A project about RSNA-Intracranial-Hemorrhage-Detection
(Kaggle:https://www.kaggle.com/c/rsna-intracranial-hemorrhage-detection)
dataset:
- (epidural/healthy/intraparenchymal/subarachnoid/subdural) each class has 1000 CT data
use:
- use preprocessing.py, filter.py to preprocessing the training dataset
- use VGGmodel.py or seresnet.py to train model
detail:
- preprocess:
- split CT data to 3 channel PNG data
- filter:
- image preprocessing by opencv(noise reduction, clipping, enhacement, binarization)
- VGGmodel:
- VGG16 model
- sersnet:
- resnet14 model
Don't forget to change the data path in code