{
  "id": 109946,
  "title": "Total number of patients in train, test and duplicated images",
  "url": "/competitions/rsna-intracranial-hemorrhage-detection/discussion/109946",
  "author_name": "Al-Khwârizmî",
  "post_date": "2019-09-23T17:03:13.160000",
  "votes": 9,
  "comment_count": 2,
  "views": 0,
  "content": "<p>The total number of patients in the train set is 17079. There are some patients with more than 100 images and others with more than 200 ! And as you can guess no \"head\" will take 200 images to scan it ! (Check for example patient ID_0369d8b8 or ID_03962d75 ).\nThese are for patients with multiple studies ! I've counted 365, maybe there are more. Some patients have 4 studies (which mean 4 scans)!</p>\n\n<p>And the total number of patients in the test is 2144 in which 285 are in the train set too.</p>",
  "messages": [
    {
      "id": 632492,
      "postDate": "2019-09-23T17:03:13.160Z",
      "content": "<p>The total number of patients in the train set is 17079. There are some patients with more than 100 images and others with more than 200 ! And as you can guess no \"head\" will take 200 images to scan it ! (Check for example patient ID_0369d8b8 or ID_03962d75 ).\nThese are for patients with multiple studies ! I've counted 365, maybe there are more. Some patients have 4 studies (which mean 4 scans)!</p>\n\n<p>And the total number of patients in the test is 2144 in which 285 are in the train set too.</p>",
      "rawMarkdown": "The total number of patients in the train set is 17079. There are some patients with more than 100 images and others with more than 200 ! And as you can guess no \"head\" will take 200 images to scan it ! (Check for example patient ID_0369d8b8 or ID_03962d75 ).\nThese are for patients with multiple studies ! I've counted 365, maybe there are more. Some patients have 4 studies (which mean 4 scans)!\n\nAnd the total number of patients in the test is 2144 in which 285 are in the train set too.\n",
      "votes": 9
    },
    {
      "id": 632679,
      "postDate": "2019-09-23T22:18:21.293Z",
      "content": "<p>Are you sure that are duplicate images. A patient can have multiple studies done at different times. </p>",
      "rawMarkdown": "Are you sure that are duplicate images. A patient can have multiple studies done at different times. ",
      "replies": [
        {
          "id": 632890,
          "postDate": "2019-09-24T07:37:05.683Z",
          "content": "<p>Agree ! those are for the same patient, but from different studies. However, I couldn't find any obvious differences ! I checked some of them and couldn't see any ! Same things for the labels !   You can check for example ID_00a3b735 who has 4 studies, but labeled as healthy ! which increase data imbalance !\nI did not check all of them though ! maybe there are some differences. \nps: I changed duplicated by multiple studies to remove the confusion !</p>",
          "rawMarkdown": "Agree ! those are for the same patient, but from different studies. However, I couldn't find any obvious differences ! I checked some of them and couldn't see any ! Same things for the labels !   You can check for example ID_00a3b735 who has 4 studies, but labeled as healthy ! which increase data imbalance !\nI did not check all of them though ! maybe there are some differences. \nps: I changed duplicated by multiple studies to remove the confusion !",
          "votes": 1
        }
      ]
    }
  ],
  "comments": [
    {
      "id": 632679,
      "author_name": "Amil Gentili",
      "author_url": "",
      "post_date": "2019-09-23T22:18:21.293000",
      "content": "<p>Are you sure that are duplicate images. A patient can have multiple studies done at different times. </p>",
      "votes": 0,
      "replies": [
        {
          "id": 632890,
          "author_name": "Al-Khwârizmî",
          "author_url": "",
          "post_date": "2019-09-24T07:37:05.683000",
          "content": "<p>Agree ! those are for the same patient, but from different studies. However, I couldn't find any obvious differences ! I checked some of them and couldn't see any ! Same things for the labels !   You can check for example ID_00a3b735 who has 4 studies, but labeled as healthy ! which increase data imbalance !\nI did not check all of them though ! maybe there are some differences. \nps: I changed duplicated by multiple studies to remove the confusion !</p>",
          "votes": 1,
          "replies": []
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "632492": "The total number of patients in the train set is 17079. There are some patients with more than 100 images and others with more than 200 ! And as you can guess no \"head\" will take 200 images to scan it ! (Check for example patient ID_0369d8b8 or ID_03962d75 ).\nThese are for patients with multiple studies ! I've counted 365, maybe there are more. Some patients have 4 studies (which mean 4 scans)!\n\nAnd the total number of patients in the test is 2144 in which 285 are in the train set too.\n",
    "632679": "Are you sure that are duplicate images. A patient can have multiple studies done at different times. "
  }
}