Breakup of datasets used during lung nodule detector training

[Rajesh-ParaxialTech]

Hello

The datasets used for training the MONAI lung nodule detector were the LUNA16 (which has around 888 datasets and is a subset of LIDC-IDRI) dataset right ? May I know the approximate breakup used for training and testing ?

Thanks
Rajesh

[ericspod]

@mingxin-zheng @tangy5 may be able to help with this one, I assume this refers to this notebook.

[mingxin-zheng]

Thanks @Rajesh-ParaxialTech for the question. I would defer the question @Can-Zhao about the dataset splitting.

[KumoLiu]

Hi @Rajesh-ParaxialTech, you can refer to the data section in this tutorial: https://github.com/Project-MONAI/tutorials/tree/main/detection#1-data.

Hope it helps, thanks.
(move to discussion for now)

[junxiant]

Hello, I was also looking through the data split. I took fold 0 for an example, used a python script to count the length, and I was getting:

training: 534
validation: 67

Is it correct to assume that there are 534 scans and 67 scans used for training and validation respectively?