Building a ConvNet for Knee injury classification from MRI exams and interpret its predictions

Поделиться
HTML-код
  • Опубликовано: 25 янв 2025

Комментарии • 18

  • @darpan810
    @darpan810 4 года назад +1

    Hi ,this is amazing project, Can you please help what are input images you take for the model ,I am confused about dataset

  • @ionicguy4919
    @ionicguy4919 4 года назад

    Hi Ahmed, I've completed all parts of your helpful guide, however when applying the logistic regression I get an error that X_val does not exist? Any idea why? I have trained 3 models (however they all give up around 12 epochs for some reason, I could never get it to 35 epochs like yours...? I do get an AUC of around 92 tho.

  • @jonitoh_theone
    @jonitoh_theone 5 лет назад

    Amazing!
    Did you build your interface using React ?

    • @AhmedBesbes
      @AhmedBesbes  5 лет назад

      No, with Dash.
      Code available on my github: github.com/ahmedbesbes/mrnet/tree/master/dash

  • @giri07
    @giri07 5 лет назад

    I go through all your 3 posts
    I have some confusion - this whole project is using alexnet alone or combination of alexnet and some other cnn. Can you clear my confusion

    • @AhmedBesbes
      @AhmedBesbes  4 года назад +1

      This final model is a logistic regression based on the stacking of three AlexNet-Based models.

  • @careymain3036
    @careymain3036 5 лет назад

    What platform did you use to do the Stanford ACL tear data set ? Im using Colab and not sure best way to pull in that dataset its running slow

    • @AhmedBesbes
      @AhmedBesbes  5 лет назад

      Hello Carey,
      I'm using my personal machine equipped with an Nvidia GPU

    • @careymain3036
      @careymain3036 5 лет назад

      @@AhmedBesbes have any suggestions using Colab? or I could switch to AWS?

    • @careymain3036
      @careymain3036 5 лет назад

      @@AhmedBesbes how did you unzip file im getting BadZipFile: Bad magic number for file header - the file header is b'PK\x03\x04' when most zip files are \x50\x4B\x03\x04"

    • @AhmedBesbes
      @AhmedBesbes  5 лет назад

      @@careymain3036 Move to AWS: you can customize your machine and increase its memory / disk. Colab is quite limited.

    • @careymain3036
      @careymain3036 5 лет назад

      @@AhmedBesbes running your code for knee tear im getting the error ---- ImportError Traceback (most recent call last)
      in ()
      17
      18 #from dataloader import MRDataset
      ---> 19 from dataloader import read_data_sets
      20 import model
      21
      /usr/local/lib/python3.6/dist-packages/dataloader/__init__.py in ()
      ----> 1 from dataloader import read_data_sets
      ImportError: cannot import name 'read_data_sets'
      Have you seen this ? I runnning in colab jupyter notebooks