Comments (9)
Hi Bardia,
Slice_level_labels.npy contains the labels in the order that is mentioned in the index.csv file.
The first 55 labeled cases are COVID-19 cases. However, they're not necessarily the first 55 patients. For example, the labeled cases 50-54 are related to P058, P104, P109, P129, P142. You can find this relation in the index.csv file.
I hope this answers your question. However, if you found other inconsistency in the labels and dicom files, please give me the case number for further review.
from covid-ct-md.
Hi, thanks for sharing the dataset
I have problem with reading labels
labels in Slice_level_labels.npy and index.csv are not match with dicom files
Could you please help me
Please also notice the following note in the readme file. I saw this mistake from many people and I guess you're doing the same mistake.
NOTE: The correct order of slices in a CT scan doesn't necessarily follow the order of the Slice-IDs. You need to sort slices based on the "slice location" parameter provided in the DICOM files when you are reading the data.
If you're using pydicom library or similar ones, you need to sort slices based on the "slice location" parameter which indicates the Z-axis location of the slices. The "slice location" parameter can be accessed when you read a .dcm file using pydicom or similar libraries. You can store them and sort the slices based on them.
from covid-ct-md.
Thanks for replying, It helped a lot
from covid-ct-md.
Hi, thanks for sharing the dataset
I have problem with reading labels
labels in Slice_level_labels.npy and index.csv are not match with dicom files
Could you please help mePlease also notice the following note in the readme file. I saw this mistake from many people and I guess you're doing the same mistake.
NOTE: The correct order of slices in a CT scan doesn't necessarily follow the order of the Slice-IDs. You need to sort slices based on the "slice location" parameter provided in the DICOM files when you are reading the data.If you're using pydicom library or similar ones, you need to sort slices based on the "slice location" parameter which indicates the Z-axis location of the slices. The "slice location" parameter can be accessed when you read a .dcm file using pydicom or similar libraries. You can store them and sort the slices based on them.
Hi, Thanks for sharing this great data set.
I still have a problem with labels. I converted dicom files to png and printed the labels. It seemed they didn't match. Could you possibly help me to find the problem?
First, for each case, I read all the dicom files by pydicom (it rotates images 90 degrees) and saved them in png format, and named them based on 'SliceLocation ' parameter. Then sorted them based on the given name.
- Shall I sort them ascending or descending? (Even though I tried them both, and failed)
- You have provided an example, that
he labeled cases 50-54 are related to P058, P104, P109, P129, P142
. Have you updated the index.csv? I find labeled cases 50-54 are related to P049-P053.
If it is necessary I will gladly share my snippet.
Thanks in advanced
from covid-ct-md.
Hello and thanks for your message,
Sorry to hear that you have still problems matching the labels and slices.
For your questions:
Q1: You should sort the slices from the upper part of the lung all the way to the bottom. Using the "Slice Location" value, I sorted them in descending order for this dataset.
Q2: Yes the index file has been updated. One difference is in the name of those cases that you mentioned, the other difference is that one labeled COVID-19 case is removed in the new version. This dataset and the related data description have been published in Nature Scientific Data and the index and label files have been updated accordingly. Please use the updated files. Also, for more information on the dataset, you can refer here.
Also, I suggest that you use a DICOM viewer software to view each case and check the labels with the slices there. You can use MicroDicom for windows and miele-lxiv for mac. They are free.
If none of them helped, please share here some sample slices with their corresponding slice locations and labels based on what you have on your side so I can better identify the source of your error.
from covid-ct-md.
Hello and thanks for your message,
Sorry to hear that you have still problems matching the labels and slices.For your questions:
Q1: You should sort the slices from the upper part of the lung all the way to the bottom. Using the "Slice Location" value, I sorted them in descending order for this dataset.Q2: Yes the index file has been updated. One difference is in the name of those cases that you mentioned, the other difference is that one labeled COVID-19 case is removed in the new version. This dataset and the related data description have been published in Nature Scientific Data and the index and label files have been updated accordingly. Please use the updated files. Also, for more information on the dataset, you can refer here.
Also, I suggest that you use a DICOM viewer software to view each case and check the labels with the slices there. You can use MicroDicom for windows and miele-lxiv for mac. They are free.
If none of them helped, please share here some sample slices with their corresponding slice locations and labels based on what you have on your side so I can better identify the source of your error.
Thanks for answering my questions.
Checking the labels on dicom viewer was a great idea. As my os is Ubuntu I used an online viewer. I suppose dicom viewers sort the images and there is no need for further attempts.
I tested P006, according to index.csv, its index is 5
therefore, to get its labels, I read Slice_level_labels[5]
for example 70-th slice is Slice_level_labels[5][70] = 1 which shows the slice is infected
and the infection location would be: lobe_level_labels[5][70] = [0. 0. 1. 1. 0.] which means the infection is located at RLL and RML
The image I got in the online viewer is:
it seems like the lobe-level-label doesn't match (I checked that slices are sorted from neck to stomach)
I would be grateful if you could help me with the problem.
from covid-ct-md.
From what you shared, I can see that everything looks fine. In this slice, RLL and RML are involved.
from covid-ct-md.
From what you shared, I can see that everything looks fine. In this slice, RLL and RML are involved.
Thank you for your immediate response.
I thought the largest lesion was in left lower lobe. So, my interpretation was wrong? or the image should be flipped?
from covid-ct-md.
I have a question:
the Label Index 43 was missing in the Index.csv but the there is still 43 index in Slice-level-labels-updated-1.npy file๏ผ or the 43 case in the npy file indicate which case ?
from covid-ct-md.
Related Issues (5)
- Acessing data on figshare HOT 2
- Test sets HOT 2
- Data download link HOT 2
- Problem about the npy file? HOT 2
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. ๐๐๐
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google โค๏ธ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from covid-ct-md.