Skip to content

Feedback on 'Cardiothoracic Abnormality Assessment' tutorial #23

@peterhcharlton

Description

@peterhcharlton

Thank you very much for the helpful resources - they're excellent.

I have some feedback on the 'Cardiothoracic Abnormality Assessment' tutorial, which I hope might be helpful:

  1. Describing the task: Currently, the task is described as follows: "In this tutorial, we demonstrate how to use PyKale to train models on two low-cost, non-invasive modalities, Chest X-ray (CXR) and 12-lead Electrocardiogram (ECG), for assessing Cardiothoracic Abnormalities." Could we provide further detail? I'll be happy to try to work on this once I've gone through the whole tutorial. Alternatively, perhaps you might have some content from the presentation which will be given during the workshop?
  2. Providing a link to the original paper: Please could we provide a link to the original paper, e.g. "This notebook is based on the work of Suvon et al. (MICCAI 2024), which introduced..."
  3. Selecting Colab runtime: In the paragraph starting 'Additional Notes for Colab', could we add: "In addition, when using Google Colab, please select the 'T4 GPU' hardware accelerator (which can be accessed by clicking the down arrow in the top right corner, and then selecting the 'Change runtime type' option in the dropdown menu)."?
  4. Creating a shortcut to the Google Drive folder: "Note: Please create a shortcut to the following Google Drive folder in your MyDrive:" I didn't know how to create a shortcut. Could more details be provided? e.g. "To create a shortcut: (i) click the link to open the Google Drive folder; (ii) click the folder name at the top to reveal a drop-down menu; (iii) from this drop-down menu, select Organise > Add shortcut; (iv) a dialog box appears titled 'Add shortcut to “EMBC_workshop_data”' - click the 'All locations' tab and then select 'My Drive'; (v) a shortcut to 'EMBC_workshop_data' should now be visible at https://drive.google.com/drive/my-drive."
  5. Clarifying the process for loading the full dataset: In the 'Pre-training Data Loading' section, could the comment # (OPTIONAL) in the first code block of this section be adjusted to provide further detail on what uncommenting this section would do, e.g. # (OPTIONAL - uncomment to load full dataset)?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions