moudle error

[cy-yong]

[TheoEst]

Hello, Thanks for your interest to my code. To launch it, you need to be in the parent folder and launch with python -m

For instance :

cd /home/home_folder/
git clone https://github.com/TheoEst/joint_registration_tumor_segmentation.git
python -m joint_registration_tumor_segmentation.main 

[cy-yong]

您好， 感谢您对我的代码的兴趣。要启动它，您需要在父文件夹中并使用python -m

例如：

cd /home/home_folder/
git clone https://github.com/TheoEst/joint_registration_tumor_segmentation.git
python -m joint_registration_tumor_segmentation.main 

I have solved this problem. Unfortunately, I have encountered a new problem. The ". NPY" file is not in the original data. Has your data been preprocessed, and the "-" only brats "on the command line does not seem to be an argument.

[TheoEst]

Hello, The data is not provided in the GitHub. You need to download both BraTS 2018 and OASIS3 before running the code. You can find the datasets on the following link : https://www.oasis-brains.org/ and https://www.med.upenn.edu/sbia/brats2018/data.html

[cy-yong]

您好， GitHub中没有提供数据。 在运行代码之前，您需要下载 BraTS 2018 和 OASIS3。 您可以在以下链接中找到数据集：https : //www.oasis-brains.org/和https://www.med.upenn.edu/sbia/brats2018/data.html I have downloaded this original data set from the official website,Unfortunately, the original data does not contain files in ". NPY" format. I want to know whether this is the preprocessed file format or the original one,thank you very much.

[TheoEst]

During my work, I save all the BraTS dataset into a numpy array. There is no preprocessing, I just load the nii.gz files and save them as .npy. You have two solutions :

• You can change the code to directly load .nii.gz
• Save all the files in .npy as I did.

To export them to .npy, you can use this code :

cohorts = ['HGG', 'LGG']
modalities = ['_flair.nii.gz', '_t1.nii.gz', '_t1ce.nii.gz', '_t2.nii.gz']

for cohort in cohorts:
        cohort_path = main_path+ cohort + '/'
        patients = [file for file in os.listdir(cohort_path) if
                    os.path.isdir(cohort_path + file)]

    for patient in tqdm(patients):
            patient_path = path + patient + '/'

            modalities_array = []

            for modalitie in modalities:

                file = patient_path + patient + modalitie
                image = sitk.ReadImage(file)
                array = sitk.GetArrayFromImage(image)
                modalities_array.append(array)

           array = np.stack(modalities_array)
           np.save(save_path + '/' + cohort + '/' + patient, array)

[cy-yong]

During my work, I save all the BraTS dataset into a numpy array. There is no preprocessing, I just load the nii.gz files and save them as .npy. You have two solutions :

• You can change the code to directly load .nii.gz
• Save all the files in .npy as I did.

To export them to .npy, you can use this code :

cohorts = ['HGG', 'LGG']
modalities = ['_flair.nii.gz', '_t1.nii.gz', '_t1ce.nii.gz', '_t2.nii.gz']

for cohort in cohorts:
        cohort_path = main_path+ cohort + '/'
        patients = [file for file in os.listdir(cohort_path) if
                    os.path.isdir(cohort_path + file)]

    for patient in tqdm(patients):
            patient_path = path + patient + '/'

            modalities_array = []

            for modalitie in modalities:

                file = patient_path + patient + modalitie
                image = sitk.ReadImage(file)
                array = sitk.GetArrayFromImage(image)
                modalities_array.append(array)

           array = np.stack(modalities_array)
           np.save(save_path + '/' + cohort + '/' + patient, array)

Sorry to bother you, I have another problem when I use it The following error occurred while NPY file format.

[TheoEst]

What did you do ? I don't understand, how the problem occurs

[cy-yong]

What did you do ? I don't understand, how the problem occurs

This problem happened because I generated it according to your way The NPY file cannot be recognized. I don't know if it is a process error in generating the file.

[TheoEst]

Normaly, if you do the training with brats file, you should open image with the following line : https://github.com/TheoEst/joint_registration_tumor_segmentation/blob/62b4b25f3c3285a016861e74f3708aca5511f796/Dataset.py#L307

And the function np.load can load .npy files without problem.

[cy-yong]

Normaly, if you do the training with brats file, you should open image with the following line :

https://github.com/TheoEst/joint_registration_tumor_segmentation/blob/62b4b25f3c3285a016861e74f3708aca5511f796/Dataset.py#L307

And the function can load files without problem.np.load``.npy

I found that my problem was generated There is an error in the NPY file. I want to know whether the code you replied to me needs to be run independently or added to a Py file to run.

[TheoEst]

You need to put in a py file and run it independently.

[cy-yong]

您需要放入py文件并独立运行它。 Senior, I think some of the errors may be caused by version problems. Can you provide keras and tensorflow versions of the source code

[HilbertMaximov]

Dear author, how can I solve this problem?

[HilbertMaximov]

您需要放入py文件并独立运行它。 Senior, I think some of the errors may be caused by version problems. Can you provide keras and tensorflow versions of the source code

Hi! How did you solve this problem?

[cy-yong]

您需要放入py文件并独立运行它。 Senior, I think some of the errors may be caused by version problems. Can you provide keras and tensorflow versions of the source code

Hi! How did you solve this problem?

I can not solve this problem

[HilbertMaximov]

您需要放入py文件并独立运行它。 Senior, I think some of the errors may be caused by version problems. Can you provide keras and tensorflow versions of the source code

Hi! How did you solve this problem?

I can not solve this problem

do you meet the problem 'DataGenerator' object has no attribute 'only_close_tumor'? Is this a question about the version?

[cy-yong]

我认为那不是版本问题，应该是代码本身问题吧

------------------ 原始邮件 ------------------ 发件人: "TheoEst/joint_registration_tumor_segmentation" @.>; 发送时间: 2022年6月22日(星期三) 下午4:06 @.>; @.@.>; 主题: Re: [TheoEst/joint_registration_tumor_segmentation] moudle error (Issue #2)

您需要放入py文件并独立运行它。 Senior, I think some of the errors may be caused by version problems. Can you provide keras and tensorflow versions of the source code

Hi! How did you solve this problem?

I can not solve this problem

do you meet the problem 'DataGenerator' object has no attribute 'only_close_tumor'? Is this a question about the version?

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

[HilbertMaximov]

我认为那不是版本问题，应该是代码本身问题吧 … ------------------ 原始邮件 ------------------ 发件人: "TheoEst/joint_registration_tumor_segmentation" @.>; 发送时间: 2022年6月22日(星期三) 下午4:06 @.>; @.@.>; 主题: Re: [TheoEst/joint_registration_tumor_segmentation] moudle error (Issue #2) 您需要放入py文件并独立运行它。 Senior, I think some of the errors may be caused by version problems. Can you provide keras and tensorflow versions of the source code Hi! How did you solve this problem? I can not solve this problem do you meet the problem 'DataGenerator' object has no attribute 'only_close_tumor'? Is this a question about the version? — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

我现在在做配准分割这类的一些东西，能加你个联系方式吗？qq vx啥的

[TheoEst]

Hello, Indeed there is a problem in the code. I will correct it soon best regards

[cy-yong]

Hello, Indeed there is a problem in the code. I will correct it soon best regards

thank you

[HilbertMaximov]

Hello, Indeed there is a problem in the code. I will correct it soon best regards

thank you very much

[TheoEst]

Fix in 2cefdeb1e00c8516bb1382697dd11ccb73cdbdb1

[HilbertMaximov]

Fix in 2cefdeb Dear author, how should I use l2 loss here?