Table 2. Ten-fold cross validation classification performance of the trained support vector machine.

	Nagoya University model	Toyama University model
Training data set	Nagoya University	Toyama University
Accuracy	73.3%	72.2%
Sensitivity	64.0%	55.1%
Specificity	82.3%	89.6%
Positive predictive value	78.0%	84.4%
Negative predictive value	70.0%	66.2%