04 Image analysis for medical visualization ぱらぱらめくる『Visual Computing for Medicine Theory, Algorithms, and Applications』
- Image segmentationの話(画像領域を分ける)
- 境界を探して決める
- 同一セグメントに帰属する範囲を定めて結果として境界を決める
- Preprocessing
- ROI (Region of Interest) selection
- 補間、規格グリッド適合
- 分布補正〜ヒストグラム変換
- Noise reduction techniques
- Static noise reduction filters
- Gaussian filter
- Separable filters
- Dynamic noise reduction filters
- -filter
- Diffusion filtering
- Filtering on boundaries
- Static noise reduction filters
- Gradient filtering for edge detection
- Segmentation
- Manual segmentation
- Threshold-based segmentation
- Threshold selection
- Probabilistic models
- Multidimensional thresholding
- Connected component analysis
- Region growing
- Watershed segmentation
- Merging basins
- Marker-based watershed
- Graph-based segmentation
- Livewire segmentation
- Cost functionを入れて探索する
- Contour-based segmentaion
- Graph cuts
- Random walker
- Livewire segmentation
- Advanced and model-based segmentation methods
- Active contour models
- Level sets and fast marching methods
- Statistical shape model
- Active appearance models
- Inteaction for segmentation
- 完全自動化は無理!なので、何かしらマニュアルな手を入れる
- Mesh-based補正
- 形の補正
- Validation方法
- 統合
- 同一個人の複数画像データの統合