search

vkola-lab / ncomms2022

Multimodal deep learning for Alzheimer's disease dementia assessment
MIT License
95 stars 25 forks source link

Question about demo #2

Closed Eros1on-Aqua closed 1 year ago

Eros1on-Aqua commented 1 year ago

Hi, I ran demo_inference.py and got these results. They are not consistent with the given label. Is this right?

filename ADD_score ADD_pred COG_score COG_pred COG ADD demo1.npy 0.99999815 1 1.0072571 1 1 0 demo2.npy 0.007306823 0 1.2147483 1 0 0 demo3.npy 0.9996529 1 0.53027755 1 2 1 demo4.npy 0.99970007 1 0.7140738 1 2 0

Thanks a lot!

shangranq commented 1 year ago

Hi, thank you for raising the issue! I reproduced the result you got and realized the demo labels may not match with the MRI samples. I have updated the demo cases in the latest commit and also submitted the output file after running demo_inference.py. If you run it one more time, you should be able to get the same output. Thanks!