Although I am using the latest version available for the dataset (v2), I continue to face errors with the DICOM images of the patient named above, both in my Kaggle notebooks and downloading the data.
\nAny recommended workarounds? Currently I'm simply ignoring this unit, which of course is less than ideal.
\nThanks in advance!
", "messages": [ { "id": "998148", "postDate": "09/04/2020 14:22:46", "content": "Although I am using the latest version available for the dataset (v2), I continue to face errors with the DICOM images of the patient named above, both in my Kaggle notebooks and downloading the data.
\nAny recommended workarounds? Currently I'm simply ignoring this unit, which of course is less than ideal.
\nThanks in advance!
", "rawMarkdown": "Although I am using the latest version available for the dataset (v2), I continue to face errors with the DICOM images of the patient named above, both in my Kaggle notebooks and downloading the data.\n\nAny recommended workarounds? Currently I'm simply ignoring this unit, which of course is less than ideal.\n\nThanks in advance!", "votes": null }, { "id": "998172", "postDate": "09/04/2020 14:45:47", "content": "Hi Douglas, I remember you told me you are using R. I didn't use R to read images so can't help I am afraid. If you can share more details (ie. what is package you are using to read images, and what is the error you see), maybe someone else here who is using R can help.
\nAs a naive workaround, I can send you a .npy file with the images of this patient. I don't know if this will help though.
", "rawMarkdown": "Hi Douglas, I remember you told me you are using R. I didn't use R to read images so can't help I am afraid. If you can share more details (ie. what is package you are using to read images, and what is the error you see), maybe someone else here who is using R can help.\n\nAs a naive workaround, I can send you a .npy file with the images of this patient. I don't know if this will help though.", "votes": null }, { "id": "998242", "postDate": "09/04/2020 15:39:30", "content": "Thank you @ahmedhshahin - what a good memory :)
\nYes, I'm using R. More specifically, package oro.dicom.
My code to read the files is the following: (note that I'm using function \"try\" because otherwise it will stop at the error with the patient named in the question.)
\nlibrary(oro.dicom)\nlibrary(tidyverse)\ntrain <- read.csv(\"../input/osic-pulmonary-fibrosis-progression/train.csv\")\ntrain <- train %>% \n group_by(Patient) %>% \n nest()\ndicom_files_base_folder <- \"../input/osic-pulmonary-fibrosis-progression/train/\"\ntrain <- train %>% \n mutate(dicom = map(Patient, function(x) try(readDICOM(paste0(dicom_files_base_folder, x))$img)))
And thank you also for the offer. I would appreciate that, if it really is no trouble for you. How would be best for you to transmit the .npy file? Many thanks for the support!
", "rawMarkdown": "Thank you @ahmedhshahin - what a good memory :) \nYes, I'm using R. More specifically, package oro.dicom.\n\nMy code to read the files is the following: (note that I'm using function \"try\" because otherwise it will stop at the error with the patient named in the question.)\n\n`library(oro.dicom)\nlibrary(tidyverse)\ntrain <- read.csv(\"../input/osic-pulmonary-fibrosis-progression/train.csv\")\ntrain <- train %>% \n group_by(Patient) %>% \n nest()\ndicom_files_base_folder <- \"../input/osic-pulmonary-fibrosis-progression/train/\"\ntrain <- train %>% \n mutate(dicom = map(Patient, function(x) try(readDICOM(paste0(dicom_files_base_folder, x))$img)))`\n\nAnd thank you also for the offer. I would appreciate that, if it really is no trouble for you. How would be best for you to transmit the .npy file? Many thanks for the support!", "votes": null }, { "id": "998408", "postDate": "09/04/2020 17:55:03", "content": "You are welcome!
\nPlease drop me an email here:
\nahmed [dot] shahin [dot] 19 [at] ucl [dot] ac [dot] uk
Hi Douglas, I remember you told me you are using R. I didn't use R to read images so can't help I am afraid. If you can share more details (ie. what is package you are using to read images, and what is the error you see), maybe someone else here who is using R can help.
\nAs a naive workaround, I can send you a .npy file with the images of this patient. I don't know if this will help though.
", "votes": null, "replies": [ { "id": 998242, "author_name": "douglaskgaraujo", "author_url": "", "post_date": "09/04/2020 15:39:30", "content": "Thank you @ahmedhshahin - what a good memory :)
\nYes, I'm using R. More specifically, package oro.dicom.
My code to read the files is the following: (note that I'm using function \"try\" because otherwise it will stop at the error with the patient named in the question.)
\nlibrary(oro.dicom)\nlibrary(tidyverse)\ntrain <- read.csv(\"../input/osic-pulmonary-fibrosis-progression/train.csv\")\ntrain <- train %>% \n group_by(Patient) %>% \n nest()\ndicom_files_base_folder <- \"../input/osic-pulmonary-fibrosis-progression/train/\"\ntrain <- train %>% \n mutate(dicom = map(Patient, function(x) try(readDICOM(paste0(dicom_files_base_folder, x))$img)))
And thank you also for the offer. I would appreciate that, if it really is no trouble for you. How would be best for you to transmit the .npy file? Many thanks for the support!
", "votes": null, "replies": [] }, { "id": 998408, "author_name": "ahmedhshahin", "author_url": "", "post_date": "09/04/2020 17:55:03", "content": "You are welcome!
\nPlease drop me an email here:
\nahmed [dot] shahin [dot] 19 [at] ucl [dot] ac [dot] uk