• Solutions to the coding assignment is in the jupyter notebook titled PNL_test_1.ipynb.
  • The notebook has detailed comments on each task and sub-task.
  • Incase there are issues with opening the note book, I have exported the file as a markdown, HTLM and a python file with the same name.
• updown.py contains wrapper functions for the Dropbox API and can be ignored.

The PNL_test_1.ipynb jupyter notebook performs the following operations:

• Task 1:

  • API to download data from Dropbox. This is authenticated using an access token from a personal application in dropbox.com/developers.
  • Anonymize:
    • The date of consent is disguised using a "day offset" that is randomly generated. The offsets are stored in enroll_data_offset_SY.csv.
    • The date of birth is disguised by calculating the age.
  • Upload files: Both csv files (anonymized enrollment data and the offsets) are uploaded back to the dropbox folder.

• Task 2:

  • Register the T1 weighted image to the T1w-atlas file using the ANTs registration tool.
  • Compute volume for each label.
    • Volume is calculated as follows.
      • For each integer label, create a binary mask.
      • Multiply the binary mask with the registered T1 image and calculate the number voxels that are non-zero. This number is defined as the volume. The units of the volume is determined by the size of the volxes in the atlas of interest.
  • Match integer labels to the freesurfer ROI names.
  • Upload the ROI volume calculations.