Atelectasis, Cardiomegaly, Effusion, Infiltration, Mass, Nodule, Pneumonia,
Pneumothorax, Consolidation, Edema, Emphysema, Fibrosis,
Pleural_Thickening and Hernia
Suppose you have an image dataset in a directory which looks like this:
FileName Atelectasis Cardiomegaly Effusion Infiltration Mass Nodule Pneumonia Pneumothorax Consolidation Edema Emphysema Fibrosis Pleural_Th Hernia
jat011.jpg 0 0 0 0 0 1 0 0 0 0 0 0 0 0
kal012.jpg 0 0 0 0 0 0 0 0 0 0 0 0 0 0
kem013.jpg 0 0 0 0 0 0 0 0 0 0 0 0 0 0
mat014.jpg 0 0 0 0 1 0 0 0 0 0 0 0 0 0
mes015.jpg 0 0 1 0 0 0 0 0 1 0 0 0 0 0
seb016.jpg 0 0 0 0 0 0 0 0 1 0 0 0 0 0
she017.jpg 0 0 0 0 0 0 0 0 0 0 0 0 0 0
.
.
.
kas028.jpg 0 0 0 0 0 0 1 0 0 0 0 0 0 0
kid029.jpg 0 0 0 0 0 0 0 1 0 0 0 0 0 0
mek030.jpg 0 0 0 0 1 0 0 0 0 0 0 0 0 0
mul031.jpg 0 0 0 0 0 0 1 0 0 0 0 0 0 0
mus032.jpg 0 0 0 0 0 1 0 0 0 0 0 0 0 0
.
.
.
elf198.jpg 0 0 0 0 1 1 0 0 0 0 0 0 0 0
get199.jpg 0 1 1 0 0 0 0 0 0 1 0 0 0 0
kas200.jpg 0 0 0 0 0 0 0 1 0 0 0 0 0 0
You can use this image dataset for training:
Programming Language
Python
All images in the dataset must have the same shape , it was originally extraced with an image resolution of 664 x 680. Also, you can validate your models using this real-world dataset