|
Diagnostic techniques for AMD |
|
Image analysis techniques for AMD | Color fundus image analysis | Texture-based segmentation | Gabor filter and wavelet analysis [33] |
Statistical structural information-based texture analysis [34] |
The Artificial Neural Network (ANN) based supervised classifiers [35] |
Expectation Maximization (EM) based unsupervised classifiers [36] |
Thresholding-based segmentation | Probability based thresholding [37] |
Otsu method for background selection and intensity thresholding [38] |
Histogram based adaptive local thresholding [39] |
Edge based thresholding [11] |
Clustering-based segmentation | Spatial histogram and similarity based classification [40] |
Distance based clustering [41] |
Wavelet based feature extraction & SVM based classification [42] |
Edge and template matching | Gaussian template matching [43] |
Gradient based segmentation [44] |
OCT image analysis techniques | Graph-based segmentation | Graph and dynamic programming [31] |
A graph-based multilayer segmentation approach [45] |
Active contour-based segmentation | Anisotropic noise suppression and deformable splines [46] |
Based on active contours and Markov random fields [47] |
Polynomial curve-fitting based technique | Local convexity condition and fitting second- or fourth-order polynomials [48, 49] |
Based on the distance between the abnormal RPE and the normal RPE floor [19, 50] |
|