Table 3.

Classification performance using cross-validation in training data from the MICCAI challenge

	Accuracy
Hang Chang [30]	85.83%
Our method [19]	97.8%
Jocelyn Barker [57]	100.0%