So I did some Medical Image Computing in my undergraduate engineering course and adored it! Since I'm interested in computer vision (and intelligence) I wanted to see how AI, namely: deep learning can be applied in this Medical computer vision. Hence, motivated by some of the work done by Nvidia as shown in GTC 2018 and deepmind medical research I wanted to apply 3D CNNs for cancer detection.
Therefore in this repository I apply 3D CNN for lung cancer detection on the Kaggle datascience bowl 2017 dataset: https://www.kaggle.com/c/data-science-bowl-2017/data
The data is not that large! Hence, a good idea might be to combine other data sources such as LUNA16 and then run the 3D CNNs. Unfortunately, I am limited to my memory space, computing power and time. Hence, running a large network on a huge dataset for me during my studies was difficult, however I hope you find this helpful.