{
  "id": 386170,
  "title": "Probably why there are seemingly mislabeled lateralities",
  "url": "/competitions/rsna-breast-cancer-detection/discussion/386170",
  "author_name": "Sawada Tsuyoshi",
  "post_date": "2023-02-11T13:41:02.154000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Regarding the MLO images, a majority of them seem to have the torso on the left/right side for the left/right breast respectively, while in the other cases the opposite is true. At first, just like some other people have discussed here, I thought these laterality values are mistakenly labeled.</p>\n<p>After some thought, I think that different medical institutions or different equipment makers have different practice regarding which side to see the images from, and that these laterality values are probably correct.</p>\n<p>Notice that each of these images can be considered as what you would get by adding up cross section images along an axis perpendicular to them. They are not images picturing the surface of something. By nature, there is no correct direction to see each image from.</p>\n<p>For the MLO images, some medical professionals may have the habit of seeing the images from outside towards the center of the body, while others may have the opposite habit. When it comes to CC images, it looks like everyone seem to see the images from the ceiling looking downwards rather than from the floor looking upwards. That is probably the reason we see seemingly-flipped images for MLO.</p>\n<p>I am no expert in this field. This is just my thought, although I think it is convincing.</p>",
  "messages": [
    {
      "id": 2140133,
      "postDate": "2023-02-11T13:41:02.153Z",
      "content": "<p>Regarding the MLO images, a majority of them seem to have the torso on the left/right side for the left/right breast respectively, while in the other cases the opposite is true. At first, just like some other people have discussed here, I thought these laterality values are mistakenly labeled.</p>\n<p>After some thought, I think that different medical institutions or different equipment makers have different practice regarding which side to see the images from, and that these laterality values are probably correct.</p>\n<p>Notice that each of these images can be considered as what you would get by adding up cross section images along an axis perpendicular to them. They are not images picturing the surface of something. By nature, there is no correct direction to see each image from.</p>\n<p>For the MLO images, some medical professionals may have the habit of seeing the images from outside towards the center of the body, while others may have the opposite habit. When it comes to CC images, it looks like everyone seem to see the images from the ceiling looking downwards rather than from the floor looking upwards. That is probably the reason we see seemingly-flipped images for MLO.</p>\n<p>I am no expert in this field. This is just my thought, although I think it is convincing.</p>",
      "rawMarkdown": "Regarding the MLO images, a majority of them seem to have the torso on the left/right side for the left/right breast respectively, while in the other cases the opposite is true. At first, just like some other people have discussed here, I thought these laterality values are mistakenly labeled.\n\nAfter some thought, I think that different medical institutions or different equipment makers have different practice regarding which side to see the images from, and that these laterality values are probably correct.\n\nNotice that each of these images can be considered as what you would get by adding up cross section images along an axis perpendicular to them. They are not images picturing the surface of something. By nature, there is no correct direction to see each image from.\n\nFor the MLO images, some medical professionals may have the habit of seeing the images from outside towards the center of the body, while others may have the opposite habit. When it comes to CC images, it looks like everyone seem to see the images from the ceiling looking downwards rather than from the floor looking upwards. That is probably the reason we see seemingly-flipped images for MLO.\n\nI am no expert in this field. This is just my thought, although I think it is convincing.",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "2140133": "Regarding the MLO images, a majority of them seem to have the torso on the left/right side for the left/right breast respectively, while in the other cases the opposite is true. At first, just like some other people have discussed here, I thought these laterality values are mistakenly labeled.\n\nAfter some thought, I think that different medical institutions or different equipment makers have different practice regarding which side to see the images from, and that these laterality values are probably correct.\n\nNotice that each of these images can be considered as what you would get by adding up cross section images along an axis perpendicular to them. They are not images picturing the surface of something. By nature, there is no correct direction to see each image from.\n\nFor the MLO images, some medical professionals may have the habit of seeing the images from outside towards the center of the body, while others may have the opposite habit. When it comes to CC images, it looks like everyone seem to see the images from the ceiling looking downwards rather than from the floor looking upwards. That is probably the reason we see seemingly-flipped images for MLO.\n\nI am no expert in this field. This is just my thought, although I think it is convincing."
  }
}