Table 2.

Benchmark dataset B used for model training and evaluation.

Dataset	Number of samples
	Positive	Negative	Total
Training set	746	746	1492
Test set	160	160	320
Total	906	906	1812