vuptran/cardiac-segmentation

Convolutional Neural Networks for Cardiac Segmentation

cardiac-segmentation's Issues

How to train the demo?

Hello, I have reead your paper, and I try to train the sunnybrook dataset in my conputer after reading your tutorial, but after 40 epochs, the loss, accuracy and the jaccard coefficient are the same as 1 epoch, how can I do to increase my accuracy by train the Sunnybrook dataset?
Thank you very much for your paper and tutorial, and I am looking forward to get your reply.

about the MICCAI 2009 dataset

Hello,I want to know that whether you need to submit the results of the segmentation of the MICCAI 2009 dataset online, just like RVSC dataset.

How to implement your method in pytorch

Since your method had got excellent result, I tried to implement your method in pytorch framework. But my result is much worse than you especially in dice coef. If you have time, could you please read my code and give me some advice

What do I need to know after change crop size to 96?

Hello, if I change the crop size from 100 to 96 in train_sunnybrook.py, fcn_model.py, and submist_sunnybrook.py, will it cause any problem? I can get the codes working well, so the question is not about if there would be execution error. I am not sure if you guys choose crop size of 100 for some reasons specific to this Sunnybrook image data?

How to achieve Hausdorff distance

Please. How to download the dataset?

About the acc value

My acc value is very high at the beginning, is it over fitting

how can I find SAX_series from the Sunnybrook database?

Regarding Downloading Dataset

Can you please check the dataset download link is not working.
Also please update the Read me regarding sax file.
How you are segregating Training and Validation Dataset.

Failed to get the number of CUDA devices

hi,bro,
when i use this code,it occurs problems like this:
"Failed to get the number of CUDA devices: CUDA driver version is insufficient for CUDA runtime version"

what should i do to solve this matter?
thks!

package not found

While trying to run the project, i am unable to install the following packages:

1. CUDA 8.0 with CuDNN 5.1
2. OpenCV 3.1

Please suggest course of action.

i/o/myo explain

Hello, I would like to take your code as a starting point for LVSC. There is something that is not clear to me and it is the meaning of i/o/ myo and exactly how the works in the submission -> tmp = otmp - itmp.

Why is not myo used? I really do not have much idea... Could not you unify all the segmentations of the same image and learn them all together or?

Thank you

The conversion between pixel and millimeter units

Hello, I find the SAX_series, but I can't display the corresponding contour coordinates on the DICOM file correctly. 
From the official document, it shows that each (Xi,Yi) pair representing the coordinates of i pixel in the contour in pixel unites. The conversion factor between pixel and millimeter units can be found by reading the PixelSpacing parameter of a DICOM file.

I obtain the PixelSpacing which is 1.367188, I don't know how to convert.
I am looking forward to your reply.

Run your code on Mac OS X?

Wonder if it's possible to run your code on Mac OS X? I don't have
an Ubuntu 14.04 installed now.

Or do you see some changes to be done in your code to do so?

Thanks

David

how did you perform multi scale center cropping

hi
you have mentioned in your paper in data preparation section,'' we choose multi resolution approach to crop ROI at multiple size that wholly contain ventricles."
where had you performed this in your code, if not how i can do this multi scale cropping of ROI.

How to use multiple graphics cards

How to realize the visualization of segmentation results

Very nice work！