences in staining and scanners are to be expected. </p>\n\n<p>Prizes will be awarded by the challenge organizers. To be eligible for winning one of the prizes you have to sign up and be selected as part of the PANDA Consortium.</p>\n\n<h1>Scientific paper on the PANDA challenge</h1>\n\n<p>Together with you, we want to write a scientific paper on the results of the competition. The aim of this paper is to directly contribute to the scientific field of prostate cancer diagnosis and computational pathology. A selection of topics we will discuss in the paper:</p>\n\n<ul>\n<li>Methods used to solve the competition;</li>\n<li>Performance of the models on the private test set;</li>\n<li>Performance of the models on unseen data (not part of the private/public leaderboard);</li>\n<li>What is needed for the clinical application of these methods?</li>\n<li>What are the pitfalls and common problems that we encountered?</li>\n</ul>\n\n<p>From all teams that signup before the deadline, we will invite up to 15 teams to join the PANDA consortium (max. 5 persons per team). The organizers will make the final selection based on the leaderboard score and the write-up of the method. Notebooks submitted will be evaluated and should be your own work. You can use methods from public notebooks but must show that you have significantly improved upon that method. Your notebook will be judged on how it is uniquely suited to solve this problem. The precise selection criteria may be subject to changes and final decisions are made by the competition hosts.</p>\n\n<p>If your team is selected, members of the team will be given authorship as part of the PANDA consortium. (see the <a href=\"https://jamanetwork.com/journals/jama/fullarticle/2665774\">CAMELYON paper</a> for an example of how we mention authors). Authorships are indexed in, e.g., PubMed. As part of the consortium, you are also in the race for the awards.</p>\n\n<h1>Present your work at our workshop</h1>\n\n<p>The PANDA workshop at the international <a href=\"https://www.miccai2020.org/en/\">MICCAI</a> 2020 conference is a great opportunity for you to showcase your work to a larger audience. The presentations will be chosen based on the method used and on competition ranking. We would like to have a wide variety of topics presented.</p>\n\n<p>Some details regarding the presentations:</p>\n\n<ul>\n<li>MICCAI 2020 will be a fully virtual conference due to COVID-19. Workshop participants will be reimbursed for one satellite-event registration fee per team.</li>\n<li>The PANDA workshop takes place on the 8th of October 2020 (specific time is to be determined). Presentations may be pre-recorded. On the workshop day itself, there will be a live moderated session with all presenters and the organizers.</li>\n<li>Each presentation is 20 minutes with an additional 5 minutes for questions. </li>\n</ul>\n\n<h1>How to sign up</h1>\n\n<p>You can signup by filling out this form: <a href=\"https://forms.gle/SL5AmMUgHDthxDeAA\">https://forms.gle/SL5AmMUgHDthxDeAA</a>. You can signup today or wait until the competition ends. After submission, you can still view and edit your information until the deadline. </p>\n\n<p><strong>The deadline to sign-up is the 31st of July</strong>. After the deadline, we will approach you through the email address you filled in on the form if we want to learn more about your approach.</p>\n\n<p><strong>Note:</strong> The inference notebook you link in the form should contain the code you used to generate the final private leaderboard score.</p>\n\n<p>Share your questions here, we hope that you will sign up and share your unique approaches in this very important work! </p>\n\n<p><strong>Update: We have extended the deadline for signup until the 31st of July to give teams some extra time to finalize their submission.</strong></p>",
  "messages": [
    {
      "id": 901263,
      "postDate": "2020-06-25T10:49:32.150Z",
      "content": "<p>With the PANDA challenge well underway, we would like to update you on the scientific part of this project. This competition is part of an ongoing research project and has an associated workshop at <a href=\"https://www.miccai2020.org/en/\">MICCAI</a>. Our goal in hosting this competition was to stimulate new ideas and focus on the clinical applicability of methods. Extend your work on this competition and you may be a co-author on our scientific paper, a presenter at MICCAI, and/or you may win additional prize money for your ideas! </p>\n\n<h1>Three new awards</h1>\n\n<p>As part of the PANDA workshop, we are introducing three additional awards on top of the competition prizes. The three new awards are:</p>\n\n<p><strong>Most innovative method award (€ 1000, runner-up € 250)</strong>\nWe want to incentivize new and innovative methods for automatically grading prostate cancer. Based on the write-ups and code we will pick one winner and a runner-up. This award is not necessarily linked to the leaderboard score. </p>\n\n<p><strong>Best video presentation award (€ 500, runner-up € 250)</strong>\nDuring the PANDA workshops, presentations will be graded based on overall presentation quality, clarity, presentation skills, etc. A jury will decide on the best presentation and a runner-up.</p>\n\n<p><strong>Best generalization performance award (€ 1000, runner-up € 250)</strong>\nAs part of the scientific paper, methods will be evaluated on external unseen data. In the end, the efforts of this competition should go towards building models that can work in clinical practice. The method that achieves the best performance across all datasets will win. This unseen data will be in the same format as the competition’s test set but differences in staining and scanners are to be expected. </p>\n\n<p>Prizes will be awarded by the challenge organizers. To be eligible for winning one of the prizes you have to sign up and be selected as part of the PANDA Consortium.</p>\n\n<h1>Scientific paper on the PANDA challenge</h1>\n\n<p>Together with you, we want to write a scientific paper on the results of the competition. The aim of this paper is to directly contribute to the scientific field of prostate cancer diagnosis and computational pathology. A selection of topics we will discuss in the paper:</p>\n\n<ul>\n<li>Methods used to solve the competition;</li>\n<li>Performance of the models on the private test set;</li>\n<li>Performance of the models on unseen data (not part of the private/public leaderboard);</li>\n<li>What is needed for the clinical application of these methods?</li>\n<li>What are the pitfalls and common problems that we encountered?</li>\n</ul>\n\n<p>From all teams that signup before the deadline, we will invite up to 15 teams to join the PANDA consortium (max. 5 persons per team). The organizers will make the final selection based on the leaderboard score and the write-up of the method. Notebooks submitted will be evaluated and should be your own work. You can use methods from public notebooks but must show that you have significantly improved upon that method. Your notebook will be judged on how it is uniquely suited to solve this problem. The precise selection criteria may be subject to changes and final decisions are made by the competition hosts.</p>\n\n<p>If your team is selected, members of the team will be given authorship as part of the PANDA consortium. (see the <a href=\"https://jamanetwork.com/journals/jama/fullarticle/2665774\">CAMELYON paper</a> for an example of how we mention authors). Authorships are indexed in, e.g., PubMed. As part of the consortium, you are also in the race for the awards.</p>\n\n<h1>Present your work at our workshop</h1>\n\n<p>The PANDA workshop at the international <a href=\"https://www.miccai2020.org/en/\">MICCAI</a> 2020 conference is a great opportunity for you to showcase your work to a larger audience. The presentations will be chosen based on the method used and on competition ranking. We would like to have a wide variety of topics presented.</p>\n\n<p>Some details regarding the presentations:</p>\n\n<ul>\n<li>MICCAI 2020 will be a fully virtual conference due to COVID-19. Workshop participants will be reimbursed for one satellite-event registration fee per team.</li>\n<li>The PANDA workshop takes place on the 8th of October 2020 (specific time is to be determined). Presentations may be pre-recorded. On the workshop day itself, there will be a live moderated session with all presenters and the organizers.</li>\n<li>Each presentation is 20 minutes with an additional 5 minutes for questions. </li>\n</ul>\n\n<h1>How to sign up</h1>\n\n<p>You can signup by filling out this form: <a href=\"https://forms.gle/SL5AmMUgHDthxDeAA\">https://forms.gle/SL5AmMUgHDthxDeAA</a>. You can signup today or wait until the competition ends. After submission, you can still view and edit your information until the deadline. </p>\n\n<p><strong>The deadline to sign-up is the 31st of July</strong>. After the deadline, we will approach you through the email address you filled in on the form if we want to learn more about your approach.</p>\n\n<p><strong>Note:</strong> The inference notebook you link in the form should contain the code you used to generate the final private leaderboard score.</p>\n\n<p>Share your questions here, we hope that you will sign up and share your unique approaches in this very important work! </p>\n\n<p><strong>Update: We have extended the deadline for signup until the 31st of July to give teams some extra time to finalize their submission.</strong></p>",
      "rawMarkdown": "With the PANDA challenge well underway, we would like to update you on the scientific part of this project. This competition is part of an ongoing research project and has an associated workshop at [MICCAI](https://www.miccai2020.org/en/). Our goal in hosting this competition was to stimulate new ideas and focus on the clinical applicability of methods. Extend your work on this competition and you may be a co-author on our scientific paper, a presenter at MICCAI, and/or you may win additional prize money for your ideas! \n\n# Three new awards\nAs part of the PANDA workshop, we are introducing three additional awards on top of the competition prizes. The three new awards are:\n\n**Most innovative method award (€ 1000, runner-up € 250)**\nWe want to incentivize new and innovative methods for automatically grading prostate cancer. Based on the write-ups and code we will pick one winner and a runner-up. This award is not necessarily linked to the leaderboard score. \n\n**Best video presentation award (€ 500, runner-up € 250)**\nDuring the PANDA workshops, presentations will be graded based on overall presentation quality, clarity, presentation skills, etc. A jury will decide on the best presentation and a runner-up.\n\n**Best generalization performance award (€ 1000, runner-up € 250)**\nAs part of the scientific paper, methods will be evaluated on external unseen data. In the end, the efforts of this competition should go towards building models that can work in clinical practice. The method that achieves the best performance across all datasets will win. This unseen data will be in the same format as the competition’s test set but differences in staining and scanners are to be expected. \n\nPrizes will be awarded by the challenge organizers. To be eligible for winning one of the prizes you have to sign up and be selected as part of the PANDA Consortium.\n\n# Scientific paper on the PANDA challenge\n\nTogether with you, we want to write a scientific paper on the results of the competition. The aim of this paper is to directly contribute to the scientific field of prostate cancer diagnosis and computational pathology. A selection of topics we will discuss in the paper:\n\n- Methods used to solve the competition;\n- Performance of the models on the private test set;\n- Performance of the models on unseen data (not part of the private/public leaderboard);\n- What is needed for the clinical application of these methods?\n- What are the pitfalls and common problems that we encountered?\n\nFrom all teams that signup before the deadline, we will invite up to 15 teams to join the PANDA consortium (max. 5 persons per team). The organizers will make the final selection based on the leaderboard score and the write-up of the method. Notebooks submitted will be evaluated and should be your own work. You can use methods from public notebooks but must show that you have significantly improved upon that method. Your notebook will be judged on how it is uniquely suited to solve this problem. The precise selection criteria may be subject to changes and final decisions are made by the competition hosts.\n\nIf your team is selected, members of the team will be given authorship as part of the PANDA consortium. (see the [CAMELYON paper](https://jamanetwork.com/journals/jama/fullarticle/2665774) for an example of how we mention authors). Authorships are indexed in, e.g., PubMed. As part of the consortium, you are also in the race for the awards.\n\n# Present your work at our workshop\n\nThe PANDA workshop at the international [MICCAI](https://www.miccai2020.org/en/) 2020 conference is a great opportunity for you to showcase your work to a larger audience. The presentations will be chosen based on the method used and on competition ranking. We would like to have a wide variety of topics presented.\n\nSome details regarding the presentations:\n\n- MICCAI 2020 will be a fully virtual conference due to COVID-19. Workshop participants will be reimbursed for one satellite-event registration fee per team.\n- The PANDA workshop takes place on the 8th of October 2020 (specific time is to be determined). Presentations may be pre-recorded. On the workshop day itself, there will be a live moderated session with all presenters and the organizers.\n- Each presentation is 20 minutes with an additional 5 minutes for questions. \n\n# How to sign up\n\nYou can signup by filling out this form: https://forms.gle/SL5AmMUgHDthxDeAA. You can signup today or wait until the competition ends. After submission, you can still view and edit your information until the deadline. \n\n**The deadline to sign-up is the 31st of July**. After the deadline, we will approach you through the email address you filled in on the form if we want to learn more about your approach.\n\n**Note:** The inference notebook you link in the form should contain the code you used to generate the final private leaderboard score.\n\nShare your questions here, we hope that you will sign up and share your unique approaches in this very important work! \n\n**Update: We have extended the deadline for signup until the 31st of July to give teams some extra time to finalize their submission.**",
      "votes": 34
    },
    {
      "id": 1639889,
      "postDate": "2022-01-06T02:53:23.603Z",
      "content": "<p><a href=\"https://www.kaggle.com/wouterbulten\" target=\"_blank\">@wouterbulten</a><br>\nHi! Have the paper been published now?</p>",
      "rawMarkdown": "@wouterbulten\nHi! Have the paper been published now?"
    },
    {
      "id": 1376868,
      "postDate": "2021-07-05T11:54:12.383Z",
      "content": "<p><a href=\"https://www.kaggle.com/wouterbulten\" target=\"_blank\">@wouterbulten</a> <br>\ndoes the scientific paper has published?</p>",
      "rawMarkdown": "@wouterbulten \ndoes the scientific paper has published?",
      "replies": [
        {
          "id": 1378101,
          "postDate": "2021-07-06T09:54:40.317Z",
          "content": "<p>Great question! Not yet, but we are working on this. Any updates we will definitely post here on the forums. </p>",
          "rawMarkdown": "Great question! Not yet, but we are working on this. Any updates we will definitely post here on the forums. "
        },
        {
          "id": 1378114,
          "postDate": "2021-07-06T10:05:43.117Z",
          "content": "<p><a href=\"https://www.kaggle.com/wouterbulten\" target=\"_blank\">@wouterbulten</a> Thank you<br>\nCan you please let me know whether I can use the dataset for research purposes?</p>",
          "rawMarkdown": "@wouterbulten Thank you\nCan you please let me know whether I can use the dataset for research purposes?"
        },
        {
          "id": 1379315,
          "postDate": "2021-07-07T08:55:39.293Z",
          "content": "<p>Unfortunately, the PANDA dataset is under embargo awaiting the publication of the paper. Currently, we are in the whole process of getting the manuscript published. Rest assured that we are fully committed to publishing the paper as soon as possible and releasing the embargo. We will make sure to post any updates on the Kaggle forum or on the challenge website at <a href=\"https://panda.grand-challenge.org/\" target=\"_blank\">https://panda.grand-challenge.org/</a>.</p>\n<p>Please also see <a href=\"https://www.kaggle.com/c/prostate-cancer-grade-assessment/discussion/201117\" target=\"_blank\">https://www.kaggle.com/c/prostate-cancer-grade-assessment/discussion/201117</a> for more information.</p>",
          "rawMarkdown": "Unfortunately, the PANDA dataset is under embargo awaiting the publication of the paper. Currently, we are in the whole process of getting the manuscript published. Rest assured that we are fully committed to publishing the paper as soon as possible and releasing the embargo. We will make sure to post any updates on the Kaggle forum or on the challenge website at https://panda.grand-challenge.org/.\n\nPlease also see https://www.kaggle.com/c/prostate-cancer-grade-assessment/discussion/201117 for more information."
        }
      ]
    },
    {
      "id": 1314333,
      "postDate": "2021-05-19T05:45:11.020Z",
      "content": "<p>Just excellent!</p>",
      "rawMarkdown": "Just excellent!"
    },
    {
      "id": 953506,
      "postDate": "2020-07-31T21:00:22.523Z",
      "content": "<p><a href=\"/wouterbulten\">@wouterbulten</a> is it to ok to submit a link to the discussion post where I detailed my method? I found it hard to include all the details in 1000 characters in the form?</p>",
      "rawMarkdown": "@wouterbulten is it to ok to submit a link to the discussion post where I detailed my method? I found it hard to include all the details in 1000 characters in the form?"
    },
    {
      "id": 943985,
      "postDate": "2020-07-24T18:16:47.103Z",
      "content": "<p>How can I sign-up for this?</p>",
      "rawMarkdown": "How can I sign-up for this?"
    },
    {
      "id": 941485,
      "postDate": "2020-07-23T08:39:46.197Z",
      "content": "<p>Good job keep up</p>",
      "rawMarkdown": "Good job keep up"
    },
    {
      "id": 928382,
      "postDate": "2020-07-14T00:05:57.077Z",
      "content": "<p>Good job</p>",
      "rawMarkdown": "Good job"
    },
    {
      "id": 916263,
      "postDate": "2020-07-05T14:16:34.210Z",
      "content": "<p>👍 </p>",
      "rawMarkdown": "👍 "
    },
    {
      "id": 916091,
      "postDate": "2020-07-05T10:53:01.253Z",
      "content": "<p><a href=\"/wouterbulten\">@wouterbulten</a> \nout of Three section you explained above can one opt for any or all of these?</p>",
      "rawMarkdown": "@wouterbulten \nout of Three section you explained above can one opt for any or all of these?",
      "replies": [
        {
          "id": 916953,
          "postDate": "2020-07-06T06:34:33.263Z",
          "content": "<p>In essence, you sign-up for the paper. From all submissions, we will select a number of teams to take part in the consortium that will be included in the paper. From this selection, we will also pick the presentations and assign the awards. I.e. you can only win one of the new awards if you sign-up and are selected for the paper. You can opt-out of the presentation by indicating this in the form. </p>",
          "rawMarkdown": "In essence, you sign-up for the paper. From all submissions, we will select a number of teams to take part in the consortium that will be included in the paper. From this selection, we will also pick the presentations and assign the awards. I.e. you can only win one of the new awards if you sign-up and are selected for the paper. You can opt-out of the presentation by indicating this in the form. "
        },
        {
          "id": 917270,
          "postDate": "2020-07-06T11:31:01.717Z",
          "content": "<p><a href=\"/wouterbulten\">@wouterbulten</a> ok thanku\n1) out of all the teams signing up 15 will be selected based on how well they perform as per 5 criteria you mentioned above ?and top 2 teams will be winning one out of 3 awards. ?\n2) which all teams will be selected for presentations only the ones out of top 15 ?</p>",
          "rawMarkdown": "@wouterbulten ok thanku\n1) out of all the teams signing up 15 will be selected based on how well they perform as per 5 criteria you mentioned above ?and top 2 teams will be winning one out of 3 awards. ?\n2) which all teams will be selected for presentations only the ones out of top 15 ?"
        },
        {
          "id": 917325,
          "postDate": "2020-07-06T12:16:31.847Z",
          "content": "<p>Teams will be selected based on leaderboard score and the write-up of the method, with a maximum of 15 teams. This does not mean that we will necessarily select the top 15 based on the leaderboard alone because we will also judge your solution. So, for example, if you have a nice and unique method and are on place 20 you could still be selected. </p>\n\n<p>From this selection, we will pick teams for the presentations and the awards.</p>",
          "rawMarkdown": "Teams will be selected based on leaderboard score and the write-up of the method, with a maximum of 15 teams. This does not mean that we will necessarily select the top 15 based on the leaderboard alone because we will also judge your solution. So, for example, if you have a nice and unique method and are on place 20 you could still be selected. \n\nFrom this selection, we will pick teams for the presentations and the awards."
        },
        {
          "id": 940592,
          "postDate": "2020-07-23T03:32:02.557Z",
          "content": "<p><a href=\"/wouterbulten\">@wouterbulten</a>  <a href=\"/petstr\">@petstr</a>  mi still eligible i have applied although,what matters  more here is it private set score or private set leader board rank.\nMy private set score is way high than public one. IMO as per essence of this subject,score on unseen data, solution methodology is what matters more. Based on my methodology many top solutions are baesd</p>\n\n<p>Rest is your teams decision. </p>",
          "rawMarkdown": "@wouterbulten  @petstr  mi still eligible i have applied although,what matters  more here is it private set score or private set leader board rank.\nMy private set score is way high than public one. IMO as per essence of this subject,score on unseen data, solution methodology is what matters more. Based on my methodology many top solutions are baesd\n\nRest is your teams decision. ",
          "votes": 1,
          "replies": [
            {
              "id": 941479,
              "postDate": "2020-07-23T08:36:33.473Z",
              "content": "<p>Same here. What is also interests me is what will be the conditions for reviewing the solutions? From one point there's a shake up of +-500 places, from the other - solution stability and robustness. For me there are 2 differences, when CV/public/private is aligned (either 0.89 or 0.92) and when is completely not (CV/public 0.87 and private 0.93) that's the frequent case here</p>",
              "rawMarkdown": "Same here. What is also interests me is what will be the conditions for reviewing the solutions? From one point there's a shake up of +-500 places, from the other - solution stability and robustness. For me there are 2 differences, when CV/public/private is aligned (either 0.89 or 0.92) and when is completely not (CV/public 0.87 and private 0.93) that's the frequent case here"
            }
          ]
        }
      ]
    },
    {
      "id": 902138,
      "postDate": "2020-06-26T00:41:48.823Z",
      "content": "<p>Great!</p>",
      "rawMarkdown": "Great!\n"
    },
    {
      "id": 928088,
      "postDate": "2020-07-13T18:41:55.463Z",
      "rawMarkdown": "",
      "isDeleted": true
    }
  ],
  "comments": [
    {
      "id": 1639889,
      "author_name": "Sky",
      "author_url": "",
      "post_date": "2022-01-06T02:53:23.603000",
      "content": "<p><a href=\"https://www.kaggle.com/wouterbulten\" target=\"_blank\">@wouterbulten</a><br>\nHi! Have the paper been published now?</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 1376868,
      "author_name": "Abinash Reddy",
      "author_url": "",
      "post_date": "2021-07-05T11:54:12.383000",
      "content": "<p><a href=\"https://www.kaggle.com/wouterbulten\" target=\"_blank\">@wouterbulten</a> <br>\ndoes the scientific paper has published?</p>",
      "votes": 0,
      "replies": [
        {
          "id": 1378101,
          "author_name": "Wouter Bulten",
          "author_url": "",
          "post_date": "2021-07-06T09:54:40.317000",
          "content": "<p>Great question! Not yet, but we are working on this. Any updates we will definitely post here on the forums. </p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1378114,
          "author_name": "Abinash Reddy",
          "author_url": "",
          "post_date": "2021-07-06T10:05:43.117000",
          "content": "<p><a href=\"https://www.kaggle.com/wouterbulten\" target=\"_blank\">@wouterbulten</a> Thank you<br>\nCan you please let me know whether I can use the dataset for research purposes?</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 1379315,
          "author_name": "Wouter Bulten",
          "author_url": "",
          "post_date": "2021-07-07T08:55:39.293000",
          "content": "<p>Unfortunately, the PANDA dataset is under embargo awaiting the publication of the paper. Currently, we are in the whole process of getting the manuscript published. Rest assured that we are fully committed to publishing the paper as soon as possible and releasing the embargo. We will make sure to post any updates on the Kaggle forum or on the challenge website at <a href=\"https://panda.grand-challenge.org/\" target=\"_blank\">https://panda.grand-challenge.org/</a>.</p>\n<p>Please also see <a href=\"https://www.kaggle.com/c/prostate-cancer-grade-assessment/discussion/201117\" target=\"_blank\">https://www.kaggle.com/c/prostate-cancer-grade-assessment/discussion/201117</a> for more information.</p>",
          "votes": 0,
          "replies": []
        }
      ]
    },
    {
      "id": 1314333,
      "author_name": "Rza Pardis",
      "author_url": "",
      "post_date": "2021-05-19T05:45:11.020000",
      "content": "<p>Just excellent!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 953506,
      "author_name": "Shujun",
      "author_url": "",
      "post_date": "2020-07-31T21:00:22.523000",
      "content": "<p><a href=\"/wouterbulten\">@wouterbulten</a> is it to ok to submit a link to the discussion post where I detailed my method? I found it hard to include all the details in 1000 characters in the form?</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 943985,
      "author_name": "Digvijay Yadav",
      "author_url": "",
      "post_date": "2020-07-24T18:16:47.103000",
      "content": "<p>How can I sign-up for this?</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 941485,
      "author_name": "Haithem Hermassi",
      "author_url": "",
      "post_date": "2020-07-23T08:39:46.197000",
      "content": "<p>Good job keep up</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 928382,
      "author_name": "",
      "author_url": "",
      "post_date": "2020-07-14T00:05:57.077000",
      "content": "<p>Good job</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 916263,
      "author_name": "Narendra Singh Parihar",
      "author_url": "",
      "post_date": "2020-07-05T14:16:34.210000",
      "content": "<p>👍 </p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 916091,
      "author_name": "Jaideep",
      "author_url": "",
      "post_date": "2020-07-05T10:53:01.253000",
      "content": "<p><a href=\"/wouterbulten\">@wouterbulten</a> \nout of Three section you explained above can one opt for any or all of these?</p>",
      "votes": 0,
      "replies": [
        {
          "id": 916953,
          "author_name": "Wouter Bulten",
          "author_url": "",
          "post_date": "2020-07-06T06:34:33.263000",
          "content": "<p>In essence, you sign-up for the paper. From all submissions, we will select a number of teams to take part in the consortium that will be included in the paper. From this selection, we will also pick the presentations and assign the awards. I.e. you can only win one of the new awards if you sign-up and are selected for the paper. You can opt-out of the presentation by indicating this in the form. </p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 917270,
          "author_name": "Jaideep",
          "author_url": "",
          "post_date": "2020-07-06T11:31:01.717000",
          "content": "<p><a href=\"/wouterbulten\">@wouterbulten</a> ok thanku\n1) out of all the teams signing up 15 will be selected based on how well they perform as per 5 criteria you mentioned above ?and top 2 teams will be winning one out of 3 awards. ?\n2) which all teams will be selected for presentations only the ones out of top 15 ?</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 917325,
          "author_name": "Wouter Bulten",
          "author_url": "",
          "post_date": "2020-07-06T12:16:31.847000",
          "content": "<p>Teams will be selected based on leaderboard score and the write-up of the method, with a maximum of 15 teams. This does not mean that we will necessarily select the top 15 based on the leaderboard alone because we will also judge your solution. So, for example, if you have a nice and unique method and are on place 20 you could still be selected. </p>\n\n<p>From this selection, we will pick teams for the presentations and the awards.</p>",
          "votes": 0,
          "replies": []
        },
        {
          "id": 940592,
          "author_name": "Jaideep",
          "author_url": "",
          "post_date": "2020-07-23T03:32:02.557000",
          "content": "<p><a href=\"/wouterbulten\">@wouterbulten</a>  <a href=\"/petstr\">@petstr</a>  mi still eligible i have applied although,what matters  more here is it private set score or private set leader board rank.\nMy private set score is way high than public one. IMO as per essence of this subject,score on unseen data, solution methodology is what matters more. Based on my methodology many top solutions are baesd</p>\n\n<p>Rest is your teams decision. </p>",
          "votes": 1,
          "replies": [
            {
              "id": 941479,
              "author_name": "Vladimir Groza",
              "author_url": "",
              "post_date": "2020-07-23T08:36:33.473000",
              "content": "<p>Same here. What is also interests me is what will be the conditions for reviewing the solutions? From one point there's a shake up of +-500 places, from the other - solution stability and robustness. For me there are 2 differences, when CV/public/private is aligned (either 0.89 or 0.92) and when is completely not (CV/public 0.87 and private 0.93) that's the frequent case here</p>",
              "votes": 0,
              "replies": []
            }
          ]
        }
      ]
    },
    {
      "id": 902138,
      "author_name": "Rashidul Hasan Hridoy",
      "author_url": "",
      "post_date": "2020-06-26T00:41:48.823000",
      "content": "<p>Great!</p>",
      "votes": 0,
      "replies": []
    },
    {
      "id": 928088,
      "author_name": "",
      "author_url": "",
      "post_date": "2020-07-13T18:41:55.463000",
      "content": "",
      "votes": 0,
      "replies": []
    }
  ],
  "raw_markdown_by_id": {
    "901263": "With the PANDA challenge well underway, we would like to update you on the scientific part of this project. This competition is part of an ongoing research project and has an associated workshop at [MICCAI](https://www.miccai2020.org/en/). Our goal in hosting this competition was to stimulate new ideas and focus on the clinical applicability of methods. Extend your work on this competition and you may be a co-author on our scientific paper, a presenter at MICCAI, and/or you may win additional prize money for your ideas! \n\n# Three new awards\nAs part of the PANDA workshop, we are introducing three additional awards on top of the competition prizes. The three new awards are:\n\n**Most innovative method award (€ 1000, runner-up € 250)**\nWe want to incentivize new and innovative methods for automatically grading prostate cancer. Based on the write-ups and code we will pick one winner and a runner-up. This award is not necessarily linked to the leaderboard score. \n\n**Best video presentation award (€ 500, runner-up € 250)**\nDuring the PANDA workshops, presentations will be graded based on overall presentation quality, clarity, presentation skills, etc. A jury will decide on the best presentation and a runner-up.\n\n**Best generalization performance award (€ 1000, runner-up € 250)**\nAs part of the scientific paper, methods will be evaluated on external unseen data. In the end, the efforts of this competition should go towards building models that can work in clinical practice. The method that achieves the best performance across all datasets will win. This unseen data will be in the same format as the competition’s test set but differences in staining and scanners are to be expected. \n\nPrizes will be awarded by the challenge organizers. To be eligible for winning one of the prizes you have to sign up and be selected as part of the PANDA Consortium.\n\n# Scientific paper on the PANDA challenge\n\nTogether with you, we want to write a scientific paper on the results of the competition. The aim of this paper is to directly contribute to the scientific field of prostate cancer diagnosis and computational pathology. A selection of topics we will discuss in the paper:\n\n- Methods used to solve the competition;\n- Performance of the models on the private test set;\n- Performance of the models on unseen data (not part of the private/public leaderboard);\n- What is needed for the clinical application of these methods?\n- What are the pitfalls and common problems that we encountered?\n\nFrom all teams that signup before the deadline, we will invite up to 15 teams to join the PANDA consortium (max. 5 persons per team). The organizers will make the final selection based on the leaderboard score and the write-up of the method. Notebooks submitted will be evaluated and should be your own work. You can use methods from public notebooks but must show that you have significantly improved upon that method. Your notebook will be judged on how it is uniquely suited to solve this problem. The precise selection criteria may be subject to changes and final decisions are made by the competition hosts.\n\nIf your team is selected, members of the team will be given authorship as part of the PANDA consortium. (see the [CAMELYON paper](https://jamanetwork.com/journals/jama/fullarticle/2665774) for an example of how we mention authors). Authorships are indexed in, e.g., PubMed. As part of the consortium, you are also in the race for the awards.\n\n# Present your work at our workshop\n\nThe PANDA workshop at the international [MICCAI](https://www.miccai2020.org/en/) 2020 conference is a great opportunity for you to showcase your work to a larger audience. The presentations will be chosen based on the method used and on competition ranking. We would like to have a wide variety of topics presented.\n\nSome details regarding the presentations:\n\n- MICCAI 2020 will be a fully virtual conference due to COVID-19. Workshop participants will be reimbursed for one satellite-event registration fee per team.\n- The PANDA workshop takes place on the 8th of October 2020 (specific time is to be determined). Presentations may be pre-recorded. On the workshop day itself, there will be a live moderated session with all presenters and the organizers.\n- Each presentation is 20 minutes with an additional 5 minutes for questions. \n\n# How to sign up\n\nYou can signup by filling out this form: https://forms.gle/SL5AmMUgHDthxDeAA. You can signup today or wait until the competition ends. After submission, you can still view and edit your information until the deadline. \n\n**The deadline to sign-up is the 31st of July**. After the deadline, we will approach you through the email address you filled in on the form if we want to learn more about your approach.\n\n**Note:** The inference notebook you link in the form should contain the code you used to generate the final private leaderboard score.\n\nShare your questions here, we hope that you will sign up and share your unique approaches in this very important work! \n\n**Update: We have extended the deadline for signup until the 31st of July to give teams some extra time to finalize their submission.**",
    "1639889": "@wouterbulten\nHi! Have the paper been published now?",
    "1376868": "@wouterbulten \ndoes the scientific paper has published?",
    "1314333": "Just excellent!",
    "953506": "@wouterbulten is it to ok to submit a link to the discussion post where I detailed my method? I found it hard to include all the details in 1000 characters in the form?",
    "943985": "How can I sign-up for this?",
    "941485": "Good job keep up",
    "928382": "Good job",
    "916263": "👍 ",
    "916091": "@wouterbulten \nout of Three section you explained above can one opt for any or all of these?",
    "902138": "Great!\n",
    "928088": ""
  }
}