Hi everyone, thanks for the cool competition :D
\nI noticed that study_id \"3008676218\" does not contain a scan with orientation Sagittal T2/STIR and study_id's \"2492114990\" and \"2780132468\" do not contain scans with orientation Sagittal T1, while every training sample has a scan of orientation Axial T2. Now to the actual question… Is it fine to ignore those 3 training samples with \"incomplete\" scan orientation? If yes, I would expect every testing/hidden sample to have at least 1 scan of every orientation. Otherwise, one appears to be somewhat limited to Axial T2 scans exclusively.
\nThanks in advance!
", "messages": [ { "id": "2836456", "postDate": "05/25/2024 22:19:02", "content": "Hi everyone, thanks for the cool competition :D
\nI noticed that study_id \"3008676218\" does not contain a scan with orientation Sagittal T2/STIR and study_id's \"2492114990\" and \"2780132468\" do not contain scans with orientation Sagittal T1, while every training sample has a scan of orientation Axial T2. Now to the actual question… Is it fine to ignore those 3 training samples with \"incomplete\" scan orientation? If yes, I would expect every testing/hidden sample to have at least 1 scan of every orientation. Otherwise, one appears to be somewhat limited to Axial T2 scans exclusively.
\nThanks in advance!
", "rawMarkdown": "Hi everyone, thanks for the cool competition :D\n\nI noticed that study_id \"3008676218\" does not contain a scan with orientation Sagittal T2/STIR and study_id's \"2492114990\" and \"2780132468\" do not contain scans with orientation Sagittal T1, while every training sample has a scan of orientation Axial T2. Now to the actual question... Is it fine to ignore those 3 training samples with \"incomplete\" scan orientation? If yes, I would expect every testing/hidden sample to have at least 1 scan of every orientation. Otherwise, one appears to be somewhat limited to Axial T2 scans exclusively.\n\nThanks in advance!", "votes": null }, { "id": "2860663", "postDate": "06/07/2024 17:50:06", "content": "To me, axial scans are very important. If there is a missing axial scan for anyone intervertebral disc, it will be very difficult to predict the stenosis at that level. But even if one has one type of Sagittal information, either T1 or T2 then its somewhat manageable to make a prediction. What are your opinions? Your response is highly encouraged.
", "rawMarkdown": "To me, axial scans are very important. If there is a missing axial scan for anyone intervertebral disc, it will be very difficult to predict the stenosis at that level. But even if one has one type of Sagittal information, either T1 or T2 then its somewhat manageable to make a prediction. What are your opinions? Your response is highly encouraged.", "votes": null }, { "id": "2861727", "postDate": "06/08/2024 10:35:47", "content": "Hello, today I am facing a problem similar to yours. Your response will be highly valuable, especially for problem 3 at https://www.kaggle.com/competitions/rsna-2024-lumbar-spine-degenerative-classification/discussion/510925
\nIn short:
\nThere are missing column values in train.csv, and correspondingly their entries are missing in train_label_coordinates.csv. how can anyone predict the probability values, especially if entries for axial (top-down) views are missing in the most expected file 'test_label_coordinates.csv', because even if a model correctly identifies the type of stenosis, which level will the model put it in? I do not know how the evaluation metric handles it. Hence, I need a lot of help.
To me, axial scans are very important. If there is a missing axial scan for anyone intervertebral disc, it will be very difficult to predict the stenosis at that level. But even if one has one type of Sagittal information, either T1 or T2 then its somewhat manageable to make a prediction. What are your opinions? Your response is highly encouraged.
", "votes": null, "replies": [] }, { "id": 2861727, "author_name": "devsya", "author_url": "", "post_date": "06/08/2024 10:35:47", "content": "Hello, today I am facing a problem similar to yours. Your response will be highly valuable, especially for problem 3 at https://www.kaggle.com/competitions/rsna-2024-lumbar-spine-degenerative-classification/discussion/510925
\nIn short:
\nThere are missing column values in train.csv, and correspondingly their entries are missing in train_label_coordinates.csv. how can anyone predict the probability values, especially if entries for axial (top-down) views are missing in the most expected file 'test_label_coordinates.csv', because even if a model correctly identifies the type of stenosis, which level will the model put it in? I do not know how the evaluation metric handles it. Hence, I need a lot of help.