Comparing the Prognostic Value of Stress Myocardial Perfusion Imaging by Conventional and Cadmium-Zinc Telluride Single-Photon Emission Computed Tomography through a Machine Learning Approach

Table 3

Machine learning analysis and statistical comparison through chi square test for proportions on the original dataset.

		Accuracy (%)	Error (%)	Recall (%)	Specificity (%)
Tree	C-SPECT	87.4	12.6	94.4	17.1
	CZT-SPECT	89.0	11.0	97.1	7.32
	value	0.471		0.057	0.177
KNN	C-SPECT	74.4	25.6	78.6	31.7
	CZT-SPECT	80.8	19.2	87.4	14.6
	value	0.021		0.001	0.067
SVM	C-SPECT	85.9	14.1	92.2	21.6
	CZT-SPECT	86.5	13.5	92.6	21.6
	value	0.773		0.597	1.000
NB	C-SPECT	83.4	16.6	89.1	26.8
	CZT-SPECT	84.1	15.9	90.1	24.4
	value	0.787		0.649	0.800
RF	C-SPECT	90.3	9.7	98.5	7.3
	CZT-SPECT	90.1	9.9	99.0	0.0
	value	0.591		0.525	0.078

Abbreviations: KNN: nearest neighbor; SVM: support vector machine; NB: Naïve Bayes; RF: random forests.