Research Article

Rapid Retrieval of Lung Nodule CT Images Based on Hashing and Pruning Methods

Table 1

Description of lung nodule feature extraction.

FeatureDescriptorsRepresentation

Gray3 gray features (gray mean, variance, and entropy)
Morphology7 invariant moments features (Hu moment invariants),
,
5 geometrical features (perimeter, area, roundness, rectangularity, and maximum diameter)
4 medical signs (calcification area and degree, cavitary area, and degree) and Fourier descriptor
Texture14 texture features from GLCM (contrast, angular second moment, entropy, inverse difference moment, etc.) with 4 different angles as well as the mean, variance of them