search

ibrahimethemhamamci / DENTEX

MICCAI 2023 & Dental Enumeration and Diagnosis on Panoramic X-rays Challenge
https://dentex.grand-challenge.org/
58 stars 8 forks source link

dataset question #6

Open gustn9609 opened 9 months ago

gustn9609 commented 9 months ago

Hi,

I'd like to pre-train a generative model using this dataset. And I just need a normal (non-disease) dataset.

If so, do I only need to use (a) and (b) in the dataset? In other words, are (a) and (b) all normal data?(what about (d)?)

(a) 693 X-rays labeled for quadrant detection and quadrant classes only, (b) 634 X-rays labeled for tooth detection with quadrant and tooth enumeration classes, (c) 1005 X-rays fully labeled for abnormal tooth detection with quadrant, tooth enumeration, and diagnosis classes. (d) unlabeled

AlexanderPeter commented 8 months ago

a, b, c and d contain diseased teeth. If you want a dataset with only healthy teeth, you have to take c and filter all images out with at least 1 diseased tooth. I think there are 27 images with healthy teeth only, which is maybe not enough for training. But for pretraining it is probably not necessary to have healthy teeth only.