{
  "id": 578409,
  "title": "Voxal spacing Importance ?",
  "url": "/competitions/byu-locating-bacterial-flagellar-motors-2025/discussion/578409",
  "author_name": "",
  "post_date": "2025-05-10T16:20:19.523062600Z",
  "votes": null,
  "comment_count": 8,
  "views": 0,
  "content": "<p>What is the use of voxal spacing in this competition? Is this to convert the coordinates to nanometers, depending on the spacing unit? How will it be helpful for this competition? How can we benefit from using this parameter? </p>",
  "messages": [
    {
      "id": "3199213",
      "postDate": "05/10/2025 16:20:19",
      "content": "<p>What is the use of voxal spacing in this competition? Is this to convert the coordinates to nanometers, depending on the spacing unit? How will it be helpful for this competition? How can we benefit from using this parameter? </p>",
      "rawMarkdown": "What is the use of voxal spacing in this competition? Is this to convert the coordinates to nanometers, depending on the spacing unit? How will it be helpful for this competition? How can we benefit from using this parameter?",
      "votes": null
    },
    {
      "id": "3199251",
      "postDate": "05/10/2025 17:19:02",
      "content": "<p>I don't see pixel spacing for a test dataset. Is it available? </p>",
      "rawMarkdown": "I don't see pixel spacing for a test dataset. Is it available?",
      "votes": null
    },
    {
      "id": "3199264",
      "postDate": "05/10/2025 17:52:59",
      "content": "<p>No, it's not available. </p>",
      "rawMarkdown": "No, it's not available.",
      "votes": null
    },
    {
      "id": "3199313",
      "postDate": "05/10/2025 19:25:38",
      "content": "<p>In general, uniform spacing helps to improve score, but we can do it only for training data in this case. So, I don't think it could be useful. </p>",
      "rawMarkdown": "In general, uniform spacing helps to improve score, but we can do it only for training data in this case. So, I don't think it could be useful.",
      "votes": null
    },
    {
      "id": "3199328",
      "postDate": "05/10/2025 19:57:38",
      "content": "<p>got it. Thanks :)</p>",
      "rawMarkdown": "got it. Thanks :)",
      "votes": null
    },
    {
      "id": "3199676",
      "postDate": "05/11/2025 10:21:12",
      "content": "<p>The th to discriminate TP is 1000 Å. So if you predict motor outside this radius from gt you will miss it. In pixels that's 1000/VS. True, you won't know test voxel spacings. But for training is helpful, you can experiment with different mask radius  for example. For local testing is a must.</p>\n<p>PD: Apparently true motor radius is about 250 Å (250/VS pixels).</p>",
      "rawMarkdown": "The th to discriminate TP is 1000 Å. So if you predict motor outside this radius from gt you will miss it. In pixels that's 1000/VS. True, you won't know test voxel spacings. But for training is helpful, you can experiment with different mask radius  for example. For local testing is a must.\n\nPD: Apparently true motor radius is about 250 Å (250/VS pixels).",
      "votes": null
    },
    {
      "id": "3199679",
      "postDate": "05/11/2025 10:22:27",
      "content": "<p>You actually can estimate voxel spacing and then use this result to impact your prediction. </p>",
      "rawMarkdown": "You actually can estimate voxel spacing and then use this result to impact your prediction.",
      "votes": null
    },
    {
      "id": "3199694",
      "postDate": "05/11/2025 10:38:15",
      "content": "<p>That is true. You can pair a median scan shape [300, 960, 960] with median spacing and then approximate test scan spacing. </p>",
      "rawMarkdown": "That is true. You can pair a median scan shape [300, 960, 960] with median spacing and then approximate test scan spacing.",
      "votes": null
    },
    {
      "id": "3199696",
      "postDate": "05/11/2025 10:45:44",
      "content": "<p>But I am not sure whether it is a better solution than resizing all scans to this same shape. </p>",
      "rawMarkdown": "But I am not sure whether it is a better solution than resizing all scans to this same shape.",
      "votes": null
    }
  ],
  "comments": [
    {
      "id": 3199251,
      "author_name": "jankowalski2000",
      "author_url": "",
      "post_date": "05/10/2025 17:19:02",
      "content": "<p>I don't see pixel spacing for a test dataset. Is it available? </p>",
      "votes": null,
      "replies": [
        {
          "id": 3199264,
          "author_name": "santusrk",
          "author_url": "",
          "post_date": "05/10/2025 17:52:59",
          "content": "<p>No, it's not available. </p>",
          "votes": null,
          "replies": [
            {
              "id": 3199313,
              "author_name": "jankowalski2000",
              "author_url": "",
              "post_date": "05/10/2025 19:25:38",
              "content": "<p>In general, uniform spacing helps to improve score, but we can do it only for training data in this case. So, I don't think it could be useful. </p>",
              "votes": null,
              "replies": [
                {
                  "id": 3199328,
                  "author_name": "santusrk",
                  "author_url": "",
                  "post_date": "05/10/2025 19:57:38",
                  "content": "<p>got it. Thanks :)</p>",
                  "votes": null,
                  "replies": []
                }
              ]
            }
          ]
        }
      ]
    },
    {
      "id": 3199676,
      "author_name": "sacuscreed",
      "author_url": "",
      "post_date": "05/11/2025 10:21:12",
      "content": "<p>The th to discriminate TP is 1000 Å. So if you predict motor outside this radius from gt you will miss it. In pixels that's 1000/VS. True, you won't know test voxel spacings. But for training is helpful, you can experiment with different mask radius  for example. For local testing is a must.</p>\n<p>PD: Apparently true motor radius is about 250 Å (250/VS pixels).</p>",
      "votes": null,
      "replies": []
    },
    {
      "id": 3199679,
      "author_name": "tom99763",
      "author_url": "",
      "post_date": "05/11/2025 10:22:27",
      "content": "<p>You actually can estimate voxel spacing and then use this result to impact your prediction. </p>",
      "votes": null,
      "replies": [
        {
          "id": 3199694,
          "author_name": "jankowalski2000",
          "author_url": "",
          "post_date": "05/11/2025 10:38:15",
          "content": "<p>That is true. You can pair a median scan shape [300, 960, 960] with median spacing and then approximate test scan spacing. </p>",
          "votes": null,
          "replies": [
            {
              "id": 3199696,
              "author_name": "jankowalski2000",
              "author_url": "",
              "post_date": "05/11/2025 10:45:44",
              "content": "<p>But I am not sure whether it is a better solution than resizing all scans to this same shape. </p>",
              "votes": null,
              "replies": []
            }
          ]
        }
      ]
    }
  ],
  "raw_markdown_by_id": {
    "3199213": "What is the use of voxal spacing in this competition? Is this to convert the coordinates to nanometers, depending on the spacing unit? How will it be helpful for this competition? How can we benefit from using this parameter?",
    "3199251": "I don't see pixel spacing for a test dataset. Is it available?",
    "3199264": "No, it's not available.",
    "3199313": "In general, uniform spacing helps to improve score, but we can do it only for training data in this case. So, I don't think it could be useful.",
    "3199328": "got it. Thanks :)",
    "3199676": "The th to discriminate TP is 1000 Å. So if you predict motor outside this radius from gt you will miss it. In pixels that's 1000/VS. True, you won't know test voxel spacings. But for training is helpful, you can experiment with different mask radius  for example. For local testing is a must.\n\nPD: Apparently true motor radius is about 250 Å (250/VS pixels).",
    "3199679": "You actually can estimate voxel spacing and then use this result to impact your prediction.",
    "3199694": "That is true. You can pair a median scan shape [300, 960, 960] with median spacing and then approximate test scan spacing.",
    "3199696": "But I am not sure whether it is a better solution than resizing all scans to this same shape."
  },
  "source": "meta"
}