Fig. 5From: Neural-network classification of cardiac disease from 31P cardiovascular magnetic resonance spectroscopy measures of creatine kinase energy metabolismThe effect of omitting variables on the ROC analysis for each disease class: control, HF; and MI and HCM without HF. The vertical axis is the net % reduction in area-under-the ROC-curve (AUC) as compared to the (first) 4-variable network that uses all of the [PCr], [ATP], kf and CK flux data. The horizontal axis denotes which of the four are included in the ROC analysisBack to article page