Skip to main content

Table 3 Performances of the 2-level SVM Classifiers for the Testing Dataset

From: Machine learning-based segmentation of ischemic penumbra by using diffusion tensor metrics in a rat model

 

Accuracy

Range

Median

IC vs. nonIC

96.0 ± 2.3%

88.2 ~ 98.9%

96.7%

IP vs. NT

80.1 ± 8.0%

61.0 ~ 92.6%

81.5%

hemisphere (IC + IP + NT) vs. PDM

88.1 ± 6.7%

69.5 ~ 96.9%

90.0%

Back to article page

Journal of Biomedical Science

ISSN: 1423-0127