Feature Selection Based on Machine Learning in MRIs for Hippocampal Segmentation

Table 6

Classification result on features selected through forward selection method using Näive Bayes Classifier in 5-fold cross validation based on confusion matrix.

36 features; Forward selection	Completeness of a class	Purity of a class	Contamination of a class
Hippocampus	82%	70%	30%
Not-hippocampus	73%	84%	16%
Efficiency	77%