search

khuhm / deep-kidney-cancer

Deep Learning for End-to-End Kidney Cancer Diagnosis on Multi-Phase Abdominal Computed Tomography
MIT License
15 stars 0 forks source link

Is there registration method code and alignment code? #4

Open Jasolicon opened 8 months ago

Jasolicon commented 8 months ago

Thank you for your work. May I get registration and alignment code? And is the classification model 2d or 3d? Thank you.

khuhm commented 8 months ago

Hi, thank you for your interest in my project. I just have uploaded the code for registration and alignment. You can find them in the registration.py file. See function register() for registration and function crop_patch() for tumor alignment. The classification model works in 2D. Hope this helps.

Best, Kwang-Hyun

Jasolicon commented 8 months ago

Thank you so much! May I ask another question that how did you choose window width and window level? Did you choose different W/L for different pase?

khuhm commented 8 months ago

Of course, I use the range of CT pixel values in [-40, 350] HU, which covers the intensity range of kidneys and tumors in the training set. And I use the same setting for all phases.