Robustness-Driven Feature Selection in Classification of Fibrotic Interstitial Lung Disease Patterns in Computed Tomography Using 3D Texture Features.
Daniel Y. ChongHyun J. KimPechin LoStefano YoungMichael F. McNitt-GrayFereidoun AbtinJonathan G. GoldinMatthew S. BrownPublished in: IEEE Trans. Medical Imaging (2016)
Keyphrases
- lung disease
- texture features
- computed tomography
- feature selection
- feature set
- ct images
- lung nodules
- feature extraction
- medical imaging
- texture analysis
- ct scans
- texture classification
- medical images
- gabor filters
- image reconstruction
- texture feature extraction
- machine learning
- three dimensional
- image features
- texture images
- feature vectors
- gray level
- support vector
- classification accuracy
- local binary pattern
- rotation invariant
- texture patterns
- computed tomography scans
- face recognition
- object recognition
- decision trees
- color features
- textural features
- multiscale
- image analysis
- mixture of gaussian distributions