Table 6. Discriminant function determining disease classification accuracy of multimodal MRI analyses differentiating Alzheimer’s disease dementia and Parkinson’s disease dementia involving dimension reduction.

Modalities	Model accuracy (%)	Features (#)	Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)	Youden’s index (J)
All modalities	95.0	11	95.7	90.5	91.7	95	0.85
T1	91.7	11	91.3	85.7	87.5	90	0.77
DTI	75.0	2	77.8	72.2	70.0	80	0.51
ASL	60.0	9	62.5	58.3	50.0	70	0.21
MRS	60.0	4	66.7	57.1	40.0	80	0.24

Based on comparison between statistical classification and gold-standard diagnosis, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and total model accuracy (Model accuracy) from each discriminant function analysis assessing the ability of each of the MRI modalities to differentiate participants with diagnosis of Alzheimer’s disease from those with diagnosis of Parkinson’s disease Dementia. MRI modalities included: arterial spin labelling (ASL), magnetic resonance spectroscopy (MRS), structural T1, and diffusion tensor imaging (DTI).