Machine-learning on XRay

Tree directories :

Project/images/ contains all original images

Project/imagesAugmented/ contains all augmented images

Project/Data/ contains all scripts to split data set between Test and Training/Validation

Project/Simple/onePathology contains all scripts to manage simple pathology detection

Project/Simple/transferLearning contains all scripts to manage transfer learning

Project/multiPathologies contains all scripts to manage multi pathologies detection

Python scripts parameters:

-p pathology name: one name between all pathologies

-t test type: between 'test' (all test dataset 11145) 'random' (2000 extracted from test dataset) 'image name' (for simple test)

-s shape value: height and weight have same size

-d dimension value: 1 for grayscale 3 for rgb

-m model name (for predict scripts)

Usage example: python AlexNetLike.py -p Cardiomegaly -s 2000 python transferPredict.py -p Cardiomegaly -t test -s 224 -d 3 -m myCardiomegaly10000.h5