Deep reinforcement learning-based image classification achieves perfect testing set accuracy for MRI brain tumors with a training set of only 30 images.
Joseph N. StemberHrithwik ShaluPublished in: CoRR (2021)
Keyphrases
- image classification
- brain tumors
- reinforcement learning
- set of training images
- image features
- magnetic resonance images
- training set
- mri images
- image data
- feature extraction
- test images
- mr images
- image analysis
- ground truth
- deformable registration
- mri scans
- endoscopic images
- test set
- machine learning
- classification accuracy
- image retrieval
- magnetic resonance
- object recognition
- image representation
- tumor segmentation
- medical image analysis
- region of interest
- segmentation method
- high resolution
- feature space
- learning algorithm